Full issue - B & S Analytik GmbH

INTERNATIONA L 

International Journal 

for Ion Mobility Spectrometry 

SOCIETY 

for 

ION 

MOBILITY 

SP ECTROMETRY 

4(2001)1 

Official publication of the 

International Society for Ion Mobility Spectrometry

International Society for 

Ion Mobility Spectrometry 

ISIMS 2001 

Second Announcement 

and Call for Papers 

10 th International 

Conference on 

Ion Mobility 

Spectrometry 

August 12 - 17, 2001 

Wernigerode 

Town Hall 

National Park Harz

Welcome to 

the 

ISIMS 2001 

10 th International Conference on Ion Mobility Spectrometry 

organised by the Institute of Spectrochemistry and Applied Spectroscopy, Dortmund, and Bruker 

Saxonia Analytik, Leipzig - under the patronage of the International Society for Ion Mobility 

Spectrometry. 

Dates: August 12-17, 2001 

Location: Town Hall of Wernigerode 

Accommodation: Gothisches Haus 

The conference will be held at the over 500 year old historic town hall of 

Wernigerode (see front page). Accommodation will be organised at the 

equally historic hotel "Gothisches Haus", built in the first half of the 15 th 

century, which is situated alongside the town hall. 

All conference correspondence should be addressed to: 

International Society for Ion Mobility Spectrometry, 

c/o Dr. Jörg Ingo Baumbach, Institute of Spectrochemistry and Applied Spectroscopy, 

P.O. Box 10 13 52, D-44013 Dortmund, Germany, 

Phone: +49 231 1392 238, FAX: +49 231 1392 438, E-Mail: Baumbach@ISAS-Dortmund.DE 

or 

Dr. Joachim Stach, Bruker Saxonia Analytik GmbH, Permoserstr. 15, 

D-04318 Leipzig, Germany, 

Phone: +49 341 2431 332, FAX: +49 341 2431 404, E-Mail: JS@BSAX.DE 

Conference announcements will be posted on the Internet http://ims.isas-dortmund.de 

Scientific Programme: 

The scientific programme covers most aspects of ion mobility spectrometry. The main subjects will 

be: 

• fundamental studies, 

• instrument development, 

• miniaturisation, 

• applications, 

• operational aspects, 

• hyphenated IMS techniques (GC-IMS, IMS-MS, ...), 

• data handling and signal processing, 

• data format standards.

Conference Schedule: 

August 12, Sunday, 16:00 - 20:00 

August 12, Sunday, 17:00 

August 12, Sunday, 19:30 

August 13, Monday, 8:30 

August 13, Monday, 8:45 - 16:45 



August 14, Tuesday, 8:30 - 17:30 

August 14, Tuesday, 19:00 

August 15, Wednesday, 8:30 - 17:30 

August 16, Thursday, 9:30 - 11:30 

August 16, Thursday, 12:30 

August 16, Thursday, 20:00 

August 17, Friday, 8:30 - 11:15 

August 17, Friday, 11.15 

Registration 

Meeting of the Steering Committee 

(closed session) 

Get-together - WelcomingReception 

Opening Address 

Lectures 

ISIMS Meeting 

Bowling* 

Lectures 

Visit of the Castle of Wernigerode* 

Lectures 

Poster Session and Exhibition 

Sightseeing Tour 

Conference Dinner 

Lectures 

Closing Remarks 

* limited number of tickets 

Conference Site and Format: 

The conference will be held between August 12 - 17, 2001, in the town hall of Wernigerode. The 

accommodation is organised in the hotel 

Gothisches Haus, 

Marktplatz 2, D-38855 Wernigerode, 

Phone: +49 3942 6750, FAX. +49 3943 675537, E-Mail: gothisches-haus@tc-hotels.de 

A limited number of rooms are reserved. After July 10, rooms not registered will be released to the 

public, late registration is therefore discouraged, as accommodation cannot be guaranteed after this 

date. Extra costs for families must be paid at the time of registration as it will be billed as part of the 

conference package. Additional room expenses, telephone, extra meals, etc. are the responsibility of 

the individual delegates. 

The conference format will be similar to previous years. Contributions presented as lectures and 

posters are welcome. All manufacturers of ion mobility spectrometers and related equipment are 

invited to participate in an industrial exhibition including special presentations of company profiles. 

The registration fee includes accommodation for five nights, Sunday to Friday, breakfast, lunch and 

dinner, welcome reception and conference banquet and the conference proceedings published as 

special issue of the International Journal for Ion Mobility Spectrometry. 

The half-day sightseeing tour is also included with a 

steam train trip up into the Harz mountains and 

the Brocken peak.

Venue: 

A Count of Wernigerode is first mentioned in 1121 in a document of the 

bishop of Halberstadt. The Counts of Wernigerode were first granted 

their city on the 27 April 1229. The town is located in the National Park 

Harz close to the peak Brocken. The town hall 

was mentioned 1277 as "gimnasio vel theatro" 

where the counts of Wernigerode held court and 

other celebrations. One of the oldest buildings in 

town is today´s Hotel "Gothisches Haus". The 

house was built in the second half of the 15th 

century and turned into a restaurant in 1848. A 

famous landmark of the town is the impressive 

castle of the counts of Stolberg-Wernigerode high 

above the city of Wernigerode. The castle was 

built on a promontory of the Agnesberg between 1110 and 1120. 

Around 1730 the castle was turned into a 

Baroque residence. 

Registration Fee: 

Registration and payment 

Regular 

Students 

before June 30, 2001 

1.600 DM 

1.300 DM 

after June 

30, 2001 

1.800 DM 

1.500 DM 

The additional price for spouses and accompanying persons is 700 DM. 

Rooms in the hotel have been reserved for both conference and the weekend before and after to 

allow some sightseeing. These rooms will be available but are not covered by the workshop costs, 

additional days lodging and meals have to be covered individually. 

The registration fee includes accommodation for five nights, Sunday to Friday, breakfast, lunch and 

dinner, coffee breaks, welcome reception and conference banquet and the conference proceedings 

published as special issue of the International Journal for Ion Mobility Spectrometry.

Three methods of payment are acceptable: 

Methods of payment: 

1. Checks made out to ISIMS´2001, payable in DM can be sent directly to ISIMS office in 

Dortmund with the registration form 

2. Payment in DM to Commerzbank Account 3232311, BLZ 44040037, SWIFT Code: 

COBADEFF440, the bank transfer should state explicitely ISIMS´2001, the name and the address 

of the participants, proof of the bank transfer should be enclosed with the registration form. 

3. Payment by VISA, MASTERCARD/EUROCARD, AMEX. 

Abstracts of Papers: 

To enhance the quality of preparation of presentations a 4 page extended abstract should be 

delivered as a PDF-File. If no PDF-Writer is availalble, camera ready files including the figures and 

tables in text in Microsoft WORD, Corel WORDPERFECT or Lotus WORDPRO could be 

accepted if readable. All the extended abstracts will be published in a special issue of the 

International Journal for Ion Mobility Spectrometry. The deadline for submission is July 10, 2001 - 

arrival at ISAS in Dortmund. E-Mail delivery would be prefered: Baumbach@ISAS-Dortmund.DE. 

But please note, that e-mails of more than 1.3 MB will be rejected by the firewall at the ISAS. 

The header of the paper should have the form as used in the International Journal for Ion Mobility 

Spectrometry, for example: 

Text should be arranged including the figures and tables in one single column per page. 

Important note: 

Please note that the submission of a 4 page extended abstract by the deadline is a precondition of 

presenting a paper or poster at the conference. 

Transportation: 

Berlin airport, Frankfurt and Düsseldorf airport are about 200 or 300 km from Wernigerode. A car 

is not essential but helpful if you wish to explore the National Park Harz. Travel service is available 

at the following internet page http://bahn.hafas.de/bin/detect.exe/bin/query.exe/e 

Frankfurt - Wernigerode 

Munich - Wernigerode 

Düsseldorf - Wernigerode 

Berlin - Wernigerode 

Distance by train 

320 km 

4 h - about 150 DM 

530 km 


380 km 


230 km 

3h - about 100 DM 

Distance by car 

335 km 

614 km 

385 km 

231 km

International Steering Committee: 

D.A. Atkinson 

J.I. Baumbach 

H. Bollan 

A. Brittain 

G.A. Eiceman 

St. Harden 

H.H. Hill 

A. Lawrence 

T. Limero 

J. Stach 

P. Thomas 

USA 

Germany 

United Kingdom 


USA 

USA 

USA 

Canada 

USA 

Germany 


Local Organising Comittee: 

J.I. Baumbach 

T. Böhme 

M. Brodacki 

St. Güssgen 

A. Rudolph 

H. Schmidt 

St. Sielemann 

J. Stach 

Important Dates: 

Early registration 

Submission of 4 pages abstract 

June 30, 2001 

July 10, 2001

10 th International Conference on Ion Mobility Spectrometry 

I wish to register for the ISIMS´2001 

Registration Form 

• Mr. 

Family Name: 

• Mrs. 

• Dr. 

• Prof. 

First Name: 

Institution/ 

Company: 

Address: 

City: 

Country: 

Phone: 

FAX: 

E-Mail 

Name(s) of accompanying persons(s): 

I plan a scientific contribution for the 

Title of contribution(s): 

• lectures 

• poster session 

by using credit cards (VISA, MASTERCARD/EUROCARD, AMEX) 

Card holder´s name: ________________________________________ 

Card number: _____________________________________________ 

Expiry date: ___________________ 

Signature: _________________

Table of Contents 

Papers presented at 

9 th International Conference on Ion Mobility Spectrometry, ISIMS 2000, 

Halifax, Nova Scotia, Canada, August 13-16, 2000 

A Miniature Ion Mobility Spectrometer with a 

Pulsed Corona-Discharge Ion Source 

Jun Xu, W. B. Whitten, T. A. Lewis, and J. M. Ramsey 

We report studies of a miniature ion mobility spectrometer (IMS) that employs a pulsed corona 

discharge ion source. IMS spectra were measured as a function of pulse width and height, drift 

field, and various corona ionization configurations. Preliminary results indicate that pulsed corona 

discharge ionization can be used for mm-scale IMS with high sensitivity and good resolution. 

3 

The development and sea trials of a prototype portable ion mobility spectrometer 7 

for monitoring monoethanolamine on board submarines 

H.R. Bollan 1 , J.L. Brokenshire 2 

Monoethanolamine (MEA) is used in equipment installed in submarines for the purpose of scrubbing 

carbon dioxide (CO 2 ) from the enclosed atmosphere. Previously, a prototype fixed-point 

continuous monitor for MEA was developed[1] and subjected to sea trials on a submarine. The 

concentration of MEA was monitored in the CO 2 scrubber compartment over a period of three 

months and found to be within the maximum permissible concentration for a continuous ninetyday 

period (MPC 90 ) for the majority of the patrol. The concentration only briefly exceeded the 

MPC 90 but did not exceed the MPC 24 (MPC for a continuous 24-hour period), and was rendered 

very quickly below the MPC 90 level. Following the success of the continuous monitor, a prototype 

portable MEA monitor was developed and also subjected to sea trials in order to investigate 

the operability of the instrument on board and to determine the atmospheric concentrations of 

MEA accumulated in various compartments within a submarine. The monitor remained fully 

functional throughout the trials performed on a submarine, and calibration was stable in excess of 

eight months spanning the calibration and trials period. The results from the sea trials indicated 

that the atmospheric MEA vapour was more concentrated in the compartment containing the 

CO 2 scrubbers. Despite frequent ventilating of the submarine, the MEA concentrations reached 

equilibrium over a relatively short period of time, but were maintained below exposure limits. 

Ion non-linear drift spectrometer (INLDS) - 

a selective detector for high-speed gas chromatography 

I.A.Buryakov, Yu. N. Kolomiets, V.B. Louppou 

Experimental data on detection of vapours of 2,4-dinitrotoluene (DNT), 2,4,6-trinitrotoluene 

(TNT), pentaerythritol tetranitrate (PETN), cocaine, crack in air with INLDS are presented in the 

report. Calculated detection limit, linearity, INLDS sensitivity on detecting vapours are given. 

Chromatographic analysis of solutions of DNT, TNT, PETN, cocaine, crack, heroin, barbital, 

phenatine dihydrophosphate with GC-INLDS is performed. Retention time of chromatographic 

peaks and peak width at half height are determined. A possibility to decrease separation time 

using selective detector is considered. 

Rapid analysis of pesticides on imported fruits by GC-IONSCAN 

R. DeBono, A. Grigoriev, R. Jackson, R. James, 

F. Kuja, A. T. Le, S. Nacson, A. Rudolph, S.Yin Loveless 

A large number of organic compounds are used today for the control of insects, weeds and 

diseases on fruits and vegetables, and consequently there is a need for analysis of residue 

products, especially on imported fruits and vegetables. Screening for banned pesticides on fruits 

and vegetables requires special methods of analysis. In this respect, gas chromatography (GC) 

has become one of the most important methods for analyzing pesticides and similar compounds, 

due to its separation capabilities and high sensitivity, employing specific detectors such as ECD, 

13 

16 

Copyright © 1998 by International Society for Ion Mobility Spectrometry

mass spectrometry and most recently ion mobility spectrometry. Standard GC investigations, 

however, often require sample work-up and may involve a lengthy analysis. This paper addresses 

the application of a novel solid phase desorption (SPD) add-on module to the Barringer 

GC-IONSCAN system for rapid screening of pesticides on imported fruits. The SPD module 

consists of a sliding tray for placing a filter and a desorber to volatilize the sample into a GC 

column through a heated six port valve. A preconcentrator sample loop traps volatile and 

non-volatile substances during the desorption cycle and releases them by resistive heating into 

the analytical column. The GC oven is temperature programmable from ambient to 300°C and 

offers ramping rates of 1 to 40°C/min. The GC is additionally fitted with a split/splitless injector 

for liquid and gaseous sample injection and a metallic megabore column, resulting in fast analyses. 

... 

The use of IMS and GC/IMS for analysis of Saliva 

Chr. Fuche, A. Gond, D. Collot, C. Faget 

Different methods for testing car drivers on illegal drugs are reviewed. Saliva analysis is preferred 

for practical reasons. Three technologies are compared for the research of narcotics in saliva : 

Ionscan 400, GC/Ionscan and GC/MS 

Strategies for smarter chemical sensors 

P. de B. Harrington, G. Chen, A. Urbas 

Advancements in computer technology are forcing a paradigm shift in the processing of data from 

analytical instrumentation. The traditional approach is to signal average data from an analytical 

instrument once it achieves a steady-state response (i.e., the signal from the instrument 

stabilizes). However, a single signal averaged spectrum from the stabilized instrument response 

excludes temporal information that may be exploited to solve significant chemical problems. This 

paper presents the opportunities furnished by modeling the dynamical or transient responses of 

ion mobility spectrometers. The transient response occurs when the sample is introduced to the 

inlet of the instrument and the sample is removed from the inlet of the instrument. In other 

cases, for which the sample is thermally desorbed, the transient is introduced by the temperature 

time profile of the desorber and sample discrimination that may occur in the transfer lines . A 

useful method that is gaining popularity for modeling dynamic or temporal changes in data is 

SIMPLISMA . Using SIMPLISMA, the ion mobility data are factored into sets of concentration 

profiles and concentration independent spectra. Each feature or IMS peak that varies independently 

with respect to the duration of the measurement will be modeled as a separate component. 

An example will be presented later. ... 

Initial Study of Electrospray Ionization-Ion Mobility Spectrometry 

for the Detection of Metal Cations 

H.M. Dion, L.K. Ackerman, H.H. Hill, Jr. 

ESI-IMS of nine inorganic cation solutions was performed for the first time. Counter ion had a 

large effect upon the sensitivity and response ion identity of the cations studied. Several salt 

solutions yielded one major cation peak including: aluminum sulfate, lanthanum chloride, strontium 

chloride, uranium acetate, uranium nitrate, and zinc sulfate. Aluminum nitrate and zinc 

acetate solutions produced multiple cation peaks, which increased the detection limits and difficulty 

of identification through comparison with the ESI-MS literature. Predicted detection limits 

ranged from 0.33 ppm to 25 ppm depending on the salt solution studied. The identity of the 

species detected is unconfirmed, but literature suggests, and drift times support the detection of 

cation-solvent or cation-solvent-anion complexes. Finally, strontium and lanthanum chloride 

were separated and detected simultaneously with a resolution of 2.2. This is the first research 

showing the use of ESI-IMS as a detection and separation method for metal ion and ion complexes 

and the results from this study warrant future development of ESI-IMS as a field technique for 

the detection of metal contaminants in the environment. 

20 

26 

31 


The use of GC-IMS to analyze high volume vapour samples from cargo containers 

P. Lafontaine, P. Pilon, R. Morrison, P. Neudorfl 

One avenue of narcotics smuggling is by concealment in cargo containers. Canada Customs statistics 

indicate that the value of drugs seized from cargo containers was greatest for cocaine. 

Approximately, a million containers enter Canada each year and a manual inspection of each one 

is impossible. A quick, inexpensive and effective method of container chemical identification 

coupled with intelligence information could target likely containers, serve to interdict drugs and 

provide a deterrent to drug smugglers. This paper describes some development of a high volume 

air sampling method with subsequent chemical analysis by GC-IMS to provide a rapid method for 

cargo container inspection. ... 

Evaluation of sample collectors for ion mobility spectrometry 

N. Mina, S.P. Hernández, F.R. Román, L.A. Rivera 

A series of commercially available filter materials and membranes were studied to investigate 

their affinity with various explosives (RDX, NG, TNT, PETN and DNT) and their 

adsorption/desorption thermal characteristics using a Barringer IonScan 400 IMS. The filters and 

membranes, that withstood the high temperatures of the IMS injector/desorber were made of 

either fiberglass or cellulose, fine or coarse porosity of various pore sizes (0.5-40 mm), and 

medium to fast flow rate. The data suggest that filter material properties such as pore size, 

surface roughness and porosity; flow rate and explosive vapor pressure are parameters that can 

influence the IMS response. The affinity of the explosives for the filter material can also influence 

IMS response. In general, the filters that showed the best responses were those with smaller 

pore size, medium to fine porosity, and medium flow rates. The explosives that showed the best 

IMS responses were those with very low vapor pressure such as PETN and RDX. However the 

data seem to suggest that the affinity of NG for the filter materials is enhancing its signal close to 

the responses seen for RDX and PETN when compared with DNT and TNT that are less volatile 

than NG. 

Principles and applications of a solid phase desorption unit 

coupled to a GC-IONSCAN system 

A. Grigoriev, R. Jackson, F. Kuja, S. Nacson and A. Rudolph* 

The Barringer Solid Phase Desorption (SPD) unit was developed to allow GC-IONSCAN analyses 

of solid samples without prior sample preparation. It attaches to the side of the 

GC-IONSCAN and interfaces with it mechanically and electronically (Figure 1). The liquid sample 

injection capability of the GC is retained for maximum flexibility. Temperatures of the SPD 

desorber, inlet, valve and transfer lines as well as the duration of each stage of the analysis can be 

programmed. The SPD-GC-IONSCAN now offers three analysis methods: Direct desorption 

IMS, liquid injection GC-IMS and solid phase desorption GC-IMS. 

Operational assessment of a handheld ion mobility spectrometry instrument 

Chih-Wu Su, St. Rigdon, T. Noble, M. Donahue, C. Ranslem 

In 1995, the U.S. Coast Guard made the decision to outfit vessel boarding teams with state of the 

art technology to assist them in detecting trace levels of narcotics. Based on an assessment 

conducted by the Research and Development Center (R&DC), ion mobility spectrometry (IMS) 

was selected as the technology best suited for operational use. Specifically, the Coast Guard 

selected the Ionscan, an IMS instrument manufactured by Barringer instruments. A limiting factor 

of all models of the Ionscan is its size and power requirements, weighing between 50 and 60 

pounds. These limitations typically require that the Ionscan be operated on the Coast Guard 

cutter where there is a ready supply of 115VAC power. As a result, samples must be transported 

from the vessel being boarded to the Ionscan for analysis. The impact of this protocol is a 

loss of time resulting in the inability of the boarding officer to quickly determine areas to resampled. 

Additionally, the requirement that the Ionscan operator remain on the cutter results in the 

boarding party being reduced by one person. 

34 

37 

41 

45 


In-situ methylation of methamphetamine during IONSCAN analysis 

Chih-Wu Su, St. Rigdon, Kim Babcock 

The Ionscan, an instrument manufactured by Barringer Instruments based on ion mobility 

spectrometry (IMS) technology, is well known for its field application in illicit drug interdiction. It 

is currently used by the U.S. Coast Guard (CG) and by the Drug Enforcement Agency (DEA) to 

support federal, state and local drug cases. The CG Research and Development Center (R&DC) 

carried out an investigation to determine the source and identity of an extra peak that appeared 

in Ionscan results for Meth, in liquid standard form, on filter paper. It was observed that the 

extra peak, tentatively named Meth-2, appeared just after Meth on the Ionscan plasmagram 

(Figure 1) and had a drift time similar to that for nicotine. In 1996, Ms. Angela DeTulleo of the 

DEA South Central Lab reported that nicotine interfered with the detection of methamphetamine 

(Meth) 1 . In 1999, Dr. Peter Harrington’s group from Ohio University in Athens, Ohio, 

reported a procedure that could pre-separate Meth and nicotine using the difference in their 

vapor pressures and the SIMPLISMA (SIMPL-to-use-Interactive Self-Modeling Mixture Analysis) 

method 2 . Although Dr. Harrington’s method solved the nicotine interference problem, it 

required the use of an additional device as well as extra sample preparation time. The CG 

Research and Development Center (R&DC) has always been involved in the investigation and 

development of fast and easy field applicable methods for CG boarding officers to confirm and 

reduce/eliminate interferences on target compounds detected by the Ionscan. The unknown 

peak called Meth-2 further complicated the observed interference of nicotine with Meth detection. 

In order to investigate the nature of Meth-2, a study was launched to isolate and identify it. 

Consequently, an interesting in-situ methylation process during Ionscan analysis was discovered. 

This paper reports the results of this study as well as some possible applications. 

Ion mobility spectrometry in helium with corona discharge ionization source 

M. Tabrizchi and T. Khayamian 

An ion mobility spectrometer is described using helium as the drift gas and corona discharge as 

its ionization source. The observed ion current is approximately seventy times larger than that 

noted for the conventional system having 63 Ni as the ionization source. The selectivity factors in 

helium are close to those in nitrogen. The relative separation of the peaks in helium is almost 

twice as much as that observed in nitrogen. This results in a major enhancement of the resolution. 

A two-fold increase in the capacity factor (k') for ions in helium is also realized. 

External exit gate Fourier transform ion mobility spectrometry 

Ed Tarver, James F. Stamps, Richard T. Jennings, William F. Siems 

Ion mobility spectrometry (IMS) has been recognized as one of the most sensitive and robust 

techniques for the detection of explosives, narcotics and chemical warfare agents. The extreme 

sensitivity and electronic simplicity of IMS instrumentation has offered tremendous theoretical 

potential for this analytical technique. While the detection limits achievable with IMS are rivaled 

by relatively few alternative methods (e.g., mass spectrometry, biochemical detection, and flame 

photometric or other chemically specific detectors interfaced to gas chromatographs), none can 

match the combination of sensitivity and speed of response. Coupled with low cost, ruggedness 

and atmospheric pressure operation, IMS has promised to be the solution for a variety of analytical 

challenges. We have recently demonstrated improved spectral resolution as well as significant 

signal-to-noise enhancement by implementing the external exit gate Fourier transform technique 

on both commercial and Sandia National Laboratory proprietary IMS instruments. ... 

Miniaturized ion mobility spectrometer 

M. Teepe, J.I. Baumbach, A. Neyer, H. Schmidt, P. Pilzecker 

Miniaturization of an ion mobility spectrometer (IMS) can lead to a handy low cost on-line and 

on-site environmental scanning system for exhausts in chemical, waste and petrochemical industries. 

Ion mobility spectrometry has changed from a not common known method in the late 

sixties to a practical method for the detection of toxic pollutants and explosives and some other 

analytes in the atmosphere at ambient pressure and temperature down to the ppb V (parts per 

billion) concentration range. The size of a common high resolution IMS (in the range of normal 

48 

52 

57 

60 


Personal Computers) makes it useful in process control for industry and in centralised measurement 

systems like gateways at airports. But ion mobility spectrometry is a more powerful detection 

method than it is now used for. Many other applications are proposed in the scientific 

literature. In the growing semiconductor industries, especially for Reactive Ion Etching an on-line 

gas control can lead to a more reliable process. New fields of application are possible like ‘intelligent’ 

on-line control for air conditions in cars and other vehicles. For these applications it is 

useful to have a small, fast, economical and powerful analytical on-site and on-line system. A 

miniaturized IMS with size of today’s common mobile phones will meet such needs. A prototype 

µ-IMS - ten times smaller than conventional ones - with components produced by microstructure-technologies 

is presented in this paper. Successful results using this spectrometer are 

reported. Finally an outlook to a modular system design with respect to mass production is 

shown. 

Evolution of IMS technology within the Australian Customs Service 

G. Webster 

Customs has been using IMS technology since 1997, in the form of the Barringer Ionscan, when 

an air cargo examination team at Sydney International Airport commenced an operational evaluation 

of the equipment. The location of this one unit and the specially selected officers to manage 

it have proven critical to the overall success of the technology. Without the initiative of these 

officers we may not have had the first taste of success that ultimately led to the introduction of 

further units. This initial success involved the seizure of 330 grams of cocaine concealed in the 

frame of a mountain bike. The presence of the cocaine “alarm” confirmed suspicions regarding 

the bike, which led to discovery of the illicit drug. 

VIP sources for ion mobility spectrometry 

H.-R. Döring, G. Arnold, V. L. Budovich 

Ion Mobility Spectrometry (IMS) for the detection of chemical warfare agents (CWAs), toxic 

industrial chemicals (TICs), drugs and explosives is mainly based on the ionization by radioactive 

sources as Ni63, Am 241 and H3, since these sources optimally meet the requirements made on 

a portable device for field use: They are small-sized and very lightweight, they have an extremely 

good mechanical stability and do not require any additional power. They are very reliable while 

displaying an excellent sensitivity with regard to the detection of quite a large number of 

compounds of interest. However, for well-known reasons (radiation safety, disposal problems) 

there is a growing interest in replacing radioactive sources by alternative ionization techniques. In 

the past the most promising candidates for replacing radioactivity were photoionization (PI) and 

corona-discharge ionization (CD). ... 

Relationships for ion dispersion in ion mobility spectrometry 

G.E. Spangler 

In recent years, the field of ion mobility spectrometry (IMS) has grown to include not only linear 

DC-field IMS, but also asymmetric RF-field IMS. These two approaches to ion mobility spectrometry 

produce two types of data that have not yet been correlated with each other. Approaches 

to achieving such a correlation is described here. 

65 

67 

71 


Regular Papers 

A novel Method for the Detection of MTBE: Ion Mobility Spectrometry coupled to 

Multi Capillary Column 

Z. Xie, St. Sielemann, H. Schmidt, J.I. Baumbach 

A combination of an ion mobility spectrometer with radioactive ionization source and equipped 

with a multi capillary column was used as a new analytical method for the detection of MTBE, a 

gasoline additive, which has become a potential water pollution problem. To extract MTBE out of 

the water a membrane extraction unit was set up, which is simple, effective and easy to automate 

with respect to further applications. The analyte was extracted on the one hand directly out of 

the water, thus the membrane was completely steeped in the water. On the other hand, the 

membrane was held in the gas phase over the surface of the water (head space). The minimum 

detectable limit for both methods was about 50 ppb vl of MTBE in water and the reproducibility 

with a standard deviation of 8.9 % (head space) and 11.5 % (aqueous phase) rather high. Finally 

the utilizability of the system for on-site and on-line measurements is briefly discussed. 

Finalisation of a IUPAC/JCAMP-DX data transfer standard 

for ion mobility spectrometry data 

A.N. Davies, J.I.Baumbach, P. Lampen, H. Schmidt 

In the last few years the rapid developments in the field of ion-mobility spectrometry many 

several different sites around the world has made it imperative to establish an agreed protocol 

for the storage and exchange of experimental data. Under the auspices and guidance of IUPAC 

through the Working Party on Spectroscopic Data Standards (JCAMP-DX) and the development 

work of the members of the International Society for Ion Mobility Spectrometry such a standard 

has been developed which is now in it’s final stages. It will be published as the Appendix to this 

paper and placed on the IUPAC web sites for final comment. An example of the usefulness of 

these standards as chemical MIME types for the display of spectroscopic data on the web will be 

described and an example is available for viewing on the JCAMP-DX web site as well as the web 

site of the ISIMS. 

77 

84 


9 th International Conference on 

Ion Mobility Spectrometry 

SOCIETY 

INTERNATIONA L 

for 

ION 

SP ECTROMETRY 

MOBILITY 

ISIMS 2000 

Halifax, Nova Scotia, Canada 

August 13-16, 2000 

Papers presented

A Miniature Ion Mobility Spectrometer with 

a Pulsed Corona-Discharge Ion Source 

Jun Xu, W. B. Whitten, T. A. Lewis, and J. M. Ramsey 

Oak Ridge National Laboratory, P. O. Box 2008, Oak Ridge, TN 37831, USA 

Abstract 

We report studies of a miniature ion mobility 

spectrometer (IMS) that employs a pulsed 

corona discharge ion source. IMS spectra 

were measured as a function of pulse width 

and height, drift field, and various corona 

ionization configurations. Preliminary results 

indicate that pulsed corona discharge ionization 

can be used for mm-scale IMS with high 

sensitivity and good resolution. 

Introduction 

There is a consensus among researchers, 

developers, and users that miniaturization of 

ion mobility spectrometry (IMS) instruments is 

desirable for the next generation of IMS [1-3]. 

We have reported a miniature IMS that has a 

drift channel of 1.7 mm in diameter and 37 mm 

in length [3]. In testing properties of the 

miniature IMS, we used a pulsed Nd:YAG laser 

to generate ions. The miniature IMS has been 

shown to have a reasonably good resolution. 

As the size of IMS instruments is reduced, 

there is a need for miniaturized ionization 

sources as well. 

Although laser ionization is a simple means for 

generating ions, it cannot be integrated with a 

tiny IMS due to the large size of most laser 

systems. For practical use, a new ionization 

source is needed for miniature IMS. The 

conventional IMS instruments use radioactive 

Ni63, which emits electrons with 67 keV kinetic 

energy. The problem associated with this 

method is the low stopping power of the 

high-energy electrons in the small volume of 

miniature IMS channel. In addition, the Ni63 

source is less likely to be accepted in the 

market place because of its radioactive nature. 

Another alternative is photoionization by 

ultraviolet light [4]. This method also shows 

reduced sensitivity in miniature spectrometers 

because the photon interaction volume is small 

in the miniaturized channel. 

Utilization of a pulsed corona discharge ion 

source is being explored in this work. The 

physics of the corona discharge has been 

presented in references [5-7]. The corona tip is 

biased to a high potential, which produces a 

highly non-uniform field that is sufficient to 

generate ions, but insufficient for electrical 

breakdown. Recent work by Tabrizchi [7] has 

demonstrated that corona discharge ionization 

is a very useful method to generate ions for 

conventional IMS instruments. It is plausible to 

use a corona discharge ion source for a 

miniature IMS as well. 

Experimental 

Our experimental setup of a miniature ion 

mobility spectrometer (IMS) that employed a 

pulsed corona discharge ion source is shown in 

Figure 1. The drift channel, 2.5 mm in diameter 

and 47 mm in length, was comprised of 10 

stacked metal electrodes separated by Teflon 

spacers. Nine miniature resistors (DIGI-KEY, 

Thief River Falls, MN), each with 2 MΩ 

resistance and 1% uncertainty, were connected 

between the electrodes to form a voltage 

divider. A power supply (Stanford Research 

Systems, Sunnyvale, CA) provided the drift 

voltage to the end electrode, with the voltage 

being distributed to the intermediate electrodes 

through these resistors. 

A Ni corona tip with end radius of approximately 

25 µm was mounted at the end of the drift 

channel, as shown in Fig.1. The tip, together 

with the second electrode of the IMS channel, 

formed a tip-ring configuration. A pulse was 

generated by a pulse generator (DEI, B1010) 

and amplified to a high voltage varying from 1-3 


J. Xu et al.: „A miniature ion ...”, IJIMS 4(2001)1,3-6, p. 4 

Potential 

H. V. Amp 

H. V. 

Drift Bias 

Pulse Gen. 

Computer 

Figure 1. 

Experimental Setup. Top illustrates the drift potential and 

the corona pulse. 

kV by a high-voltage pulse amplifier (DEI, 

GRX-3K). The pulse width varied from 400 ns 

to 400 µs with a repetition frequency of 20 Hz. 

The high-voltage pule was applied to the 

corona tip, as illustrated in the top of Fig. 1, and 

was superimposed with the drift potential. 

During the period of the high voltage pulse, 

ions were generated and confined in the 

vicinity of the tip. After the pulse, the ions 

moved in the drift field. The corona 

discharge pulse also served as the ion 

injection process to start the mobility 

measurement. 

Corona discharge ions were separated 

according to their mobilities, and reached a 

detector plate (OFHC) located at the end of 

the IMS channel. The ion current was sent 

to a homemade current amplifier coupled 

with a digital oscilloscope (TDS410A, 

Tektronix, Wilsonville, OR). The current 

amplifier was based on an Analog Devices 

preamplifier (Model 549LH). The current 

amplifier had a 1-KHz bandwidth with a 

gain of 0.08 V/nA. The oscilloscope was 

triggered by the corona pulse, which 

served as a start signal for the mobility 

spectra. The oscilloscope digitized and 

averaged the mobility spectra. The spectra 

were subsequently stored on an Apple 

Macintosh computer running Labview 

software. 

Copyright © 2000 by International Society for Ion Mobility Spectrometry 

TDC/ 

O’scope 

I (nA) / V 

Results and Discussions 

The initial current generated by 

the corona tip was studied by 

measuring the current passing 

through the ring electrode which 

was grounded through a 

picoammeter (Keithley, 486). A 

blue glow was observed around 

the tip in the dark room. The 

ratio of the current to the tip bias 

(V), I/V, has been plotted as a 

function of the tip bias, and 

shown in Figure 2. The plot 

shows two linear regions. In the 

low bias region (< 2250 V), the 

corona threshold was 1860 V 

with a slope of 6.85x10 -5 nA/V 2 . 

In the high bias region, the 

threshold was found to be 1960 

V with a slope of 9.23x10 -5 

nA/V 2 . A graph of I/V versus V 

has been referred to as a 

Townsend plot [5, 6], with a 

linear relationship derived for a low current self 

sustained corona discharge. The presence of 

two linear regions with different slope may be 

due to imperfections of the tip, such as an 

asymmetric physical structure. For large tip to 

ring separation, these imperfections may be 

negligible, resulting in a linear Townsend 

0.09 

0.08 

0.07 

0.06 

0.05 

0.04 

0.03 

0.02 

0.01 

0 

1700 1900 2100 2300 2500 2700 2900 

Bias (V) 

Figure 2. 

Townsend plot of corona discharge with tip-ring


0.012 -600V 

8 

Current (V-Eqv.) 

-800V 

-900V 

0.009 

-1000V 

-1100V 

-1300V 

0.006 

-1500V 

-1700V 

0.003 

Velocity (m/sec) 

7 

6 

5 

4 

3 

2 

1 

0 

0 5 10 15 20 25 30 35 40 

E-Field (V/mm) 

0.000 

0.000 0.005 0.010 0.015 0.020 

Time (Sec) 

Figure 3. 

Mobility spectra of negative ions generated by pulsed corona discharge as a function of drift potentials. 

The upper corner plots the velocity of the ions as a function of the drift field. 

relationship. We speculate that the asymmetric 

structure may be more important when the 

distance between the tip and the ring is small (2 

mm in the present case). It should be noted 

that, if the corona electrode is too close to the 

corona electrode, arcing occurs, which may 

ultimately limit how small the IMS can be. 

Ion mobility spectra of negative ions produced 

in air were measured as a function of drift bias, 

as shown in Figure 3. For these 

measurements, the pulse width was fixed at 

1.08 µs and the corona tip bias was +2600V. 

Air at atmospheric pressure and room 

temperature served as both drift and sample 

gas. The velocity of the ion packet was plotted 

as a function of the drift electric field, as shown 

in the upper corner of Fig. 3. The plot clearly 

shows a linear relation between the mobility 

velocity and the drift field, as expected. The 

current detected at the IMS collection electrode 

was found to have a quadratic relation with the 

drift field. This is a typical property for IMS that 

uses corona discharge as ion source, as 

discussed in Ref. 7. 

The mechanism responsible for negative ion 

formation by pulsed corona discharge is 

interpreted as follows: The high field near the 

tip generates both electrons and ions through 

field assisted ionization in air. The electrons 

are captured by O 2, which has an electron 

- 

affinity of 0.44 eV. O 2 ions then react with 

water molecules to form cluster ions, such as 

(H 2O)O 2- . This prediction is consistent with 

mass spectrometry studies which indicate that 

if oxygen is present the dominant reactant ions 

produced in corona discharge are the 

- 

superoxide ion: O 2 and its water clusters 

(H 2O)O 

- 2 [8]. 

Measurement of IMS charge as a function of 

pulse width showed that the yield of negative 

ions detected was optimized at a very narrow 

pulse width and then decreased as the pulse 

width was increased. These data are 

interpreted as indicative of a competitive 

process between the formation of negative ions 



and ion capture by the positively biased tip. A 

resolution of R=13 was achieved with the 

corona discharge miniature IMS under the 

described conditions. 

Conclusions 

A miniature ion mobility spectrometer with a 

pulsed corona discharge ionization source has 

been successfully demonstrated. A linear 

relation between the drift field and ion velocity 

was observed, with resolution comparable to 

that obtained with laser ionization in a drift cell 

of similar dimensions [3], with the potential for 

construction of an instrument package of 

considerably smaller dimensions. 

Acknowledgement 

This research was sponsored by the US DOE, 

Office of Research and Development. Oak 

Ridge National Laboratory is managed by 

UT-Battelle, LLC for the U.S. Department of 

Energy under contract DE-AC05-00OR22725. 

References 

[1] Baumbach, J. I. and Eiceman, G. A. Applied 

Spectrosc. 1999, 53, 338A-355A, and references 

therein. 

[2] R. A. Miller, G. A. Eiceman, E. G. Nazarov, and T. A. 

King, Technical Digest, Solid-State Sensor and 

Actuator Workshop, Hilton Head Island, SC, 2000, 

page 120. 

[3] Jun Xu, William B. Whitten, and J. Michael Ramsey, 

Anal. Chem. In Press 

[4] M. A. Baim, R. L. Eatherton, and H. H. Hill, Jr., Anal. 

Chem. 55, 1761 (1983). 

[5] J. M. Meek and J. D. Craggs, Electric Breakdown of 

Gases(Wiley, Chichster, 1978). 

[6] Yu. P. Raizer, gas Discharge Physcis, (Springer, 

Berlin, 1991). 

[7] M. Tabrizchi, T. Khayamian, and N. Taj, Rev. Sci. 

Instrum., 71, 2321 (2000). 

[8] C. J. Proctor and J. F. J. Todd, Org Mass Spectrom. 

18, 509 (1983). 


The development and sea trials of a prototype portable ion mobility 

spectrometer for monitoring monoethanolamine on board submarines 

H.R. Bollan 1 , J.L. Brokenshire 2 

1 

Sea Technology Group/Submarines, Defence Procurement Agency, MOD Abbey Wood, Bristol, UK 

2 

Graseby Dynamics Ltd., Bushey, Watford, Herts, UK 

ABSTRACT 

Monoethanolamine (MEA) is used in equipment 

installed in submarines for the purpose of 

scrubbing carbon dioxide (CO 2) from the 

enclosed atmosphere. Previously, a prototype 

fixed-point continuous monitor for MEA was 

developed[1] and subjected to sea trials on a 

submarine. The concentration of MEA was 

monitored in the CO 2 scrubber compartment 

over a period of three months and found to be 

within the maximum permissible concentration 

for a continuous ninety-day period (MPC 90) for 

the majority of the patrol. The concentration 

only briefly exceeded the MPC 90 but did not 

exceed the MPC 24 (MPC for a continuous 

24-hour period), and was rendered very quickly 

below the MPC 90 level. Following the success 

of the continuous monitor, a prototype portable 

MEA monitor was developed and also 

subjected to sea trials in order to investigate 

the operability of the instrument on board and 

to determine the atmospheric concentrations of 

MEA accumulated in various compartments 

within a submarine. The monitor remained fully 

functional throughout the trials performed on a 

submarine, and calibration was stable in excess 

of eight months spanning the calibration and 

trials period. The results from the sea trials 

indicated that the atmospheric MEA vapour was 

more concentrated in the compartment 

containing the CO 2 scrubbers. Despite frequent 

ventilating of the submarine, the MEA 

concentrations reached equilibrium over a 

relatively short period of time, but were 

maintained below exposure limits. 

INTRODUCTION 

An instrument designed for the continuous 

fixed-point monitoring of MEA was developed 


[1], evaluated, calibrated, and subjected to sea 

trials in the CO 2 scrubber compartment of a 

submarine. The purpose of the trial was 

two-fold. It was necessary to determine the 

atmospheric concentrations of MEA and to 

prove the functionality of the prototype 

instrument. As reported previously[1] the 

monitor was capable of detecting MEA over a 

range of temperatures and without interference 

from other contaminant species normally found 

in the atmosphere of a submarine, at their 

respective MPC 90s. Due to the success of the 

project, the Ship Support Agency / Marine 

Auxiliary Environmental and Steam Integrated 

Project Team MAES5c (SSA/MAES5c) 

sponsored the Defence Evaluation and 

Research Agency (DERA) Bridgwater 

Laboratories with the development of a 

prototype portable MEA monitor. The purpose 

of this project was to evaluate the operation 

and suitability of the instrument for monitoring 

on board of a different class of submarine, and 

to determine the concentrations not only in the 

scrubber compartment but also in as many 

other compartments as possible. 

The prototype portable MEA monitor was based 

upon the hand-held Chemical Agent Monitor 

(CAM). A CAM was modified by replacement of 

the normal membrane inlet system with a direct 

inlet system, and the dopant chemical was 

changed from acetone to 4-heptanone in order 

to improve resolution, response and recovery 

characteristics. The range of the instrument 

was set up to encompass the MPC 90 and , if 

possible the MPC 24. As well as not being 

affected by interference from normal 

atmospheric contaminants within the submarine 

(already shown possible with the continuous 

MEA monitor), the portable monitor would have 

© British Crown Copyright 2000/MOD - 

Published with the permission of the Controller of Her Britannic Majesty’s Stationery Office

H.R. Bollan and J.L. Brokenshire: „The development and sea trials...”, IJIMS 4(2001)1,7-12, p. 8 

to function under various conditions of airflow, 

pressure, and relative humidity. 

TEST METHODS AND EQUIPMENT 

MEA vapour generation and calibration 

Vapour generators were set up with permeation 

sources installed, to deliver a range of 

concentrations of MEA vapour under conditions 

of dynamic air flow. The MEA vapour was 

sampled from the exhaust of the generator 

using an impinger containing 18 mN methane 

sulphuric acid as the sampling medium. The 

collected samples were then analysed using ion 

chromatography. 

Prototype MEA monitor 

In order to provide the required performance for 

MEA monitoring, including sensitivity, speed of 

response and recovery, and immunity to false 

alarms, a number of changes to a CAM were 

required. These included the replacement of 

the membrane inlet system (which results in 

very long response and recovery times), by a 

capillary inlet system. The recirculating system 

of CAM provides a slight negative pressure 

within the detector, which is used to produce a 

continuous inward flow through the capillary 

inlet. The dimensions of the capillary were 

selected such that the net flow resulted in an 

appropriate rate of sample ingress, while 

maintaining a reasonably dry air stream in the 

detector and hence an acceptable sieve pack 

life. The chosen capillary dimensions result in 

high sample velocity and minimal residence 

time giving rapid response and recovery 

characteristics compared with the membrane 

inlet. 

In order to provide the necessary selectivity, in 

particular the ability to operate in the presence 

of ammonia, the ion molecule chemistry had to 

be changed. The acetone permeation source, 

used in CAM, was removed and a special 

adaptor fitted in its place. This adaptor carried 

inlet and outlet flow tubes to a much larger 

permeation source containing 4-heptanone, 

external to the sieve pack. This higher 

molecular weight ketone generated at greater 

concentration than acetone results in the 

dynamic range and degree of selectivity 

required for the MEA monitor. The MEA 

monitor was programmed to give the required 

response characteristics by removing the 

existing data from the CAM EEPROM and 

replacing with appropriate mobility windows and 

look up tables for MEA. The look up tables are 

used to convert the signal from the collector 

electrode to display output. 

Prototype portable MEA monitor testing 

The MEA monitor was connected to a personal 

computer, programmed with the necessary 

software to record trials data, according to the 

instructions provided. Details of each test were 

recorded into a note file and the computer set 

to collect data. The MEA monitor was exposed 

to the MEA vapour source until a stable MEA 

product ion peak was observed in the ion 

mobility spectrum; the spectrum was stored 

and displayed on the PC. With the nozzle 

protective cap in place, collection of ion mobility 

spectra was continued until full recovery of the 

response was observed. 

EXPERIMENTAL 

Pre-trials evaluation 

Response and recovery characteristics were 

determined from ion mobility spectra collected 

during exposure cycles to the calibrated levels 

of MEA ranging from 0.16 vpm to 2.69 vpm. 

The data recorded included RIP and MEA peak 

drift times and amplitudes for exposure to a 

nominal concentration of 0.5 vpm of MEA 

vapour (equivalent to the MPC 90), determined 

as 0.47 vpm MEA. 

Sea trials 

The monitor and portable computer were taken 

on board an attack submarine. Readings and 

spectra were recorded in the various 

compartments in the submarine. The data was 

returned to the laboratory for post trials 

analysis, 

Post trials calibration 

The monitor was returned to DERA Bridgwater 

for calibration checks. 

RESULTS 

Pre-trials evaluation 

The MEA peak amplitudes as a function of 

time, when sampling 0.47 vpm MEA, are shown 

in figure 1. The MEA peak amplitude stabilised 

very quickly. The ratio of MEA peak to RIP drift 

times varied over the range 1.222 to 1.179; in 

figure 2 it can be seen that the ratio is relatively 

constant throughout the monitor operation. 



MEA peak amplitude 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 100 200 300 400 500 

Elapsed time/mins 

Figure 1: MEA peak amplitude as a function of time 

Response and recovery 

The rates of response and recovery of the 

monitor to MEA as a function of concentration 

are shown in figures 3 and 4. The MEA peak 

amplitudes for the pre-trials evaluation 

concentrations are summarised in table 1 and 

the calibration curve plotted in figure 5. 

Dynamic range 

Figure 6 shows the equilibrium response 

spectrum 2.69 vpm MEA. At 2.69 vpm the RIP 

was not depleted, showing that the dynamic 

range could be extended to a higher 

concentration of MEA monitoring, but that at 

the higher level the sensitivity was decreased 

i.e. the change in amplitude became smaller 

with increase in MEA concentration. 

Sea trials 

The monitor remained functional throughout the 

trials and showed that the levels of MEA varied 

throughout the submarine, the highest 

concentration was in the compartment housing 

the CO 2 scrubbers, diminishing with distance 

from this compartment, and showed a 

considerable homogeneity away from the 

immediate vicinity of the scrubber 

compartment. There were some areas where 

no MEA peak was recorded in the spectra. 

2 

Ratio of MEA to RIP drift times 

1.5 

1 

0.5 

0 

0 100 200 300 400 500 

Elapsed time/mins 

Figure 2: Ratio of MEA peak to RIP drift times as a function of time 




90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 50 100 150 200 250 300 

Elapsed time/secs 

0.16 vpm 

0.25 vpm 

0.47 vpm 

0.82 vpm 

1.14 vpm 

2.69 vpm 

Figure 3: Rate of monitor response to MEA as a function of concentration 


90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 50 100 150 200 250 300 

Elapsed time/secs 

0.16 vpm 

0.25 vpm 

0.47 vpm 

0.82 vpm 

1.14 vpm 

2.69 vpm 

Figure 4: Rate of monitor recovery from MEA as a function of concentration 

Table 1: Summary of equilibrium responses to MEA as a function of concentration 

MEA concentration 

(vpm) 

0.16 

0.25 

0.47 

0.82 

1.14 

2.69 

Equilibrium MEA 

peak amplitude 

19 

25 

41 

52 

64 

89 

Time to 

equilibrium 

response (min) 

2½ 

1½ 

2 

2 

1½ 

2 

Time to recovery 

(sec) 

45 

180 



90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 0.5 1 1.5 2 2.5 3 

Figure 5: MEA peak amplitude as a function of concentration 

Post trials calibration check 

The calibration check of the monitor after 

completion of the trials indicated that the 

calibration was satisfactory. 

DISCUSSION 

The MEA Monitor is a hand-held instrument of 

dimensions 390 mm x 145 mm x 80 mm, 

weighing approximately 2.5 Kg, including a 6V 

Ni/Cd rechargeable battery. The replacement of 

the membrane, which requires heating, with a 

capillary inlet results in less power being 

drained from the battery , thus extending 

battery life compared with CAM. The life of the 

battery under these conditions has not yet been 

determined. There is an automatic battery 

check so if the voltage becomes too low a 

warning symbol is displayed. 

The monitor was designed to determine the 

concentration of monoethanolamine vapour 

and is ideally suited for use in submarine 

environments. The MEA monitor incorporates 

the widely proven detection system from the 

original CAM, with enhanced sampling and 

detection systems, which provide improved 

response and recovery characteristics and 

greater interference rejection. 

Amplitude 

200 

150 

100 

RIP: drift time 10.282 

ms amplitude 115 

units 

MEA: drift time 12.093 ms 

amplitude 89 units 

50 

0 

1 8 

15 22 29 36 43 50 57 64 71 78 85 92 99 106113120127134141148155162169176183190197204211218225232239246253 

Drift time 

Figure 6: Ion mobility spectrum for equilibrium response to 2.69 vpm MEA 



The specification drawn from the trials is as 

follows: limit of detection 0.05 vpm; dynamic 

range 0.05 – 2.5 vpm; 90 % response < 2 

minutes (typical); 90 % recovery < 4 minutes 

(typical); environmental temperature 0C to + 

40C operating and - 30C to + 70C storage. 

The response and recovery characteristics 

emphasise the difficulties of monitoring high 

surface active materials such as MEA where 

prolonged exposure to high concentrations 

results in longer recovery times, and exposure 

to lower concentrations results in longer initial 

response times. 

The analysis of the trials data showed that the 

bar responses were always equivalent to 

between four and seven bars, out of a possible 

maximum of eight bars, equivalent to a range 

of 0.5 vpm to 2.5 vpm in the scrubber 

compartment. The MEA concentration in one 

other compartment reached 0.25 vpm on one 

occasion, but generally only trace levels of 

MEA vapour, between 0.05 vpm and 0.1 vpm, 

were detected in the other compartments 

where MEA vapour was registered. This 

indicated that MEA vapour spread into other 

parts of the submarine adjacent to or linked to 

the scrubber compartment, or that some MEA 

is being carried over from the scrubber. 

CONCLUSIONS 

The development of the MEA monitor as a 

convenient, hand-held robust means of 

detecting MEA vapour in real time was 

successful. The monitor was capable of 

detecting the required dynamic range and had 

an appropriate limit of detection. 

As MEA has been recorded at various 

concentrations in the submarine atmosphere 

there is a requirement for localised real-time 

monitoring of the vapour. 

The Monitor was not only useful for determining 

MEA concentrations but also the length of time 

required for MEA concentrations to accumulate. 

REFERENCES 

[1] H R Bollan, D J West, J L Brokenshire ”Assessment 

of ion mobility spectrometry for monitoring 

monoethanolamine in recycled atmospheres” Int. J. 

for IMS 1(1998)1 

ACKNOWLEDGEMENTS 

The preparation and presentation of this paper 

was sponsored by the Defence Procurement 

Agency, Sea Technology Group/Submarines. 

The work was sponsored by the Ship Support 

Agency, Marine Auxiliary Environmental and 

Steam Integrated Project Team MAES5c 

(SSA/MAES235c), MoD Foxhill, Bath, under 

contract to the Defence Evaluation and 

Research Agency (DERA), Bridgwater 

Laboratories, with sub contract development 

assigned to Graseby Dynamics Ltd., Bushey, 

Watford. The sea trials were performed by Lt 

Cdr M H Lunn RN, MAES5c, and laboratory 

work was carried out by Tim Kelley and David 

West of DERA Bridgwater. 


Ion non-linear drift spectrometer (INLDS) - a selective detector for 

high-speed gas chromatography 

I.A.Buryakov, Yu. N. Kolomiets, V.B. Louppou 

The Design & Technological Institute of Instrument Engineering for Geophysics and Ecology, the Siberian Branch of 

RAS, 3/6 Pr. Ak. Koptyuga, 630090, Novosibirsk, Russia 

Abstract 

Experimental data on detection of vapours of 

2,4-dinitrotoluene (DNT), 2,4,6-trinitrotoluene 

(TNT), pentaerythritol tetranitrate (PETN), 

cocaine, crack in air with INLDS are presented 

in the report. Calculated detection limit, 

linearity, INLDS sensitivity on detecting vapours 

are given. Chromatographic analysis of 

solutions of DNT, TNT, PETN, cocaine, crack, 

heroin, barbital, phenatine dihydrophosphate 

with GC-INLDS is performed. Retention time of 

chromatographic peaks and peak width at half 

height are determined. A possibility to decrease 

separation time using selective detector is 

considered. 

DESCRIPTION OF THE TECHNIQUE 

The most efficient means for high-speed 

detection of chemical compounds in air are gas 

chromatographs. In the field of analytical 

devices the trend today is to decrease 

separation time. Such fast chromatographic 

separation is completed with a relatively low 

resolution (∼10 3 theoretical plates). Therefore, 

to ensure a highly reliable detection detectors 

with a high selectivity are required. Ion 

non-linear drift spectrometer (INLDS) meets 

these requirements. 

INLDS operation consists in sampling, sample 

ionization, ion separation in a carrier gas 

stream in a strong electric field and registration 

of separated ions [1,2]. A mixture of ions of 

different types is separated in INLDS by the 

electric field strength dependence of the 

mobility coefficient. Ion drift velocity V d, caused 

by an action of electric field is [3]: 

V d = K(0) (1+α(E)) E, (1) 

where K(0) is the mobility coefficient in a weak 

field, E is electric field strength, α(E) is 

normalized function which describes the electric 

field dependence of the mobility. 

Under the action of periodic alternating 

asymmetric waveform field E d (t) that meets the 

conditions: 

T 

∫ 

0 

2b 

, < E (t) >≠0, 

+ 1 

E (t)dt E (t) > = 0 

d 

≡< 

d 

(2) 

(b>1 is an integer) ions executing oscillatory 

motions (period T) drift with velocity V i [1]. 

The drift is compensated by constant field: 

E c = /(1++) (3). 

With E c changing and E d(t) given spectrum of a 

mixture of ions of all types is recorded [2]. 

High selectivity permits the use of INLDS as a 

detector in a gas chromatograph and allows 

selective detection of chromatographic fraction, 

which reduces separation time and 

requirements placed upon the column 

efficiency. This is revealed with the use of GC 

with a high-speed multicapillary column (MCC) 

[4]. 

EXPERIMENTAL 

Block diagram of GC-INLDS is given on Fig.1. 

INLDS comprises an ionization chamber with 

63 

Ni, an ion separation chamber formed by two 

electrodes which are coaxial cylinders 7 cm in 

length, 1 and 0.68 cm in diameter, an ion 

detection system and a voltage generator. Ions 

with purified air with flow rate Q t=50 cm 3 /sec 

are transported through the separation 

chamber. The generator has the following 

parameters: voltage form 

f(t)= (sin[π⋅(t-mT) /τ]-2τ / πT) / (1-2τ / πT), with 

mT ≤ t ≤ (mT+τ), f(t)=- (2τ / πT) / (1-2τ / πT), 

with (mT+τ) ≤ t ≤ (m+1)T (m ≥ 0 - is an integer), 

d 


Buryakov et al.: „Ion non-linear drift spectrometer...”, IJIMS 4(2001)1, 13-15, p. 14 

8 1 2 3 4 

9 

10 

11 

12 

± 

7 

5 

U d 

U c 

6 

I 

temperature T c was 175°C, column gas 

flow rate was Q c = 0.8 cm 3 /s. 

Values of saturated vapour 

concentration C eq of explosives [5]: 

DNT-55.7ppb (v/v), TNT -9.4ppb, 

PETN-18ppt and cocaine-300ppt. For 

chromatographic studies we used 

acetone solutions of (g/l): DNT-3x10 -3 , 

TNT-2,5x10 -4 , PETN-1,4x10 -4 , 

trotyl-3x10 -4 , ethanol solutions of: 

barbital 

((C 2H 5) 2CCONHCONHCO)-1.41, 

crack-1, heroin-0.88, phenatine 

dihydrophosphate 

(C 15H 16N 2O.2H 3PO 4)-1.07, methanol 

solution of cocaine-1 of ”Alltech”. 

Figure1: 

Block-diagram of GC-INLDS : 

1-ionizer, 2- 63 Ni, 3-separator, 4-electrodes, 5-generator, 

6-collector, 7-electrometer, 8-syringe injector, 9-MCC, 

10-oven, 11-pneumatic switch, 12-compound. 

high-voltage pulse duration τ = 1.9 µsec, T = 5,9 

µsec, dispersion voltage amplitude U d= 1÷4 kV. 

Sampling flow rate Q ex=5 cm 3 /s. 

GC comprising a syringe injector (volume of 

1µl), a chromatographic oven and MCC 0.2 m 

in length containing 1000 capillaries 4x10 -5 m in 

diameter coated with phase SE-30 was used. 

As a carrier-gas purified air was used. The time 

of sample injection was 0.3 s. Column 

Ion Current, r.a. 

-15 -5 5 15 25 35 

U c , V 

Figure 2: 

Drift-spectra of air containing vapours of: a) 

phenatine, b) barbital, c) crack (positive mode) and 

d) DNT, e) TNT, f) PETN (negative mode). 

a) 

b) 

c) 

d) 

e) 

f) 

RESULTS AND DISCUSSION 

Selective detection of explosives 

and drug vapours in air with INLDS. 

Drift spectra of air containing vapours 

of compounds: DNT, TNT, PETN, 

phenatine, barbital obtained under 

negative mode and vapours of cocaine and 

crack under positive mode are given on Fig.2. 

No difference between the drift spectra of pure 

air and air containing heroin vapours was 

revealed. Reference to Fig. 2 shows that peaks 

of explosives are completely separated from 

one another and from background. On 

analysing phenatine and barbital several types 

of ions are detected. Peaks of cocaine and 

crack completely coincide with one another and 

partially overlap with the background. It should 

be mentioned that to detect explosives, 

barbital, phenatine positive U c is to be applied 

(α(E) >0). U c is negative for ions of cocaine. 

Analytical characteristics of INLDS. The 

concentration dependence of the detector 

signal is expressed by: I=D×C k , where D is a 

constant, k refers to sensitivity coefficient, C is 

the compound concentration. Calculated 

detection limit (L) and the concentration 

dependence I(C) are presented in Table 2 

(columns 2, 3). The double width of zero line 

was 4 fA. With the given k values the range 

width was over 10 3 on detecting DNT and TNT 

and equal to ~ 300 on detecting cocaine. The 

detection time of one compound did not exceed 

1 s. 

Chromatographic analysis of solutions. 

Retention times (t R) and peak width at half 

height of (µ 1/2) detected with GC-INLDS along 


Buryakov et al.: „Ion non-linear drift spectrometer...”, IJIMS 4(2001)1, 13-15, p. 15 

r.a 

1 

0 

1.7 

U c , V 

2.6 

3.5 

0 

10 

20 

t, s 

a) 

r.a. 

1 

0 

U c ,V 

-10 -5 

0 

10 

50 

40 

30 

20 

t, s 

b) 

Figure 3: 

Chromato-drift-spectra of solutions of: a) military trotyl consisting of DNT and TNT (U d=-1.8 kV. Q c=0.5 

cm 3 /s and T c=150°C), b) criminal crack. (U d=-4 kV). 

with the fundamental parameters of INLDS, 

such as: ion sign, compensation voltage (U c), 

peak width at half height of (DU c1/2), current 

amplitude on detecting chromatographic 

compound fraction (I) are presented in Table 1. 

One can see from the tabulated data that the 

lower is the retention time, the lower is the 

column efficiency. 

The results obtained with GC-INLDS upon 

scanning voltage U c over a chosen range of 

values at regular intervals are presented as 

three-dimensional chromato-drift-spectra, 

where U c of INLDS is plotted on the abscissa, t R 

is laid off as ordinate and I is plotted on Z axis. 

Chromato-drift-spectra of solutions of: a) 

military trotyl consisting of DNT and TNT, b) 

criminal crack are given on Fig.3. The time of 

recording one drift-spectrum is 1 sec. It is 

evident from Fig.3à that peaks of DNT and TNT 

are completely separated both in retention time 

and compensation voltage U c. Fig.3b shows 

sufficient degree of separation between peaks 

of cocaine and background. 

The results of testing have shown a possibility 

for selective detection of vapours DNT, TNT, 

PETN, cocaine, crack, barbital, heroin, 

phenatine with INLDS and GC(MCC)-INLDS. 

REFERENCES 

[1] M.P. Gorshkov. Inventor’s Certificate of USSR, No. 

966583 (1982). 

[2] I.A. Buryakov, E.V. Krylov, E.G. Nazarov, U.Kh. 

Rasulev. Int. J. of Mass Spec. and Ion Pros. 128: 

143-48 (1993). 

[3] E.A. Mason and E.W. McDaniel. John Wiley & Sons, 

New York, 1988, p. 160. 

[4] V.P. Soldatov, I.I. Naumenko, A.P. Ephimenko. Pat.of 

Russia, No. 16512000 (1991).. 

[5] V.C. Dionne, D.P. Roundbehler, E.K. Achter, R.J. 

Hobbs, D.H. Fine. J. of Ener. Mater. 4: 447 (1986). 

Table 1: Parameters of GC-INLDS on detecting solutions of explosives and drugs. 

Compound 

DNT 

TNT 

PETN 

Cocaine 

Crack 

Barbital 

Heroin 

Phenatine 

L, (ppt) 

2 

0.4 

0.6 

0.1 

I(fA) =D·C k (ppt) 

I =3×C 0.6 

I =12×C 0.7 

I =9×C 0.85 

I =40×C 0.85 

t R, c 

4.3 

7.2 

10 

51 

15 

206 

49 

m 1/2, c 

0.65 

0.8 

1 

3.2 

3 

9.4 

4.7 

Sign 

- 

- 

- 

+ 

- 

+ 

- 

U c, B 

11.5 

7.2 

2.1 

-3.3 

31 

3.85 

3.3 

DU c1/2, B 

0.9 

0.85 

0.8 

1.7 

6.5 

3.1 

1.4 

I, fA 

680 

580 

160 

1700 

2200 

30 

260 

60 


Rapid analysis of pesticides on imported fruits by GC-IONSCAN 

R. DeBono 2 , A. Grigoriev 1 , R. Jackson 1 , R. James 1 , F. Kuja 1 , A. Loveless 1 

T. Le 2 , S. Nacson 1 , A. Rudolph 1 and S.Yin 2 

1 

Barringer Research Ltd., 1730 Aimco Boulevard, Mississauga, Ontario L4W 1V1, Canada 

2 

Barringer Instrument Inc., 30 Technology Drive, Warren, New Jersey 07059, USA 

Introduction 

A large number of organic compounds are used 

today for the control of insects, weeds and 

diseases on fruits and vegetables, and 

consequently there is a need for analysis of 

residue products, especially on imported fruits 

and vegetables. Screening for banned 

pesticides on fruits and vegetables requires 

special methods of analysis. In this respect, 

gas chromatography (GC) has become one of 

the most important methods for analyzing 

pesticides and similar compounds, due to its 

separation capabilities and high sensitivity, 

employing specific detectors such as ECD, 

mass spectrometry and most recently ion 

mobility spectrometry. Standard GC 

investigations, however, often require sample 

work-up and may involve a lengthy analysis. 

This paper addresses the application of a novel 

solid phase desorption (SPD) add-on module to 

the Barringer GC-IONSCAN system for rapid 

screening of pesticides on imported fruits. The 

SPD module consists of a sliding tray for 

placing a filter and a desorber to volatilize the 

sample into a GC column through a heated six 

port valve. A preconcentrator sample loop 

traps volatile and non-volatile substances 

during the desorption cycle and releases them 

by resistive heating into the analytical column. 

The GC oven is temperature programmable 

from ambient to 300°C and offers ramping 

rates of 1 to 40°C/min. The GC is additionally 

fitted with a split/splitless injector for liquid and 

gaseous sample injection and a metallic 

megabore column, resulting in fast analyses. 

Sample investigation involves wiping the 

exterior of a fruit with a Teflon or fiberglass filter 

and placing the filter onto the sliding tray of the 

SPD module. The analysis is initiated by 

sliding the tray into the SPD module. Total 

cycle time varies depending on the number of 

pesticides being examined and is in the range 

of 5 to 10 minutes per sample. 

Alternatively, the collected filter or a sample 

directly can be extracted with 1mL of acetone 

or other appropriate solvents, and an aliquot, 

usually 1-5µL, is injected into the heated 

injector of the GC-IONSCAN system. The 

instrument also enables direct thermal 

desorption of the filter into the IMS (non-GC 

mode) for rapid sample screening when the 

sample matrix is not too complex. 

Most pesticides examined in this study were 

detected in the negative ion mobility mode 

where they were ionized through an 

electron-capture process in the reaction region 

by a 63 Ni ionization source. Other pesticides 

may be examined in the positive ion mode, 

where ionization occurred by proton transfer 

from the chemical ionization reagent 

nicotinamide. 

Detection limits at picogram levels for 

pesticides listed in the EPA 8081 method were 

demonstrated. The identification of pesticides in 

the samples was based on their GC retention 

time and specific reduced mobility K o. 

Experimental 

Figure 1 shows a schematic plumbing diagram 

of the instrument with the three modes of 

sample introduction. 

The GC-IONSCAN was operated in negative 

ion mode with a drift tube temperature of 

105°C. The drift flow was 300cc/min, the inlet 

temperature was 240°C and the desorber 

temperature was 225°C. 

The SPD operating conditions were as follows: 

Inlet temperature 240°C, desorber temperature 

260°C, transfer line temperature from desorber 


R. DeBono et al.: „Rapid analysis of pesticides...”, IJIMS 4(2001)1,16-19, p. 17 

Solid Phase Desorption GC-IMS 

N 2 Ccarrier Gas 

Filter Desorber 

Preconcentration Loop 

GC Oven 

Megabore Column 

Drift Gas + Calibrant 

6 Port Switching Valve 

Exhaust 

N 2 Carrier Gas 

Heated Injector 

IMS 

Modes of Operation 

Direct Sample Desorption 

into IMS 

1) Direct thermal desorption into IMS 

2) Liquid/gas injection through GC column 

Purge Gas + CI Reagent 

3) Filter desorption onto the GC column 

Figure 1: Schematic Diagram of Barringer’s SPD-GC-IONSCAN 

to valve 240°C, valve temperature 220°C, 

transfer line temperature from valve to GC 

220°C, desorption time 15s, purge time 5s, loop 

cooling time and back purge 100s, loop heating 

time 6.5s. 

The GC. operating conditions were as follows: 

GC column 10m DB-5, 0.53 mm i.d., 1µm film 

thickness, nitrogen carrier gas at 10psi, 

temperature programming 80°C for 40s, 

ramping at 15°C/min to 240°C. 

Results 

A typical analysis of a mixture of pesticides is 

shown in Figure 2. The cis and trans isomers 

of chlordane are resolved under the GC 

conditions employed as their different collision 

cross sections result in different interactions 

between the molecules and the stationary and 

mobile GC phases. The cis and trans isomers 

also result in different IMS spectra 

(plasmagrams), where the trans isomer has a 

longer drift time due to its larger collision cross 

section and increased interaction with the drift 

gas. ß-BHC and γ-BHC were not resolved by 

the GC; however, their respective reduced 

mobilities and IMS drift times were significantly 

different to allow identification. 

Each chromatographic peak shown in Figure 2 

produces a specific plasmagram profile and an 

example of this feature is shown in Figure 3 for 

cis- and trans-chlordane. 

Pesticides were detected by wiping spiked and 

unspiked apples and oranges, using both 

fibreglass and Teflon filters. γ-BHC residues 



Figure 2 Chromatogram of a Mixture of Pesticides 

Figure 3 Plasmagram of cis-Chlordane (left) and trans-Chlordane (right) 



1.0% for parathion and 5.9% for ß-BHC. With a 

fibreglass filter, efficiencies were 4.1% and 

2.8%, respectively. 

Table 1 lists the reduced mobilities, limit of 

detection (LOD) and the dynamic range of the 

pesticides investigated. The linear range can 

be easily extended by simply using the split 

feature of the GC-IONSCAN, a split ratio of 5:1 

would extend the upper limit to 30 ng. 

Detection limits for pesticides analyzed in the 

negative ion mode ranged from 10 to 300 pg, 

with linear ranges of 10 to 6000 pg. 

Figure 4: Analysis of Orange 

Conclusions 

A rapid method for screening fruits for 

pesticides has been developed using a novel 

solid phase desorption for the GC-IONSCAN 

system. Banned pesticides in Canada, US and 

Pesticide 

ß-BHC 

γ-BHC 

cis-Chlordane 

trans-Chlordane 

Endosulfan 

Endrin Ketone 

Parathion 

Mol. Mass 

291 

291 

410 

410 

407 

381 

291 

were found on the skin of unspiked oranges 

purchased in local supermarkets (Figure 4). 

Evaluation of the transfer efficiencies of two 

pesticides from spiked apples were determined. 

Using a Teflon filter, the transfer efficiency was 

Table 1: Pesticides in This Study 

Red. Mobility 

1.2650 

1.2550 

1.0840 

1.0616 

1.0930 

1.0820 

1.2740 

Ion 

Observed 

M − 

M − 

M − 

M − 

M − 

M − 

M − 

LOD 

10 pg 

100 pg 

50 pg 

25 pg 

300 pg 

200 pg 

20 pg 

Linear Range 

10-300 pg 

100-5000 pg 

50-5000 pg 

25-1000 pg 

300-6000 pg 

200-1500 pg 

20-300 pg 

Europe can be programmed into the analyzer, 

allowing routine screening of these pesticides 

in imported fruits and vegetables. Processing 

time per sample is typically under 10 minutes 

with detection limits in the low picogram levels. 


The use of IMS and GC/IMS for analysis of Saliva 

Chr. Fuche 1 , A. Gond 1 , D. Collot 1 and C. Faget 2 

1 

CREL, 168 rue de Versailles F78150 Le CHESNAY, France 

2 

Laboratoire VERGER, 53 rue Corbier Thiébaut, F60270 GOUVIEUX, France 

ABSTRACT 

Different methods for testing car drivers on 

illegal drugs are reviewed. Saliva analysis is 

preferred for practical reasons. Three 

technologies are compared for the research of 

narcotics in saliva : Ionscan 400, GC/Ionscan 

and GC/MS 

INTRODUCTION 

The CREL (Centre de Recherche et d'Etudes 

de la Logistique) is the french police research 

centre. The aim of this centre is to find new 

technologies and new methodologies for the 

police squad. Police forces need rugged 

equipments and methods that can be used by 

any officer after minimum training. 

The number of deaths increases in traffic 

accidents. For example, the first day of 2000 

year between 4 and 6 am, 90 persons died on 

the roads. Most of the drivers involved were 

between 18 and 26 years old. Besides, the 

percentage of drivers that could be positive on 

drugs is not insignificant. Statistics allow to say 

that this percentage lies between 3 and 10 %. 

These figures may well be underestimated, as 

it has been once established in Belgium that 

43% of the drivers leaving a night club were 

tested positive on narcotics. Litterature shows 

that smoking one single marijuana cigarette 

has the same effects as drinking 3 glasses of 

wine on empty stomach. It means a marijuana 

cigarette reduces the reflexes and the 

capacities to evaluate speed and distances. 

As a consequence, in june1999, the french 

government passed a new law stating that any 

driver involved in a fatal accident must be 

tested for alcohol and narcotics. Today 

nevertheless in case of body injuries, the state 

prosecutor requires a narcotics detection too 

even if there is no death. In France, in any road 

accident with body injuries or death, the 

breathalyser is used automatically. 

The aim of this study is to find a simple 

technology for the detection of narcotics in 

saliva. This technology is indeed for use on site 

on the road, not in a laboratory. Five main 

narcotics are studied: THC, Cocaine, MDMA 

(methylenedioxymethamphetamine) or ecstasy, 

Heroin and its metabolite 6MAM 

(6monoacetylmorphin). Three technologies are 

compared : Ionscan 400, GC/Ionscan and 

GC/MS. The latter is used for checking the 

results. 

BIOLOGICAL MATRICES 

How can we detect the presence of narcotics ? 

We know that every biological matrix contains a 

chemical print of narcotics: 

• the first biological matrix used is blood 

because it remains the only reference 

matrix for a court of justice and because of 

its medical interest. But, only a doctor in 

medecine or a qualified nurse is legally 

qualified to take blood samples. 

• the second is urine, which offers many 

analytical interests but which is difficult to 

sample on site. Its collection requires a 

special equipement and we know it ’s 

possible to alter the sample, as some 

problems with sportsmen have shown. 

• another solution lies in sweat but the 

quantity collected is often unsufficient and 

collection is difficult to perform. 

• the last solution seems to be saliva, which 

is suitable for the scope and easy to collect. 

We believe this is the simplest matrix to 

collect for police forces. Drugs contents in 

saliva can be used as a good biological 

indicator matrix. To support this view Pichini 

and coll. (1) and Cone and coll. (2) have 


Chr. Fuche et al.: „The use of IMS and GC/IMS...”, IJIMS 4(2001)1, 20-25, p. 21 

shown that saliva testing for abuse of drugs 

provides both qualitative and quantitative 

information on drug status and that drug 

concentrations reflect the free fraction of 

drug in blood. 

Saliva testing detects the narcotics physically 

present inside the mouth as well as those 

biologically absorbed. The quantity detected is 

higher when the collection is performed closer 

to the comsumption when the narcotics are 

absorbed by inhalation or smoking. 

Some pharmacokinetics data for illegal 

substances, expressed as halflives in saliva are 

24 hours for THC, Heroin and 6MAM and 4 

days for Cocaine. It is unknown for MDMA. 

For comparison, it is possible to find a chemical 

print of THC in urine 1 month after the last 

consumption. This is an another reason why we 

prefer saliva rather than urine because the print 

inside saliva better reflects the real state of the 

driver. 

METHODOLOGY 

We use a piece of the same absorbent cotton 

as is used in dentist surgery. The cotton is 

placed inside the mouth by the patient. The 

cotton is masticated as a chewing gum thus it 

absorbs both saliva and mouth cells which may 

contain twice as much of the narcotics print as 

the saliva does. 

The process is not invasive. It does not expose 

the patient to discomfort because he collects 

his saliva himself. And there is no skin irritation 

or risk of infection because the cotton is sterile 

and stored inside a test tube. The cotton is put 

in a syringe and compressed to recover the 

saliva. The volume of is between 0 and 1ml. 

IONSCAN 400 

The standard operating conditions of IMS 

detector are summarized in table 1. The 

Ionscan 400 was used without any hardware 

modifications. 

Table 1 : standard conditions 

Temperature : tube : 240 °C 

inlet : 290 °C 

desorb : 295 °C 

Flow: drift : 300cc/mn 

sample : 200 cc/mn 

The temperatures are higher than usual 

because the desorption is higher particularly for 

THC. 

GC/IONSCAN 

For the GC/IONSCAN, we replaced 2 parts, the 

injector and the column. The injector is a Ross 

injector and the column is a capillary column. 

The Ross injector is composed of two 

concentric glass tubes, the capillary being the 

inner one. The vector gas is helium. The 

advantages are that it is possible to set any 

required concentration of solute, and that the 

evaporation of solvent prevents the signal from 

Chemical treatment. 

The saliva can be analysed directly or it can be 

"chemically treated" or "extracted". In this case 

narcotics are extracted three times from both 

cotton and saliva by a mixture of hexane and 

ethyl acetate (50:50). The extracts from cotton 

and saliva are combined and concentrated. 

INSTRUMENTS : 

The Ionscan 400 is able to analyse both the 

pure saliva, that is to say saliva without 

chemical treatment, and the extracted saliva. 

With the GC/ionscan and the GCMS it is only 

possible to analyse the extracted saliva. The 

GC/MS is used as a reference technique in this 

study. 



GC 

oven: 

initial Temp. :190°C 

rate 1 : 8,75°C/min up to 260°C 

rate 2 : 40°C/min 

final temp. : 290°C 

delay of analysis : 645s 

transfer line : 285 °C 

IONSCAN: 

tube : 236°C 

inlet : 285°C 

desorb : 285°CC 

shifting. The disadvantage is that the injector is 

very fragile because it ’s outside the machine 

and it is made of glass. 

The capillary column offers a good resolution 

and requires smaller quantity of sample. This 

column is not standard. It is custom- made 

according to our specifications. 

The temperatures are lower than for the 

Ionscan 400 because there are no problems in 

the column desorption. 

All narcotics signals are clearly separared (fig 

1). With the GC /MS, the reference 

chromatogram is the same as with the 

GC/Ionscan (same chromatographic order and 

a good resolution). 

EXPERIMENTAL 

We use reference samples made from 

narcotics standards and clean saliva. For this 

purpose, we collect saliva from people who do 

not take drugs and who neither smoke nor 

drink coffee because caffeine and nicotine have 

a higher detection rate than narcotics. Each 

sample is analysed 5 times. 

Narcotics Standards 

We test four narcotics solutions. Table2 

summarizes the quantities dropped on the filter 

for Ionscan400. For comparison purposes, the 

quantity dropped on the filter is the same as for 

the GC/Ionscan and GC/MS analyser. 

Direct analysis (Ionscan 400) 

The volume analysed is always 200 µl. 

Samples are dropped on the teflon filter. When 

the teflon filter is dry, the deposit is analysed 

directly. 

Extracted saliva (Ionscan 400, GC/Ionscan, 

GC/MS) 

300 µl of saliva are always extracted and 2 µl 

are dropped on the filter or injected. 


IONSCAN 400 

From a quantitative point of vue, quantities 20 

times lower are present in the extracted saliva 

than in the unextracted saliva. 

For the three mixtures, the increase in 

compounds concentration is not observed for 

unextracted saliva(figure 2). The results for 

these narcotics are that the amplitude of the 

signal is not correlated with the quantity of 

narcotics. 

Heroin and 6MAM are detected in pure 

saliva(figure3). We assume that 6MAM arises 

from a degradation of heroin by enzymatic 

process. 

In the case of the extracted saliva, we have 

only the signal of heroin perhaps because the 

quantity of 6MAM is insufficient. 

Figure 1 : 

reference 

chromatogram 

from 

GC/ 

IONSCAN 



Table 2 : Quantities dropped in each sample 

Drug 

Pure 

saliva 

Extracted 

Saliva 

Drug 

Pure 

saliva 

Extracted 

Saliva 

Mixture 1 

Cocaine: 0.4 µg 

MDMA: 10 µg 


MDMA: 0.6 µg 

Thc1 

57.6 µg 

3.5 µg 

Mixture 2 


MDMA: 20 µg 


MDMA: 1.2 µg 

THC2 

230.4 µg 

13.8 µg 

Mixture 3 


MDMA: 40 µg 


MDMA: 2.4 µg 

THC3 

576 µg 

35 µg 

Heroin 

9.4 µg 

0.54 µg 

THC cannot be detected inside pure saliva at 

any concentration with this process. The 

chemical method is necessary to obtain the 

THC detection with the IONSCAN 400 (figure 

4).We test a fourth reference solution which is 

a mixture of cocaine and THC. It is run to check 

if the signal of cocaine would mask that of 

THC. Again, for pure saliva, only cocaine is 

detected and in extracted saliva, both cocaine 

and THC are detected (figure 4); their signals 

are clearly separated. 

In conclusion, pure saliva cannot be used for 

the detection of illegal drugs because on the 

one hand THC is not detected and on the other 

hand, false amphetamine signals are created 

by the degradation of enzymes. 

We calculate the response of IMS with 

differents concentrations of narcotics. All 

equations are quadratic. The correlation 

coefficent shows a relatively strong relationship 

between the variables (cocaine: R=0,93; MDMA 

and THC: R=0,96). 

GC/IONSCAN. 

The quantities injected are the same as for the 

IONSCAN 400 in order to compare the results. 

The three mixtures are injected via the Ross 

injector. As explained earlier, only extracted 

saliva is used with this instrument. MDMA is 

detected in each case and cocaine is detected 

only in the first and third mixture. But the signal 

for cocaine is lower in the third than in the first 

chromatogram. This is a nonsense because the 

quantity of cocaine is four times higher than in 

the first one. The sensitivity is not linear (figure 

5). 

In the 5 samples, heroin is never detected 

whereas it is detected in each case with the 

IONSCAN 400 and GC/MS. 

The GC/Ionscan detects THC in 2 of 3 

reference solutions, whereas the ionscan 400 

detects it in each case. For the mixture , once 

more , only cocaine is detected. The quantity of 

THC is too low. These results are checked with 

the GC/MS. 

CONCLUSIONS 

The most sensitive and reliable method for 

saliva analysis is GC/MS, but this technology is 

expensive and it is difficult to use. 

The analysis of saliva requires a chemical 

treatment in order to obtain the best sensitivity 

for all narcotics. 

We obtain better results with the IONSCAN 400 

than with the GC/ionscan for many narcotics. 

The problem with Ionscan 400 is the overlap of 

the signals. For example, heroin and THC, V3 

calibrant sample and methadone. 

For police controls, Ionscan has the capability 

to analyze not only the extracted saliva but also 

to analyze sample taken from the vehicle, 

objects and clothes with the same settings on 

the instrument. 

ACKNOWLEDGMENTS 

The volunteers who donated their saliva, Dr 

Claude Guionnet, Barringer Europe for the 

technical help, Dr Olivier Mauzac from CREL 

for the help in writing this document. 

REFERENCES 

[1] PICHINI and coll : Drug Monitoring in Non 

conventional Biological Fluids and Matrices ; Clin. 

Phamacokinet, 1996 Mar, 30 (3) 211-228 

[2] Cobe et al. cocaine disposition in saliva following 

intravenous, intrasanal and smoked administration, J. 

of Analytical Toxicology 21(1997) 465-475 





Mixture 

1 

Mixture 

2 

Mixture 

3 


Strategies for smarter chemical sensors 

P. de B. Harrington, G. Chen, A. Urbas 

Ohio University Center for Intelligent Chemical Instrumentation, Department of Chemistry and Biochemistry, Ohio 

University, Athens OH 45701-2979, USA 

Advancements in computer technology are 

forcing a paradigm shift in the processing of 

data from analytical instrumentation. The 

traditional approach is to signal average data 

from an analytical instrument once it achieves a 

steady-state response (i.e., the signal from the 

instrument stabilizes). However, a single signal 

averaged spectrum from the stabilized 

instrument response excludes temporal 

information that may be exploited to solve 

significant chemical problems. 

This paper presents the opportunities furnished 

by modeling the dynamical or transient 

responses of ion mobility spectrometers. The 

transient response occurs when the sample is 

introduced to the inlet of the instrument and the 

sample is removed from the inlet of the 

instrument. In other cases, for which the 

sample is thermally desorbed, the transient is 

introduced by the temperature time profile of 

the desorber and sample discrimination that 

may occur in the transfer lines . A useful 

method that is gaining popularity for modeling 

dynamic or temporal changes in data is 

SIMPLISMA . Using SIMPLISMA, the ion 

mobility data are factored into sets of 

concentration profiles and concentration 

independent spectra . Each feature or IMS 

peak that varies independently with respect to 

the duration of the measurement will be 

modeled as a separate component. An 

example will be presented later. 

The problem associated with dynamic modeling 

methods is that individual spectra must be 

stored instead of a single signal averaged 

spectrum. The ion mobility measurement is 

fast enough that a large volume of spectra may 

be acquired. Therefore, data compression 

becomes a useful tool. Ion mobility spectra are 

amenable to two-dimensional compression for 

which both the drift time and spectrum 

acquisition time dimensions are reduced . If 

wavelet compression is used, SIMPLISMA can 

directly model the compressed data . In 

addition, SIMPLISMA has been written as a 

real-time program in LabVIEW and can build 

and refine models as the data are collected . 

A single set of data will be used to illustrate the 

modeling of 2D compressed spectra. The 

measurement spectra are given in . The data 

were collected from a 1:1 mix (v/v) of 

dicyclohexylamine (DCHA) and diethylmethyl 

phosphonate (DEMP) that was sampled with a 

Graseby (Graseby Ionics, Ltd, Watford, UK) 

Chemical Agent Monitor (CAM) operating in 

positive ion mode. DCHA (99%, Lot 

#02810MX, Aldrich Chemical Company, WI) 

and DEMP (98%, Batch # 10010573, 

Lancaster, U.K.) were mixed equally by volume 

to furnish the liquid mixture. The CAM was 

modified in that the acetone dopant was 

removed and the instrument was cleaned of all 

residual acetone. The analog output was 

interfaced with a National Instruments 

AT-MIO-16X board, which provided the 

triggered analog-to-digital conversion and 

timing of the ion mobility spectra. The board 

interfaced to the ISA bus of a 200 MHz Pentium 

Pro Personal Computer that operated under 

Windows 98 Second Edition. The acquisition 

software was a virtual instrument (VI) that was 

written in LabVIEW 5.1. The spectra were 

sampled at an 80 kHz acquisition rate that 

furnished 1500 points per spectrum and 

collected at a rate of approximately 10 spectra 

per second. 


P. de B. Harrington: „Strategies for smarter...”, IJIMS 4(2001)1,26-30, p. 27 

The data set comprised 1560 spectra. Each 

spectrum was baseline corrected by subtracting 

the average calculated from the points between 

1.5 and 3.0 ms. For each spectrum, 1024 data 

points that ranged between 3.0 and 15.8 ms 

were selected for constructing the SIMPLISMA 

model. The SIMPLISMA program was written 

in C++ by Harrington and applied to the data 

post-run. The alpha value for damping noise 

was set to 5% of the maximum peak of the 

average spectrum calculated from the data set. 

The SIMPLISMA model was set to 5 

components. SIMPLISMA was applied to the 

2D wavelet compressed data. This data set 

was compressed using the Daubechies family 

of wavelets. Each spectrum was compressed 

to 64 points using the partial daublet 22 filter. 

The columns of the compressed data matrix 

were then compressed to 64 points using the 

partial daublet 12 filter. The original data size 

was 1560´1024 points and was compressed to 

64´64 points or 0.26% of its original size and a 

compression ratio of 99.7% was obtained. The 

concentration profiles and spectra from the 

SIMPLISMA model were transformed back to 

their corresponding sizes using the inverse 

wavelet transform. 

Figure 2 gives the concentration profiles that 

were obtained from the full and compressed 

data for the SIMPLISMA components that 

model the DEMP dimer peak. The sample vial 

was presented and then removed from the 

CAM inlet at approximately 20 s into the 

experiment. The information that is lost during 

compression corresponds to high frequency 

noise. 

Figure 3 gives the spectra obtained from 

SIMPLISMA from the compressed and 

uncompressed data. Note that the dimer peak 

is modeled well by SIMPLISMA and there are 

no other peaks contained in these spectra. In 

addition, the model from the compressed data 

Figure 1: Positive ion CAM measurement of a 1:1 (v/v) mix of DCHA and DEMP 



Sample Acquisition Time (s) 

0 20 40 60 80 100 120 140 

0.7 

Unprocessed Data 

99.7% 2D-Wavelet Compressed 

0.23 

0.18 



0.5 

Intensity (V) 

0.3 

Relative Intensity 

0.13 

0.08 

0.1 

0.03 

-0.1 

-0.02 

50 300 550 800 1050 1300 1550 

Scan Number 

Figure 1: 

Concentration profiles for the DEMP dimer peak. 

0 2 4 6 8 10 12 14 16 

Drift Time (ms) 

Figure 2: 

Spectra for the DEMP dimer peak 


0 20 40 60 80 100 120 140 

2.5 

0.24 

0.19 



Intensity (V) 

2.0 

1.5 




0.14 

0.09 

0.04 

1.0 

-0.01 

50 300 550 800 1050 1300 1550 

Scan Number 

Figure 4: 

Concentration profiles for the water RIP 

0 2 4 6 8 10 12 14 16 


Figure 3 

Water RIP spectra 


0 20 40 60 80 100 120 140 

0.5 



0.25 

0.20 



Intensity (V) 

0.3 

0.1 


0.15 

0.10 

0.05 

0.00 

-0.1 

50 300 550 800 1050 1300 1550 

Scan Number 

Figure 66: 

Concentration profiles for the DEMP 

monomer peak 

-0.05 

0 2 4 6 8 10 12 14 16 


Figure 5: 

Spectra for the DEMP monomer peak 



shows no attenuation in peak height. Figure 4 

and Figure 5 give the SIMPLISMA 

concentration profiles and spectra for the water 

reactant ion peak (RIP). Figure 6 and Figure 7 

give the concentration profiles and the spectra 

for the DEMP monomer peak. All three 

SIMPLISMA spectra contained only one peak, 

because these peaks varied throughout the 

measurement. 

This chemical system is interesting because 

DEMP will suppress the DCHA peaks through 

competitive charge inhibition. However, DCHA 

may be detected and resolved due to sample 

discrimination that occurs during the CAM 

measurement. In the concentration profiles 

given in , the DCHA component appears and 

rapidly disappears as the concentration of 

DEMP increases during the experiment. 

Because the DCHA peaks are short-lived 

during the measurement process, the DCHA 

monomer, the dimer, and the DEMP-DCHA 

mixed dimer all appear as one component in 

the spectrum in Figure 9. 

Figure 10 and Figure 11 give examples of a 

spurious component. This last component is 

modeling noise, which is evident in the 

unprocessed and compressed data. 

SIMPLISMA would be run again with the 

number of components specified as 4. 

Dynamic modeling of IMS data can be used to 

solve analytical problems. For the 

DCHA-DEMP mixture, a single time-averaged 

spectrum would not reveal the presence of 

DCHA. Using 2D wavelet compression a 

volume of data may be compressed, so that 

SIMPLISMA can exploit trends that occur 

during the measurement process. The 

SIMPLISMA models obtained from the 

compressed data can be transformed back to 

the their original sizes without altering the data. 


0 20 40 60 80 100 120 140 

0.5 



0.18 



Intensity (V) 

0.3 

0.1 


0.13 

0.08 

0.03 

-0.1 

50 300 550 800 1050 1300 1550 

Scan Number 

Figure 8: 

Concentration profiles for DCHA 

-0.02 

0 2 4 6 8 10 12 14 16 


Figure 7: 

Spectra that correspond to DCHA 


0 20 40 60 80 100 120 140 

0.2 



0.10 



0.1 

Intensity (V) 

0.0 


0.05 

-0.1 

0.00 

-0.2 

-0.05 

50 300 550 800 1050 1300 1550 

Scan Number 

Figure 10 

Spurious concentration profiles 

0 2 4 6 8 10 12 14 16 

Figure 9 

Spurious spectra 




References 

[1] Reese, E.S.; Harrington, P B. J. Forens. Sci. 

1999, 44, 68-76. 

[2] Shaw, L.A.; Harrington, P B. in press, Nov. 2000. 

[3] Windig, W.; Guilment, J. Anal. Chem. 1991, 63, 

1425-1432. 

[4] Harrington, P.B.; Reese, E.S.; Rauch, P J.; Hu, 

L.; Davis, D.M. Appl. Spectrosc. 1997, 51, 

808-816. 

[5] Cai, C.; Harrington, P.B.; Davis, D.M. Anal. 

Chem. 1997, 69, 4249-4255. 

[6] Harrington, P.B.; Rauch, P.J.; Cai, C. Submitted 

2000. 

[7] Chen, G.; Harrington, P.B. Submitted 2000. 


Initial Study of Electrospray Ionization-Ion Mobility 

Spectrometry for the Detection of Metal Cations 

H.M. Dion, L.K. Ackerman, H.H. Hill, Jr.* 

1 

Department of Chemistry, P.O. Box 644630,Washington State University, Pullman, WA 99164-4630, USA 

2 

Center for Multiphase Environmental Research, P.O. Box 642710, Washington State University, Pullman, 

WA 99164-2710, USA 

Abstract 

ESI-IMS of nine inorganic cation solutions was 

performed for the first time. Counter ion had a 

large effect upon the sensitivity and response 

ion identity of the cations studied. Several salt 

solutions yielded one major cation peak 

including: aluminum sulfate, lanthanum 

chloride, strontium chloride, uranium acetate, 

uranium nitrate, and zinc sulfate. Aluminum 

nitrate and zinc acetate solutions produced 

multiple cation peaks, which increased the 

detection limits and difficulty of identification 

through comparison with the ESI-MS literature. 

Predicted detection limits ranged from 0.33 

ppm to 25 ppm depending on the salt solution 

studied. The identity of the species detected is 

unconfirmed, but literature suggests, and drift 

times support the detection of cation-solvent or 

cation-solvent-anion complexes. Finally, 

strontium and lanthanum chloride were 

separated and detected simultaneously with a 

resolution of 2.2. This is the first research 

showing the use of ESI-IMS as a detection and 

separation method for metal ion and ion 

complexes and the results from this study 

warrant future development of ESI-IMS as a 

field technique for the detection of metal 

contaminants in the environment. 

Results and Discussion 

Several metal salts of differing ionic radius, 

charge, and anionic composition were 

investigated for response and sensitivity using 

ESI-IMS. Additionally, several salts with the 

same metal cation and different anion were 

examined. A complete list of metal salts 

studied, drift times, calculated peak resolution, 

reduced mobility values, and predicted 

detection limits are reported in Table 1. 

The ESI-IMS spectra for aluminum sulfate (10 

ppm Al) and aluminum nitrate (100 ppm Al) are 

illustrated in figure 1. It appears that the 

aluminum sulfate is much more sensitive to 

detection than the aluminum nitrate. This may 

be explained by the fact that for the aluminum 

nitrate the available charge is divided among 

several peaks (7.5, 9.4, 11.4 ms) whereas the 

aluminum sulfate salt only displays one 

predominant peak (11.1 ms). The peak at 11.4 

ms for the nitrate and the 11.1 ms peak for the 

sulfate can be predicted as the same 

complexed ion since the difference in drift times 

is only 2.6%. It has been shown in the ESI-MS 

literature that aluminum can form hydrolysis 

ions as well as complex with small organic 

anions [1]. From the ESI-IMS work completed, 

it is reasonable to predict that the 11.4 and 11.1 

ms peaks for aluminum are either a charged 

hydration or acetate complex while the 

additional peaks in the nitrate spectra may be a 

mixture of nitrate complexes such as 

Al(NO 3)(OH)(H 2O) +2 or Al(NO 3)(OH)(H 2O) +3 as 

detected by ESI-MS [1]. 

Additionally, in Figure 1, zinc salts show 

distinctly different ion peaks depending on the 

anions present in solution. It should be noted 

that while every sample was carried by a 

mobile phase with 5% acetic acid zinc sulfate 

did not produce response peaks similar to the 

acetate counterpart. 

ESI-MS literature, while very specific in its 

metal species identification, is not readily 

transferable to ESI-IMS due to the lack of CID 

region. It should be noted that the ESI-MS 

literature only twice observed a bare cation, 

and even with high-energy CID regions, 

cation-solvent, and cation-solvent-anion 

complexes were the norm. Thus, it is very 

likely that the species being detected here are 


H.M. Dion: „Initial Study of Electrospray...”, IJIMS 4(2001)1, 31-33, p. 32 

Table 1: 

List of metal salts studied, drift times, calculated resolving power of IMS for each peak, and reduced 

mobility values in cm 2 v -1 s -1 . (nd = not determined) 

Salt 

Aluminum 

Nitrate 

Aluminum 

Sulfate 

Lanthanum 

Chloride 

Strontium 

Chloride 

Strontium 

Nitrate 

Uranium 

Acetate 

Uranium 

Nitrate 

Zinc Acetate 

Zinc Sulfate 

Chemical Form 

Al(NO 3) 3 

Al 2(SO 4) 3*18 H 2O 

LaCl 3*H 2O 

SrCl 2*6 H 2O 

Sr(NO 3) 2 

UO 2(CH 3COO) 2*2 

H 2O 

UO 2(NO 3) 2*6 H 2O 

Zn(CH 3COO) 2*2 

H 2O 

ZnSO 4*7 H 2O 

Drift 

Time 

(ms) 

7.5, 9.4, 

11.4 

11.1 

10.8 

9.8 

10.0, 

12.4 

12.3, 

16.3 

10.7 

12.0, 

14.4, 

18.8, 

20.1, 

24.5 

10.2 

Resolving 

Power 

22.1, 23.4, 

30.7 

41.6 

31.3 

38.0 

24.3, 28.1 

47.6 

31.4 

28.8, 34.6, 

36.6, 34.5, 

36.8 

30.2 

Observed 

K o 

2.3, 1.8, 

1.5 

1.5 

1.6 

1.8 

1.7, 1.4 

1.4, 1.1 

1.6 

1.4, 1.2, 

0.9, 0.9, 

0.7 

1.7 

Corrected 

K o 

2.2, 1.7, 

1.4 

1.5 

1.5 

1.7 

1.6, 1.3 

1.3, 1.0 

1.5 

1.3, 1.1, 

0.9, 0.8, 

0.7 

1.6 

Predicted 

Detection 

Limit (ppm) 

nd 

0.33 

3.4 

8.9 

nd 

1.8 

25 

nd 

5.4 

cation-solvent or cation-solvent-anion 

complexes, whose exact identity is a function of 

concentration, charge-competition at the 

needle, and gas-phase stability. With further 

refinement it may be possible to control the 

complexation so that the majority of the current 

is carried by a single species, making it 

possible to develop a sensitive method for the 

direct field analysis of inorganic cations and 

anions in water. 

ACKNOWLEDGEMENTS 

Heather Dion was supported by the National 

Science Foundation's Integrative Graduate 

Education and Research Training Grant to 

Washington State University under Grant 

#9972817. The National Science Foundation’s 

Integrative Graduate Education and Research 

Training Grant also provided support for Luke 

Ackerman through the form of an 

undergraduate summer research experience 

fellowship. The authors also wish to 

acknowledge INEEL for supplies and operating 

expenses. 

References 

[1] Agnes, G.R.; Horlick, G. Appl. Spectrosc. 1995, 49, 

324-334. 

[2] Agnes, G.R.; Horlick, G. Appl. Spectrosc. 1992, 46, 

401-406. 


H.M. Dion: „Initial Study of Electrospray...”, IJIMS 4(2001)1, 31-33, p. 33 

3.4 

4.25 

3.2 

3.0 

a 

4.00 

3.75 

c 

2.8 

3.50 

3.25 

2.6 

3.00 

2.4 

0 2 4 6 8 10 12 14 16 18 20 22 24 

0 2 4 6 8 10 12 14 16 18 20 22 24 26 

0.8 

4.4 

4.2 

4.0 

b 

0.7 

d 

3.8 

3.6 

0.6 

3.4 

3.2 

0.5 

3.0 

0 2 4 6 8 10 12 14 16 18 20 22 24 

0 2 4 6 8 10 12 14 16 18 20 22 24 26 

Figure 1: 

a. Aluminum nitrate (10 ppm Al), 

b. Aluminum sulfate (100 ppm Al), 

c. Zinc sulfate (100 ppm Zn), 

d. Zinc acetate (100 ppm Zn). 

All spectra collected at 0.300 ms pulse width and averaged 500 times. 


The use of GC-IMS to analyze high volume vapour samples 

from cargo containers 

P. Lafontaine, P. Pilon, R. Morrison, P. Neudorfl 

Canada Customs and Revenue Agency, Laboratory and Scientific Services Directorate 

Research and Development Division 

One avenue of narcotics smuggling is by 

concealment in cargo containers. Canada 

Customs statistics indicate that the value of 

drugs seized from cargo containers was 

greatest for cocaine. Approximately, a million 

containers enter Canada each year and a 

manual inspection of each one is impossible. A 

quick, inexpensive and effective method of 

container chemical identification coupled with 

intelligence information could target likely 

containers, serve to interdict drugs and provide 

a deterrent to drug smugglers. This paper 

describes some development of a high volume 

air sampling method with subsequent chemical 

analysis by GC-IMS to provide a rapid method 

for cargo container inspection. 

The envisioned cocaine detection method by air 

sampling will consist of three steps: sampling of 

the containers; trapping of the vapors and 

analysis of the trapped vapors. The sampling 

of the containers is done with a head assembly 

which fits over container vents, a filter 

assembly which contains specially coated filters 

for collecting vapors and a vacuum pump for 

sampling the air. Separate collection and 

analysis steps will likely be necessary for well 

concealed narcotics. The trapping medium will 

contain a complex matrix of materials collected 

from the containers as well as any narcotics 

vapors contained in the sampled air volume. 

The analysis method employed in the field 

needs to be portable, easy to use, fairly fast 

and analytically effective for the trace analysis 

of cocaine in a complex matrix. Laboratory 

investigations have used GC-MS/MS(gas 

chromatography- ion trap mass spectrometry) 

with SPME (solid phase microextraction) 

analysis of separated cocaine and also direct 

injection into the GC( gas chromatography) 

column. Although this method is effective, it is 

not field portable. The various analytical 

methods which are being evaluated are: Ion 

mobility spectrometry (IMS), GC coupled with 

IMS (GC-IMS) and portable GC- mass 

spectrometry(GC-MS). 

Canada Customs employs the Barringer 400 

and 400B IMS instruments at most points of 

entry into Canada. These are dedicated to the 

detection of narcotics. The use of these 

instruments for the chemical analysis method 

for cargo container air analysis would be a 

convenient and low cost option. Unfortunately, 

the IONSCAN analysis of the air sampled 

materials suffer from signal suppression. This 

signal suppression is observed from air 

sampled materials when a known amount of 

cocaine added to the sampling medium will not 

give an signal even though an identical non-air 

sampled medium sample will give a positive 

identification. The use of a gas 

chromatography coupled with ion mobility 

spectrometry for dealing with signal 

suppression was studied. 

A GC-IMS system was assembled in the 

laboratory and consisted of two parts: A Varian 

3400 CX gas chromatograph coupled to 

Barringer IONSCAN Model 250. The sample 

interface was effected by pushing the GC 

column through a heated metal tube into the 

IMS inlet. The sample introduction system into 

the chromatography column (DB-15) was either 

a solid phase desorber (SPD) via a six port 

value or direct desorption with a Varian 

ChromatoProbe accessory. The data analysis 

software was developed in house. The 

apparatus is shown in Figure 1. 

The results for cocaine and EDME (ecgonidine 

methyl ester, a cocaine degradation product or 

cocaine identifier) deposited on Teflon filter with 

this apparatus are shown in Figure 2. In the left 


P. Lafontaine et al.: „The use of GC-IMS...”, IJIMS 4(2001)1, 34-36, p. 35 

3D plasmagraph (digital units versus retention 

time versus drift time), a GC- IMS experiment 

using SPD was performed on a coated Teflon 

filter with added cocaine and EDME and the 

peaks are easily seen. In the right hand 

plasmagraph, the results using an air sampled 

spiked coated Teflon filter and the same 

conditions are shown. In this plasmagraph, 

there are many peaks and some of which 

obscure the cocaine and EDME peaks. 

In a further set of experiments, a Barringer 

SPD-GC-IONSCAN was used to examine the 

air sampled filters which were spiked with 

cocaine and EDME. The signal processing 

capabilities of the commercial instrument 

allowed better separation and identification of 

the cocaine and EDME(Fig. 4). Some 

experiments demonstrating the better recovery 

results for the SPD-GC-IONSCAN compared to 

the IONSCAN are shown in Table 1. The 

control blanks also gave good results. 

Further work is necessary to verify whether the 

sensitivity of the analytical method and the 

deployment characteristics of the 

SPD-GC-IONSCAN will make this method a 

viable field deployable method for vapor 

sampling of cargo containers. 

Figure 1: Laboratory GC-IMS 


Figure 2: Plasmagraphs of coated filters from SPD GC-IMS

P. Lafontaine et al.: „The use of GC-IMS...”, IJIMS 4(2001)1, 34-36, p. 36 

Figure 3: 

Plasmagraph and data reduction of a spiked air sampled coated Teflon filter 

using the SPD- GC-IONSCAN 

Table 1 Analysis of spiked samples of various Air Background 

by IONSCAN, SPD-GC-IONSCAN 

Samples were spiked with Vapour source set for EDME and Cocaine generation 

(equivalent to app. 1 ng of each) 

IONSCAN GC-IONSCAN 

Container Cocaine EDME Cocaine EDME 

Coated filter no air 1953 1109 1134 2279 

C1 0 214 507 853 

C1 0 231 

C2 PASS 0 994 

C2 292 0 434 978 

C3 527 0 0 475 

C4 PASS 478 1883 

C5 1004 0 0 1174 

C6 PASS 0 573 

C4 Not spiked PASS 

C6 Not spiked PASS 


Evaluation of sample collectors for ion mobility spectrometry 

N. Mina, S.P. Hernández, F.R. Román, L.A. Rivera 

Energetic Materials Research Laboratory, Department of Chemistry, University of Puerto Rico Mayagüez Campus, 

P O Box 9019, Mayagüez, PR, 00681 

Abstract 

A series of commercially available filter 

materials and membranes were studied to 

investigate their affinity with various explosives 

(RDX, NG, TNT, PETN and DNT) and their 

adsorption/desorption thermal characteristics 

using a Barringer IonScan 400 IMS. The filters 

and membranes, that withstood the high 

temperatures of the IMS injector/desorber were 

made of either fiberglass or cellulose, fine or 

coarse porosity of various pore sizes 

(0.5-40mm), and medium to fast flow rate. The 

data suggest that filter material properties such 

as pore size, surface roughness and porosity; 

flow rate and explosive vapor pressure are 

parameters that can influence the IMS 

response. The affinity of the explosives for the 

filter material can also influence IMS response. 

In general, the filters that showed the best 

responses were those with smaller pore size, 

medium to fine porosity, and medium flow 

rates. The explosives that showed the best IMS 

responses were those with very low vapor 

pressure such as PETN and RDX. However the 

data seem to suggest that the affinity of NG for 

the filter materials is enhancing its signal close 

to the responses seen for RDX and PETN 

when compared with DNT and TNT that are 

less volatile than NG. 

Introduction 

Ion Mobility Spectrometry is an analytical 

technique known for over 25 years. The 

principles, instrumentation, technological 

advances, and applications are very well 

described elsewhere 1-3 . IMS detectors are very 

sensitive and limits of detection in the ppt or 

picogram range are reported in the literature 4,5 . 

Thus, the main challenge so far is not to lower 

the detection limits but how to transfer the 

sample from a real scenario to the IMS 

injector/desorber. In this regard, a series of 

filters and membranes were investigated, in 

order to identify materials with the optimum 

physical and chemical properties to be used as 

IMS sample collectors for explosives. Tests 

were performed on various sampling filter 

materials to identify the filter properties that 

enhance the collection stage of the explosives 

and their adsorption and thermal desorption 

properties. 

Experimental details 

An Ionscan 400 Barringer Instruments, Inc. 

Murray Hill, NJ, ion mobility spectrometer with 

the thermal desorption injector operating at 220 

0 

C was used for the experiments. In order to 

establish baseline measurements, the 

instrument was calibrated for its response to 

the explosives selected for the experiment, 

using fiberglass-sampling filters. The filters 

were obtained from Barringer and are ordinarily 

used for this purpose. Explosive standard 

solutions were obtained from Radian 

International, Analytical Reference Materials, 

Austin, TX. All of the standards were used as 

the acetonitrile (ACN) solutions and had an 

initial concentration of 1000 parts per million 

(ppm), except DNT which had a concentration 

of 100 ppm. Table 1 shows the properties of 

the energetic material tested: RDX, PETN, 

TNT, NG, and DNT. Working solutions were 

prepared at least on a weekly basis to avoid 

degradation 6 by successive dilutions of the 

stock solutions in High Pressure Liquid 

Chromatography (HPLC) grade ACN, obtained 

from Aldrich Chemical Co., Milwaukee, WI. 

Other materials tested included Barringer’s rag 

swabs, and a variety of filters commercially 

obtained from VWR Scientific Supplies, 


N. Mina et al.: „Evaluation of sample collectors...”, IJIMS 4(2001)1, 37-40, p. 38 

Cataño, PR. Mass deposits in the moderately 

low picogram (> 50 pg) to the 100-nanogram 

mass range were utilized. A Hamilton Co. 

Microliter # 701 microsyringe was used to 

dispense the solutions containing the mass of 

the explosive being tested. Before performing 

the ion mobility analysis, filter materials were 

subjected to two important resistance tests: 

solvent compatibility and heat resistance. For 

the solvent compatibility tests, small aliquots of 

ACN and methanol were utilized and then 

visual inspection under microscope followed. 

An Olympus BH-2 designed for mineralogical 

applications and equipped with infinity 

corrected 20x, 50x, 80x, and 100x objectives 

was used. For the heat treatment tests, a 

bench-top, laboratory oven was used to heat 

watch glasses (Corning Glass, 75 mm nominal 

diameter) to 230-260 °C. Once the watch 

glasses reached thermal equilibrium, a sample 

of each sampling material was placed on the 

watch glass and left in contact for 60, 180, and 

300 seconds. The tests were run in duplicate. 

Then they were observed visually and by 

inspection under microscope. 

The materials that withstood the stability tests 

were: Whatman-Balston # 1 (# 1), 

Whatman-Balston # 41 (41), Gelman A/E (A/E), 

Gelman Metrigard (MG), Scientific Products- 

Baxter-VWR (S-P) and VWR # 417 (VWR). 

From here on, the manufacture names or their 

abbreviation, which appear above in 

parenthesis, will be used. 

The Barringer instrument was calibrated on a 

daily basis, qualitatively using a “dip-stick” 

impregnated with a mixture of explosives 

(supplied by Barringer). On a quantitative 

mode, daily runs with certified standards of 

RDX on Barringer fiberglass filter were 

performed. Instrument pressure was monitored 

constantly as well as position of internal 

standard. 

Results and discussion 

In a series of experiments different filter 

materials were spiked with increasing amounts 

of various explosives (one at the time) in order 

to test for the affinity of the explosives and their 

adsorption-desorption characteristics. In these 

experiments the Barringer SWAB and Barringer 

fiberglass filter, and filters # 1, MG, VWR and 

A/E were spiked with explosives in the 

picogram range. 

Barringer Filter Spiked with RDX, PETN, NG, 

TNT, and DNT 

The order of IMS response was the following: 

RDX > NG > PETN > TNT >>> DNT 

RDX gives the overall best response followed 

by NG. PETN starts to respond when 50 pg are 

deposited and TNT starts to respond at 200 pg. 

However, PETN has a tendency to level off 

while the TNT response becomes sharper 

specially, above 400 pg. In order to get a 

response from DNT 10,000 pg must be 

deposited on the surface of the filter. 

Barringer Swab Spiked with RDX, PETN, 

TNT, and DNT 

The order of IMS response was the following: 

RDX >> PETN > TNT >>> DNT 

RDX gives a very sharp response. It is much 

more responsive than the other explosives 

tested. PETN and TNT show a response that 

does not differ significantly. DNT required 30 

ng to give an average area response of 193 

units and as the mass deposited increased 

(30-300 ng range) the response did not change 

significantly. In both the Barringer Filter and 

Table 1: PROPERTIES OF EXPLOSIVE COMPOUNDS USED 

EXPLOSIVE 

RDX 

PETN 

TNT 

NG 

DNT 

MW 

222 

316 

227 

227 

182 

VAPOR 

PRESSURE 

(Ppb) 

0.0015 

0.0005 

6 

300 

145 

CONCENTRATION 

STANDARD 

SOLUTION (ppm) 

1000 

1000 

1000 

1000 

100 

TYPE OF 

MOLECULAR 

STRUCTURE 

ALIPHATIC 

NITROAMINE 

ALIPHATIC 

NITROESTER 

AROMATIC 

NON-POLAR 

ALIPHATIC NITRO 

ESTER 

AROMATIC 

POLAR 



INSTRUMENT RESPONSE 

Figure 1. IMS AVE RESP TO # 1 FILTER SPIKED WITH EXPLOSIVES 

1800 

1500 

NG RDX PETN TNT 

1200 

900 

600 

300 

0 

0 200 400 600 800 1000 1200 1400 

MASS DEPOSITED: (pg) 

Swab, RDX is the most responsive and DNT is 

the less responsive. 

Whatman–Balston #1 Filter Spiked with NG, 

RDX, PETN, TNT, and DNT 

The overall order of IMS response is the 

following: NG > PETN > RDX > TNT >> 

DNT 

The #1 filter seems to have good 

absorption-desorption characteristics. All the 

explosive tested gave good sharp responses 

except DNT. The values for DNT are not in the 

range therefore do not appear in Fig 1. 

VWR Filter with RDX, PETN, NG, and TNT 

When the VWR filter is spiked with varying 

amounts of explosives (500-1500) in the pg 

range it is observed that 500 pg PETN gives an 

area of 156 compared to 117 for NG and 80 for 

RDX. As the mass increases (600 pg and 

above) the PETN and NG graphs have a 

tendency to overlap and are probably within the 

statistical error. As the mass increases (700 to 

1250 pg) the RDX curve becomes slightly more 

responsive and levels off at 1500 pg. In general 

terms, this filter, with some exceptions at low 

and mid points for RDX. There is not a 

clear-cut difference in response for the above 

explosives and thus an order of IMS response 

will not be established. The response for TNT is 

very poor. This filter requires over 500 pg of the 

explosives in order to respond and as mass 

increases, the increase in response is not 

significant. (Maximum response is ~ 400 area 

units). 

S-P Filter Spiked with RDX, PETN, NG, and 

TNT 

TNT and NG gave a better response than RDX 

and PETN in this material. The order of 

responsiveness is the following: NG > TNT > 

RDX > PETN. TNT is the only explosive that 

gives a response when 200 pg are deposited 

(peak area of 104). When 300 pg is deposited 

TNT continues to be the more responsive with 

a peak area of 132 compared with 88 for NG. 

At mass deposited above 600 pg NG becomes 

more responsive and the response of the TNT 

flattens off. The standard deviations are 

significantly higher than those of others. As with 

the VWR filter, the increase in response is not 

significant. (Maximum response is ~ 400 area 

units). 

Conclusions 

Various commercially available filter materials 

were tested for optimum explosive collection 

and release properties. Of the different filter 

materials tested for thermal stability, only 

cellulose and fiberglass resisted the high 

temperatures necessary for their utilization at 

the desorption chamber of the IMS. This finding 

is consistent with the materials presently used 

as IMS sample collectors. The various filters 

tested showed different IMS responses for the 

combination of explosives and collection 

materials tested. The data suggest that filter 

material properties such as: pore size, surface 

roughness and porosity, flow rate and explosive 

vapor pressure of explosive were parameters 

that can influence the IMS response. The 

affinity of the explosives for the filter material 

can also influence IMS response. This affinity 



can be explained in terms of interactions of the 

filter with the explosive molecules. Such 

interactions can be hydrogen bonding, 

dipole-dipole, and van der Waals or weak 

interactions between partial charges induced in 

the carbon skeleton of cellulose molecules. In 

general, the filters that showed the best 

responses were those with smaller pore size, 

medium to fine porosity, and medium flow 

rates. On the other hand, the explosives that 

showed the best IMS responses were those 

with very low vapor pressure such as PETN 

and RDX. However the data seem to suggest 

that the affinity of NG for the filter materials is 

enhancing its signal close to the responses 

seen for RDX and PETN when compared with 

DNT and TNT that are less volatile than NG. 

The ideal sampling material should have the 

following desirable properties. Be able to 

withstand the high temperatures necessary for 

thermal desorption: thermal stability. It should 

have good adsorption-desorption properties for 

all explosives: rapid adsorption of the explosive 

during sampling and quick release of explosive 

while heated at the desorption chamber of the 

IMS. Finally, it should be inert: do not react with 

the explosive or the environment. 

Acknowledgments 

This work was supported by the FAA under 

project No. 99-G-029. 

References 

[1] R. H. St. Louis and H. H. Hill Jr., Crit. Rev. Anal. 

Chem., 21, 321, (1990). 

[2] J. V. Haase Ewin, Asian Defense Journal, 6, 56, 

(1993). 

[3] D. D. Fetterolf, in Advances in Analysis and 

detection of Explosives, J. Yynon (Ed.), 

Dordrecht, Netherlands, 1993. 

[4] F. Garofolo, F. Marziali, V. Migliozzi and A. 

Stama, Rapid Commun. In Mass 

Spectrom.10,1321, (1996). 

[5] F. Garofolo, V. Migliozzi and B. Roio, Rapid 

Commun. In Mass Spectrom.8, 527, (1994) 

[6] Jankowski, P.Z., Jappinga, E.M., and Butler, E., 

“Operational Deploymentof Explosives Detection 

Equipment During the 1996 Olympics. 


Principles and applications of a solid phase desorption unit 

coupled to a GC-IONSCAN ® system 

A. Grigoriev, R. Jackson, F. Kuja, S. Nacson and A. Rudolph* 

Barringer Instruments Ltd., 1730 Aimco Boulevard, Mississauga, ON L4W 1V1, Canada 

Introduction 

The Barringer Solid Phase Desorption (SPD) 

unit was developed to allow GC-IONSCAN 

analyses of solid samples without prior sample 

preparation. 

It attaches to the side of the GC-IONSCAN and 

interfaces with it mechanically and 

electronically (Figure 1). The liquid sample 

injection capability of the GC is retained for 

maximum flexibility. Temperatures of the SPD 

desorber, inlet, valve and transfer lines as well 

as the duration of each stage of the analysis 

can be programmed. The SPD-GC-IONSCAN 

now offers three analysis methods: Direct 

desorption IMS, liquid injection GC-IMS and 

solid phase desorption GC-IMS. 

Method of Operation 

During loading, samples are thermally 

desorbed through a heated transfer line and a 

heated six port valve into the air cooled sample 

collection loop where they condense. 

For the sample analysis, the six port valve 

switches position and the loop is rapidly heated. 

GC carrier gas purges the loop and transfers 

the collected sample into the column. After the 

purging is complete, the valve switches back 

into the loading position and the loop is then air 

cooled for the next analysis. Figure 2 shows 

the timing of events during sample loading and 

analysis. 

GC Injector 

GC Module 

Cooling Fan 

For Sample Loop 

SPD Desorber 

IONSCAN 

Desorber 

SPD Module 

IONSCAN 

Module 

Figure 1: 

Barringer GC-IONSCAN 


A. Grigoriev et al.: „Principles and applications of solid phase ...”, IJIMS 4(2001)1, 41-44, p. 42 

2ng Amphetamine 2ng PCP 

2ng Methamphetamine 2ng Cocaine 

2ng MDA 

10ng THC 

2ng MDMA 

10ng Heroin 

2ng MDEA 

Figure 2: Timing of Events 

Examples of Analyses 

Mixtures of narcotics or explosives were 

analyzed both by direct (splitless) GC injection 

and by SPD injection (from Teflon filter) on the 

standard 15m MXT-1 column (0.53mm i.d., 

1µm film thickness), using nitrogen as carrier 

gas. Figure 3 shows the analysis of a mixture 

of nine narcotics, using the splitless liquid 

injection method (2µL injection), whereas 

Figure 4 shows the results from the solid phase 

desorption analysis of the same mixture (2µL 

on a Teflon filter). 

Figure 5 shows the analysis of a mixture of six 

explosives, using the splitless liquid injection 

method (2µL injection), and Figure 6 shows the 

results from the solid phase desorption analysis 

of the same mixture (2µL on a Teflon filter). 

Figure 7 illustrates the separation of EDME and 

cocaine from a Teflon filter, using the SPD 

analysis method with a 10m DB-5 column 

(0.53mm i.d., 1µm film thickness) and nitrogen 

as carrier gas. This method was developed for 

the vapour screening of cargo containers for 

the presence of cocaine. 



2ng Amphetamine 2ng PCP 

2ng Methamphetamine 2ng Cocaine 

2ng MDA 

10ng THC 

2ng MDMA 

10ng Heroin 

2ng MDEA 

Figure 3: Splitless Injection Figure 4 Solid Phase Desorption 

(Oven Temperature Profile: 120°C for 30s, 15°C/min to 140°C, 30°C/min to 280°C) 

Figure 5: Splitless Injection 

Figure 6: Solid Phase Desorption 

(Oven Temperature Profile: 120°C for 60s, 40°C/min to 240°C) 

Conclusion 

The Solid Phase Desorption add-on module to 

the Barringer GC-IONSCAN performed well 

and gave good results in the analyses of 

mixtures of narcotics and explosives without 

prior sample preparation. Even high boiling 

analytes such as heroin or tetryl were observed 

to pass efficiently through the SPD system. 

Sample loss through degradation or 

condensation is estimated to be less than 20% 

for most analytes investigated. 

Further experiments will be carried out to 

exactly determine analyte losses, to optimize 

the system and to apply the method to various 

other analytical problems. 



Figure 7: 

Analysis of a EDME / Cocaine Mixture 


Operational assessment of a handheld ion mobility spectrometry 

instrument 

Chih-Wu Su 1 , Steve Rigdon 2 , Tim Noble 2 , Mike Donahue 2 , Corey Ranslem 3 

1 

United States Coast Guard Research & Development Center, 1082 Shennecossett Road, Groton, CT 06340, USA 

2 

Anteon Corporation, Rte. 2, P.O. Box 220, North Stonington, CT 06359 USA 

3 

United States Coast Guard, Taclet South, 15000 NW 42nd Ave, Opa Locka Airport, Opa Locka, FL 33052, USA 

INTRODUCTION 

In 1995, the U.S. Coast Guard made the 

decision to outfit vessel boarding teams with 

state of the art technology to assist them in 

detecting trace levels of narcotics. Based on an 

assessment conducted by the Research and 

Development Center (R&DC), ion mobility 

spectrometry (IMS) was selected as the 

technology best suited for operational use. 

Specifically, the Coast Guard selected the 

Ionscan, an IMS instrument manufactured by 

Barringer instruments. A limiting factor of all 

models of the Ionscan is its size and power 

requirements, weighing between 50 and 60 

pounds. These limitations typically require that 

the Ionscan be operated on the Coast Guard 

cutter where there is a ready supply of 115VAC 

power. As a result, samples must be 

transported from the vessel being boarded to 

the Ionscan for analysis. The impact of this 

protocol is a loss of time resulting in the inability 

of the boarding officer to quickly determine 

areas to resampled. Additionally, the 

requirement that the Ionscan operator remain 

on the cutter results in the boarding party being 

reduced by one person. 

THE SABRE 2000 HANDHELD IMS 

In early 2000, Barringer Instruments released 

the Sabre 2000, a new instrument based on 

IMS. The Sabre 2000 weighs less than 6 

pounds and is capable of operating on battery 

power. The advantage to the Sabre 2000 is that 

it can be transported to the vessel being 

boarded. Should the Sabre 2000 provide the 

operator with an equivalent level of confidence 

in the analytical result as the Ionscan, then the 

physical advantages afforded by the Sabre 

2000 will enable the Coast Guard to conduct 

analysis of samples on the vessel being 

boarded. 

This protocol offers many advantages to the 

boarding party over the existing protocol. 

Some advantages this device would bring 

include the capability of having near real-time 

analysis of the samples. Since samples will be 

transferred in batches, and not at the end of the 

sampling phase, analytical results should be 

available minutes after a sample is taken. 

Second, with near real-time analysis, the ability 

to quickly determine areas of interest requiring 

additional resampling will be possible. If each 

batch of samples correlates to a specific area 

of the vessel, then an alarm in that batch may 

indicate the requirement to resample that area. 

Third, since the samples are being analyzed on 

the vessel, there is no longer a need to leave a 

member of the boarding team behind to be the 

Ionscan operator. Larger boarding teams 

results in increased safety and an increase in a 

team’s ability to more thoroughly search a 

vessel. 

From an analytical standpoint, there is a great 

difference between the Sabre 2000 and the 

Ionscan 400B. The Sabre 2000 operates using 

a closed loop IMS architecture while the 

Ionscan 400B operates using an open loop IMS 

architecture. One repercussions of employing 

a closed loop IMS architecture is that a 

semi-permeable membrane is necessary to 

physically separate the inside of the instrument 

from the outside environment. Another 

difference between the Sabre 2000 and the 

Ionscan 400B is the operating temperatures. 

The Sabre 2000 operates at a much lower 

temperature than the Ionscan. The vaporization 

rate of organic compounds from a sampling 

[The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the 

views of the Commandant or the Coast Guard at large. Any instrument, material or chemical referred to by its brand name does not 


constitute an endorsement by the authors, the Commandant or the Coast Guard at large.]

Chih-Wu Su et al.: „Operational assessment of a handheld...”, IJIMS 3(2000)1, 45-47, p. 46 

matrix is proportional to the desorption 

temperature. The higher the desorption 

temperature, the more mass of a sample can 

be vaporized over a shorter time period. At the 

lower operating temperatures of the Sabre 

2000, it will take longer to evaporate the same 

mass of sample than the Ionscan. Since the 

IMS analysis is a nearly continuous process, 

the maximum amplitude of a peak detection will 

be lower. The overall result is a decrease in 

the detection sensitivity. The lower operating 

temperatures also mean that the instrument 

may be more prone to sample hang-up if high 

levels of contamination are encountered. 

When comparing the Sabre 2000 and the 

Ionscan, it is evident that they are two totally 

different instruments from both a physical as 

well as an analytical perspective. As a result, 

the assessment of the Sabre 2000 was 

approached from the standpoint of it being a 

totally new and different instrument. An 

intensive evaluation was conducted ranging 

from the laboratory to the field. 

U.S. COAST GUARD’S APPROACH TO 

ASSESSING THE SABRE 2000 

The U.S. Coast Guard’s approach to assessing 

the Sabre 2000 was a three-step process. The 

assessment began with a laboratory evaluation. 

The purpose of the laboratory evaluation was to 

determine the baseline performance of the 

instrument under ideal conditions. The second 

step in the evaluation process was a controlled 

field evaluation. The purpose of this phase was 

to determine the Sabre 2000 performance 

under controlled real world conditions. 

Following the controlled field evaluation, a 

go/no-go decision will be made as to whether 

the assessment should be continued on to its 

final phase, an at-sea evaluation. 

Laboratory Evaluation 

The laboratory evaluation was conducted in the 

narcotics detection laboratories at the Coast 

Guard Research and Development Center. 

The evaluation was conducted alongside an 

Ionscan 400B for comparison purposes. There 

were several tests conducted in the laboratory 

evaluation, all centered around determining the 

Sabre 2000 analytical capabilities under ideal 

conditions. 

• Minimum Detection Limit (MDL) Testing. 

This testing was conducted to determine 

the Sabre 2000 sensitivity for cocaine, 

heroin, 

methamphetamine, 

tetrahydrocannabinol (THC), and 

ecgonidine methyl ester (EDME). 

• Instrument Response Testing. This testing 

was conducted to determine the Sabre 

2000 response to a wide range of 

concentrations of liquid standards of each 

of the narcotics used for the MDL testing. A 

series of ”calibration curves” were 

developed. 

• Instrument Response to Wipe Testing. 

This series of tests was similar to the 

instrument response testing with the 

exception that liquid standards were 

deposited on glass plates, allowed to dry, 

and then wipe sampled. This test more 

closely replicated how the instrument will be 

used in the field. Only cocaine was 

analyzed for this test. 

• Dirty Sample Matrix Testing. Since ”clean” 

samples are very rare in the field, a series 

of tests to determine the effect of a dirty 

sample matrix on the ability of the Sabre 

2000 to detect cocaine were conducted. 

• Environmental Testing. In its intended 

application, the Sabre 2000 will be utilized 

during boardings being conducted on the 

high seas under conditions of elevated 

temperature and high relative humidity. A 

series of tests were conducted in an 

environmental chamber to test the ability of 

the Sabre 2000 to operate under these 

conditions. 

• Finally, several qualitative parameters 

concerning the Sabre 2000 performance 

were assessed. They were: battery life, 

battery recharge cycle, dryer life, and 

membrane life. 

Controlled Field Evaluation 

The controlled field evaluation was conducted 

in several areas around Miami. This location 

was chosen for its high marine vessel traffic as 

well as its amenability to conducting testing on 

live targets in a pierside environment. 

Throughout the controlled field evaluation, a 

total of three samples were taken from each 

test location. Two of the samples were taken 

with the Sabre 2000 swipes (swipes specifically 

manufactured by Barringer Instruments for use 

with the Sabre 2000) for analysis on the two 

instruments being utilized for the test. The third 

swipe was taken on traditional Coast Guard 

sample swipe paper (Schleicher & Schuell filter 


Chih-Wu Su et al.: „Operational assessment of a handheld...”, IJIMS 3(2000)1, 45-47, p. 47 

paper grade 404) for analysis on an Ionscan 

400B. 

The first area of testing was at a U.S. Customs 

seizure yard. This area had several boats 

impounded for smuggling drugs. This location 

enabled samples to be taken from areas of 

known contamination which would provide 

valuable side-by-side comparison data between 

the Sabre 2000s and the Ionscan. As 

mentioned, each location that was sampled 

was analyzed on two Sabre 2000’s and on an 

Ionscan 400B. 

The second area of testing was at a Florida 

Highway Patrol (FHP) station. Approximately 

one year earlier, FHP had seized a van with a 

hidden compartment containing a significant 

amount of cocaine. Once again, this location 

enabled samples to be taken from areas of 

known contamination which would provide 

valuable side-by-side comparison data between 

the Sabre 2000s and the Ionscan. 

The final area of testing was on actual boats 

arriving in Miami and mooring in the Miami 

River. This area was chosen because it 

enabled testing to be conducted on actual 

vessels similar to those that are boarded on the 

high seas. Considering that these were live 

targets, it was not known if these targets would 

alarm, nor did we have any prior knowledge of 

the contaminants´ make-up or their levels. This 

phase of the testing provided information on 

how the Sabre 2000 will perform on an average 

target of interest. 

Go/No-Go Decision 

At this stage of the assessment, the Sabre 

2000’s physical characteristics and analytical 

capabilities have been thoroughly investigated. 

Based on these findings, a decision is being 

drafted which will govern whether the 

assessment should progress to its final phase, 

an at-sea evaluation. The purpose of holding 

this review is to ”weed out” instruments that are 

unlikely to be fully employable under CG 

operating scenarios regardless of technical 

achievement at this point. In broad terms, the 

assessment and decision to proceed is based 

on one of two criteria being met: 

• The instrument provides vastly superior 

analytical performance with similar 

physical characteristics. 

• The instrument provides similar analytical 

performance with vastly superior physical 

characteristics. 

At-Sea Evaluation 

The final step in the evaluation of the Sabre 

2000 will be deploying the instrument on a 

Coast Guard cutter for use during live 

boardings. At the completion of this period, 

several qualitative determinations will be made: 

• Transportation Requirements – How 

difficult is it to transfer the Sabre 2000 from 

the cutter to the rigid hull inflatable boat 

(RHIB) and then to the vessel being 

boarded? How waterproof/water resistant 

is the Sabre 2000 during this transfer? Do 

special precautions have to be taken to 

insure the instrument stays dry? How 

capable is the Sabre 2000 of withstanding 

an impact during the transfer as well as 

while being lifted to and from the RHIB? 

• Battery Life – Does the internal battery on 

the Sabre 2000 provide sufficient operating 

time to conduct a boarding or is an 

extended-life external battery pack 

required? 

• Is the bridge of the ship the best place to 

set the Sabre 2000 up? If not, where is the 

best place? 

Finally, the lessons learned during the at-sea 

evaluation will provide valuable insight into the 

development of operational protocols. 

Conclusions 

The Coast Guard has successfully used IMS 

technology for several years. Should the Sabre 

2000 provide a similar level of analytical 

confidence as the Ionscan, the ability of the 

Coast Guard to conduct a boarding will be 

revolutionized. To this point, the Coast Guard 

has never had the ability to conduct on-site 

analyses of samples due to the lack of 

handheld, high temperature IMS instruments 

that would accept wipe samples. In the 

ever-changing world of drug smuggling, this 

capability is a must if the Coast Guard is to 

continue to provide a real deterrent to 

smugglers trying to bring narcotics to the 

shores of the United States. 


In-situ methylation of methamphetamine during IONSCAN analysis 

Chih-Wu Su 1 , Steve Rigdon 2 and Kim Babcock 2 

1 

United States Coast Guard Research & Development Center, 1082 Shennecossett Road, Groton, CT 06340 USA 

2 

Anteon Corporation, Rte. 2, P.O. Box 220, North Stonington, CT 06359 USA 

INTRODUCTION 

The Ionscan, an instrument manufactured by 

Barringer Instruments based on ion mobility 

spectrometry (IMS) technology, is well known 

for its field application in illicit drug interdiction. 

It is currently used by the U.S. Coast Guard 

(CG) and by the Drug Enforcement Agency 

(DEA) to support federal, state and local drug 

cases. 

complicated the observed interference of 

nicotine with Meth detection. In order to 

investigate the nature of Meth-2, a study was 

launched to isolate and identify it. 

Consequently, an interesting in-situ methylation 

process during Ionscan analysis was 

discovered. This paper reports the results of 

this study as well as some possible 

applications. 

The CG Research and Development Center 

(R&DC) carried out an investigation to 

determine the source and identity of an extra 

peak that appeared in Ionscan results for 

methampetamine (Meth), in liquid standard 

form, on filter paper. It was observed that the 

extra peak, tentatively named Meth-2, 

appeared just after Meth on the Ionscan 

plasmagram (Figure 1) and had a drift time 

similar to that for nicotine. In 1996, Ms. Angela 

DeTulleo of the DEA South Central Lab 

reported that nicotine interfered with the 

detection of Meth 1 . In 1999, Dr. Peter 

Harrington’s group from Ohio University in 

Athens, Ohio, reported a procedure that could 

pre-separate Meth and nicotine using the 

difference in their vapor pressures and the 

SIMPLISMA (SIMPL-to-use-Interactive 

Self-Modeling Mixture Analysis) method 2 . 

Although Dr. Harrington’s method solved the 

nicotine interference problem, it required the 

use of an additional device as well as extra 

sample preparation time. The CG Research 

and Development Center (R&DC) has always 

been involved in the investigation and 

development of fast and easy field applicable 

methods for CG boarding officers to confirm 

and reduce/eliminate interferences on target 

compounds detected by the Ionscan. The 

unknown peak called Meth-2 further 

EXPERIMENTAL RESULTS AND 

DISCUSSION 

An Ionscan Model 400B with its drift tube, inlet 

tube and desorber operated at 235ºC, 290ºC, 

and 280ºC, respectively, was used to study 

possible sources of Meth-2 formation. A 

Finnigan MAT Model INCOS 50 quadrupole 

mass spectrometer (MS) coupled to a 

Hewlett-Packard Model 5890A gas 

chromatograph (GC) and UltraXL data system 

were used to identify the chemical structure of 

Meth-2. Schleicher and Schuell grade 404 filter 

paper (S&S404), Barringer Teflon discs and 

140-mesh stainless steel screen were used as 

sample holders for the Ionscan 400B. 

Investigation of sources of Meth-2: Meth exists 

as two optical isomers, d-Meth and l-Meth. If 

the Meth-2 was due to the separation of d-Meth 

and l-Meth in the Ionscan, the amplitudes of 

Meth and Meth-2 should correspond to 

plasmagrams of pure samples of each of these 

isomers. Analysis of the Ionscan plasmagrams 

for d-Meth and l-Meth indicated they have 

identical drift times. Therefore, these two 

optical isomers of Meth are not separable by 

Ionscan analysis and neither can cause the 

Meth-2 peak. The possibility that an impurity in 

the Meth standard would cause Meth-2 to form 

was ruled out after Ionscan tests with Meth (5 

[The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the 

views of the Commandant or the Coast Guard at large. Any instrument, material or chemical referred to by its brand name does not 


constitute an endorsement by the authors, the Commandant or the Coast Guard at large.]

Chih-Wu Su et al.: „In-situ methylation of ...”, IJIMS 3(2000)1, 48-51, p. 49 

Meth Meth Meth-2 

Meth-2 

Figure 1. 20 ng Meth on S&S404 

Figure 2. Plasmagram of Meth-2 Extract 

10 

75 

50 

Meth 

Meth-2 

Ethamph 

Ethamph-2 

25 

0 

Figure 3. GC/MS TIC of Meth-2 

Figure 5. 80 ng Ethamph on S&S404 

72 

91 

Figure 4. Mass Spectrum of Meth-2 (m/z 45 to 200) 



ng to 1000 ng) on stainless steel screens or 

Teflon consistently indicated a single peak (i.e., 

no Meth-2 formation). Since the Meth-2 peak 

maximum amplitude increased with increasing 

amounts of Meth during Ionscan analysis, 

impurities in the S&S404 filter paper were also 

ruled out as the sole source of Meth-2 

formation. 

Isolation and identification of possible chemical 

structures for Meth-2: The formation of Meth-2 

was clearly due to the heating of Meth 

deposited on S&S404 in the Ionscan desorber. 

Large amounts of Meth-2 were extracted from 

two S&S404 filter papers pre-spotted with a 

total of 1 mg of Meth in methanol after being 

heated at 270ºC for 2.5 min. After centrifuging 

and concentration, the extract was analyzed by 

both Ionscan and GC/MS. The Ionscan 

plasmagram for the extract on stainless steel 

screen showed a major peak for Meth-2 and a 

trace amount of Meth (Figure 2). The total ion 

chromatogram (TIC) from the GC/MS analysis 

of this extract also showed a small Meth peak, 

and a large peak assigned as Meth-2 with a 

retention time of 9 min 40 sec (Figure 3). Meth 

has a spectrum with m/z=58 (CH 3CH(NHCH 3)) + 

as the base peak, m/z=91 from the tropylium 

ion with an amplitude about 10% as high as the 

base peak and an undetectable molecular ion 

at m/z=149. Meth-2 has a spectrum with 

m/z=72 as the base peak, m/z=91 about 10% 

as high as the base peak and an undetectable 

molecular ion (Figure 4). The overall 

characteristic of the mass spectrum of Meth-2 

resembles that of Meth except its base peak is 

14 amu larger. It is logical to assume that this 

increase can result from replacing any 

hydrogen in the m/z=58 fragment ion with a 

methyl group. It was hypothesized, therefore, 

that Meth and Meth-2 are homologs. The 

subtle difference (about 0.063) between their 

reduced mobilities supports this assumption. 

Four possible structures can be derived from 

Meth by replacing any one of the hydrogen 

atoms with a methyl group in the N-methyl 

aminoethyl moiety to form Meth-2. They are 

ethyl amphetamine (Ethamph), N,a,a,-trimethyl 

phenethylamine, N-methyl-1-phenyl propylamine, 

and dimethyl amphetamine (Dimeth). 

Dimeth is a tertiary amine and the other three 

are secondary amines. The NIST/EPA/NIH 

mass spectra database shows that all four 

compounds have the same mass spectrum as 

that of Meth-2. Ethamph was the only one that 

was commercially available for this study. The 

GC/MS TIC retention times of Ethamph and 

Meth-2 were different, indicating they are 

different compounds. Although Ionscan 

analyses showed that both Ethamph and 

Meth-2 have identical drift times, the 

plasmagram of Ethamph revealed an extra 

peak, Ethamph-2 (Figure 5). This plasmagram 

is, therefore, similar to the pattern in the 

plasmagram of Meth (Figure 1) where a 

primary peak appears for the compound 

deposited as well as a secondary peak with a 

slightly longer drift time for a related, product 

compound. Figure 2 shows that Meth-2 does 

not produce this secondary peak. Both Meth 

and Ethamph are secondary amines. Other 

secondary amines such as dioctylamine (DOA) 

and 3,4-methylene dioxymethamphetamine 

(MDMA) also show the same two-peak pattern. 

It was thus concluded that Meth-2 was not a 

secondary amine but a tertiary amine. Dimeth 

is the only chemical of the four possible 

chemicals that met this requirement. 

Synthesis of Dimeth: Two methods are 

commonly used to synthesize Dimeth via 

methylation of Meth in a liquid solution. The 

first is the reaction of Meth with methyl iodide. 

The second is the reaction of Meth with formic 

acid and formaldehyde (F/FA), known as the 

Eschweiler-Clarke reaction. Both methods 

were quickly checked by adding the proper 

reagent onto an S&S404 filter paper 

pre-spotted with Meth followed by analysis in 

the Ionscan. Meth-2 was observed in the 

reaction of Meth with the solution of F/FA but 

not with methyl iodide. Dimeth was then 

synthesized using the Eschweiler-Clarke 

reaction. The Ionscan plasmagram and the 

GC/MS TIC retention time matched those of 

Meth-2. Based upon these results, Meth-2 was 

identified as Dimeth. 

Formation of Meth-2 in the vapor phase: 

Samples from the real world contain a myriad 

of matrices such that the Meth will most likely 

not have direct contact with the cotton fabric as 

is the case of pre-spotting a liquid standard 

solution. To verify whether Meth has to be in 

contact with S&S404 filter paper in the Ionscan 

for the methylation reaction to occur, three 

pieces of material were used for this 

investigation: a) a stainless steel screen 

pre-spotted with a Meth standard solution; b) a 



Barringer Teflon O-ring; and c) a blank S&S404 

filter paper. These three materials were 

inserted into the Ionscan desorber together in a 

sandwich formation with the Teflon O-ring in 

the middle which prevented the outer layers 

from touching. Analyses were conducted with 

the blank S&S404 filter paper either over or 

under the pre-spotted screen. Results showed 

that Meth-2 was observed in all these tests. It 

was concluded that methylation of Meth by 

vapor emitted from S&S404 filter paper did 

occur. 

Application of the Eschweiler-Clarke reaction to 

Ionscan analyses: By applying a mixture of 

F/FA either directly to where the Meth was 

spotted on the filter paper or on a Teflon layer 

above or below the Meth (i.e., in the sandwich 

configuration) a high yield of Dimeth was 

achieved. This indicated that the Eschweiler-Clarke 

reaction methylated the Meth 

almost completely either in the liquid or gas 

phase conditions, thereby becoming a potential 

confirmation method for field Meth detection. 

This confirmation method can also be applied 

to other secondary amines such as the 

methylation of MDMA to 3,4-methylene 

dioxydimethylamphetamine (MDDMA), and the 

methylation of DOA to N-methyl DOA. A 

mixture of F/FA in Ionscan analyses will 

methylate primary amines and secondary 

amines but not tertiary amines. This 

phenomenon can be used to confirm whether 

an amine detected by the Ionscan is a tertiary 

amine or not. 

CONCLUSION 

The peak that appears in the Ionscan 

plasmagram just after the Meth peak, when 

Meth is spotted on S&S404 filter paper, has 

been identified as dimethyl amphetamine. The 

mechanism for its formation is still unresolved 

since it is still unknown whether the paper 

contains any formaldehyde or formic acid, 

regardless of whether they are present in free 

form or as by-products from heating the filter 

paper during analysis. 

Introduction of a mixture of F/FA into the 

Ionscan during analysis can be used as a 

confirmation method to identify primary and 

secondary amines. If an amine does not 

respond to this treatment, it is likely a tertiary 

amine. 

REFERENCES 

[1] DeTulleo-Smith, A.M., “Methamphetamine 

Versus Nicotine Detection on the Barringer Ion 

Mobility Spectrometer,” Proceedings of the 5 th 

International Workshop on Ion Mobility 

Spectrometry, Jackson, WY, August 20-22, 

1996. 

[2] Reese, E.S., Harrington, P.B., “The Analysis of 

Methamphetamine Hydrochloride by Thermal 

Desorption Ion Mobility Spectrometry and 

SIMPLISMA”, Journal of Forensic Sciences, Vol. 

14, No. 1, 1999, pp. 68 - 76. 


Ion mobility spectrometry in helium with corona discharge ionization source 

M. Tabrizchi* and T. Khayamian 

Chemistry Department, Isfahan University of Technology, Isfahan 84154, Iran 

ABSTRACT 

An ion mobility spectrometer is described using 

helium as the drift gas and corona discharge as 

its ionization source. The observed ion current 

is approximately seventy times larger than that 

noted for the conventional system having 63 Ni 

as the ionization source. The selectivity factors 

in helium are close to those in nitrogen. The 

relative separation of the peaks in helium is 

almost twice as much as that observed in 

nitrogen. This results in a major enhancement 

of the resolution. A two-fold increase in the 

capacity factor (k') for ions in helium is also 

realized. 

INTRODUCTION 

IMS has been considered as a relatively 

low-resolution technique. Many attempts, 

including the use of long drift tubes [1] and 

increasing the electric field [2], have been 

made to enhance the resolution of IMS. It is 

commonly believed that the resolving power will 

not be improved by changing the drift gas [3]. 

We report here the use of helium instead of 

air/nitrogen in IMS with corona discharge 

ionization source and show that, this not only 

enhances the resolution of IMS but also 

increases its ion current considerably. 

EXPERIMENTAL 

The corona discharge ion mobility spectrometer 

used in this study was constructed in our 

laboratory at Isfahan University of Technology 

and has been described elsewhere [4,5]. A 

corona discharge ionization source with a 

point-to-plane geometry was used in this 

instrument. The electric field strength was 310 

V/cm and typically a 30 µs pulse was used 

during the measurements in helium and a 150 

µs pulse in nitrogen. Flow rates of the carrier 

and drift gases were about 100 and 500 ml/min 

respectively. The gases used in this work was 

chromatographic grade helium and nitrogen 

(99.9995% Internmar B.V., Netherlands). Eight 

chemicals were used in this study including; 

acetophenone, n-butyl acetate, ethyl acetate, 

methyl methacrylate, dimethyl sulfoxide 

(DMSO), dimethyl methylphosphonate (DMMP), 

acetone, and methyl iso-butyl ketone (Fluka). 


The total ion current observed for helium 

corona was a quadratic function of the drift field 

and was about 40 nA at about 300 V/cm, which 

is almost four times greater than that of 

nitrogen corona and almost fifty times greater 

than that of 63 Ni source. The mobility spectra of 

all 8 compounds were obtained in helium and 

nitrogen at room temperature. As an example, 

the positive ion mobility spectra of acetone and 

DMMP recorded in helium and nitrogen are 

shown in Figure 1. Although the drift times are 

considerably shorter in helium, the number of 

peaks and the general shape of IMS spectra 

are similar to those observed in nitrogen. 

Therefore a correlation between each pair of 

spectra can be speculated. The dominant 

reactant ions in nitrogen are (H 2O) nH + . It has 

also been shown that the main reactant ions in 

helium are (H 2O) nH + [6]. The product ions are 

generated by a reaction between sample 

molecules and reactant ions. Since the reactant 

ions are the same, the product ions are also 

expected to be the same for a specific sample, 

provided the buffer gas does not attach to ions. 

Hill and his coworkers have observed the same 

product ions (protonated molecules) in different 

drift gases, nitrogen, helium, argon and carbon 

dioxide [3]. Assuming the existence of the 

same ions in helium and nitrogen for each 


M. Tabrizchi and T. Khayamian: „Ion mobility Spectrometry...”, IJIMS 4(2001)1, 52-56, p. 53 

sample, their corresponding spectra can be 

compared with each other, just as the 

comparing chromatograms, which are obtained 

from two different columns. The correlation 

between the two sets of spectra will be 

explored based on the selectivity factors, the 

resolution and the capacity factors. 

The selectivity factor 

The selectivity factor, α, for two species A and 

B is defined in chromatography as [7]: 

tB 

−tM 

α = 

(1) 

tA 

−tM 

where, t A and t B are the retention times for 

species A and B, and t M is the retention time of 

a species that is not retained by the stationary 

phase. If t M is replaced in IMS by t R, the drift 

time of the reactant ions peak (RIP), then the 

selectivity factor will be; 

tB 

−tR 

α = 

(2) 

tA 

−tR 

The drift times of the main product ions, usually 

referred to the monomer and dimer protonated 

molecules, in helium and nitrogen are displayed 

in Table1. The selectivity factors are calculated 

using a reactant ion drift time of 1.95 ms and 

10.70 ms in helium and nitrogen respectively. 

As Table 1 shows, apart from DMSO, the 

selectivity factors for each compound in helium 

and nitrogen are almost the same. 

The resolution 

The generally accepted resolving power for IMS 

is based on a single-peak [8], which is defined 

by R = t d /w 1/2 , where t d is the ion drift time and 

w 1/2 is the ion pulse duration at the detector 

measured at half maximal intensity. Because 

the drift time of the reactant ions in helium is 

typically 1/5 of that in nitrogen, an injection 

pulse of 30 µs was selected for helium as 

compared to 150 µs for nitrogen, for a 

reasonable comparison of resolving power. 

Generally, the resolving power for spectra in 

helium is slightly less than that in nitrogen 

mainly due to the shorter drift times. However 

for slower moving ions the resolving power 

increases. 

The resolution can also be defined on the basis 

of separation of pairs of adjacent peaks, as 

follows [7]: 

tB−t 

A 

RS 

= 

wA 

+ wB 

(3) 

where, t A and t B are the drift times of the two 

peaks, and w A and 

R w B are their 

s = 2(t B−t a ) 

w a −w b b respective widths. 

A high improvement in resolution (R S) is 

observed when helium is used. In order to 

compare the resolution, spectra were 

normalized by dividing them by t R, the reactant 

ion drift time. The normalized spectra of methyl 

iso-butyl ketone in helium and nitrogen are 

demonstrated in Figure 2. Apart from the peak 

widths, which are approximately the same, the 

separation in helium is about twice, resulting in 

a better resolution. The two-fold increase in 

peak separations was observed for all spectra. 

The IMS spectra of acetophenone and n-butyl 

acetate are presented in Figure 3. The 

separation between the first product ion peaks 

is increased in helium and the second product 

peaks are completely resolved where as in 

nitrogen they are strongly overlapped. Similar 

behavior is observed for DMMP and methyl 

iso-butyl ketone. 

The capacity factor 

The capacity factor, is an important 

experimental parameter in chromatography [7]. 

It is defined as, k' = (t/t M)-1, where t M is the 

retention time of a species that is not retained 

by the stationary phase. If again, t M is replaced 

in IMS by t R, the drift time of the reactant ion 

peak (RIP), then the capacity factor will be k' = 

(t/t R)-1. A two-fold increase in the capacity 

factor for all product ions was observed by 

switching from nitrogen to helium. This is 

demonstrated in Table 2 where capacity factors 

are calculated and compared for all spectra. 

The increase in the capacity factor of the 

product ions of each sample is generally the 

same. Therefore, for the selected compounds, 

the IMS spectra in helium and nitrogen match 

together if they are plotted versus the capacity 

factors and the nitrogen spectrum is multiplied 

by a factor of about 2. 

CONCLUSIONS 

It is generally believed that there is no 

fundamental advantage in terms of resolving 

power for the different drift gases [3]. However, 

the resolution depends not only on the 

resolving power but also on the capacity factor, 

(8) 



Table 1: Drift Times and Selectivity Factors in Helium and Nitrogen 

Sample 

Acetophenone 

n-Butyl acetate 

Ethyl acetate 

Methyl methacrylate 

t A (ms) a 

2.670 

2.836 

2.351 

2.437 

In Helium 

t B (ms) a 

4.110 

4.358 

3.424 

3.529 

Selectivity 

Factor b (α) 

3.00 

2.72 

3.67 

3.24 

t A (ms) a 

12.70 

13.15 

11.65 

11.95 

In Nitrogen 

t B (ms) a Selectivity 

Factor b (α) 

16.60 2.95 

17.00 2.57 

14.06 3.55 

14.60 3.12 

DMSO 

DMMP 

Acetone 

Methyl iso-butyl 

ketone 

2.054 

2.475 

2.110 

2.605 

2.946 

3.647 

2.490 

3.885 

9.59 

3.23 

3.37 

2.95 

11.05 

12.00 

11.24 

12.49 

13.00 

14.89 

12.20 

15.51 

6.57 

3.22 

2.77 

2.68 

a. A and B are the first and the second product ions respectively. 

b. Calculated using Equation 2 with a drift time of 1.95 ms and 10.7 ms for the reactant ion (t R) in helium 

and nitrogen respectively. 

Table 2. Capacity factors (k') for the first (A) and the second (B) product ions of the selected compounds 

Sample 

k’ A in N 2 

k’ A in He 

k’ B in N 2 

k’ B in He 

k’ A in He 

k’ A in N 2 

k’ B in He 

k’ B in N 2 

Acetophenone 

n-Butyl acetate 

Ethyl acetate 

Methyl methacrylate 

DMSO 

DMMP 

Acetone 

Methyl iso-butyl 

ketone 

0.187 

0.229 

0.088 

0.117 

0.033 

0.122 

0.051 

0.167 

0.369 

0.454 

0.206 

0.250 

0.053 

0.269 

0.084 

0.336 

0.551 

0.589 

0.314 

0.364 

0.215 

0.391 

0.140 

0.449 

1.108 

1.235 

0.756 

0.810 

0.511 

0.870 

0.281 

0.992 

1.9 

1.9 

2.3 

2.1 

1.6 

2.2 

1.65 

2.0 

2.0 

2.1 

2.4 

2.2 

2.3 

2.2 

2.0 

2.2 

k'. This work shows that, using helium as the 

drift gas instead of nitrogen or air in IMS 

increases the capacity factor and consequently 

improves the resolution. 

ACKNOWLEDGMENT 

The authors would like to thank Dr. M. K. Amini 

and Professor M. Amirnasr for valuable 

discussion and their help in this work. Isfahan 

University of Technology and the National 

Researches Council of Iran supported this 

work. 

REFERENCES 

[1] G. A. Eiceman, Ion Mobility Spectrometry, Boca 

Raton, CRC Press, (1993). 

[2] W. F. Siems, C. Wu, E. E. Tarver, H.H. Jr. Hill, P. R. 

Larsen, D. G. McMinn, Anal.Chem., 66 (1994) 4195. 

[3] G.R.Asbury and H. H. Jr. Hill, Anal. Chem., 72 (2000) 

580. 

[4] M. Tabrizchi, T. Khayamian, and N.Taj, Rev. Sci. 

Inst., l 71 (2000), 2321. 

[5] M. Tabrizchi, T. Khayamian and N. Taj, "Corona 

Discharge Ion Mobility Spectrometry" The 8th 

International Conference on Ion Mobility 

Spectrometry, Buxton,UK August 8-12, (1999). 

[6] D.R. Kojiro, M.J. Cohen, R.M. Stimac, R.F. Wernlund, 

D.E. Humphry, N. Takeuchi, Anal. Chem., 63, (1991), 

2295. 

[7] D. A. Skoog, D. M. West and F. J. Holler, Analytical 

Chemistry, Philadelphia, Saunders Golden Sunburst 

Series, 1994. 

[8] C. Wu, W. F. Siems, G. R. Asbury, H.H. Jr. Hill, Anal. 

Chem., 70 (1998) 4929. 



DMMP 

in He 

in N 2 

Acetone 

0 5 10 15 20 


Figure 1. Selected ion mobility spectra in helium and nitrogen. 

RIP 

A 

in N 2 

DMMP in He 

B 

RIP 

A 

DMMP in He 

B 

DMMP in N 2 

0.5 1.0 1.5 2.0 

Relative Drift Time (t / t R ) 

Figure 2. The detectore responce versus the relative drift time (t/t R ) 

for methyl-iso-buthyl ketone. The t R is the reactant ion drift time. 



RIP 

In Helium 

a 

b 

0 1 2 3 4 5 6 7 

RIP 

In Nitrogen 

a 

b 

0 5 10 15 20 25 

Drift Time (msec) 

Figure 3. IMS spectra of a) acetophenone and b) n-buthyl acetate, 

recrded in helium (top spectra) and in nitrogen (bottom) spectra. 


External exit gate Fourier transform ion mobility spectrometry 

Ed Tarver*, James F. Stamps, Richard T. Jennings, William F. Siems 

Sandia National Laboratories, Livermore, California 94551-0969, USA 

Ion mobility spectrometry (IMS) has been 

recognized as one of the most sensitive and 

robust techniques for the detection of 

explosives, narcotics and chemical warfare 

agents. The extreme sensitivity and electronic 

simplicity of IMS instrumentation has offered 

tremendous theoretical potential for this 

analytical technique. While the detection limits 

achievable with IMS are rivaled by relatively few 

alternative methods (e.g., mass spectrometry, 

biochemical detection, and flame photometric 

or other chemically specific detectors interfaced 

to gas chromatographs), none can match the 

combination of sensitivity and speed of 

response. Coupled with low cost, ruggedness 

and atmospheric pressure operation, IMS has 

promised to be the solution for a variety of 

analytical challenges. We have recently 

demonstrated improved spectral resolution as 

well as significant signal-to-noise enhancement 

by implementing the external exit gate Fourier 

transform technique on both commercial and 

Sandia National Laboratory proprietary IMS 

instruments. 

The problem in IMS has always been resolution 

of the complex mobility spectra resulting from 

the inherent sensitivity of the method. The 

broad, tailing and sometimes overlapping 

peaks of the spectra have necessitated the 

development of complex peak de-convolution 

and recognition algorithms or the introduction of 

pre-separation methods, which lead to loss of 

sensitivity and increased analysis time. The 

challenge in IMS is to maintain the chemical 

sensitivity and response time while increasing 

the spectral resolution to aid in accurate 

identification of target analytes. In common 

IMS instruments, spectra are generated by 

pulsing open the entrance to the drift region for 

0.2 ms and then monitoring the ion current after 

a 20 ms drift time period. This brief entrance 

pulse represents only 1% of the 20 ms duty 

cycle. A longer ”gate” pulse would allow more 

ions to be collected and increase the signal, but 

short entrance pulse duration is necessary to 

prevent unacceptable peak broadening and 

decreased resolution. The poor resolution 

typical of conventional IMS instruments is the 

result of this trade-off for increased sensitivity. 

Even so, acceptable signal-to-noise (S/N) is 

achieved by repeating and storing many scans 

and then summing the signals with a computer. 

Signal averaging IMS has limited ability to 

resolve adjacent peaks in a complex spectrum, 

which is critical to unambiguous sample 

identification and elimination of false positives. 

An alternative to the signal-averaging 

methodology is Fourier transform ion mobility 

spectrometry (FT-IMS). In this mode of 

operation two ion gates are employed, an 

entrance gate and an external exit gate. The 

entrance gate admits ions into the drift region 

of the spectrometer in a manner similar to 

signal-averaging IMS with the distinction being 

the 50% duty cycle of the gates in FT-IMS. The 

downstream exit gate is pulsed synchronously 

with the entrance gate at increasing frequency 

to interact with the flowing ion stream and 

generate a frequency domain interferogram. 

This interferogram which contains velocity 

information about all of the ions in the spectrum 

is then fast Fourier transformed to recover the 

complete time domain mobility spectrum. We 

have demonstrated the increased S/N resulting 

from the 50% duty cycle of the gates (S/N 

increases as v50). Also, ion-molecule reactions 

and labile clustering during ion transit through 

the drift tube results in peak broadening due to 


E. Tarver: „External exit gate Fourier transform...”, IJIMS 3(2000)1, 57-59, p. 58 

random variations in ion velocities. We have 

demonstrated the phasing action of the gates in 

FT-IMS eliminates this noise signal and greatly 

improves spectral resolution. 

Signal Averaged Ion Mobility Spectrometry 

Ion mobility spectrometry (IMS) is an important 

trace analytical method used for the vapor 

phase detection of drugs, explosives, chemical 

warfare agents, and hazardous workplace 

pollutants. The low cost and relative simplicity 

of the electronics as well as atmospheric 

pressure operation have allowed this 

instrument to be miniaturized for use in the field 

by the military as sensor/alarm for chemical 

warfare agents. Because IMS responds to 

numerous organic functionalities, the resulting 

spectra are complicated and often poorly 

resolved when complex sample matrices are 

introduced into the spectrometer. These 

spectra may be simplified by modification of the 

reagent ion chemistry (similar to chemical 

ionization mass spectrometry), selective ion 

gating, or chromatographic separation prior to 

introduction into the detector. To better identify 

target compounds from mobility spectra, 

improvements in resolution and signal-to-noise 

ratio are critical. In typical IMS instruments, 

where the entrance gate is open for ion 

collection for 1% of the analytical cycle, the 

short entrance pulse duration is necessary to 

prevent unacceptable peak broadening and 

decreased resolution. (1) Acceptable S/N ratio is 

achieved by repeating many scans and 

averaging with a computer. In addition, 

significant contributions to peak broadening 

result from random ion-molecule reactions and 

labile cluster formation occurring during ion 

transit through the drift tube. When events such 

as these occur in the drift region on the time 

scale of the separation, peak broadening is 

observed in the signal averaged spectra due to 

the summation of the variations in ion 

velocities. This dynamic peak broadening is 

recorded in the signal-averaged spectrum and 

serves to compound diffusional broadening and 

further degrade the peak resolution. Figure 1 

shows the signal-averaged trace produced from 

the unmodified Barringer IonScan 400 by the 

major background reactant ion peak at 8.5 ms 

and the calibrant peak at 11.3 ms. 

Fourier Transform Ion Mobility 

Spectrometry 

An alternative method of signal acquisition 

used in IMS involves the Fourier transformation 

of a mobility interferogram generated when the 

entrance and exit gates of a two gate IMS are 

simultaneously opened and closed by a 

frequency sweeping generator. (2) The entrance 

gate creates a continuous stream of ion pulses, 

which drift down the tube with their 

characteristic drift time, t d. As the square wave 

frequency is varied, ions of different drift time 

pass in and out of phase with the exit gate 

generating a mobility interferogram. With this 

Fourier Transform IMS (FT-IMS) technique the 

hardware gates inside the drift tube are pulsed 

opened (50% of the time) and closed (50% of 

the time) at increasing frequency over the total 

duty cycle. This allows half of the initial ion 

population approaching the first gate to be 

admitted into the drift tube and of this ion 

population half passes through the second gate 

used to generate the interferogram. This 

interferogram is Fourier transformed to recover 

the complete time dispersive mobility spectrum. 

As a result a 25% duty cycle is achieved and 

the S/N is improved. By the removing second 

hardware gate and performing its function, 

outside of the drift tube, in the electronics, the 

ion loss at the second hardware gate is 

eliminated. Utilizing this ”external second gate” 

technique we have developed a rapid scanning 

FT-IMS system capable of achieving a 50% 

duty cycle. We have also demonstrated the 

phasing action of the dual FT-IMS gates 

eliminates the peak broadening caused by 

random variation in ion velocities and greatly 

improves spectral resolution. In order for an ion 

to contribute to the maximum measured signal 

it must maintain a constant velocity to be in 

phase with a corresponding gate opening 

frequency. If the ion randomly changes velocity 

it loses the necessary phase correspondence 

with the gates and does not contribute to the 

measured signal. 

The electronic and software modifications we 

have developed enable us to adapt common 

single gate instruments to this external second 

gate FT-IMS method. A working bench-top 

prototype has been constructed and the 

performance demonstrated. No major 

alterations or fabrication of new hardware was 

required to adapt the instrument (Barringer 

Instruments IonScan 400) to achieve external 


E. Tarver: „External exit gate Fourier transform...”, IJIMS 3(2000)1, 57-59, p. 59 

Figure 1: 

Typical IMS spectrum of reactant ion peak (8.5 ms) 

and calibrant peak (11.3 ms) produced from 

Barringer Ionscan 400 operating in normal mode 

(multiple scans, signal averaged). The peak tailing 

is due to variations in ion velocity (mobilities)from 

scan to scan caused by ion-molecule reactions. 

The contributions from each individual scan are 

recorded in the final signal averaged spectrum. 

Figure 2: 

FT-IMS spectrum of Reactant Ion Peak (8.5 ms) 

and Calibrant Peak (11.3 ms) produced from a 

Barringer Ionscan 400 modified for FT-IMS 

operation. The S/N is is improved because the 

gates are operating at 50% duty cycle allowing 

more ions to be collected in FT mode. The peak 

tailing, caused by variations in ion velocity, is 

eliminated due to the phasing action of the gates 

which require constant ion velocity to produce a 

signal. 

All other instrument conditions identical to normal 

(signal-averaging) mode. 

second gate FT-IMS. While the ease of 

adaptability of the commercial IMS was 

demonstrated, the limitations of the 

(signal-averaging) instrument design did not 

allow us to demonstrate the optimal theoretical 

performance expected from external second 

gate FT-IMS (e.g. theoretical increase in S/N = 

√50 = 7).Operating the same Barringer IonScan 

400 under identical experimental conditions 

following the Fourier transform modifications 

we produced the trace shown in Figure 2. The 

significant gain in signal-to-noise is the result of 

the increased ion transmission efficiency 

produced by the 50% duty cycle of the FT-IMS 

gate system. Also apparent is the dramatic 

improvement in spectral resolution resulting 

form the phasing action of the ion gates using 

the external second gate Fourier transform 

method. 

References 

[1] Siems, W.F.; Ching Wu; Tarver, E.E.; Hill, H.H.; Anal 

Chem (1994) 4195-4201 

[2] Knorr, F.J.; Eatherton, R.L.; Siems, W.F.; Hill, H.H.; 

Anal Chem (1985) 57, 402-406 


Miniaturized ion mobility spectrometer 

M. Teepe 1 , J.I. Baumbach 2 , A. Neyer 1 , H. Schmidt 2 , P. Pilzecker 3 

1 

University of Dortmund - Dep. Micro-Structure-Technologies, Otto-Hahn-Str. 6, D-44227 Dortmund, Germany 

2 

Institut für Spektrochemie und Angewandte Spektroskopie (ISAS), Bunsen-Kirchhoff-Str. 11, D-44139 Dortmund, 

Germany 

3 

G.A.S. Gesellschaft für analytische Sensorsysteme mbH, TechnologieZentrumDortmund, Emil-Figge-Str. 76-80, 

D-44227 Dortmund, Germany 

Introduction 

Miniaturization of an ion mobility spectrometer 

(IMS) can lead to a handy low cost on-line and 

on-site environmental scanning system for 

exhausts in chemical, waste and petrochemical 

industries. 

Ion mobility spectrometry [1] has changed from 

a not common known method in the late sixties 

[2,3] to a practical method for the detection of 

toxic pollutants [3] and explosives [4, 5] and 

some other analytes [1] in the atmosphere at 

ambient pressure and temperature down to the 

ppb V (parts per billion) concentration range. 

The size of a common high resolution IMS (in 

the range of normal Personal Computers) 

makes it useful in process control for industry 

[6] and in centralised measurement systems 

like gateways at airports. But ion mobility 

spectrometry is a more powerful detection 

method than it is now used for. Many other 

applications are proposed in the scientific 

literature [7]. In the growing semiconductor 

industries, especially for Reactive Ion Etching 

an on-line gas control can lead to a more 

reliable process. New fields of application are 

possible like ‘intelligent’ on-line control for air 

conditions in cars and other vehicles. For these 

applications it is useful to have a small, fast, 

economical and powerful analytical on-site and 

on-line system. A miniaturized IMS with size of 

today’s common mobile phones will meet such 

needs. A prototype µ-IMS - ten times smaller 

than conventional ones - with components 

produced by micro-structure-technologies is 

presented in this paper. Successful results 

using this spectrometer are reported. Finally an 

outlook to a modular system design with 

respect to mass production is shown. 

Concept 

Portability is one of the main aspects for 

miniaturization of an IMS. It leads to an on-site, 

on-line and real-time monitoring system for 

volatile organic compounds, regardless of the 

origins of these compounds. A size not bigger 

than a normal cellular phone should be 

reachable with the results of this work. Using a 

short - in optimum infinitesimal - gate opening 

time a diffusion controlled domain in ion 

transport can be reached [8] which then will 

lead to a resolution of the IMS independent 

from gate opening time. The problem in this 

case is to have decreasing amounts of ions to 

amplify and the need for direct amplification 

nearby the Faraday plate. The connection 

between the Faraday plate and the 

pre-amplifier should be as short as 

implemental. A ”single chip” solution should 

lead to the best results in ion mobility 

spectrometry. As a result of miniaturization the 

costs for the measurement system will be 

reduced dramatically because of mass 

production techniques and a decreased use of 

resources. 

A first prototype for the proof of a possible 

miniaturization has been designed without 

decreasing the size of the supporting electronic 

except the amplifier to obtain a direct 

amplification. In respect to a flexible drift length 

modification and field shaping a 

sandwich-design has been chosen. Figure 1 

shows an example of a working miniaturized 

drift tube in comparison to a conventional one. 

In miniaturization imprecise alignment of the 

free space electrodes of the ion gate leads to 

high electric field forces and results in 

deformation of the grid. In worst case a short 

circuit is the consequence. Thus a stiff 

conducting material has to be selected for the 


M. Teepe et al.: „Miniaturized ion mobility spectrometer”, IJIMS 4(2001)1, 60-64, p. 61 

Figure 1: 

Conventional drift tube and drift tube 

used in a µ-IMS 

Figure 2: 

Conventional ion gate and a 

laser-structured component 

ion gate. A highly precise microstructuring of 

such a material is possible by etching, 

micro-milling, 

micro-electro-discharge-machining (micro-edm) 

or laser-ablation. Micro-milling and micro-edm 

are no real batch processes and therefore not 

cheap enough for mass fabrication. Etching 

always comes along with masking techniques 

which usually are (expensive) photolithography 

processes or equivalent procedures. 

Laser-ablation and laser cutting are well known 

techniques for micro-structure-technologies and 

are available at low costs considering the sizes 

and the number of units needed here. In Figure 

2 a conventional ion gate is compared to a 

laser-structured component. Further progress 

has been made for the stability of the prototype. 

Figure 3 shows a complete set-up of a working 

device with the electronic for the ion gate, the 

drift tube with the Faraday plate, the ion gate 

Figure 3: µ-IMS-Prototype 

and the connectors. The coin in the front has a 

diameter of 15 millimeters. 

The total length of this device is less than 6 cm. 

The total ion drift length in this unit is 3.6 cm, 

the inner diameter is 3 mm. A 63 Ni-ß-radiation 

source is modularly implemented in order to 

enable a quick change of two different sources 

with 400 kBq and 16 MBq. The laser-structured 

Bradbury-Nielsen ion gate (Fig. 2) is also built 

in. As mentioned before in this phase of the 

project the supporting unts like high voltage 

supply and flow controllers have not been 

minimized. 

The shown arrangement (Fig. 3) is meant to 

compare characteristics of a conventional IMS 

with the miniaturized prototype and therefore 

was built compatible to existing big supporting 

units. 

Experimental Results 

Successful results obtained with the described 

prototype are shown in Fig. 4 - 5. The graphic 

in Fig. 4 shows the Reactant Ion Peak in 

nitrogen at different ion gate times down to 

100 µs. The applied drift voltage is in the range 

of 2 kV. As expected from a conventional 

device the intensity of the ion signal of the 

µ-IMS decreases with decreasing opening time 

of the ion gate. Even in the 100 µs range an 

adequate signal to noise ratio is obtained. 

Therefore the opportunity to reach the diffusion 

controlled drift domain mentioned before should 

be achievable. 

The resolution (drift time / full width at half 

maximum) is approximately 5 at 1 ms up to 16 

at 300 µs. In Fig. 5a a spectrum of acetone in 



Signal / a.u. 

-0.30 µ-IMS 

-0.25 

-0.20 

-0.15 

-0.10 

-0.05 

0.00 

5 10 15 20 

Drift Time / ms 

Ion Gate Time / µs 

100 

300 

1000 

Figure 4: Reactant ion peak in nitrogen 

nitrogen is shown. The applied gas flow for the 

analyte gas inlet was 5 ml/min. The 

concentration level of this measurement is in 

the range of 1 ppm v. The respective resolution 

is discernable to 20. 

In Fig. 5b and 5c the spectra with higher ion 

gate times of 300 µs and 1 ms are added. As 

one would expect a higher intensity of the ion 

signal is reached according to increasing ion 

gate time. In order to allow a better comparison 

of the spectra values are normalized to the 

maximum signal of 1 ms. Though every 

measurement is shutter time corrected a 

Gaussian curve analysis shows a slight 

difference between the peak positions of 

0.2 ms (Fig. 5b). Considerable are two possible 

reasons for this effect: a changing ambit and 

the ion gate. 

No pressure correction was performed in these 

measurements and in this phase of the 

development the µ-IMS is not stabilized in 

temperature so that the ambient temperature 

Signal / a.u. 

0.0104 

0.0078 

0.0052 

0.0026 

0.0000 

µ-IMS 

5 6 7 8 9 10 

Driftzeit / ms 


300 

100 

Figure 5b: 

Gaussian curve analysis of the peak 

shape 

Gate Time Corrected 

and pressure may have changed on different 

days of measurements. 

Furthermore the Bradbury-Nielsen shutter and 

its electronics have been optimised for a 100 µs 

pulse width so that the results with other 

opening times may differ. It can be seen clearly 

that the Gaussian curve analysis shows an 

expected regular peak shape at 100 µs ion 

gate time but not in the case of the peak with 

the 300 µs (Fig. 5b). These results show the 

importance for further studies on ion gate 

design in miniaturization. 

The resulting resolution of the ion mobility 

spectrum obtained using a gate time of 1 ms 

(Fig. 5c) is insufficient for successful measurements 

and a complete characterisation of the 

compounds to be detected. 

Modular System Design 

The next step in this miniaturization project is 

asystem design illustrated in Fig. 6. It shows a 

5,00 

-0,025 

3,75 

-0,020 

Acetone 

Signal / a.u. 

2,50 

1,25 

Acetone 

Signal / a.u. 

-0,015 

-0,010 


1000 

300 

100 

-0,005 

0,00 

0,000 

5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 


5,0 7,5 10,0 12,5 15,0 


Figure 5a: 

Ion mobility spectrum of acetone 

in nitrogen at 100 µs ion gate time 

Figure 5c: 

Ion mobility spectra of acetone in 

nitrogen at different ion gate times 



Signal / a.u. 

0,4 

0,3 

0,2 

0,1 

0,0 

-0,1 

Figure 6: 

A modular system design 

Figure 7: 

Prototypes of molecular design 

4 5 6 7 


Trichloroethene 

Tetrachloroethene 

Perfluorohexane 

Figure 8: 

Gaussian curve analysis of 

ion mobility spectra from different volatile 

organic compounds (negative ions) 

modular arrangement for a Micro-IMS (µ-IMS) 

with respect to mass fabrication, optimized field 

shaping and a higher signal to noise ratio. The 

three different groups building the functional 

parts of a onventional IMS are still maintained 

i.e. the ionisation chamber, the drift tube and 

the Faraday plate. This concept has certain 

advances considering the miniaturized IMS 

prototype: first of all the whole concept is 

worked out taking mass fabrication into 

account. The different electrical components 

are compatible to printed circuit board 

technology which is a commonly known and 

very low cost fabrication technique. The 

micro-structured grids can be changed quickly 

which leads to a flexible design modification for 

optimisation of ion gate design. The new drift 

tube design is one solid and therefore stable 

component. Standard fittings for gas inlets are 

implemented. Modular ionisation sources for 

63 

Ni, UV-ionisation and partial discharges have 

been taken into consideration and are partly 

realised today. 

The Fig. 7 shows the final arrangement of this 

concept with a size comparison. 

First results from this spectrometer are given in 

Fig. 8. Spectra of trichloroethene, tetrachloroethene, 

perfluorohexane are presented to 

show the feasibility of such a device. The 

applied settings for this measurement are 

summarized in Fig. 9. A Gaussian curve 

analysis show characteristic differences 

between fluorinated and chlorinated 

compounds. As expected from conventional 

IMS both two chlorinated substances show 

comparable ions assuming three Gauss curves 

beneath. 

Summary and Outlook 

The results mentioned above prove a 

successful miniaturization of an IMS. A 

demonstrator has been built and successful 

results obtained by this demonstrator show an 

adequate performance. Further developments 

have to be made on ion gate design and in field 

shaping for higher resolutions. Tests with 

ionisation sources different to the 63 Ni source 

like UV-light and partial discharges are taken 

into consideration for further developments. 

With miniaturized electronics and high voltage 

supplies this demonstrator leads to a working 

device in the size of a common cellular phone. 




The financial support of the Bundesministerium 

für Bildung, Wissenschaft, Forschung und 

Technologie and the Ministerium für 

Wissenschaft und Forschung des Landes 

Nordrhein-Westfalen is gratefully acknowledged. 

References 

[1] J.I. Baumbach.; G.A. Eiceman,.: Ion mobility 

spectrometry: Arriving on-site and moving beyond a 

low profile. - Applied Spectroscopy 53(1999)9, 

338A-355A 

[2] M.J. Cohen, F.W. Karasek: Plasma chromatography - 

a new dimension for gas chromatography and mass 

spectrometry, J. Chromatogr. Sci.8, 330-337, 1970. 

[3] G.A. Eiceman, Z. Karpas: Ion Mobility Spectrometry, 

CRC Press Inc., Boca Raton - Ann Arbor - London - 

Tokyo, 1994. 

[4] F.W. Karasek, D.W. Denney: Detection of 

2,4,6-trinitrotoluene vapours in air using plasma 

chromatography, J. Chromatogr. 93, 141-147, 1974. 

[5] M.J. Cohen, R.F. Wernlund, R.M. Stimac: The ion 

mobility spectrometer for high explosive vapor 

detection, Nucl. Mater. Manage. XIII, 220-240, 1984. 

[6] W.J. Holzapfel, K.J. Budde: Ultratrace analysis for 

volatile organic compounds in semiconductor 

industry, Fresenius J. Anal. Chem. 343, 769770, 

1992. 

[7] T. Kotiaho, F.R. Lauritsen, H. Degn, H. Paakkanen: 

Membran inlet ion mobility spectrometry for on-line 

measurements of ethanol in beer and in yeast 

fermentation, Anal. Chim. Acta 309, 317-325, 1995. 

[8] J.I. Baumbach, D. Klockow,: Diffusion controlled 

transport of ion swarms in gases and its relevance for 

the resolution of ion mobility spectrometers, 

Proceedings of XXI International Conference on 

Phenomena in Ionized Gases, Bochum, 

19.-24.9.1993 


Evolution of IMS technology within the Australian Customs Service 

G. Webster 

INTRODUCTION: 

Customs has been using IMS technology since 

1997, in the form of the Barringer Ionscan, 

when an air cargo examination team at Sydney 

International Airport commenced an operational 

evaluation of the equipment. The location of 

this one unit and the specially selected officers 

to manage it have proven critical to the overall 

success of the technology. Without the 

initiative of these officers we may not have had 

the first taste of success that ultimately led to 

the introduction of further units. This initial 

success involved the seizure of 330 grams of 

cocaine concealed in the frame of a mountain 

bike. The presence of the cocaine “alarm” 

confirmed suspicions regarding the bike, which 

led to discovery of the illicit drug. 

IMS TECHNOLOLGY INTRODUCED 

The success of the trial paved the way for the 

introduction of the first 20 Ionscan units 

purchased by Australian Customs. Eighteen 

months later a further 21 units were deployed 

and in May this year our total number of 

Ionscans rose to 51. These are deployed in all 

Australian seaports and airports, from Hobart, 

Tasmania in the far south to Darwin in the 

tropical north to some of our more remote 

localities. These Ionscans form part of the 

arsenal of drug detection tools used by 

Customs. These include drug detector dogs, 

X-ray systems and Closed Circuit TV. 

EDUCATION OF IONSCAN USERS 

One of the greatest challenges following the 

introduction of the Ionscan was to get our 

officers to use it proactively. Internal marketing 

and training has ensured the best use of these 

resources. Late last year saw the development 

of a national training program for Ionscan 

users. We found that the training program 

resulted in the units being better maintained, 

with better results. The following program was 

implemented in March this year: 

v A fully trained Competency Assessment 

Training Officer, or “CATO”, oversees the 

National Ionscan Training Program in each 

region. These officers are qualified to 

deliver training in their area; 

v The program caters for two levels of 

operator. The Basic Level Operator course 

is a one-day course designed for those 

officers who use the Ionscan as part of their 

everyday duties. The Advanced Level 

Operator course is another day long course 

and is for those officers who are required to 

conduct more complex operations and 

routine Ionscan maintenance; 

v The training program has ensured national 

consistency in the use of the Ionscan and a 

strong network of users and information 

sharing has been established. 

The message in education of users is that 

whilst IMS technology, in the form of the 

Ionscan, is not difficult to use, interpretation of 

results is the key to more effective utilisation. 

THE EVOLUTIONARY PROCESS: 

There are numerous examples of how the 

Ionscan has assisted Australian Customs in the 

detection of significant illicit drug seizures and 

the education program has provided our 

officers with the skills to use this technology in 

its most innovative and aggressive fashion. For 

example, IMS technology was instrumental in 

the detection and seizure of 115 kilograms of 


G. Webster: „Evolution of IMS technology...”, IJIMS 4(2001)1, 65-66, p. 66 

black cocaine – the first seizure of its type in 

Australia. 

Customs has, until very recently, been the only 

law enforcement user of IMS technology in 

Australia. As we are geographically quite 

isolated from those countries which have been 

using this tool for quite some time, we built 

upon the technical skills and knowledge 

provided by the manufacturer in developing our 

own procedure and policies for use. We are 

now in a position to assist other agencies in 

Australia and the Pacific who are embarking on 

an IMS program. We have trained officers 

from Australian State Police Forces and Fiji 

Customs, and provided New Zealand Customs 

with details of our training program. 

IMS TECHNOLOGY IS ONLY A TOOL 

Despite its reliability and proven success in 

narcotic interdiction, the Ionscan is only a tool 

and is to be used in the same way we use our 

other tools, such as x-ray and dogs. As with all 

technology an important fact remains: 

“Ionscan’s’ don’t find drugs, Customs Officers 

find drugs.” 

FUTURE OF IMS TECHNOLOGY: 

As the term “evolution” suggests, we must 

continue to learn and grow if we are to 

accomplish what we believe is possible with this 

technology. The future of the program will see 

the following: 

♦ 

Continued development of our national 

Ionscan training program to cater for 

scientific and operational changes & 

training of further regional Competency 

Assessment Training Officers to allow for 

succession planning; 

♦ Liaison with other law enforcement 

agencies using the technology in Australia 

and overseas to attempt to create an “IMS 

User’s Network” to workshop ideas and 

theories; 

CONCLUSION: 

Until quite recently the very latest technology to 

assist a customs officer with the search of 

vessels, cargo and aircraft was a torch and a 

mirror. Today officers have a vast array of drug 

detection tools at their disposal – the Drug Dog, 

the X-ray, access to a national CCTV network 

and of course, the Ionscan. 

In just three years Australian Customs has 

introduced a total of 51 Ionscan units to all 

Australian air and seaports. The acceptance 

and successful use of the technology by such a 

large number of officers is proof that this drug 

detection initiative – IMS - is the outstanding 

success. The enthusiasm for the technology 

has spread to other Australian law enforcement 

agencies that regularly ask us to provide 

assistance, either in the form of training or 

operational support. 

One of the exciting things about this technology 

is that we really have only just begun to tap its 

potential. I believe that the use of the Ionscan 

in Australia will continue to develop and 

“evolve”. 


VIP sources for ion mobility spectrometry 

H.-R. Döring 1) , G. Arnold 1) , V. L. Budovich 2) 

1 

Bruker Saxonia Analytik GmbH, Permoserstr. 15, 04318 Leipzig, Germany 

2 

Fa. CHROMDET Ekologija, Pleteshkovskij per. 22, Moscow 107005, Russia 

Ion Mobility Spectrometry (IMS) for the 

detection of chemical warfare agents (CWAs), 

toxic industrial chemicals (TICs), drugs and 

explosives is mainly based on the ionization by 

radioactive sources such as 63 Ni, 241 Am and 3 H, 

since these sources optimally meet the 

requirements made on a portable device for 

field use: They are small-sized and very 

lightweight, they have an extremely good 

mechanical stability and do not require any 

additional power. They are very reliable while 

displaying an excellent sensitivity with regard to 

the detection of quite a large number of 

compounds of interest. 

However, for well-known reasons (radiation 

safety, disposal problems) there is a growing 

interest in replacing radioactive sources by 

alternative ionization techniques. In the past the 

most promising candidates for replacing 

radioactive ionization sources were 

photoionization (PI) and coronadischarge 

ionization (CD). 

Developments of Ultra-Violet (UV) IMS have 

shown that the analytical performance of an 

UV-IMS is only insignificantly higher than that 

of a simple photoionization detector (PID). An 

improved performance could be reached by 

adding special dopants [1,2]. Nevertheless, the 

conclusions were that there was no full 

replacement found for radioactive sources. 

CD sources provide an excellent response for 

TICs and CWAs whenever they are detected in 

the positive mode (i. e. where positive ions are 

involved). Detection limits are not sufficient for 

TICs and CWAs which need to be monitored in 

the negative IMS operation mode. Even after 

several years of reasearch and development 

the insufficient overall detection performance 

as well as the stability and reliability of CD 

sources are, according to today’s knowledge, 

an unsolvable outstanding problem [3]. 

The new non-radioactive variable ionization 

potential (VIP) source [4, 5], also called 

„electron lamp“, consists of a small evacuated 

glass tube in which electrons emitted by a 

heated cathode are accelerated in the direction 

of a special window (Photograph 1). By 

variation of the accelerating voltage and 

selection of the window (material, thickness), 

the electrons can pass through the window into 

the reaction chamber of the IMS and ionize 

there as in the case of a 63 Ni source or a CD 

source, or they get absorbed in the window 

causing in addition to secondary electrons, an 

X-ray radiation typical for the window material. 

In the first case the lamp is usually fitted with a 

mica window having a thickness of 5...7 µm. If 

the primary electrons are accelerated with such 

a high voltage that they are able to penetrate 

through the window e. g. higher than 15 kV, a 

large number of substances including 

hydrocarbons can be ionized. Thus, they may 

be detected by the IMS with a relatively good 

sensitivity. However, if this broad-band 

detectibility is not intended the accelerating 

voltage must be reduced resulting in a poorer 

sensitivity as may be seen from Fig. 2. In this 

case ionization of substances is only effected 

by the characteristic X-ray radiation of the 

elements in the mica window (Si, K, Al,...). 

But if it is necessary to detect such substances 

like organic phosphorous compounds or 

halogenated thioethers selectively and with a 

high sensitivity, a beryllium foil is used as 


H.-R. Döring et al.: „VIP sources for ion mobility spectrometry...”, IJIMS 4(2001)1, 67-70, p. 68 

glass tube 

heated cathode 

Wehnelt 

cylinder 

Figure 1: Electron lamp 

window, e. g. a foil having a thickness of 25 µm 

that is coated with an aluminium anode at the 

vacuum side. The typical radiation emitted by 

aluminium is utilized then for the ionization of 

the compounds. In addition, this window has 

the following advantages: i) the heat that is 

generated during 

slowing down of 

electrons is conducted 

well, ii) the larger 

thickness of the 

window ensures a 

higher lifetime of the 

lamp, iii) it is possible 

to increase the energy 

transfer through the 

window that results in 

stronger IMS signals. 

An experimental setup 

was constructed 

based on the 

window 

well-known RAID 

technology. An 

electron lamp was 

fixed axially on the 

IMS tube. This 

experimental setup 

was used for the 

detection of acetates, 

organic phosphorous 

compounds, 

halogenated and non-halogenated 

hydrocarbons like hexane (see Fig. 3), of toxic 

industrial chemicals like HCN, COCl2, Cl2, SO2, 

NO2, NH3 and all known nerve and blister 

agents (see Fig. 4 and 5). The normalized ion 

mobilities (so-called K0-values) determined 

peak area [pA•ms] 

100 

80 

60 

positive 

reactant 

ion peak 

acetone 

(10 ppm) 

40 

ethyl acetate 

(150 ppb) 

20 

0 

n-hexane 

(50 ppm) 

0 5 10 15 20 25 

anode voltage [kV] 

Figure 2: 

Ionization of selected compounds in dependence on the anode voltage 



AIR 

2.09 

acceleration 

voltage 24 kV 

HEXANE 

1.90 

2.39 

Fig. 3: Detection of 50 ppm hexane 

NH 3 

2.23 RIP 

2.08 

acceleration 

voltage 2,5 kV 

5s 

1.67 

1.54 

1.28 

1.16 

acceleration 

Fig. 4: Detection of 130 µg/m³Sarin 

Cl - 

2.80 

(HD)O 2 

- 

2.80 

voltage 2,5 kV 

5s 

Fig. 5: Detection of 150 µg/m³ HD 



during these measurements are absolutely 

identical with those gained with radioactive ion 

sources. 

With an accelerating voltage of 2 kV and an 

anode current of 5 µA sensitivities were 

reached which were equal to those gained 

using a 63 Ni source with an activity of 555 Mbq. 

Moreover, the investigations have shown that 

the detection limit of the IMS can be 

considerably reduced by increasing the anode 

current of the lamp (100 µA are possible). The 

lifetime of the electron lamp is equal to that of 

PID lamps. 

Summary: 

A new nonradioactive ion source, the so-called 

„electron lamp“, has been developed. It 

consists of a small evacuated glass tube in 

which electrons are generated and accelerated 

in direction towards a window. Depending on 

the design of the window and the value of the 

accelerating voltage, this lamp either emits a 

very soft X-ray radiation or electrons by help of 

which the substances to be detected are 

ionized in the IMS. It was proved that an IMS 

equipped with such a lamp has shown 

absolutely the same analytical performance as 

a 63 Ni IMS with an activity of the source of 555 

MBq or even a better one. 

References: 

[1] Photoionization Ion Mobility Spectrometer, Glenn E. 

Spangler et al., US Patent 5, 338,931, Int. Cl. 301 D 

59/44 H01 J 49/00, August 16, 1994 

[2] Photoionization Ion Mobility Spectrometry, H.-R. 

Döring, G. Arnold, US Patent 5,968,837, Int.Cl. g01 

N33; H0149/00; October 19, 1999 

[3] J. Adler, H.-R. Döring, G. Arnold, V. Starrock, E. 

Wülfing, First Results with the Bruker Corona 

Discharge IMS, Paper held on the 6. International 

IMS Workshop in Dresden, August 1997 

[4] Ion Mobility Spectrometer, V. Budovich, A. Mikhailov, 

G. Arnold, US Patent No. 5,969,349, Int. Cl. 

H01J49/00 

[5] Ionization Chamber with Electron Source, H.-R. 

Döring, Patent Application in July, 1999 in Europe, 

USA, Canada, Japan, Korea 


RELATIONSHIPS FOR ION DISPERSION IN 

ION MOBILITY SPECTROMETRY 

Glenn E. Spangler 

Technispan LLC, 1133C Greenwood Road, Pikesville, MD 21208, USA 

INTRODUCTION 

In recent years, the field of ion mobility 

spectrometry (IMS) has grown to include not 

only linear DC-field IMS, but also asymmetric 

RF-field IMS. 1-2 These two approaches to ion 

mobility spectrometry produce two types of 

data that have not yet been correlated with 

each other. Approaches to achieving such a 

correlation is described here. 

DISCUSSION 

The equation describing the motion of ions in 

”local equilibrium” in the gas phase is the 

continuity equation 3-5 

∂n 

v t v 

T 

v v 

−∑∇• xD 

i i•∇ nT + ∑xv i di•∇ nT + ∑xiα 

inT= 

0 

∂t 

i i i 

(1.1) 

where n T is the total ion concentration, x i is the 

mole fractional contribution to the i-th ion 

species, D i is the diffusion tensor, v di is the drift 

velocity and α i is the rate of ion loss during 

migration; and the momentum transfer 

equation 5 2v 

v 

⎛dr 

⎞ ⎛dr 

⎞ v v 

mi⎜ 

+ µν 

2 i i( εi) ⎜ ⎟ − qE 

i ( rt , ) = 0 

dt 

⎟ 

⎝ ⎠ 

dt 

i 

⎝ ⎠i 

(1.2) 

where m i is the mass, µ i is the reduced mass, ν i 

is the collision frequency, ε i is the energy, q i is 

the charge on the i-th ion; r v is the most 

probable location of the ion cloud in electric 

field E v 

; and the first derivative of equation 1.2 

corresponds to the drift velocity in equation 1.1. 

The momentum transfer equation describes the 

ion trajectory, and the continuity equation 

describes the dispersion of the ion cloud as it 

drifts through the drift region. 

Since in a linear DC-field IMS the ions travel 

with a constant velocity, the acceleration or 

second-derivative term of equation 1.2 is zero, 

and the drift velocity is 

v 

v ⎛ dr ⎞ q v 

i v v v 

di 

= ⎜ ⎟ = E( r, t) ≡ KE 

i ( r, 

t) 

⎝ dt ⎠i µν 

i i( εi) 

(1.3) 

where K is the mobility of the ion. Furthermore 

since the ions in an asymmetric RF-field IMS do 

not travel with a constant velocity, the 

acceleration term of equation 1.2 is not zero 

and the momentum transfer equation remains a 

second order non-linear equation. For an 

oscillatory electric field, the momentum transfer 

equation corresponds to the Mathieu equation 

of quadrupole mass spectrometry. 6 The 

first-order term can be handled by adding a 

second order harmonic to the oscillatory field. 

Perturbation theory allows the drift velocity for 

the ions to be written as 

v = xv v +δ 

v 

where 

∑ xv 

v 

i di 

i 

(1.4) 

is the low-field drift velocity for 

the ions under the thermal conditions of the 

IMS, and δ v is the change in that drift velocity 

v d 

∑ 

d i di d 

i 

due to the action of the electric field. The 

electric field increases the drift velocity by 

declustering the ions, and decreases the drift 


G.E. Spangler: „Relationships of ion dispersion...”, IJIMS 4(2001)1, 71-74, p. 72 

velocity by increasing the collision cross 

section. These ideas are contained in 

v 

v v δ T 

δ d 

= f ( xK 

i i) E( rt , ) 

T 

(1.5) 

where f(x iK i) is defined by Spangler, 5 and δT is 

the increase in the effective temperature of the 

ion over the drift temperature T. f(x iK i) may be 

positive or negative depending upon whether 

declustering or the change in the collision cross 

section dominates. 

While the relationship that describes the 

change in effective temperature is not firmly 

established, Wannier’s work with high electric 

fields suggests that 7-9 

(1.6) 

where M is the total mass of the colliding 

particles and k is Boltzmann’s gas constant. 

When equation 1.6 is substituted in equation 

1.7, δ v becomes proportional to the cube of 

v d 

the electric field. For the asymmetric RF-field 

IMS, this electric field is the dispersion field. 

The complementary field E C is related to δ v 

through 5 τ 

v v 

− KECτ 

=∫δvdt 

d 

0 

(1.7) 

where K is the mobility of the ion cloud and τ is 

one complete cycle of the exciting RF 

waveform. Differentiation of equation 1.7 yields 

τ 

dEC 

E ⎡ 

3 

⎤ 

C 

= dK + Ed 

C 

ln ⎢ f ( xK 

i i) Edt 

dK K 

∫ 

⎥ 

⎣0 

⎦ 

(1.8) 

that is similar to 

dt 

M 

δ T = v 

3k 

d 

2 

d 

td 

=− dK 

K 

d 

(1.9) 

for the linear DC-field IMS. The 

complementary field E C of equation 1.8 is 

functionally similar to the drift time t d of 

equation 1.9 provided the differential 

logarithmic term is small. Equation 1.8 

contains a positive sign, and equation 1.9 a 

negative sign. 

Equation 1.8 seems to describe the 

compensation versus dispersion voltage data of 

Riegner, et al. with a small differential 

logarithmic term. 10 An exception is the DMMP 

data that apparently deviates from the trend 

due to declustering activity. The compensation 

versus dispersion voltage data of Buryakov, et 

al., however, does not satisfy equation 1.8 

without the differential logarithmic term. 11 Since 

the RF potentials used to excite the two 

systems were different, the method used to 

exite ion motion may directly impacts the ability 

of an asymmetric RF-field IMS to produce data 

that can be compared to a linear DC-field IMS. 

Work is continuing on this issue. 

It has been noted that the declustering of ion 

clusters in a high electric field can lead to 

changes in drift velocity. A question related to 

this issue is whether the asymmetric RF-field 

IMS is a useful device, or only an academic 

curiosity. For example, do the ions form 

adducts with the drift gas to frustrate the 

measurement of ion mobilities? Very little data 

exists to answer this question, and the facility of 

a mass spectrometer is needed to address the 

topic. 

Over the years, a variety of data has been 

collected on the nature of the reactant ions in 

IMS using IMS / MS. 12-14 More recently, a more 

comprehensive data set has been collected 

under controlled conditions of water, and 

compiled and compared to other previously 

published data. 15 All the data suggest that the 

ions arriving at the detector of the mass 

spectrometer are ions clustered with water and 

drift gas molecules (e.g., N 2 for air, and Ar for 

argon) 

(1.10) 

where n, m and p are integers. However when 

± ± 

M ( HO) ( N) ( Ar) HO M ( HO)( N) ( Ar) 

l 

2 n− 1 2 m p+ 

2 2 n 2 m p 

amplitude ratios are computed and compared 

to published thermodynamic data, the 

calculated water concentration does not agree 

with the experimentally measured concentration 

(1 to 3 ppm) in the IMS, and the free energies 



associated with the drift gas adducts are 

endothermic. These results suggest that the 

expansion region immediately behind the ion 

sampling pinhole significantly alters the 

distribution of the ions as they are sampled. 

The gas dynamics for the expanding gas is free 

jet expansion. 16 The density of the gas and the 

gas kinetic temperature decreases until the 

speed of the flow eventually exceeds that of the 

local speed of sound. The expanding zone 

(sometimes referred to as the zone of silence) 

is surrounded by a concentric barrel shock 

terminated by a perpendicular Mach disc. As 

the gas adiabatically expands within this barrel 

shock, the equilibria that previously existed in 

the IMS tend to shift to lower temperature. This 

continues until there are an inadequate number 

of collisions to support the equilibrium 

dynamics, at which point the equilibrium 

freezes. 17 Since the relative concentrations 

(assumed to be in excess) in the expanding 

gas remain constant, it is possible to deduce 

the temperature where freezing occurs by 

tallying amplitude ratios in the mass 

spectrometer data. For the water adducts, the 

freezing temperature is found to decrease 

linearly with the number of adducts. 

When the above arguments are applied to 

adducts of drift gas, a different result is 

obtained. The equilibria associated with these 

adducts appear to shift continuously through 

the zone of silence and reach a new equilibrium 

with the background gas beyond the Mach disk. 

The free energy of reaction ∆G n-1,n /RT is on the 

order of –12.9 ± 0.6. 15 Hundreds of amplitude 

ratios were considered before arriving at this 

value that is both independent of the nature of 

the adduct (N 2 or Ar), number of adducts, etc. 

Since a free energy with this order of 

magnitude arises from a very weak interaction 

(such as an ion-induced dipole interaction 

possibly involving a third body for complex 

stabilization), the enthalpy of reaction can be 

estimated from the polarizability of the adducts 

and the temperature of the gas within the 

vacuum system. When the thermodynamics 

0.693 ⎛ ∆G 

1, 

2 exp 

n− 

n ⎞ 

τ 

r 

= 

kfP 

⎜ − 

RT ⎟ 

STP ⎝ ⎠ 

are extrapolated to the atmospheric pressure 

conditions of the IMS, the lifetime of the 

adducts can be found from 

(1.11) 

that is on the order of 0.013 nanoseconds, 

assuming a forward rate constant of 

2.6x10 -30 cc 2 /molecule 2 sec. Since this 

lifetime is less than the collision time 

between molecules under atmospheric 

pressure conditions (0.6 nanoseconds), 

evidence for adduct formation within an 

IMS is lacking. One might continue to 

argue for momentary adduct formation, but 

these clusters impact minimally the 

measurement of mobility values due to their 

instability. 

CONCLUSIONS 

The ability to correlate IMS data generated 

by an asymmetric RF-field IMS with data 

generated by a linear DC-field IMS 

depends upon the waveform used to excite 

the RF-field IMS. Other methods of 

excitation can provide variable resolution, 

but such methods destroy the mobility 

information contained in the ionogram. 

Momentary adduct formation between ion 

and drift gas molecules may momentarily 

occur under the atmospheric pressure 

conditions of an IMS. However, it is 

doubtful that these adducts significantly 

change drift velocities and associated 

mobility values. This is particularly true for 

the asymmetric RF-field IMS where the 

effective temperature of the ion is 

increased by the high electric field. This 

comment may not apply to water adducts 

that have higher associative bond energies. 

REFERENCES 

[1] Cohen, M.J.; Karasek, F.W. J. Chrom. Sci. 

1970, 8(6), 330-337. 

[2] Buryakov, I.A.; Krylov, E.V.; Nazarov, E.G.; 

Rasulev, U. Kh. Int. J. Mass Spectrom. Ion 

Proc. 1993, 128, 143-148. 

[3] Moseley, J.T.; Gatland, I.R.; Martin, D.W.; 

McDaniel, E.W. Phys. Rev. 1969, 178, 234-239. 

[4] Spangler, G.E.; Collins, C.I. Anal. Chem. 1975, 

47, 403-407. 

[5] Spangler, G.E., Field Analytical Chemistry and 

Technology, 4/5(2000)255-267. 

[6] Quadrupole Mass Spectrometry and Its 

Applications, P.H. Dawson, Ed.; Elsevier, 

Amsterdam, 1976. 



[7] Wannier, G.H. Phys. Rev. 1951, 83, 281-289. 

[8] Wannier, G.H. Bell Syst. Tech. J. 1953, 32, 

170-254. 

[9] Mason, E.A. McDaniel, E.W. Transport 

Properties of Ions in Gases, Wiley: NY, 1988, 

pp. 233, 280, 298, 358. 

[10] Riegner, D.E.; Harden, C.S.; Shoff, D.B.; Ewing, 

R.G. Proc. 1997 ERDEC Scientific Conf. 

Chemical & Biological Defense Research, July 

1998. 

[11] Buryakov, I.A.; Krylov, E.V.; Nazarov, E.G.; 

Rasulev, U.Kh. Int. J. Mass Spectrom. Ion Proc. 

1993, 128, 143-148. 

[12] Kim, S.H.; Betty, K.R.; Karasek, F.W. Anal. 

Chem. 1978, 50, 2006-2012. 

[13] Kim, S.H.; Karasek, F.W.; Rokushika, S. Anal. 

Chem. 1978, 50, 152-155. 

[14] Kim, S.H. Ph.D. Dissertation, University of 

Waterloo, Ontario, Canada, 1977. 

[15] Spangler, G.E., accepted for publication, Int. J. 

Mass Spectrom. (2001) 

[16] Douglas, D.J.; French, J.B. J. Anal. Atom. 

Spectrom. 1988, 3, 743-747. 

[17] Bray, K.N.C. In GasDynamics, Volume I: 

Nonequilibrium Flows, Part II, P.P. Wegener, 

Ed.; Marcel Dekker: New York, 1970, chapter 3. 


P. de B. Harrington: „Strategies for smarter...”, IJIMS 3(2000)1, -, p. 75

P. de B. Harrington: „Strategies for smarter...”, IJIMS 3(2000)1, -, p. 76

A novel Method for the Detection of MTBE: 

Ion Mobility Spectrometry coupled to Multi Capillary Column 

Z. Xie 1 , St. Sielemann 2 , H. Schmidt 1 , J.I. Baumbach 1 

1 

Institut für Spektrochemie und Angewandte Spektroskopie (ISAS), 

Bunsen-Kirchhoff-Str. 11, D-44139 Dortmund, Germany 

2 

G.A.S. Gesellschaft für analytische Sensorsysteme mbH, 

TechnologieZentrumDortmund, Emil-Figge-Str. 76-80, D-44227 Dortmund,Germany 

Abstract 

A combination of an ion mobility spectrometer 

with radioactive ionization source and equipped 

with a multi capillary column was used as a 

new analytical method for the detection of 

MTBE, a gasoline additive, which has become 

a potential water pollution problem. To extract 

MTBE out of the water a membrane extraction 

unit was set up, which is simple, effective and 

easy to automate with respect to further 

applications. The analyte was extracted on the 

one hand directly out of the water, thus the 

membrane was completely steeped in the 

water. On the other hand, the membrane was 

held in the gas phase over the surface of the 

water (head space). The minimum detectable 

limit for both methods was about 50 ppb vl of 

MTBE in water and the reproducibility with a 

standard deviation of 8.9 % (head space) and 

11.5 % (aqueous phase) rather high. Finally the 

utilizability of the system for on-site and on-line 

measurements is briefly discussed. 

Introduction 

Methyl tert-butyl ether (MTBE, CAS 1634-04-4) 

has been used as an octane-enhanced 

replacement for lead, primarily in mid- and 

high-grade gasoline at concentrations as high 

as 8 % (by volume) since 1979. As part of two 

US EPA programs it is also used as 

oxygenates at higher concentrations (11 to 15 

% per volume) to reduce ozone and carbon 

monoxide levels in polluted areas of the country 

[1-3]. 

MTBE is highly soluble in water (4.8 weight %), 

does not really degrade in the environment, 

and most public water systems are not 

equipped to completely remove it from drinking 

water. MTBE is considered a potential human 

carcinogen and the Comprehensive 

Environmental Response, Compensation, and 

Liability Act (CERCLA), commonly known as 

Superfund, listed MTBE as a hazardous 

substance. Due to the use of MTBE the number 

of reports concerning its detection in 

groundwater and surface water increase. 

Current methods to detect MTBE are based on 

capillary gas chromatography with flame 

ionization detection (FID) or mass spectrometry 

(MS) [4-7]. In other studies FT near-IR and FT 

Raman spectroscopy are used as 

non-destructive methods to quantify the amount 

of MTBE in gasoline [8, 9]. 

As a new analytical method for the detection of 

MTBE in air and water a method based on ion 

mobility spectrometry is introduced in this 

paper. Ion mobility spectrometers (IMS) are 

very fast, sensitive and inexpensive devices 

which permits the efficient analysis and 

characterization of gaseous analytes [10-20]. 

IMS can operate at ambient temperature and 

pressure without requiring a vacuum. During 

the last 10 years dramatic changes have 

occurred in the practical aspects of ion mobility 

spectrometry and the understanding of 

underlying principles of response. IMS, in the 

early 1970 mostly used for the detection of 

chemical warfare agents [10, 11, 21, 22], are 

more and more applied for the use in industrial 

and environmental venues, not only because of 

the simplicity of the instrumentation but also 

due to the astonishing detection limits in the 

range of some µg/L, sometimes even down to 

ng/L [23-27]. 

For the detection of MTBE in mixtures the IMS 

was coupled to a multi capillary column (MCC, 

28-32] to realize an effective pre-separation. 

Such columns consists of about 1000 parallel 

capillaries with an inner diameter of 43 µm. 

Because of a total diameter of 3 mm of the 


Z. Xie et al.: „A novel method for the detection of MTBE...”, IJIMS 4(2000)1, 77-83, p. 78 

Table 1: 

Main parameters of the ISAS custom designed 

63 

Ni-IMS coupled to a multi capillary column 

(MCC) 

Parameter 

Ionization source 

Length of the drift tube 

Electrical field strength 

Shutter opening time 

Drift gas 

Drift gas flow 

Sample gas 

Sample gas flow 

Temperature (IMS) 

Stationary phase of the MCC 

Length of the column 

Carrier gas of the MCC 

Carrier gas flow 

Temperature (MCC) 

MCC- 63 Ni-IMS 

63 

Ni (510 MBq) 

120 mm 

326 V/cm 

100 µs 

N 2 (99.999%) 

100 mL/min 

N 2 (99.999%) 

100-1500 mL/min 

23 °C 

nonpolar 

25 cm 

N 2 (99.999%) 

400 mL/min 

23 °C 

whole column it is possible to operate the 

column with carrier gas flows between 100 and 

800 mL/min, which are the optimum flow rates 

to introduce gaseous samples into IMS. In the 

past the connection of MCC with IMS with a 

UV-ionization source was successfully used for 

the fast separation and sensitive detection of 

mixtures of different volatile organic substances 

(VOC) [33, 34]. 

In this present short communication the 

arrangement of an IMS with MCC was used for 

the detection of MTBE in gaseous and aqueous 

matrix and the detection limits and calibration 

curves will be presented. The results of the 

experiments on separations of different 

mixtures will be finished soon and published 

in a following paper. 

The extraction of MTBE from the water was 

realized in a single step by using a 

membrane sampling device, which is a 

frequently used technique as separator in a 

wide variety of analytical determinations 

[35-47]. Blanchard and Hardy [48] 

introduced this separation method based on 

the permeation of volatile organic 

compounds through a silicone poly 

carbonate membrane from an aqueous 

sample matrix into an inert gas stream. 

Extraction using this kind of membranes 

achieve high selectivity, because the 

solubility of the organic molecules in silicone 

is higher than that in water. In such non 

porous structures, the molecules first 

dissolve in the membrane and then diffuse 

due to a concentration gradient. 

The purpose of this paper is to illustrate that ion 

mobility spectrometry in combination with a 

membrane extraction unit is a fast and sensitive 

method for the on-line and near real-time 

detection of MTBE in the gas phase and in 

aqueous sample matrix. 

Experimental Section 

Instrumentation 

For the measurements a system consisting of 

an ISAS custom designed ion mobility 

spectrometer with a radioactive ionization 

source ( 63 Ni, 555 MBq) coupled to a rod-sharp 

multi capillary column (Institute of Applied 

carrier gas 

inlet 

stainless steel 

connection 

carrier gas 

outlet 

stainless 

steel needle 

teflon seal 

stainless steel ring 

stainless steel seal 

stainless steel wire 

membrane 

Figure 1: 

Schematic diagram of the membrane extraction unit 



Table 2: 

Drift times t d, peak half time width w and resolving power 

R for monomer and dimer ions of MTBE using the 

63 

Ni-IMS (see figure 2) 

Peak 

t d / ms w / ms R 

Physics, Novosibirsk, Russia) was used. The 

relevant parameters of the system are 

summarized in table 1. The sample is 

introduced on the MCC using an electric 

six-port-valve (Valco, Supelco, Deisenhofen, 

Germany) and a stainless steel loop with a 

volume of 1 mL. The IMS was driven only in the 

positive mode. 


Monomer ion peak 

Dimer ion peak 

Signal / V 

-0.06 

-0.04 

-0.02 

0.00 

MTBE 

RIP 

monomer ion peak 

MIP 

1.72 cm 2 /Vs 

dimer ion peak 

DIP 

1.45 cm 2 /Vs 

15 20 25 30 


Figure 2: 

Ion mobility spectrum of the positive ions of MTBE 

using the MCC- 63 Ni-IMS 

25 

20 

25 

20 

25 

20 

RIP 

RIP 

RIP 

DIP 

DIP 

MIP 

MIP 

DIP 

1.981 

2.355 


positive ions 

300 ppb vg 

1510 ppb vg 

0 2 4 6 8 10 

Retention Time / s 

MIP 

0.4 

0.6 

160 ppb vg 

50 

39 

Figure 3: 

2D-IMS-Chromatogramm of MTBE test gas of 

concentrations of 160, 300 and 1510 ppb vg 

(MCC- 63 Ni-IMS, RIP: reactant ion peak, MIP: 

monomer ion peak, DIP: dimer ion peak) 

Sample preparation and data processing 

To generate the gaseous samples of MTBE 

(99,5%, Fischer Scientific, Schwerte, 

Germany) with different concentrations a 

permeation system was used. The pure and 

fluid substance was filled into a 2 mL glass 

flask (permeation tubes) which was covered 

with a PTFE coated caoutchouc membrane 

(1 mm). The tube was kept at a constant 

temperature of 23 °C with the help of a 

thermostat to achieve reproducible and stable 

permeation rates through the membrane. The 

sample gas stream to carry the gaseous 

substance was kept constant by the use of flow 

controllers. The permeation rate of the 

substance was determined by the 

weight loss in a defined time period. By 

changing the sample carrier gas stream 

different concentrations were achieved. 

The test gas was flowing continuously to 

the 1 mL loop of the six-port-valve. 

For the detection of MTBE in water an 

extraction unit was set up which is 

shown in figure 1. The extraction unit 

consists of a stainless steel vessel 

covered gas tightly with a stainless steel 

cap with a seal. In the cap there are two 

stainless steel screws pierced by a 

needle (outer diameter: 700 µm; 

Hamilton Deutschland GmbH, 

Darmstadt, Germany). A silicon high 

temperature tube (length: 18 cm; inner 

diameter: 300 µm; outer diameter: 700 

µm; Reichelt Chemietechnik, 

Heidelberg, Germany) is used as 

membrane. The gas tight connection of 

the membrane with the needle was 

realized by submerging the end of the 

membrane in toluene and after it had 

swollen to insert the needle into the 

hollow fibre. When the toluene 

evaporated, the membrane shrunk and 

a gas tight connection between needle 

and membrane is formed. Additional 

there is a wire to hold the membrane. 

MTBE was extracted from the water by 

two different ways: On the one hand 50 

mL of the solution with different 

concentrations of MTBE were filled into 

the vessel, so that the membrane was 

only in contact with the gas phase 

('head space') above the aqueous 

phase, on the other hand 200 mL were 

filled into the vessel and the whole 

membrane was covered with water. The


Area / a.u. 

0,1 

C 

D 

E 

The solutions with different 

concentrations were prepared 

by adding pure MTBE to 

distilled water and to dilute it. 

The areas of the peaks were 

calculated using Origin 6.0 

(Microcal, Northampton, MA, 

USA). The standard deviations 

were determined by analyzing 

the same solution with a 

concentration of 200 ppb vl nine 

times [vl: concentratation of 

MTBE in the liquid phase]. 

100 1000 

Concentration / ppb vg 

Figure 4: 

Peak areas of the monomer, dimer and the sum of the monomer and 

dimer ions vs the concentrations of the MTBE test gas samples in 

ppb vg (MCC- 63 Ni-IMS) 

Table 3: 

Polynomial functions and correlation coefficients of the calibration 

curves (see figure 4) for the peak areas of monomer, dimer and the 

sum of the monomer and dimer ions of MTBE test gas 

Function: y=A+B1*x+B2*x 2 

Monomer peak area 

Dimer peak area 

Sum of both 

Table 4: 

Peak areas and relative standard deviations (RSD) for a concentration 

of 200 ppb vl 

MTBE in water (the extraction of the substances was 

performed in the gas phase over the water (head space) and directly in 

the aqueous phase) 

Measurement Gas phase Aqueous phase 

1 

0.03 

0.07 

2 

0.03 

0.07 

3 

0.03 

0.07 

4 

0.04 

0.05 

5 

6 

7 

8 

9 

Mean value 

RSD (%) 

-4.38 

-6.46 

-4.09 

0.04 

0.04 

0.03 

0.03 

0.04 

0.03 

8.9 

B1 

2.26 

3.37 

2.07 

continuous and constant gas flow of 50 mL of 

nitrogen (5.0, 99,999%, Messer Griesheim, 

Dortmund, Germany) through the membrane 

transported the extracted analytes to the valve. 


A 

B2 

-0.34 

-0.50 

-0.36 

0.07 

0.05 

0.07 

0.05 

0.05 

0.06 

11.5 

Correlation 

coefficient 

0.986 

0.991 

0.847 

Results and Discussion 

Calibration of the 63 Ni-IMS using 

the permeation system: 

A single spectrum of the 

positive ions of MTBE displayed 

in figure 2 is showing two sharp 

peaks with reduced mobilities of 

1.72 and 1.45 cm 2 /Vs. 

Analogous to ethers, esters, 

ketones and alcohols to form 

monomer and dimer ions, the 

peaks can be ascribed to be the 

peaks is equivalent to the sum of 

the functions for the monomer 

and dimer ions. Corresponding to 

what is shown in the 2D-graphs, 

for low concentrations the area of 

the monomer peak is larger than 

for the dimer one. This changes 

when the concentration of the 

substance is higher than 350 ppb vg, 

and the area of the dimer ion peak 

increased while the amount of 

monomer ions seems to be nearly 

constant. The minimum detectable 

limit (MDL) is lower than 160 ppb vg 

(figure 4) and the saturation level is 

reached at a 1.5 ppm vg. 

Calibration of the 63 Ni-IMS for the 

detection of MTBE in water: 

For the detection of MTBE directly in 

water and in the head space over the 

contaminated water, the 

reproducibility was determined 

analyzing the same solution with a 

concentration of 200 ppb vl nine 

times. The results for the sum of the monomer 

and dimer peak areas as well as the mean 

values and the relative standard deviations 

(RSD) are summarized in table 4. For the gas


Signal Area / a.u. 


0.1 

0.01 

Signal / a.u. 

RIP 

Head space 

50 ppb vl -0.1 

750 ppb vl 

MIP 

0.0 

-0.06 

DIP 

15 20 25 

MIP 

RIP 


-0.03 

0.00 

15 20 25 

Figure 5: 

Peak areas as the sum of the monomer and dimer ions vs the 

concentrations of MTBE in water extracted directly from the gas phase 

over the water (head space) 

RIP: reactant ion peak, MIP: monomer ion peak, DIP: dimer ion peak 

0.1 

0.01 

Signal / a.u. 

-0.10 

-0.05 

0.00 

RIP 

phase (head space) the RSD was about 9 %, 

for the aqueous phase about 12 %, which is 

rather low. Thus, the extraction process can be 

regarded as a reproducible step for both 

phases. 

The calibration curve in figure 5 shows the 

result of the extraction of MTBE out of the gas 


100 1000 

Concentration / ppbvl 

Aqueous Matrix 

Signal / a.u. 

50 ppb vl 

MIP 

750 ppb vl 

20 25 

DIP 

-0.050 RIP 


MIP 

-0.025 

0.000 

20 25 



100 1000 

Concentration / ppbvl 

Figure 6: 

Peak areas as the sum of the monomer and dimer ions vs the 

concentrations of MTBE in water extracted directly from the aqueous 

phase - RIP: reactant ion peak, MIP: monomer ion peak, 

DIP: dimer ion peak 

Signal / a.u. 

phase (head space) over 

the surface of water 

containing different 

concentrations (50 up to 

750 ppb vl) of MTBE. Single 

Spectra are displayed for 

the lowest and highest 

measured concentration. 

The illustrated curve 

shows the sum of the 

monomer and dimer peak 

areas, and just as for the 

calibration curves for the 

test gas measurements a 

polynomial function was 

used (table 5). As 

demonstrated in the 

graph, 50 ppb vl was the 

lowest detectable 

concentration of MTBE in 

water. 

Larger 

concentrations than 750 

ppb vl have not been 

investigated. 

For the extraction of 

MTBE directly out of water 

the calibration curve for the 

sum of the monomer and 

dimer peak area and the 

graphs for concentrations of 

50 and 750 ppb vl are shown if 

figure 6. The minimum 

detectable limit is as for the 

extraction out of the head 

space about 50 ppb vl and also 

the correlation coefficient for 

the polynomial function is with 

0.996 similar to the one 

achieved for the head space 

measurements. 

A comparison (figure 7) of the 

calibration curves for the 

aqueous matrix and the head 

space exhibits larger peak 

areas for the direct extraction 

of MTBE out of water, 

whereby the calibration curve 

for the extraction from the 

aqueous matrix reached a saturation level. For 

further measurements, the membrane will be 

introduced completely into the aqueous matrix. 

With the set up system it is possible to detect 

MTBE in aqueous as well as in gaseous 

environmental matrices. With the possibility to


use the extraction unit also as a 

flow through unit, the system can 

be applied for a nearly continuous 

monitoring e.g. of groundwater, 

so that sudden appearance of 

contaminations or changes in the 

concentration can be detected 

immediately. Also, the connection 

with the MCC should enable a 

separation of MTBE and different 

by-products, e.g. Benzene or 

other gasoline components, which 

will be investigated in the future. 


0,1 

0,01 

MTBE 

Aqueous Matrix 

Head space 

Conclusion 

It has been demonstrated that the 

63 

Ni-IMS coupled to a MCC is a 

fast (time of analysis of a few 

seconds) and a sensitive 

(detection limit about 50 ppb vl of 

MTBE in water) method for the 

detection of the gasoline additive 

MTBE, which provides in addition 

with the membrane extraction unit 

the advantage of on-line and on-site 

measurements. 


The financial support of the 

Bundesministerium für Bildung und 

Forschung and the Ministerium für 

Schule, Wissenschaft und Forschung 

des Landes Nordrhein-Westfalen are 

gratefully acknowledged. The first impulse to 

detect MTBE using IMS by Peter Popp, 

Department of Analytical Chemistry of the UFZ 

Centre for Environmental Research 

Leipzig-Halle should be mentioned thankfully. 

References 

[1] Agency, U. S. E. P. MTBE Fact Sheet 2 January 

1998, EPA 510-F-97-015, 1-5. 


1998, EPA 510-F-97-014, 1-3. 


1998, EPA 510-F-97-014, 1-5. 

[4] Prazen, B. J.; Bruckner, C. A.; Synovec, R. E.; 

Kowalski, B. R. Anal. Chem. 1999, 71, 1093-1099. 

[5] Quach, D. T.; Ciszkowski, N. A.; Finlayson, B. J. J. 

Chem. Educ. 1998, 75, 1595-1598. 

[6] Cassada, D. A.; Zhang, Y.; Snow, D. D.; Spalding, R. 

F. Anal. Chem. 2000, 72, 4654-4658. 

[7] Miller, M. E.; Stuart, J. D. Anal. Chem. 2000, 72, 

622-625. 

Table 5: 

Polynomial functions and correlation coefficients of the 

calibration curves (see figures 5 and 6) for the peaks areas 

(as a sum of the monomer and dimer ions) for MTBE 

extracted from the head space and the aqueous phase 

Function: y=A+B1*x+B2*x 2 

Head space 

Aqueous phase 

100 1000 

Concentration / ppb 

Figure 7: 

Comparison of the peak areas (sum of the monomer and 

dimer ions) vs the concentrations of MTBE in water extracted 

from the gas phase over the water (head space) and the 

aqueous phase 

A 

-3.44 

-4.65 

B1 

1.11 

2.20 

B2 

-0.10 

-0.32 

[8] Choquette, S. J.; Chesler, S. N.; Duewer, D. L.; 

Wang, S.; O'Haver, T. C. Anal. Chem. 1996, 68, 

3525-3533. 

[9] Hansen, S.; Berg, R. W.; Stenby, E. H. Asian 

Chemical Letter 2000, 4, 65-74. 

Correlation 

coefficient 

0.999 

0.996 

[10] Cohen, M. J.; Karasek, F. W. J. Chromatogr. Sci. 

1970, 8, 330-337. 

[11] Karasek, F. W. Anal. Chem. 1971, 43, 1982-1986. 

[12] Baumbach, J. I.; Eiceman, G. A. Appl. Spectrosc. 

1999, 53, 338A-355A. 

[13] Sielemann, S.; Soppart, O.; Baumbach, J. I.; v.Irmer, 

A.; Walendzik, G.; Klockow, D. Recycling Waste 

Management Remediation of Contaminated Sites 

1995, 8. 

[14] Sielemann, S.; Baumbach, J. I.; Eiceman, G. A.; 

Jauzein, M.; Walendzik, G.; Klockow, D. Field 

Screening Europe 1997. 

[15] Sielemann, S.; Baumbach, J. I.; Pilzecker, P.; 

Walendzik, G. Int. J. Ion Mobility Spectrom. 1999, 2, 

15-21. 

[16] Leonhardt, J. W.; Bensch, H.; Berger, D.; Nolting, M.; 

Baumbach, J. I. 3rd International Workshop on IMS, 

Galveston, TX, Oct. 16-19, 1994 1994, 49-56. 



[17] Eiceman, G. A.; Snyder, A. P.; Blyth, D. A. Int. J. 

Environ. Anal. Chem. 1989, 38, 415-425. 

[18] Eiceman, G. A.; Karpas, Z. CRC Press, Boca Raton, 

Ann Arbor, London, Tokyo 1994, 1-228. 

[19] Eiceman, G. A.; Bell, S. E.; Wang, Y. F. 3rd 

International Workshop on IMS, Galveston, TX, Oct. 

16-19, 1994 1994. 

[20] Eiceman, G. A.; Garcia-Gonzalez, L.; Wang, Y. F.; 

Pittman, B.; Burroughs, G. E. Talanta 1992, 39, 

459-467. 

[21] Karasek, F. W.; Denney, D. W., J. Chromatogr . 

1974, 93, 141-146. 

[22] Preston, J. M.; Karasek, F. W.; Kim, S. H. Anal. 

Chem. 1977, 49, 1746-1750. 

[23] Pilzecker, P.; Baumbach, J. I.; Trindade, E. Proc. of 

the IEEE International Symposium on Electrical 

Insulation, Anaheim, April 2-5 2000. 

[24] Kotiaho, T.; Lauritzen, F. R.; Degn, H. Anal. Chim. 

Acta 1995, 309, 317-325. 

[25] Hatano,H.; Rokushika, S.; Ohkawa,T. Instrum. Trace 

Org. Monit. 1992, 27-48. 

[26] Bacon, A. T. Pittsburgh Conference 1992, Paper 258 

1992. 

[27] Lawrence, A. H.; Barbour, R. J.; Sutcliffe, R. Anal. 

Chem. 1991, 63, 1217-1221. 

[28] Aleksander, J. J. Chromatogr. Sci. 1976, 14, 589. 

[29] Golay, M. J. E. Chromatographia 1975, 8, 421. 

[30] Hawkes, S. J. J. Chromatogr. Sci. 1977, 15, 89. 

[31] Pierce, H. D.; Unrau, J. A. M.; Oelschlager, A. C.; 

Cutteridge, A. M. J. Chromatogr. Sci. 1979, 17, 297. 

[32] Schmitt, V. O.; Pereiro, I. R.; Lobinski, R. Anal. 

Commun. 1997, 34, 141-143. 

[33] Baumbach, J. I.; Eiceman, G. A.; Klockow, D.; 

Sielemann, S.; Irmer, A. v. Int. J. Environ. Anal. 

Chem. 1997, 66, 225-239. 

[34] Baumbach, J. I.; Sielemann, S.; Soppart, O. Fifth Int. 

Workshop on IMS, Jackson, Wyoming 1996, 431-436. 

[35] Wan, C.; Harrington, P.; Davis, D. M. Talanta 1998, 

46, 1169-1179. 

[36] Yang, M. J.; Pawliszyn, J. Anal. Chem. 1993, 65, 

1758-63 CODEN ANCHAM; ISSN 0003-2700. 

[37] Wong, P. S. H.; Cooks, R. G.; Cisper, M. E. Environ. 

Sci. Technol. 1995, 29, 215-218. 

[38] Spangler, G. E.; Carrico, J. P. Int. J. Mass Spectrom. 

Ion Phys. 1983, 52, 267-287. 

[39] Silvon, L. E.; Bauer, M. R.; Ho, J. S.; Budde, W. L. 

Anal. Chem. 1991, 63, 1335-1340. 

[40] Slivon, L. E.; Ho, J. S.; Budde, W. L. ACS Symp. Ser. 

1992, 508, 169-77 CODEN ACSMC8; ISSN 

0097-6156. 

[41] Rautenbach, R.; Welsch, K. J. Membrane Sci. 1994, 

87, 107-118. 

[42] Pratt, K. F.; Pawliszin, J. Anal. Chem. 1992, 64, 

2101-2106. 

[43] Motomizu, S.; Yoden, T. Anal. Chim. Acta 1992, 261, 

461-469. 

[44] Marquardt, K. Stuttg. Ber. Abfallwirtsch. 1987, 24, 

187-234. 

[45] Kotiaho, T.; Ketola, R. A.; Ojala, M. Am. Environ. Lab. 

1997, 9, 19--21. 

[46] Hauser, B.; Popp, P.; Paschke, A. Int. J. Environm. 

Anal. Chem. 1999, 74, 107-121. 

[47] Harland, B. J.; Nicholson, P. J. D.; Gillings, E. Water 

Res. 1987, 21, 107-13 CODEN WATRAG; ISSN 

0043-1354. 

[48] Blanchard, J. B.; Hardy, J. K. Anal. Chem. 1986, 58, 

1529-1532. 

[49] Siems, W. F.; Wu, C.; 

Tarver, E. E.; Hill, H. H. Anal. Chem. 1994, 66, 

4195-4201. 

[50] Rokushika, S.; Hatano, H.; Baim, M. A.; Hill, H. H. J. 

Anal. Chem. 1985, 57, 1902-1907. 

[51] Wu, C.; Siems, W. F.; Asbury, G. R.; Hill, H. H. Anal. 

Chem. 1998, 70, 4929-4938. 


Finalisation of a IUPAC/JCAMP-DX data transfer 

standard for ion mobility spectrometry data 

A.N. Davies*, J.I.Baumbach, P. Lampen, H. Schmidt 

ISAS, Institute of Spectrochemistry and Applied Spectroscopy, Bunsen-Kirchhoff-Str.11, Postfach 10 13 52, 44013 

Dortmund, Germany 

ABSTRACT 

In the last few years the rapid developments in 

the field of ion-mobility spectrometry many 

several different sites around the world has 

made it imperative to establish an agreed 

protocol for the storage and exchange of 

experimental data. Under the auspices and 

guidance of IUPAC through the Working Party 

on Spectroscopic Data Standards (JCAMP-DX) 

and the development work of the members of 

the International Society for Ion Mobility 

Spectrometry such a standard has been 

developed which is now in it’s final stages. It 

will be published as the Appendix to this paper 

and placed on the IUPAC web sites for final 

comment. An example of the usefulness of 

these standards as chemical MIME types for 

the display of spectroscopic data on the web 

will be described and an example is available 

for viewing on the JCAMP-DX web site as well 

as the web site of the ISIMS. 

INTRODUCTION 

The ability to compare and contrast 

experimental data is a pre-requisite in a 

modern rapidly developing field like ion-mobility 

spectrometry. Although the exchange of the 

data points via spreadsheet programs or simple 

x-y ASCII tables is possible it is often only the 

inclusion of subsidiary supporting information 

which makes a data file valuable. Work on this 

critical data dictionary has been ongoing for 

several years including presentations at 

international conferences in order to gain the 

widest possible input and has been reported 

previously in this journal . The drafts have also 

been available for comment on the Web site of 

the co-ordinating IUPAC working party. It was 

decided that the form which the standard 

should take was to follow the successful ASCII 

data exchange protocols of the Joint 

Committee on Atomic and Molecular Physical 

Data commonly known as the JCAMP-DX 

standards. These standards have also been 

adopted for use for the transport and display of 

spectroscopic data on the Internet as the 

Chemical Multipurpose Internet Mail Extension 

or Chemical MIME type and examples of their 

implementation are described below. 

The development of a JCAMP-DX standard file 

format for Ion Mobility Spectrometry (IMS) 

follows the development and implementation of 

such standards for Infrared Spectroscopy in 

1988 , Chemical Structures in 1991 , Nuclear 

Magnetic Resonance Spectrometry (NMR) in 

1993 and Mass Spectrometry in 1994 . 

Following the transfer of responsibility of the 

development and maintenance of these 

standards to the International Union of Pure 

and Applied Chemistry (IUPAC) a new 

extension to the protocols which includes 

additional standardisation of audit trail 

information and a long date format as well as 

new NMR labels was published in 1999 . 

RESULTS 

The results of the work to date can be seen in 

the appendix which forms the final draft of the 

standard protocol and is reproduced here to 

encourage last-minute comments or corrections 

before being adopted as a IUPAC 

Recommendation in 2001. 

The appendix comes with the usual warnings in 

that this has currently the status of a final draft 

for comment only and should under no account 

be implemented by manufacturers or end users 

until the final document is adopted by IUPAC in 

the Spring of 2001. 


A.N. Davies et al.: „Finalisation of a IUPAC/JCAMP-DX...”, IJIMS 4(2001)1,84-108, p. 85 

Figure 11: 

An IMS JCAMP-DX spectroscopic data file displayed directly 

Use in building Web Pages 

The use of JCAMP-DX spectroscopic data files 

within the environment of a web page has been 

described elsewhere and there are some 

excellent examples authored by the pioneer of 

this method of publishing spectroscopic data, 

Robert Lancashire, on the web site of the 

University of West Indies, Jamaica and can be 

viewed under the URL: 

http://wwwchem.uwimona.edu.jm:1104/spectra/ 

chime/chime.html. Much 

information is available on the 

pages on this site. In order to 

use the Chemical MIME types it 

is necessary to have your web 

browser enabled for these 

chemical file formats and this 

can be easily achieved by 

installing the CHIME plug-in 

freely available from the web 

site of MDL Informations 

systems Inc. at 

http://www.mdli.com/download/ 

chime/index.html. Further links 

for test sites for other chemical 

data MIME types are also to be 

found at the home page of the 

IUPAC Working Party at 

http://jcamp.isas-dortmund.de. 

Figure 1 shows an example of 

an ion mobility spectrometry 

data set in the draft file format displayed in a 

web browser with the CHIME plug-in installed. 

The data is not displayed a fixed graphic file but 

as real data and as such may be analysed 

more closely by zooming or rescaling for 

example. The spectrum is loaded directly into 

the browser from the JCAMP-DX file. 

In a more creative use of the options available 

for implementing JCAMP-DX files into HTML 

pages for display on the internet one of the 

early examples created by 

Robert Lancashire has been 

adapted to display the same 

file as shown in figure 1 but 

this time the chemical 

structure is also displayed as 

well as some explanatory text. 

The graphic shown in Figure 2 

is also available ‘live’ following 

the IMS links on the IUPAC 

working party home page. 

Figure 12: 

An IMS file this time displayed as part of an HTML web page 

including additional explanatory text and the associated chemical 

structure. 

CONCLUSIONS 

As has been described above 

the new file format for ion 

mobility spectrometry has now 

been all but completed. This 

short article is intended 

principally to encourage any 

last-minute requests for 

changes and to show some of 

the possible uses to which the 

new format can be put once it 


A.N. Davies et al.: „Finalisation of a IUPAC/JCAMP-DX...”, IJIMS 4(2001)1,84-108, p. 86 

has been officially adopted by IUPAC in the 

Spring of 2001. 


The financial support of the Bundesministerium 

für Bildung, Wissenschaft, Forschung und 

Technologie and the Ministerium für 

Wissenschaft und Forschung des Landes 

Nordrhein-Westfalen is gratefully acknowledged. 

REFERENCES 

[1] J.I. Baumbach, P. Lampen and A.N. Davies, IUPAC / 

JCAMP-DX: An International Standard for the 

Exchange of Ion Mobility Spectrometry Data. 

International Journal for Ion Mobility Spectrometry, 

1998, 1(1) p.64-67. 

[2] R.S. McDonald and P.A. Wilks Jr., JCAMP-DX: A 

Standard Form for Exchange of Infrared Spectra in 

Computer Readable Form. Applied Spectroscopy, 

1988, 42(1) p.151-162. 

[3] J. Gasteiger, et al., JCAMP-CS: A Standard 

Exchange Format for Chemical Structure Information 

in Computer-Readable Form. Applied Spectroscopy, 

1991 45(1) p.4-11. 

[4] A.N. Davies and P. Lampen, JCAMP-DX for NMR. 

Applied Spectroscopy, 1993 47(8) p.1093-1099. 

[5] P. Lampen, H. Hillig, A.N. Davies and M. Linscheid, 

JCAMP-DX for Mass Spectrometry. Applied 

Spectroscopy, 1994 48(12) p.1545-1552. 

[6] P. Lampen, et al., An Extension to the JCAMP-DX 

Standard File Format, JCAMP-DX V.5.01. Pure and 

Applied Chemistry, 1999 71(8) p.1549-1556. 


J.I. Baumbach, A.N. Davies, P. Lampen, H. Schmidt - JCAMP-DX for Ion Mobility Spectrometry 

Draft manuscript for Pure and Applied Chemistry 

INTERNATIONAL UNION OF PURE 

AND APPLIED CHEMISTRY 

COMMITTEE ON PRINTED AND 

ELECTRONIC PUBLICATIONS 

Working Party on Spectroscopic Data Standards (JCAMP-DX) 

JCAMP-DX – A Standard Format for 

the Exchange of Ion Mobility 

Spectrometry Data 

(IUPAC Recommendations 2000) 

Prepared for publication by 

Jörg Ingo Baumbach * , Antony N. Davies, Peter Lampen, 

Hartwig Schmidt 

ISAS, Inst. für Spektrochemie und Angewandte Spektroskopie, Bunsen-Kirchhoff-Str.11, 44139 Dortmund, 

Germany 

* Author to whom comments should be sent. 

Members of the International Society for Ion Mobility Spectrometry Working Party, IUPAC Working Party on 

Spectroscopic Data Standards (JCAMP-DX) and others during the development of this standard: 

J.I. Baumbach, O. Soppart, A. Von Irmer, A.N. Davies, T. Fröhlich, P. Lampen, R. Lancashire, 

R.S. McDonald, P.S. McIntyre, D.N. Rutledge, G.A. Eiceman, H.H. Hill 

Thursday, December 5, 2000 ims dx 5g.doc Page 1 of 22



JCAMP-DX – A Standard Format for 

the Exchange of Ion Mobility 

Spectrometry Data 

J.I. Baumbach, A.N. Davies, P. Lampen, H. Schmidt 

Institut für Spektrochemie und Angewandte Spektroskopie, 

Bunsen-Kirchhoff-Str. 11, 44139 Dortmund, Germany 

ABSTRACT 

The relatively young field of ion mobility spectrometry has now advanced to the stage where 

the need to reliably exchange the spectroscopic data obtained world-wide by this technique 

has become extremely urgent. To assist in the validation of the various new spectrometer 

designs and to assist in inter-comparisons between different laboratories reference data 

collections are being established for which an internationally recognized electronic data 

exchange format is essential. 

To make the data exchange between users and system administration possible, it is important 

to define a file format specially made for the requirements of ion mobility spectrometry. The 

format should be computer readable and flexible enough for extensive comments to be 

included. In this document we define a data exchange format, agreed on a working group of 

the International Society for Ion Mobility Spectrometry at Hilton Head Island (1998) and 

Buxton U.K. (1999). 

This definition of this format is based on the IUPAC JCAMP-DX protocols, which were 

developed for the exchange of infrared spectra [1] and extended to chemical structures [2], 

nuclear magnetic resonance data [3] and mass spectra [4]. This standard of the Joint 

Committee on Atomic and Molecular Physical Data is of a flexible design. The International 

Union of Pure and Applied Chemistry have taken over the support and development of these 

standards and recently brought out an extension to cover year 2000 compatible date strings 

and good laboratory practice [5]. The aim of this paper is to adopt JCAMP-DX to the special 

requirements of ion mobility spectra [6]. 

KEY WORDS 

Data exchange, ion mobility spectrometry, JCAMP-DX, data standards 




1 INTRODUCTION 

JCAMP-DX is an electronic file based format using ASCII characters (American Standard 

Code of Information Interchanging) reduced to the printable character. This guarantees the 

acceptance on all computer systems. The main components to describe the ion mobility 

spectrometer (IMS) and the spectra are shown in Figure 1. 

##.carrier gas= 

##sample description= 

##cas name= 

##cas registry no= 

##concentrations= 

##.reduced mobility= 

##.electric field= 

##.ionization chamber= 

##.ionization mode= 

##.ionization energy= 

##.ims temperature= 

##.ims pressure= 

##.drift gas= 

##.drift chamber= 

##.repetition rate= 

Figure 1. The main ion mobility spectrometry specific and generic terms to be stored in a 

JCAMP-DX data file are shown here. 

2 DEFINITIONS 

The following definitions are important for the understanding of the JCAMP-DX protocol: 

2.1 Labeled-Data-Records 

Labeled-Data-Records (LDR) consist of a flagged data-label and an associated data-set. An 

LDR begins with a data-label-flag (##) and ends with the next data-label-flag. LDRs are 

divided into lines of 80 or fewer characters, terminated by . End-of-line is equivalent 

to a blank, except for certain special cases. Each LDR occupies as many lines as necessary. A 

data-label is the name of an LDR. It is delimited by a data-label-flag (##) and a data-labelterminator 

(=), for example ##TITLE= is a data-label for the definition of the working title of 

the following spectrum. A line contains no more than one data-label. When labels are parsed, 

alphabetic characters are converted to upper case, and all spaces, dashes, slashes, and 

underlines are discarded. Thus, XUNITS and xunits are equivalent. 

There are two kinds of LDR: Core and Note. Core LDRs are required. Notes are optional. 

Every definition of an LDR should include if it is a core or a note. 

In part 2.2 - 2.5 we discuss the forms used for defined data-sets. 




2.2 TEXT 

The LDR contain descriptive information for humans, not normally intended to be parsed by 

computers, i.e., title, comments, origin, etc. 

2.3 STRING 

Some LDR contain predefined text fields intended to be parsed by computers and read by 

humans. The format of each string field is specified under the LDR in which it is used. 

2.4 AFFN 

The easiest way to write the data is to use the ASCII free format numeric. This format is 

important to simplify direct user input. It is a format similar to freeform input in BASIC. A 

field starts either with +, -, decimal point or digit. E is the only allowed character to give the 

power of 10 by which the field must be multiplied. It is followed by + or - and two or three 

digits. The numeric field is terminated either by E, comma or blank. 

2.5 ASDF 

This is the ASCII squeezed difference form. Tabular data using JCAMP-DX data 

compression scheme (see section 3.4.1). 

2.6 Comments 

##=. Comments may be specified by a data-label-flag plus a data-label-terminator, with a null 

data-label. Such comments may continue for more than one line, terminating at the next datalabel-flag. 

$$. Comments may be entered at any point in a line by prefixing the first word of the 

comment by $$. Such comments continue only to the end of the current line, and they do not 

terminate an LDR. 

3 THE CORE 

This section describes the data labels for the JCAMP-DX core data. 

The Core consists of four parts. The first part, called the 'Fixed Header Information', contains 

generic LDRs which are required for all JCAMP-DX files and which appear at the beginning 

of each file in a given order. The 'Variable Header Information' contains records which are 

data type specific (in this case IMS specific) or which are only used in special types of 

JCAMP-DX files (e.g. compound files). Whether a particular is required or not depends on 

the application. The third section 'Core Data' contains the relevant parameters for the fourth 

section the 'Data Table'. The type of data in the data table determines the parameters which 

must appear in the core data. Only one data table may appear per JCAMP-DX block (a block 

being a part of the JCAMP-DX file starting with ##TITLE= and ending with ##END=). 




3.1 Core Fixed Header Information 

3.1.1 ##TITLE= (TEXT) 

Title and/or reason of the measurement. 

(Required) 

3.1.2 ##JCAMP-DX= (STRING) 

The version number of this protocol is 5.01. A description of the software used to generate 

the file follows the version number as a comment. 

For example: 

##JCAMP-DX= 5.01 $$ JCAMP-DX for IMS (NT4 version 1.0 ISAS Germany) 

(Required) 

3.1.3 ##DATA TYPE= (STRING) 

Keywords: 

ION MOBILITY SPECTRUM or 

IMS PEAK TABLE or 

IMS PEAK ASSIGNMENTS or 

LINK 

Distinguishes between different kinds of spectra such as IMS-, NMR- or IR-data. The string 

ION MOBILITY SPECTRUM reports a continuous spectrum located at the ##XYDATA= 

LDR and IMS PEAK ASSIGNMENTS reports distinct and analyzed peaks located in the 

##PEAK ASSIGNMENTS= LDR. 

(Required) 

3.1.4 ##DATA CLASS= (STRING) 

Keywords: 

XYDATA or 

XYPOINTS or 

PEAK TABLE or 

ASSIGNMENTS 

This label defines the type of tabular data within the data block and is not to be used for link 

blocks. 

(Required) 

3.1.5 ##ORIGIN= (TEXT) 

Here the name of organization, address, telephone number, name of individual contributor, 

email, etc., as appropriate must be added. This information is not optional. 

(Required) 




3.1.6 ##OWNER= (TEXT) 

It is possible to set here a copyright linked to the spectrum that has the form: "COPYRIGHT 

(C) by ". If ##OWNER= contains "PUBLIC DOMAIN", the implication is 

that the data may be copied without permission on the authority of whoever is named under 

##ORIGIN= 

(Required) 

3.1.7 ##END= 

It is important to have a mark at the end of file in the data format to know that transfer has 

been complete and to distinguish between the blocks of a multi spectrum file. 

(Required) 

3.2 CORE VARIABLE HEADER INFORMATION 

The JCAMP-DX standard is easy to understand and expand. Many LDRs are already defined 

in previous JCAMP-DX protocols and they should be used for Ion Mobility Spectrometry. 

However, in the case of the equipment parameters the particular requirements of this 

technique call for some special LDRs. Data-type specific LDRs start with "##." (see 4.2 

below). In the following part definitions are given for LDRs which will allow a precise 

description of the equipment parameters. 

3.2.1 ##BLOCKS= (AFFN) and ##BLOCK_ID= (AFFN) 

Blocks are used as suggested in the JCAMP-DX 4.24 protocol (3.2) [1] with the extensions 

defined in the JCAMP-CS protocol (5.13) [2] to provide inter-block referencing. 

For ease of use it is recommended to use separate files and to avoid the use of blocks. 

However, when, for example, XYDATA and a PEAKTABLE must be stored in the same file, 

these must be written as a compound file [1]. The use of compound data files is, of course, 

optional but when more than one block per file is stored the ##BLOCKS= LDR has the 

priority REQUIRED. 

A compound JCAMP-DX file consists of a LINK block (##DATA TYPE=LINK) 

surrounding the DATA blocks and the total number of DATA blocks (N) must appear in the 

LINK block header (##BLOCKS=N). A unique positive integer (n) should be assigned to 

each DATA block inside the compound file (##BLOCK_ID=n) . Linking of the blocks can 

then be achieved using ##CROSS REFERENCE= (see example 1). These records are to be 

used in compound JCAMP-DX files to provide inter-block referencing. 

(Required only for compound files) 

EXAMPLE 1. 

##TITLE= example compound data file $$ title of the whole compound file 

##JCAMP-DX= 5.01 

$$ Name & Version No. of JCAMP-DX software 

##DATA TYPE= LINK 

##BLOCKS= 2 

$$ number of data blocks 

##ORIGIN= 

$$ name of contributor, organization, address, telephone etc. 

##OWNER= 

$$ COPYRIGHT (C)’year’ by ‘name’ or PUBLIC DOMAIN 

##TITLE= 

$$ title of the first data block 


##DATA TYPE= ION MOBILITY SPECTRUM 




##DATA CLASS= XYDATA 

##BLOCK_ID= 1 

##ORIGIN= 


##OWNER= 


##CROSS REFERENCE= IMS PEAK TABLE:BLOCK_ID=2 

. . . 

##END= $$ end of first data block 

##TITLE= 

$$ title of the second data block 


##DATA TYPE= IMS PEAK TABLE 

##DATA CLASS= PEAK TABLE 

##BLOCK_ID= 2 

##ORIGIN= 


##OWNER= 


##CROSS REFERENCE= ION MOBILITY SPECTRUM: BLOCK_ID=1 

. . . 

##END= $$ end of second data block 

##END= $$ end of link block and file 

3.2.2 ##.IMS PRESSURE= (AFFN) 

Pressure inside the IMS-System. (in SI units – Kilo Pascal). 

(Required) 

3.2.3 ##.CARRIER GAS= (TEXT) 

A description of the carrier gas is required here. 

(Required) 

3.2.4 ##.DRIFT GAS= (TEXT) 

A description of the drift gas is required when used in addition to the carrier gas. 

(Required) 

3.2.5 ##.ELECTRIC FIELD= (AFFN, AFFN) 

The electric field description is divided in two parts: The field in the ionization chamber and 

the one in the drift chamber (in Volt / cm) separated by commas, including the polarity (+ for 

positive ions detected and – for negative ions respectively). 

(Required) 

3.2.6 ##.ION POLARITY= (STRING) 

The polarity of the measured ions. Allowed values are POSITIVE or NEGATIVE. 

(Required) 

3.2.7 ##.IONIZATION MODE= (STRING) 

The following keywords define how the system ionizes the gas: 

UV ultraviolet source 

BR beta radiation source 

AL alpha radiation source 

PD partial discharge 




CD corona discharge 

ESI electrospray ionization 

LI laser ionization 

LD laser desorption 

SY synchrotron radiation 

For more detailed information it is possible to give a comment like $$Ni63-source. 

(Required) 

3.2.8 ##.IMS TEMPERATURE= (AFFN[, AFFN]) 

Here the first value is characterizing the temperature in the drift chamber, the second the 

temperature in the ionization chamber in °C. The second value is optional and separated by a 

comma. 

(Required) 

3.2.9 ##.SHUTTER OPENING TIME= (AFFN) 

Shutter opening time in microseconds. 

Without knowing the time the shutter was open it is not possible to calculate the individual 

mobilities. 

(Required) 

3.3 CORE DATA 

3.3.1 ##XUNITS= (STRING) and ##YUNITS= (STRING) 

Here the units of the axis can be given. The following keywords are defined: 

For ##XUNITS=: SECONDS, MILLISECONDS, MICROSECONDS and 

NANOSECONDS, 

For ##YUNITS=: MICROAMPERES, NANOAMPERES, and PICOAMPERES. 

(Required) 

3.3.2 ##FIRSTX= (AFFN) and ##LASTX= (AFFN) 

First and last actual abscissa values of ##XYDATA=. First tabulated abscissa times 

##XFACTOR= should equal ##FIRSTX=. 

(Required for ##DATA CLASS=XYDATA) 

3.3.3 ##FIRSTY= (AFFN) 

Here the actual Y-value corresponding to ##FIRSTX= is meant. ##FIRSTY= should be equal 

##YFACTOR= times the first Y-value in ##XYDATA=. 





3.3.4 ##XFACTOR= (AFFN) and ##YFACTOR= (AFFN) 

The values of a spectrum may be converted to integer to save space and allow the DIFDUP 

format (see 3.4.1). It is important to select a convenient scaling to keep the file within 

reasonable limits but to store all significant digits. In such a case the ##XFACTOR= and 

##YFACTOR= LDR contain a floating point number to be multiplied by the values in 

##XYDATA= to arrive at the original data point value. 

In most cases ±32767 is sufficient, therefore this is the recommended ordinate scaling. If a 

larger scaling is necessary it is required to give the actual unscaled maximum and minimum 

of the ordinates in the records ##MAXY= and ##MINY=. This avoids a 2-byte integer 

overflow in the program reading the data table. 

For example if a Y-value with 9 significant figures (e.g. 0.002457194) needs to be converted 

to an integer value for ASDF coding then: 

a) divide by the maximum of the absolute Y-value (say 0.346299765) 

b) and then multiply by the largest integer value (MAXINT) necessary to place all significant 

figures left of the decimal point 

c) convert to integers 

for this example - 

Y integer =INTEGER((Y real / MAX(ABS(Y real )))*MAXINT) 

=INTEGER(0.002457194/0.346299765)*1.0E+11 

=INTEGER(709556935.4487) 

=709556935 

then ##YFACTOR= (MAX(ABS(Y real ))/MAXINT) 


3.3.5 ##NPOINTS= (AFFN) 

The numbers of points are required for all cases: IMS-peak tables, peak assignments and ion 

mobility spectra. 

(Required) 

3.4 CORE DATA TABLE 

Data must be stored in one of the following data formats (3.4.1 – 3.4.4). Only one of these 

formats is allowed per DATA block and the selected data table is given in the header e.g. 

##DATA CLASS=XYDATA. 

3.4.1 ##XYDATA= (AFFN or ASDF). 

This LDR contains a table of spectral data with abscissa values at equal intervals specified by 

parameters defined in section 3.3. The label is followed by a variable list, 

(X++(Y..Y)) where .. indicate indefinite repeat of Y-values until the end of line and ++ 

indicates that X is incremented by (LASTX-FIRSTX)/(NPOINTS-1) between two Y-values. 

For discrete point the AFFN form is allowed where each Y value is written out in full. This 

form creates large files, but is easily human readable. The following ASDF forms produce 

smaller files at the cost of reduced human readability. 




SQUEEZED FORM (SQZ) 

To have a better data compression it is possible to use the squeezed form (SQZ) in which the 

delimiter, the leading digit and sign are replaced by a pseudo-digit from Table 1. 

For example the Y-values 30, 32 would be represented as C0C2. 

DIFFERENCE FORM (DIF) 

For a better compression, it is possible to use the difference form (DIF) where the delimiter, 

leading digit and sign of the difference between adjacent values are transformed in a pseudodigit 

from Table 1. 

For example the Y-values 30, 32 would now be represented as C0K. 

To ensure data coding is correct a Y-value check is built-in. Each line starts with the absolute 

X- and Y-value, which is the same as the last calculated value of the previous line. The last 

line of a block of DIF data contains only the abscissa and ordinate for a Y-value check of the 

last ordinate. 

DUPLICATE SUPPRESION (DUP) 

Another possible variation is to include duplicate suppression (DUP) replacing two or more 

adjacent and identical numbers with pseudo-digits as given in Table 1. This can be used with 

all ASDF forms. 

For example 50 50 50 50 becomes E0V when combining DUP with SQZ. 

The best compression (but the least human-readable form) can be achieved when DIF and 

DUP are combined to provide the form called DIFDUP. In this format the duplicate count is 

obtained by counting identical differences. 

The example above becomes E0%%% in DIF form and E0%U in DIFDUP form. 

Table 1: 

Pseudo digits used to compress spectra data in SQZ-, DIF- and DUP-format 

ASCII digits 0 1 2 3 4 5 6 7 8 9 

Positive SQZ @ A B C D E F G H I 

Negative SQZ a b c d e f g h i 

Positive DIF % J K L M N O P Q R 

Negative DIF j k l m n o p q r 

DUP S T U V W X Y Z s 

Example for uncompressed data storage 

##TITLE= Incomplete example file for uncompressed data! 

.......... 

##XUNITS= MILLISECONDS 

##YUNITS= NANOAMPERES 

##XFACTOR= 1 

##YFACTOR= 0.1 

##FIRSTX= 4 

##LASTX= 56 

##NPOINTS= 53 




##FIRSTY= 0 

##XYDATA= (X++(Y..Y) 

4 0 0 0 0 2 4 4 4 7 

13 5 4 4 5 5 7 10 11 11 

22 6 5 7 6 9 9 7 10 10 

31 9 10 11 12 15 16 16 14 17 

40 38 38 35 38 42 47 54 59 66 

49 75 78 88 96 104 110 121 128 

##END= 

Example for DIFDUP-form 

##TITLE= Incomplete example file for DIFDUP data form! 

...... 

##XUNITS= MILLISECONDS 

##YUNITS= NANOAMPERES 

##XFACTOR= 1 

##YFACTOR= 0.1 

##FIRSTX= 4 

##LASTX= 56 

##NPOINTS= 53 

##FIRSTY= 0 

##XYDATA= (X++(Y..Y) 

4@VKT%TLkj%J%KLJ%njKjL%kL%jJULJ%kLK1%lLMNPNPRLJ0QTOJ1P 

56A28 

##END= 

3.4.2 ##XYPOINTS= (AFFN) 

This LDR contains a table of spectral data with unequal abscissa increments. The label is 

followed by a variable list, (XY..XY). X and Y are separated by commas, data pairs are 

separated by semi-colons or blanks. This LDR should not be used for peak tables. 

3.4.3 ##PEAK TABLE= (STRING) 

It is recommended to store peak information using ##PEAK ASSIGNMENTS=, however, for 

backward compatibility this definition is included. This data table contains a table of peaks 

where the peak data starts on the following line. The label is followed by the variable list 

(XY) or (XYW) for peak position, intensity and width, where known, on the same line. The 

function used to calculate the peak width should be defined by a $$ comment in the line 

below the label. The peak groups are separated by semi-colon or space, components of a 

group are separated by commas. 

3.4.4 ##PEAK ASSIGNMENTS= (STRING) 

Variable list: (XA), (XYA) or (XYWA) 




After this LDR a list of peaks and their assignments for each components are given in the 

following form: 

(X 1 [, Y 1 ][, W 1 ], ) 

. . . . 

(X i [, Y i ][, W i ], ) 

X and Y indicates the location and height of each peak in units given by ##XUNITS= and 

##YUNITS=. W stands for width in ##YUNITS= and A represents a string describing the 

assignment enclosed in angle brackets. 

The parentheses provide a start and end flag of each assignment. Square brackets indicate 

optional information. It is important for the technical readability to have the same format for 

the whole peak assignment table and describe it after the ##PEAK ASSIGNMENTS= LDR 

with (XA), (XYA) or (XYWA). This LDR should be followed by a comment, which gives 

the method of finding the peak. 

For example – 

##PEAK ASSIGNMENTS= (XYWA) 

(15, 20.0, 1.0 ,) 

(30, 40.0, 2.0,) 

(60, 45.0, 4.0,) 

...... 

4 THE NOTES 

The notes portion of a JCAMP-DX file or block complements the core. Notes describe an 

experiment in greater detail than does ##TITLE=, including descriptions of equipment, 

method of observation, and data processing, as appropriate. Notes may contain information 

which is not found in the native file in which data is originally collected by an instrument. 

Notes are placed before the core data section to permit them to be viewed without listing the 

whole file. The contents of the notes depend on the user as well as the technique or 

application. Notes will vary for different samples, sites, data systems, and applications. 

4.1 Global Notes 

These have been already defined in JCAMP-DX and are common to all spectroscopy types. 

The file headers, spectral and sample parameters are often the same for different analytical 

techniques. This allows us to implement many of the standard LDRs from the existing 

JCAMP-DX protocols. The list given below is only a selection from those allowed. A 

complete list can be found in the references [1-5]. 

At least one of the optional LDRs described in section 4.1.4 - 4.1.8 should be included in 

each JCAMP-DX file. This is important for later archiving as these fields will yield more 

detailed information on the content of the data stored than a simple ##TITLE= field. 




4.1.1 ##LONG DATE= (STRING) 

Date of measurement is required by many agencies and recommended in the year 2000 form: 

YYYY/MM/DD [HH:MM:SS[.SSS] [±XXXX]]. YYYY is the long format of the year, MM 

the number of the month, DD the number of the day, HH the hour, MM the minutes, SS.SSS 

the seconds and fractions of a second of the measurement, ±XXXX is the difference to the 

UTC (e.g. +0100 is one hour difference to UTC). 

(Optional) 

4.1.2 ##SOURCE REFERENCE= (TEXT) 

Here an identification of the original spectrum file in native format or library name and serial 

number is possible for example. 

(Optional) 

4.1.3 ##CROSS REFERENCE= (TEXT) 

Used to link additional data for the same sample such as other types of spectra or chemical 

structures for example: 

##CROSS REFERENCE= 

ION MOBILITY SPECTRUM: EXTERNAL_FILE= FILENAME.DX 

IMS PEAK TABLE: BLOCK_ID=16 

STRUCTURE: BLOCK_ID=4 

..... 

(Optional) 

4.1.4 ##SAMPLE DESCRIPTION= (TEXT) 

If the sample is not a pure compound, this field should contain its description, i.e., 

composition, origin, appearance, results of interpretation, etc. If the sample is a known 

compound, the following LDRs specify structure and properties, as appropriate. 

(Optional) 

4.1.5 ##CAS NAME= (STRING) 

Name according to Chemical Abstracts naming conventions as described in the CAS Index 

Guide is required here. Examples can be found in Chemical Abstracts indices or the Merck 

Index. Greek letters are spelled out, and standard ASCII capitals are used for small capitals, 

Sub-/Superscribts are indicated by prefixes / and /\. Example: alpha-D-glucopyranose, 1- 

(dihydrogen phosphate). 

(Optional) 

4.1.6 ##NAMES= (STRING) 

Here the common, trade or other names are allowed. Multiple names are placed on separate 

lines. 

(Optional) 




4.1.7 ##MOLFORM= (STRING) 

Another possibility of describing the sample is to write down the molecular formula. 

Elemental symbols are arranged with carbon first, followed by hydrogen, and then remaining 

element symbols in alphabetic order. 

The first letter of each elemental symbol is capitalized. The second letter, if required, is lower 

case. One-letter symbols must be separated from the next symbol by a blank or digit. Sub- 

/Superscripts are indicated by the prefixes: / and /\, respectively. Sub- and superscripts are 

terminated by the next non digit. Slash may be omitted for subscripts. 

For readability, each atomic symbol may be separated from its predecessor by a space. For 

substances that are represented by dot disconnected formulas (hydrates, etc.), each fragment 

is represented in the above order, and the dot is represented by *. Isotopic mass is specified 

by a leading superscript. D and T may be used for deuterium and tritium. 

(Optional) 

Examples: 

C2H4O2 or C2 H4 O2(acetic acid) 

H2 /\17O (water, mass 17 oxygen) 

4.1.8 ##CONCENTRATIONS= (STRING) 

The list of the known components and their concentrations has the following form, where N 

stands for the name and C for the concentration of each component in units given with U in 

the form: 

##CONCENTRATIONS= (NCU) 

(N 1 , C 1 , U 1 ) 

. . . 

(N i , C i , U i ) 

The group for each component is enclosed in parentheses. Each group starts a new line and 

may continue on following lines. 

(Optional in JCAMP-DX but in this case strongly recommended) 

4.1.9 ##SPECTROMETER/DATA SYSTEM= (TEXT) 

This LDR contains manufacturers’ name, model of spectrometer, software system, and 

release number, as appropriate in the form used by the manufacturer. 

(Optional) 

4.1.10 ##DATA PROCESSING= (TEXT) 

Here all mathematical procedures used before storing the data in the JCAMP-DX file are 

described. This LDR is also important in peak assignments. 

(Optional) 

4.1.11 ##XLABEL= (TEXT) and ##YLABEL= (TEXT) 

These LDRs give the possibility of labeling the axis. 

(Optional) 




4.2 Data-type specific notes 

DATATYPE-SPECIFIC-LABELS are RESERVED labels which are defined by qualified 

user groups for a particular DATATYPE. A DATATYPE-SPECIFIC-LABEL is 

distinguished by a DATATYPE-SPECIFIC-LABEL-NAME which starts with a period e.g. 

(##.REDUCED MOBILITY=). Choice of period as distinguishing character is by analogy 

with the convention for data-structure names in Pascal and C. Effectively, the full LABEL- 

NAME is the concatenation of the DATATYPE NAME and the LABEL NAME, with a 

period in between, i.e., ##ION MOBILITY SPECTRUM.REDUCED MOBILITY=. 

4.2.1 ##.REDUCED MOBILITY= (STRING) 

After this LDR a list of reduced mobilities and their assignments for each of the components 

where K is the reduced mobility and A a string describing the assignment in closed angle 

brackets in the form: 

(K 01 , ) 

(K 02 , ) 

.... 

(K 0i , ) 

where the K 0 values are the known reduced mobilities in cm 2 V -1 s -1 given according to the 

formula: 

P T 0 

K 0 = K. 

. 

P0 

T 

The variable list (KA) MUST be given in the labeled-data-record and followed on the next 

line by the list of reduced mobilities each on its own line as shown in the example below: 

##.REDUCED MOBILITY=(KA) 

(2.38, ) 

(2.27, ) 

(2.06, ) 

(1.83, ) 

....... 

(Optional) 

4.2.2 ##.IONIZATION ENERGY= (AFFN) 

The ionization energy in eV. 

(Optional) 

4.2.3 ##.IONIZATION CHAMBER= (STRING, AFFN, AFFN[, AFFN]) and 

4.2.4 ##.DRIFT CHAMBER= (STRING, AFFN, AFFN[, AFFN]) 

Different geometrical parameters of the ionization or drift chamber are distinguished by the 

following keywords: 

RECT 




This means a rectangular chamber. It is followed by the size of the three dimensions length, 

width and height in mm. 

CYL 

Stands for a cylindrical form of the chamber. It is followed by its length and radius in mm. 

(Optional) 

4.2.5 ##.CARRIER GAS FLOW= (AFFN, AFFN) 

This LDR is divided in two flow values. The first value is linked to the ionization chamber 

and the second to the drift chamber. All values have the unit liter/minute. 

(Optional) 

4.2.6 ##.DRIFT GAS FLOW= (AFFN) 

This LDR defines the drift gas flow into the drift chamber when a separate gas is used here to 

the carrier gas. All values have the unit liter/minute. 

(Optional) 

4.2.7 ##.CARRIER GAS MOISTURE= (AFFN) 

The water concentration in the carrier gas is relevant to the different ionization processes. It is 

strongly recommended that this value be reported. This label has the units parts per million 

volume (ppmv). 

(Optional) 

4.2.8 ##.IONIZATION SOURCE= (TEXT) 

A description of the relevant ionization source details. 

For partial discharges (##.ionization mode=PD) this would include for example the materials 

used in the construction of the source and the needle/plate gap in the form; 

##.IONIZATION SOURCE= Partial Discharge: Needle = Steel, Plate = Silver, Gap = 5 mm, 

Voltage = 5000 V 

or for an ultraviolet lamp as source the energy and the lamp mounting position (on or off 

axis); 

##.IONIZATION SOURCE= UV Lamp: 10.6 eV, off-axis 

or for a radioactive source the type nature and intensity such as 

##. IONIZATION SOURCE= 63 Ni, 555 MBq 

or 

##. IONIZATION SOURCE= 3 H, 150 kBq 

for laser ionization the type of laser used, the wavelength selected and any relevant pulse 

information in the form 

##. IONIZATION SOURCE= Nd:YAG laser, 266 nm, pulse energy 10 mJ, pulse width 30 ps 

(Optional) 




4.2.9 ##.SHUTTER GRID POTENTIAL= (AFFN) 

The potential difference between the wires of the Bradbury-Nielsen shutter in Volts. 

e.g. 

##.SHUTTER GRID POTENTIAL=100 

(Optional) 

4.2.10 ##.REPETITION RATE= (AFFN) 

The time between subsequent shutter openings in milliseconds. 

e.g. 

##.REPETITION RATE=100 

(Optional) 




5 SUMMARY 

The following tables list the LDRs discussed with their basic parameters shown. 

THE CORE 

Core Fixed Header Information 

LDR Data-form Keyword Status Location 

##TITLE= (TEXT) Required 3.1.1 

##JCAMP-DX= (STRING) 5.01 Required 3.1.2 

##DATA TYPE= (STRING) ION MOBILITY Required 3.1.3 

SPECTRUM, 

IMS PEAK TABLE, 

IMS PEAK 

ASSIGNMENTS or 

LINK 

##DATA CLASS= (STRING) XYDATA, 

Required 3.1.4 

XYPOINTS, PEAK 

TABLE, or 

ASSIGNMENTS 

##ORIGIN= (TEXT) Required 3.1.5 

##OWNER= (TEXT) Required 3.1.6 

##END= Required 3.1.7 

Core Variable Header Information 


##BLOCK_ID= (AFFN) Required for 3.2.1 

compound files 

##BLOCKS= (AFFN) Required for 3.2.1 

compound files 

##.IMS PRESSURE= (AFFN) Required 3.2.2 

##.CARRIER GAS= (TEXT) Required 3.2.3 

##.DRIFT GAS= (TEXT) Required 3.2.4 

##.ELECTRIC FIELD= (AFFN, AFFN) Required 3.2.5 

##.ION POLARITY (STRING) Required 3.2.6 

##.IONIZATION MODE= (STRING) UV, BR, AL, PD, CD, Required 3.2.7 

ESI, LI, LD, SY 

##.IMS TEMPERATURE= (AFFN[,AFFN]) Required 3.2.8 

##.SHUTTER OPENING 

TIME= 

(AFFN) Required 3.2.9 

Core Data 


##XUNITS= (STRING) SECONDS, 


MILLISECONDS, 

MICROSECONDS; 

NANOSECONDS 

##YUNITS= (STRING) MICROAMPERES, 

NANOAMPERES; 

PICOAMPERES 





##FIRSTX= (AFFN) Required for 3.3.2 

##DATACLASS= 

XYDATA 

##LASTX= (AFFN) Required for 3.3.2 

##DATACLASS= 

XYDATA 

##FIRSTY= (AFFN) Required for 3.3.3 

##DATACLASS= 

XYDATA 

##XFACTOR= (AFFN) Required for 3.3.4 

##DATACLASS= 

XYDATA 

##YFACTOR= (AFFN) Required for 3.3.4 

##DATACLASS= 

XYDATA 

##NPOINTS= (AFFN) Required 3.3.5 

Core Data Table 


##XYDATA= (AFFN or ASDF) (X++(Y..Y)) 3.4.1 

##XYPOINTS= (AFFN) (XY..XY) 3.4.2 

##PEAK TABLE= (STRING) (XY) or (XYW) 3.4.3 

##PEAK ASSIGNMENTS= (STRING) (XA), (XYA) or 

(XYWA) 

3.4.4 

THE NOTES 

Global Notes 


##LONG DATE= (STRING) Optional 4.1.1 

##SOURCE REFERENCE= (TEXT) Optional 4.1.2 

##CROSS REFERENCE= (TEXT) Optional 4.1.3 

##SAMPLE DESCRIPTION= (TEXT) Optional 4.1.4 

##CAS NAME= (STRING) Names defined by CAS Optional 4.1.5 

Index Guide 

##NAMES= (STRING) Optional 4.1.6 

##MOLFORM= (STRING) see Def. Optional 4.1.7 

##CONCENTRATIONS= (STRING) (NCU) Optional but strongly 4.1.8 

recommended 

##SPECTROMETER/ (TEXT) Optional 4.1.9 

DATA SYSTEM= 

##DATA PROCESSING= (TEXT) Optional 4.1.10 

##XLABEL= (TEXT) Optional 4.1.11 

##YLABEL= (TEXT) Optional 4.1.11 

Data type Specific Notes 


##.REDUCED MOBILITY= (STRING) (KA) Optional 4.2.1 

##.IONIZATION ENERGY= (AFFN) Optional 4.2.2 

##.IONIZATION 

(STRING, AFFN, RECT or CYL Optional 4.2.3 

CHAMBER= 

AFFN[, AFFN]) 

##.DRIFT CHAMBER= (STRING, AFFN, RECT or CYL Optional 4.2.4 

AFFN [, AFFN]) 

##.CARRIER GAS FLOW= (AFFN, AFFN) Optional 4.2.5 




##.DRIFT GAS FLOW= (AFFN) Optional 4.2.6 

##.CARRIER GAS (AFFN) Optional 4.2.7 

MOISTURE= 

##.IONIZATION SOURCE= (TEXT) Optional 4.2.8 

##.SHUTTER GRID (AFFN) Optional 4.2.9 

POTENTIAL= 

##.REPETITION RATE= (AFFN) Optional 4.2.10 




EXAMPLE JCAMP-DX-FILE FOR IMS DATA EXCHANGE 

##TITLE=EXAMPLE JCAMP-DX FILE FOR IMS 

##JCAMP-DX=5.01 

$$ ISAS JCAMP-DX program for IMS (V.1.0) 

##DATA TYPE=ION MOBILITY SPECTRUM 

##DATA CLASS=XYDATA 

##ORIGIN=J.I.Baumbach, H. Schmidt, ISAS Dortmund, Germany 

##OWNER=COPYRIGHT (C) 2000 by ISAS Dortmund, Germany 

##LONG DATE=2000/09/28 14:24:38 +0100 

##SOURCE REFERENCE=P:\Hartwig\Messung\000928\H1092801 

##SPECTROMETER/DATA SYSTEM=ISAS/TE/TE-lang 

##NAMES=Tetrachloroethene 

##CAS REGISTRY NO=127-18-4 

##CONCENTRATIONS=(NCU) 

(TETRACHLOROETHENE, 0.46, ppmv) 

##.IMS PRESSURE=101 

##.CARRIER GAS=NITROGEN 

##.DRIFT GAS=NITROGEN 

##.ELECTRIC FIELD=-158,-294 

##.ION POLARITY=NEGATIVE 

##.IONIZATION MODE=PD 

##.IMS TEMPERATURE=25.0 

##.SHUTTER OPENING TIME=300 $$ microseconds 

##.REPETITION RATE= 100 

$$ milliseconds 

##.IONIZATION CHAMBER=CYL,30,7.5 

##.IONIZATION SOURCE= Partial Discharge, Needle= Stainless Steel, Gap = 4.6 mm 

##.DRIFT CHAMBER=CYL,120,7.5 

##.CARRIER GAS FLOW=0.2 

##.DRIFT GAS FLOW=0.11 

##.CARRIER GAS MOISTURE=0.03 

##.SHUTTER GRID POTENTIAL=100 

##.REDUCED MOBILITY=(KA) 

(2.38, ) 

(2.27, ) 

(2.06, ) 

(1.83, ) 

##XUNITS=MILLISECONDS 

##YUNITS=PICOAMPERES 

##XLABEL=Drift Time / ms 

##YLABEL=Ion Current / pA 

##FIRSTX=0 

##LASTX=59.975 

##FIRSTY=0. 4491087E+01 

##XFACTOR=0.1830348E-02 

##YFACTOR=0.1037643E-01 

##NPOINTS=2400 

##XYDATA=(X++(Y..Y)) 

0D33k31J0844J6532N189j595k202m79K87J50R4r4k02m3J42m808j8005q458M14K093Q14 

273H75l8j74q4j41m8K64r9j27J22p5l3M3%J27m3j03J9k4N6R5m3j13%Q5O6r4j18N2J4% 

683D61J41q5m7o1Q5K3J0q5Q5j27n2J93l8J04j22r0M3P9j8Q0j13N6P5j17N2l3Rk3j23J79 

... 

32426C53L7K45m2r0j17M2J04q5J41l3j93J46rJ8K4o6K8J0j41Q9o1J13n2j4O6 

32767E13 

##END= 




6 REFERENCES 

1. R.S. McDonald, and P.A. Wilks Jr., JCAMP-DX: A Standard Form for Exchange of Infrared 

Spectra in Computer Readable Form. Applied Spectroscopy, 1988. 42: p. 151-162. 

2. J Gasteiger, B.M.P. Hendriks, P. Hoever, C. Jochum and H. Somberg., JCAMP-CS: A Standard 

Exchange Format for Chemical Structure Information in Computer-Readable Form. Applied 

Spectroscopy, 1991. 45: p. 4-11. 

3. A.N. Davies, and P. Lampen, JCAMP-DX for NMR. Applied Spectroscopy, 1993. 47: p. 1093- 

1099. 

4. P. Lampen, H. Hillig, A.N. Davies and M. Linscheid, JCAMP-DX for Mass Spectrometry. Applied 

Spectroscopy, 1994. 48: p. 1545-1552. 

5. P. Lampen, J. Lambert, R.J. Lancashire, R.S. McDonald, P.S. McIntyre, D.N. Rutledge, T. Fröhlich 

and A.N. Davies, An Extension to the JCAMP-DX Standard File Format, JCAMP-DX V.5.01 (IUPAC 

Recommendations 1999). Pure and Applied Chemistry, 1999. 71 p. 1549-1556. 

6. J.I. Baumbach, P. Lampen, A.N. Davies, IUPAC / JCAMP-DX: An International Standard for the Exchange 

of Ion Mobility Spectrometry Data. International Journal for Ion Mobility Spectrometry 1998 1 p. 64-67. 


1Compound 1 

2Compound2 

3Compound3 

4Compound4 

5Compound 5 

6Compound 6 

7Compound 7 

8RIPSignal 

9Compound8 

10Compound9 

11Compound10 

12Compound11 

13Compound12 

14Compound13 

15Compound14 

16RIPSignal 

THE PROTECTIONOF CIVIL FACILITIES 

BY MEANS OF ASTATIONARY IMS 

H. Bensch,J.W.Leonhardt,IUTLtd.Berlin, G.Bendisch,E.Wolters,Dräger SicherheitstechnikGmbH,K.M.Baether, DrägerwerkAG 

THEPROBLEM 

Terrorists attacks are directed to various civil and 

public facilities: parliament, governmental buildings, 

banks, hotels, subway stations and other facilities. 

Chemical warfare agents and industrial poisons were 

used frequently and have an increasing importance. 

Therefore the monitoring of air flows in the 

conditioning systems of the facilities wide range of 

possible chemical compounds, which are used in 

buildings there is ahigh potential of false alarms by 

crosssensitivities. 

GermanParliament 

covered by Christo and protected by IUT-IMS 

SCHEMEOFTHESTATIONARYIMS 

AMBIENT 

AIR 

INPUT 

DUST 

FILTER 

EXIT 

SAM- 

PLING 

I 

SAM- 

PLING 

II 

P 

P 

P 

FILTER 

IMS 

(+)Mode 

IMS 

(-)Mode 

FILTER 

Ampl. 

HV 

Pulse 

generator. 

Ampl. 

HV 

pulse 

Generator 

M-P 

MEMORY 

M-P 

MEMORY 

EXIT 

4-20 mA 

VOLTAGE 

HEATING 

FLOW 

RIP 

VOLTAGE 

HEATING 

FLOW 

RIP 

SPECIFICATIONS 

The IMS-stations being used as monitors under these conditions have to 

meet the following conditions: 

! Highreliability - theuserdemandsan5yearsnonstop 

operation 

! Highsensitivity - aslowertheMDCthanbetter 

! Nofalsealarms - theevacuationofpeoplehavetobedone 

Immediately 

! Reducedcrosssensitivities - the matrix of the ambient air has 

to be analysed due to painting, cleaning, 

contraction(weldingprocedures) 

PARAMETERS 

1 Minimal detectable concentrations in microgram per cubmeter are 

arrived: 

The Stationary IMS inEXPROOF Version 

GB0,7/GA0,4/GD4,2/VX 0,4/L15,0/HD2,1/HN3 

19,0/ 

COCl 

2 

GC-IMS FOR INDUSTRIAL USE 

H.Bensch,J.W.Leonhardt,IUTLtd. Berlin,Germany 

IUT 

Institut für 

Umwelttechnologien 

GmbH 

Introduction 

PICTUREOFTHEGC-IMS 

The analysis of mixturesby means ofan IMS as detectorcan only be avery difficultor even 

impossibletask. 

Depending on thematrix composition the driftspectra can change. Mixed ions can be 

created or charge transfer takesplace. 

The preseparation of gas samples by gaschromatographiccolumnsgives some 

improvement. 

The chromatographicseparation also depends on thecomposition of gas samples.The 

analysis of an unknown mixture is problematicand time consuming. 

An adapted GC can be helpfulin cases when only afew compoundsshould be measuredin 

acomplex matrix. 

For process controlin industry thecondition oflow complexity issometimes given. Agood 

reproducibility of the results is demanded during months. 

The influence ofvarying water concentration can be reduced with GC preseparation 

dramatically. 

The use of externalcarrier gas is helpfuldepending on the special application. 

Software 

SCHEMEOFTHE GC-IMS 

The internalmicrocomputeris programmable fordifferentconfigurationsand applications via 

a 

scriptlanguage. 

Thereare twoprograms (on lineand off line) running on externalPC. 

P 

Collector 

Filter 

Emitter 

On line version 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

Compound specificcalibration data 

rel. position to the RIP 

polarity 

declaration of monomer,dimer,trimeror 

dissoziated products 

zerovalue 

offset 

11 approximation points of calibration curve 

internalcalibration due totheRIP drifttime 

showing allpeaksrelated to one compound 

different reports and averaging 

• 

• 

Offlineversion 

• 

• 

• 

• 

• 

Library of300 compounds with 

monomer,dimer.. 

Showing 25 spectraatonce for 

preselection 

Chromatographic cut through the 

collected spectra maximumpeak 

heightin an defined window 

Automated signalprocessing 

denoising with wavelets 

deconvolution with modified van Cittert 

iteration 

Filter 

Tube 

P 

Column 

Valve 

samp. 

loop 

CHROMATOGRAMOFAMMONIA 

P 

Devicedescription 

Adjustable temperature fordetector,inletsystem and column 

2adjustable temperatures for the column 

Internalpump forcolumn carrier gas pressure 

Multicapillarorpacked column 

Radioactivetritium sourceof

G.A.S. Gesellschaft 

für analytische Sensorsysteme mbH 

Ion Mobility Spectrometer 

GAS-MPD-SA 

INSTRUMENT TO MONITOR ONLINE AND 

ON-SITE SF 6 -DECOMPOSITION CAUSED BY ARCING 

OR PARTIAL DISCHARGES 

PORTABLE 

STANDALONE VERSION FOR INSPECTIONS 

IN COMBINATION WITH LAPTOP 

(NOT SHOWN) 

DEVELOPED AND MANUFACTURED BY: 

G.A.S. Gesellschaft für analytische Sensorsysteme mbH 

TechnologieZentrumDortmund, Emil-Figge-Str. 76-80, 


Phone: +49 231 9742 288 - FAX: +49 231 9742 289 

E-Mail: GAS@GAS-DORTMUND.DE 

WWW.GAS-DORTMUND.DE, WWW.GAS-DORTMUND.COM 

Technical Information, 24.06.01




Changes without further notice 

Function 

The basic principle of the instrument is the measurement of the drift time of ions provided by an 

ionisation source (in this cause a partial discharge). Decomposition Products causes changes in the 

ion swarm. Thus a peakshift arises and the peakshift is used as signal about the quality of SF 6 gas 

within the compartment under investigation. 

Application Fields 

Gas Insulated Substations, Curcuit Breakers, Gas Insulated Lines, Gas Insulated Transformers, 

Bus Bars, etc. 

Concentration Area 

500 ppb v - 1000 ppm v concentration of decomposition products (1ppb v = 1 ng/l, 1 ppm v = 1 µg/l) 

Installation 

Easily connectable to existing 

valves, automatic shut down. 

Included Software 

GASoszi2001 

Currentime 

Viewer13 

Morepeaks 

Instrumentation Parameter 

Power Supply 

Power Consumption 

Output Signal 

Output Trigger 

Weight 

Size 

Pressure 

115/230 V AC // 50-60 Hz 

30 W 

0 - 10 V DC 

0 - 10 V DC 

8,8 kg 

185 x 340 x 320 mm 

Max. 14 bar 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289 


WWW.GAS-DORTMUND.DE, WWW.GAS-DORTMUND.COM





GAS-MCC-UV-IMS 

SEPARATION AND IDENTIFICATION 

OF MIXTURES OF THE GASOLINE COMPOUNDS 

METHYL -TERT-BUTYLETHER (MTBE) NEXT TO 

BENZENE, TOLUENE AND XYLENE (BTX) 

MTBE 

Signal / a.u. 

-1,0 

-0,5 

0,0 

Chromatogram 

Benzene 

20 40 60 


Toluene 

m-Xylene 


30 

20 

10 

0 

10 20 30 40 50 60 70 


PID 

IMS-Chromatogram 


25 

20 

15 

MTBE MCCUV-IMS 

Toluene 

m-Xylene 

Benzene 

10 20 30 40 50 60 70 


Peakheightdiagram 

IMS-Chromatogram of a mixture of MTBE, Benzene, Toluene and m-Xylene 

APPLICATION NOTE # MTBE-BTX 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289 







Compounds 

Mixture of Methyl -tert-Butylether (MTBE) next to Benzene, Toluene and Xylene (BTX) 


Detection limits in the µg/L-Range 

(e.g. MTBE 2 µg/L) 

1 

MCC-UV-IMS 

- MTBE-Testgas - 


Ionization source 

UV (10.6 eV) 

Length of the drift tube 

120 mm 

Electrical field strenght 

326 V/cm 

Shutter opening time 

100 µs 

Drift gas 

N 2 (99.999%) 

Drift gas flow 

200 mL/min 

Sample gas 

N 2 (99.999%) 

Sample gas flow 

100 mL/min 

Temperature (IMS) 

23 °C 

Stationary phase of the MCC nonpolar 

Length of the column 

25 cm 

Carrier gas of the MCC 

N 2 (99.999%) 

Carrier gas flow 

21.5 cm 

Temperature (MCC) 

23 °C 

Peak Area / a.u. 

0,1 

0,01 

-0,02 

0,00 

MTBE: 2.0 µg/L 

Peak 1 

15 20 

Sum Peak 1 and Peak 2 

Peak 1 

Peak 2 

1 10 100 1000 

Concentration / µg/L 

Peak 2 

-0,2 Peak 1 

-0,1 

0,0 

MTBE: 2150 µg/L 

20 30 

Signal / V 

-1,0 

-0,5 

MTBE 

Benzene 

Toluene 

Chromatogram 

m-Xylene 


0,0 

25 

20 

15 

MTBE 

Benzene 

IMS-Chromatogram Ion Mobility Spectra 

MTBE 

m-Xylene 

Toluene 

Benzene 

MTBE 

20 40 60 0,0 -0,1 -0,2 


Signal / V 

Toluene 

m-Xylene 

IMS-Chromatogram, 

Chromatogram of a mixture and 

Ion Mobility Spectra of the single 

substances of MTBE, Benzene, 

Toluene and m-Xylene recorded 

using the GAS-MCC-UV-IMS 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289 







GAS-MCCUV-IMS 

INSTRUMENT TO MONITOR ONLINE AND ON-SITE 

MIXTURES OF VOLATILE ORGANIC COMPOUNDS 

PORTABLE STANDALONE VERSION 

FOR MONITORING 


(NOT SHOWN) 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289 




Signal / V 




Characteristics 

• Fast (spectra are provided within 50 ms) 

• Sensitive (detection limits in the pg or 

ppb v range and partly even below) 

• Selective (UV-lamps with ionization 

energies from 8.4 - 11.8 eV) 

• Non-Radioactive Ionization Source 

• Cheap and Environmentally Favorable 

• Portable 


• Environmental Analysis 

(air, gas, water, soil) 

• Process Control 

(water treatment, energy distribution) 

• Work Place Control 

(organic compounds) 

Spectra 

Signal / V 

1,5 

1,0 

0,5 

0,0 

PID 

Acetone 



0 20 40 


Software 

GASoszi2001 

-0,6 

-0,4 

-0,2 

0,0 

10 

Acetone 

20 


Chromatogram and IMS-Chromatogram of a mixture of 

Acetone, Trichloroethene and Tetrachloroethene 

30 

40 


50 

0 

10 

UV-IMS 


20 


Compounds 

Compound detectable with GAS-UV-IMS 

• Alkanes (e.g. Hexane) 

• Alcohols (e.g. Ethanol, Ethoxypropanol) 

• Halogenated Hydrocarbons 

(e.g. Tetrachloroethene) 

• Aromatic Hydrocarbons 

• Aromatoc Hydrocarbons with Chains 

(e.g. BTX) 

• Halogenated Aromatic Hydrocarbons 

• Halogenated Biphenyls (e.g. PCP) 

• Carbonyl Compounds 

• Aldehyds (e.g. Formaldehyd) 

• Ketons (e.g. Aceton) 

• Ether (e.g. MTBE, 2,2 Dichloroethylether) 

• Phosphorous Compounds 


Power Supply 

Ionization Source 

Available Energies 

of UV-Lamps 

Column 


Output Signal 


Weight 

Size 

115/230 V AC // 50-60 Hz 

Photoionization 

8.4 / 9.6 / 10.0 / 10.6 / 11.8 eV 

Muli-Capillary Column 

30 W 

0 - 10 V DC 

0 - 10 V DC 

8,8 kg 

185 x 340 x 320 mm 


1 ppb v - 1000 ppm v 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289 







GAS-UV-IMS 

INSTRUMENT TO MONITOR ONLINE AND ON-SITE 

VOLATILE ORGANIC COMPOUNDS 

PORTABLE STANDALONE VERSION 

FOR CONTINUOUS MONITORING 


(NOT SHOWN) 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289 







Characteristics 

Spectra 

• Fast (spectra are provided within 50 ms) 

• Sensitive (detection limits in the pg or ppb v 

range and partly even below) 

• Selective (UV-lamps with ionization 

energies from 8.4 - 11.8 eV) 

• Non-Radioactive Ionization Source 

• Cheap and Environmentally Favorable 

• Portable 

Signal / a.u. 

-0,06 

-0,04 

-0,02 

0,00 

-0,05 

K 0 =1.99 cm 2 /Vs 

K 0 =1.43 cm 2 /Vs 

10 20 30 

K 0 =1.90 cm 2 /Vs 

MTBE 

Toluene 

-0,05 

0,00 

-0,05 

Benzene 

K 0 =2.00 cm 2 /Vs 

10 20 30 

m-Xylene 

K 0 =1.81 cm 2 /Vs 


• Environmental Analysis 

(air, gas, water, soil) 

• Process Control 

(water treatment, energy distribution) 

• Work Place Control 

(organic compounds) 

0,00 

0,00 

Software 

GASoszi2001 

10 20 30 


10 20 30 

Compounds 

Compound detectable with GAS-UV-IMS 

• Alkanes (e.g. Hexane) 

• Alcohols (e.g. Ethanol, Ethoxypropanol) 

• Halogenated Hydrocarbons 

(e.g. Tetrachloroethene) 

• Aromatic Hydrocarbons 

• Aromatoc Hydrocarbons with Chains 

(e.g. BTX) 

• Halogenated Aromatic Hydrocarbons 

• Halogenated Biphenyls (e.g. PCP) 

• Carbonyl Compounds 

• Aldehyds (e.g. Formaldehyd) 

• Ketons (e.g. Aceton) 

• Ether (e.g. MTBE, 2,2 Dichloroethylether) 

• Phosphorous Compounds 


Power Supply 

Ionization Source 

Available Energies 

of UV-Lamps 


Output Signal 


Weight 

Size 

115/230 V AC // 50-60 Hz 

Photoionization 

8.4 / 9.6 / 10.0 / 10.6 / 11.8 eV 

30 W 

0 - 10 V DC 

0 - 10 V DC 

8,8 Kg 

185 x 340 x 320 mm 


1 ppb v - 1000 ppm v 





Phone: +49 231 9742 288 - FAX: +49 231 9742 289

Full issue - B & S Analytik GmbH

Create successful ePaper yourself

Delete template?

Save as template?