Full issue - B & S Analytik GmbH

INTERNATIONA L 

International Journal 

for Ion Mobility Spectrometry 

SOCIETY 

for 

ION 

MOBILITY 

SP ECTROMETRY 

3(2000)1 

Official publication of the 

International Society for Ion Mobility Spectrometry

Table of Contents 

Regular Papers 

Comparative ion mobility measurements of isomeric nitrogenous aromatics 

using different ionization techniques 

Helko Borsdorf, Mathias Rudolph 

We investigated the ion mobility spectra of 8 isomeric nitrogenous aromatic compounds using 

various ionization techniques in order to estimate the influence of structural differences on the 

ionization pathways and drift behavior. The positive ion mobility spectra of benzylamine, 3 different 

methylanilines and 4 different dimethylpyridines with a molecular weight of 107 amu were 

measured using 63 Ni ionization, corona discharge ionization and photoionization. A distinct influence 

of molecular structure on the ion mobility spectra can be identified depending on the 

ionization technique applied. 

Detection of chlorinated and fluorinated substances using 

partial discharge ion mobility spectrometry 

H. Schmidt, J.I. Baumbach, P. Pilzecker, D. Klockow 

Radioactive nickel foils ( 63 Ni), so far most frequently used in ion mobility spectrometry, are 

expected to be partly replaced by nonradioactive alternatives for ionization of analyte molecules, 

for instance partial discharge sources. A partial discharge ion mobility spectrometer was used for 

the detection of selected volatile halogenated compounds containing chlorine as well as fluorine 

atoms. Spectra of trans-1,2-dichloroethene, trichloroethene, tetrachloroethene and perfluorohexane 

measured with the developed ion mobility spectrometer (IMS) will be presented and 

discussed. 

A micro-machined ion mobility spectrometer-mass spectrometer 

G.A. Eiceman, E.G. Nazarov, R.A. Miller 

An ion filter for APCI mass spectrometry was constructed using a micro-machined radio 

frequency (RF) ion mobility spectrometer (IMS) operating at ambient pressure. The ion filter was 

positioned between the ion source and the flange of the mass spectrometer and allowed 

pre-separation of ions before mass spectrometry measurements. The micro-machined RF-IMS 

drift tube was fabricated from two glass plates separated by 0.5 mm thick silicon strips providing 

a drift tube with dimensions of 30 mm long X 10 mm wide X 2 mm thick. Ions are swept, using a 

clean gas flow of about 2 liters/min nitrogen or air, through the drift tube containing metal 

electrodes deposited on the glass plates. During passage in the drift region, ions enter an RF 

field created using a 2 MHz asymmetric waveform with high (20,000 V/cm) and low (-1000 V/cm) 

electric fields; a superimposed DC electric field of –400 to +400 V/cm can be adjusted to select 

and pass ions through the filter and into the mass spectrometer. In this miniature design, ions 

that pass through the filter are deflected into a pinhole inlet on the vacuum flange of a mass 

spectrometer. Binary mixtures of volatile organic compounds were used to demonstrated 

continuous monitoring with a photo-discharge lamp ion source and ion pre-separation before 

mass analysis. 

1 

8 

15 

Copyright © 2000 by International Society for Ion Mobility Spectrometry

Coupling of Multi-Capillary Columns with two Different 

Types of Ion Mobility Spectrometer 

J.I. Baumbach, S. Sielemann, P. Pilzecker 

Pre-separation using gas chromatographic Multi-Capillary Columns is used to overcome resolution 

problems if complex mixtures of volatile organic compounds are analyzed by ion mobility 

spectrometry. It is shown that inter-molecular charge transfer reactions take place and the interpretation 

of the acquired spectra is improved. Mixtures of selected volatile organic compounds 

(butanol, ethylmethylketone, pentane, propanol, pyridine, tetrachloroethene, toluene and trichloroethene) 

with masses up to 1 ng are separated in a few seconds at ambient temperature. 

Detecting heroin in the presence of cocaine using ion mobility spectrometry 

A.M. DeTulleo, P.B. Galat, M.E. Gay 

The South Central Laboratory has been using the Barringer field portable Ion Mobility 

Spectrometer (IMS) to support federal, state, and local drug cases for over seven years. In the 

course of hundreds of field operations, difficulty has been encountered in detecting heroin when 

cocaine is present. Cocaine and heroin, when present individually, are easily detected on the 

Barringer IMS at concentrations as low as 1 nanogram. However, depending on relative concentrations 

of mixtures of cocaine and heroin, the heroin may not be detected at levels as high as 25 

nanograms. In addition, heroin samples routinely contain other opium derivatives, such as acetylcodeine 

and 6-monoacetylmorphine. These two compounds exhibit peaks that have similar 

reduced mobilities to cocaine. The South Central Laboratory has encountered numerous situations 

where cocaine is initially detected. However, when using Gas Chromatography-Mass 

Spectrometry (GC/MS) for confirmation, both cocaine and heroin are detected. Therefore, the 

operator must be able to recognize this potential problem in law enforcement field situations. 

A high resolution ims for environmental studies 

J. W. Leonhardt, W. Rohrbeck, H. Bensch 

IUT Ltd. develops various IMS devices for environmental purposes. The analysis of mixtures of 

aromatics has shown that there are problems to separate benzene and toluene ions properly by 

means of a low resolution cell (R=25). Similar problems exist in the negative mode for various 

halocarbons, for example trichloroethylene and dibromomethane. Therefore a new basic equipment 

was developed in order to improve resolution and transmission. New detector cells have a 

50 or 100 mm long drift tube with diameters of 20 or 30 mm respectively. Ionisation is 

produced by tritium -sources or by UV-lamps. The trigger pulse can be varied in the range of 10 - 

350 microseconds, the drift field is 400 - 700 V/cm. A resolution better than 100 was achieved. 

This value could be improved up to 200 by use of a deconvolution program. The simultaneous 

detection of chlorine and bromine ions produced in a sample of bromochloromethane is demonstrated. 

Further applications are discussed for benzene, toluene, xylene, halothan, isofluorene, 

formaldehyde etc. 

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

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

Programme 

Abstracts of Papers 

Abstracts of Posters 

28 

38 

43 

50 

54 

78 


Comparative ion mobility measurements 

of isomeric nitrogenous aromatics using different ionization techniques 

Helko Borsdorf 1 and Mathias Rudolph 2 

1 

UFZ-Centre for Environmental Research Leipzig-Halle, Department of Analytical Chemistry, PF2, D-04301 Leipzig, 

Germany 

2 

Chemnitz University of Technology, Faculty of Electrical Engineering and Information Technology, Chair of System 

Theory, D-09107 Chemnitz, Germany 

Abstract 

We investigated the ion mobility spectra of 8 

isomeric nitrogenous aromatic compounds 

using various ionization techniques in order to 

estimate the influence of structural differences 

on the ionization pathways and drift behavior. 

The positive ion mobility spectra of 

benzylamine, 3 different methylanilines and 4 

different dimethylpyridines with a molecular 

weight of 107 amu were measured using 63 Ni 

ionization, corona discharge ionization and 

photoionization. A distinct influence of 

molecular structure on the ion mobility spectra 

can be identified depending on the ionization 

technique applied. 

The methods of atmospheric-pressure chemical 

ionization (corona discharge ionization and 63 Ni 

ionization) provide product ion peaks with 

nearly identical reduced mobility values for the 

compounds investigated. Although the spectra 

consist of up to three product ion peaks, their 

intensities in ion mobility spectra vary 

depending on the ionization technique used. 

The difficult interpreted spectra obtained by 

photoionization exhibit considerable differences 

for the compounds investigated. 

Introduction 

Ion mobility spectrometry (IMS) is based on 

determining the drift velocities which ionized 

sample molecules attain in the weak electric 

field of a drift tube at atmospheric pressure. 

Therefore, determining ion mobilities initially 

requires the formation of ions from neutral 

sample molecules. Subsequently, the ions 

formed are separated within the drift tube and 

the drift velocities are determined. Both the 

ionization processes and drift behavior can be 

affected by structural differences of the 

analytes. 

Various techniques can be applied to ionize 

neutral sample molecules [1]. The methods in 

this study used include the application of 63 Ni 

ionization, corona discharge (CD) ionization 

and photoionization (PI). When using 

63 

Ni 

ionization, the formation of positive product ions 

is mainly initiated by proton-transfer reactions 

[2]. The intensities of these reactions strongly 

depend on the gas-phase basicities of the 

investigated compounds as well as on the 

temperature and composition of the carrier gas 

[3,4]. The most probable ionization pathway 

using photoionization provides [M] + product ions 

[5]. Using CD ionization, positive product ions 

may be formed via different processes. 

Electron impact, photoionization and 

proton-transfer reactions may be initiated the 

formation of product ions depending on the 

field strength around the corona needle [6]. 

Subsequent ion-molecule reactions and cluster 

formations can be expected for the ionization 

processes mentioned [7]. 

The reaction course of product ion formation 

and the intensities of these reactions are 

affected by the ionization technique applied and 

by the physicochemical properties of the 

compounds investigated. Therefore, different 

product ions may be formed due to the different 

ionization pathways. 

The influence of structural differences on drift 

behavior can be derived from the expression of 

general theory for the mobility of ions in a weak 

electric field according to Equation (1) 

1 

⎛ 

3q 

⎞ ⎛ 

2π 

⎞ 

2 ⎛ 

(1 + α 

) ⎞ 

K = ⎜ ⎟∗⎜ 

⎟ ∗ 

⎜ 

⎟ (1) 

⎝ 

1 6N 

⎠ ⎝ µkT ⎠ ⎝ ΩD 

⎠ 

where q = charge of the ion, N = density of drift 

gas molecules, µ = (m*M)/(m+M) = reduced 

mass of the ion (m) and drift gas molecule (M), 

Received for review January 25, 2000, Accepted July 10, 2000 


H. Borsdorf and M. Rudolph: „Comparative ion mobility measurements...”, IJIMS 3(2000)1,1-7, p. 2 

k = Boltzmann constant, T = temperature, α = 

correction factor (α < 0.02 for m > M), Ω D = 

average ionic collision cross-section. The term 

(Ω D) includes geometrical parameters (physical 

size and shape), along with electronic factors 

describing the ion-neutral interaction forces 

(polarizability) [8-11]. The average ionic 

collision cross-section expresses the influence 

of structural parameters on drift behavior in 

IMS. According to Equation 1, ion mobility is 

considerably influenced by the mass of 

analytes and their collisional cross sector, as 

well as parameters resulting from the 

measuring conditions (temperature, drift gas 

used, ionic charge). The application of identical 

operational parameters permits variables 

influencing the ion mobilities of different 

compounds to be limited to the ionic masses 

(m) and the structures (Ω D), enabling the 

comparison of ion mobility spectra. Unlike ionic 

mass, the influence of structural features on the 

reduced mobility values is not completely 

understood. The objective of our study was to 

contribute to the investigation of the 

relationship between structural features and the 

ion mobilities determined. 

Therefore, 8 constitutional isomers of 

nitrogenous aromatics with a molecular weight 

of 107 amu were investigated by ion mobility 

spectrometry equipped with the above 

mentioned ionization sources. The influence of 

structural differences on the ionization pathway 

and drift behavior was estimated by comparing 

the spectra obtained. 

Previously, studies of isomeric compounds 

were only performed using 63 Ni ionization and 

only included substances containing polar 

functional groups (branched and unbranched 

ketones, thiocyanates, isothiocyanates [12], 

phthalic acids [13], dihalogenated benzenes 

[14], halogenated nitrobenzenes [15], amines 

[16], anilines [17], E/Z isomers [18], diamines 

[19] and nitrogenous heterocycles, as well as 

aromatics substituted by functional groups 

[20]). 

Experimental 

The details of the applied sample introduction 

system are shown in Fig. 1. About 300 µl of 

liquid samples (benzylamine, o- and 

m-toluidine, 2,4-, 2,6-, 3,4- and 3,5-lutidine) 

were sealed in permeation tubes consisting of 

Measuring arrangement for detection of ion mobility spectra 

moisture 

sensor 

gas washing bottles 

(slica gel with moisture indicator and charcoal) 

dew point 

pressure 

compensation 

25 - x l/h 

flowmeters 

pump (ambient air) and 

pressure compensation 

25 l/h 

25 l/h 

x l/h 

permation vessel 

IMS 

25 - x l/h 

Figure 1: 

Sample introduction system for ion mobility measurements 



polyethylene. The solid sample of p-toluidine 

was positioned in an open cup. The substances 

had a purity of about 99% and were obtained 

from Fluka. Purified and dried ambient air was 

pumped through a glass column containing the 

permeation tubes at constant flow. This sample 

gas stream was split using flow controllers. A 

defined amount of sample gas stream was 

rarefied with purified and dried ambient air. The 

concentration of the compounds in the sample 

gas stream was calculated using the weight 

loss of the permeation vessels over certain 

time. The moisture content of the gas streams 

was controlled by a moisture sensor AMX1 

(Panametrics). Gas-drying by silica gel and 

purification by charcoal yielded a relative 

humidity of about 2.4% (-25°C dew point). 

The measurements were performed with ion 

mobility spectrometers by BRUKER. With the 

exception of ionization, all the measuring 

parameters were kept constant. The 

photoionization source was equipped with a 

krypton lamp (10eV). The basic features of the 

corona discharge ionization source used in this 

study are described in detail elsewhere [6]. For 

β ionization, a 555 MBq 63 Ni source was used 

as the electron source. 

The spectrometers are equipped with a 

membrane inlet and operate with a 

bi-directional flow system. The operational 

parameters used to obtain the spectra were: 

inlet system temperature: 80°C; carrier gas flow 

rate: 25 l/h; drift gas flow rate: 25 l/h; electric 

field: about 245 V/cm; temperature of drift tube: 

80°C; pressure: atmospheric pressure. Air was 

used as the carrier gas and drift gas. 

The reduced mobility values (K 0 values) were 

calculated according to the conventional 

equation [21]: 

⎛ d ⎞ ⎛ p ⎞ ⎛ 

2 7 3 ⎞ 

2 

K 

0 

= ⎜ ⎟ 

* ⎜ ⎟ 

* ⎜ ⎟ = 

( cm 

/ Vs 

) (2) 

⎝ t 

* E ⎠ ⎝ 

7 6 0 ⎠ ⎝ T ⎠ 

where d = drift length (cm); t = drift time (sec); 

E = field strength (V/cm); p = pressure (torr) 

and T = temperature (K). 

Results and discussion 

The investigated isomeric compounds, their 

structure and significant physicochemical 

Table 1: Properties of investigated substances and measuring results 

C 7 

H 9 

N 

m/z=107 

ionization energy 

(eV) 

proton affinity 

(kJ/mol) 

dipole moment 

(D / neat liquid) 

benzyl- o- m- p- 2,4- 2,6- 3,4- 3,5- 

amine toluidine toluidine toluidine lutidine lutidine lutidine lutidine 

NH 2 NH2 

CH 3 

NH 2 

CH 3 

NH 2 

8,49 7,47 7,54 7,60 8,85 8,86 9,22 9,25 

913,3 890,9 895,8 896,7 

1,18 1,59 1,49 1,52 

CH 3 

CH 3 

N CH 3 

N 

962,9 963,0 957,3 955,4 

2,24 1,78 1,85 2,35 

N 

CH 3 

CH 3 

N 

K 0 

values CD 1,86 / 1,70 1,89 / 1,75 1,89 / 1,73 1,89 / 1,71 1,89 1,89 1,87 1,86 

(cm 2 /Vs) [1,36] [1,37] [1,34] [1,32] 

measuring range 0,3 - 1,3 0,8 - 1,2 0,3 - 1,7 0,3 - 1,7 0,4 - 2,2 0,2 - 1 0,7 - 3,7 0,4 - 2,1 

K 0 

values 63 Ni 1,86 / 1,70 1,88 / 1,74 1,88 / 1,72 1,88 / 1,71 1,88 1,89 1,86 1,86 

(cm 2 /Vs) [1,36] [1,32] [1,37] [1,33] [1,32] 

measuring range 0,1 - 0,6 0,05 - 0,25 0,06 - 0,3 0,3 - 1,3 0,1 - 0,6 0,07 - 0,32 0,01 - 0,13 0,1 - 0,5 

K 0 

values PI 1,89 1,90 1,90 1,88 

(cm 2 /Vs) 1,41 - 1,31 [1,32] [1,32] [1,32] [1,37] [1,34] [1,37 / 1,33] [1,37 / 1,30] 

measuring range 7,4 - 92 4 - 26,5 5 - 30 0,3 - 2,5 15 - 130 20 - 150 7 - 61 15 - 125 

bold print: major peaks 

63 

Ni: 63 Ni ionization; CD: corona discharge ionization; PI: photoionization 

concentration unit: µg/l (g) 



63 

Ni ionization Corona discharge ionization 

0,5 µg/l benzylamine 

1,2 µg/l 

0,2 µg/l 

2-methylaniline 

1 µg/l 

0,25 µg/l 


0,8 µg/l 

0,5 µg/l 


1,2 µg/l 

Figure 2: 

Ion mobility spectra of benzylamine and isomers of toluidine 



Hierarchical agglomerative cluster analysis (WARD method) 

2,4-lutidine 

3,5-lutidine 

2,6-lutidine 

3,4-lutidine 

p-toluidine 

m-toluidine 

o-toluidine 

benzylamine 

Variables: Physicochemical values (ionization potential, proton affinity, 

dipole moment) 

0 2 4 6 8 10 

EUKLIDean distance 

Figure 3: 

Dendrogram for structural parameters of investigated compounds 

properties are summarized in Tab. 1 and in the 

measuring results obtained, which include the 

reduced mobility values calculated and the 

concentration ranges detected. It can been 

seen from Tab. 1 that the methods of 

atmospheric-pressure chemical ionization (CD 

ionization and 63 Ni ionization) provide nearly 

identical reduced mobility values for the 

substances investigated. For the nitrogenous 

isomers examined, up to three peaks appear in 

the spectra. These ionization methods permit 

the more sensitive detection of 

these compounds in 

comparison to PI. 

Benzylamine provides three 

peaks with reduced mobility 

values of 1.86 cm 2 /Vs, 1.70 

cm 2 /Vs and 1.36 cm 2 /Vs. Using 

63 

Ni ionization, a weak 

dependence on the 

concentration is observed for 

the peak at 1.86 cm 2 /Vs. The 

intensity of the peak at 1.70 

cm 2 /Vs increases with 

enhanced concentration up to 

the incipient detection of the 

peak at 1.36 cm 2 /Vs. A similar 

intensity distribution is observed 

using CD ionization depending 

on the concentration. 

Generally, a peak at 1.89 or 1.88 cm 2 /Vs is 

detected for the isomers of toluidine in all the 

spectra obtained by CD ionization and 63 Ni 

ionization. For these compounds, an additional 

peak is observed between 1.75 cm 2 /Vs and 

1.71 cm 2 /Vs. In Fig. 2, the K 0 values of these 

peaks depend on the position of substituents. A 

significant shift to lower reduced mobilities is 

observed for o- (1.75 cm 2 /Vs), m- (1.73 cm 2 /Vs) 

and p-toluidine (1.71 cm 2 /Vs). Using CD 

Mass to mobility correlation curve for p-toluidine 

2,35 

2,3 

lg m = -0,52 K 0 + 3,02 

[M 2 H] + [M(H 2 O)H] + 

2,25 

2,2 

lg m 

2,15 

2,1 

2,05 

[MH] + 

2 

1,2 1,4 1,6 1,8 2 

reduced mobility values [cm 2 /Vs] 

Figure 4: 

Mass to mobility correlation curve for product ions of p-toluidine 



ionization, a constant intensity ratio between 

both peaks (1:0.9) is obtained in the 

concentration range detected for all isomers of 

toluidine. Using 63 Ni ionization, the intensity 

ratio between both peaks varies for m-toluidine 

(1:1.8-1.3) due to the stronger concentration 

depending on the peak at 1.88 cm 2 /Vs and for 

p-toluidine (1:2.9-2.3) due to the affection of 

intensity ratio by the formation of dimeric ions. 

A third peak was obtained for p-toluidine at a K 0 

value of 1.32 cm 2 /Vs. A constant intensity ratio 

(1:2.3) was observed for o-toluidine within the 

concentration range detected. 

With the exception of 2,6-lutidine, CD ionization 

and 63 Ni ionization provide two peaks for the 

isomers of lutidine investigated. However, a 

negligible shift of reduced mobility values is 

observed depending on the position of the 

substituents. The major peak in the spectra of 

2,4- and 2,6-lutidine is detected at 1.89 cm 2 /Vs. 

This peak appears in spectra of 3,4- and 

3,5-lutidine at a reduced mobility value of 1.86 

cm 2 /Vs. The second detected peak in these 

spectra (1.37–1.32 cm 2 /Vs) evidently results 

from the formation of dimeric ions. 

The deviating ion mobility spectra of 

benzylamine and isomers of toluidine on the 

one hand and isomers of lutidine on the other 

hand obtained by 

63 

Ni ionization and CD 

ionization clearly reflect the influence of 

structural features. Therefore, the structural 

parameters of the compounds investigated 

were analyzed using cluster analysis. This 

method of pattern recognition is suitable for 

finding and highlighting structures within given 

data. The physicochemical data for 

characterizing structural features include the 

dipole moment for characterizing cluster 

formation, the ionization energy and proton 

affinity as the determinant of ionization 

processes (Tab. 1). The data were obtained 

from [22]. The output of cluster analysis, the 

dendrogram, is depicted in Fig. 3. The distance 

of cluster formation indicates the similarity 

between the given data (physicochemical 

parameters) of objects (chemical compounds) 

[23]. As shown in Fig. 3, similar structural 

features are obtained for compounds which 

provide similar ion mobility spectra. The low 

level of cluster formation for the summarized 

physicochemical data of toluidines and 

benzylamine as well as for isomers of lutidine 

and the high distance of union of formed 

clusters indicate significant differences in the 

structural features. 

Using PI, defined spectra are only observed for 

isomers of toluidine. Two peaks are detected at 

comparable reduced mobility values (1.90 

cm 2 /Vs and 1.32 cm 2 /Vs) for all isomers of 

toluidine. Only a broad peak with a maximum 

intensity at 1.41 cm 2 /Vs is obtained for 

benzylamine. With the exception of 2,6-lutidine, 

the PI provide peaks with a low resolution 

between 1.37 cm 2 /Vs and 1.30 cm 2 /Vs for 

isomers of lutidine investigated. 

2,4-lutidine, 3,4-lutidine, benzylamine and 

m-toluidine were investigated in previous 

studies [24-27] using 63 Ni ionization and PI with 

different measuring conditions. Comparison of 

the reduced mobility values obtained with 

literature data provides partial conformity. 

However, the data available from the literature 

exhibit considerable variation. The detected 

reduced mobility values are evidently affected 

by the measuring conditions and the ionization 

technique used. 

Regarding the peak assignment, we supposed 

the formation of [M] + and [MH] + product ions for 

the peaks between 1.86 cm 2 /Vs and 1.90 

cm 2 /Vs due to the comparison of spectra 

obtained by PI with those detected by CD 

ionization and 63 Ni ionization. The theoretically 

most probable ionization pathway using PI 

provides [M] + product ions. The peaks in the 

range between 1.32 cm 2 /Vs and 1.37 cm 2 /Vs 

were assigned to dimeric ions due to their 

typical intensity distribution depending on the 

concentration. We attributed the peak 

additionally detected in the case of 

benzylamine and isomers of toluidine to the 

formation of ion-water clusters. As illustrated in 

Fig. 4, this peak assignment permits the 

derivative of a mass to the mobility correlation 

curve. 

However, Karpas et al. [27] investigated the 

structure of product ions obtained from anilines 

by IMS/MS studies. The appearance of two 

peaks in the spectra of o-toluidine were 

assigned to different sites of protonation in 

anilines. The higher reduced mobility peak in 

this spectrum is similar to the single peak 

detected for 2,4-lutidine and is assigned to 

carbon-protonated product ions, while the 

peaks with lower reduced mobility result from 

nitrogen-protonated product ions. The single 

peak obtained for benzylamine corresponds to 

the peak obtained for nitrogen-protonated 

toluidine. Our investigations established a 

similar distribution of product ion peaks. 

However, deviating intensity ratios were 



detected in comparison to the investigations by 

Karpas et al. [27]. Therefore, the information 

available from the comparison of ion mobility 

spectra obtained by different ionization 

techniques are insufficient to ascertain the 

exact mode of product ion formation. 

Summary 

For isomeric nitrogenous aromatics, significant 

differences in ion mobility spectra are obtained 

depending on the ionization technique used 

and on the structure of compounds 

investigated. Differences in the ion mobility 

spectra of atmospheric-pressure chemical 

ionization and PI as well as different intensities 

of product ion formation indicate different 

ionization pathways. The deviating ion mobility 

spectra of the constitutional isomers studied 

reflect the influence of structural features. 

References 

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

Spectroscopy 53 (1999) 338A. 

[2] Sunner, J.; Nicol, G.; Kebarle, P.: Anal. Chem. 60 

(1988) 1300. 

[3] Eiceman, G. A.; Nazarov, E. G.; Rodriguez, J. E.; 

Bergloff, J. F.: Int. J. of IMS 1 (1998) 28. 

[4] Karpas, Z.; Berant, Z. J.: Phys. Chem. 93 (1989) 

3021. 

[5] Stach, J.: Analytiker Taschenbuch 16 (1997) 119. 

[6] Adler, J.; Arnold, G.; Döring, H.-R.; Starrock, V.; 

Wülfling, E.: Recent Developments in Ion Mobility 

Spectrometry, ed.: J. I. Baumbach and J. Stach, Int. 

Society for Ion Mobility Spectrometry, Dortmund, 

Germany, 1998, ISBN 3-00-003676-8, p. 110. 

[7] Eiceman, G. A.; Kremer, J. H.; Snyder, A. P.; Tofferi, 

J. K.: Intern. J. Environ. Anal. Chem. 33 (1988) 161. 

[8] Lin, S. N.; Griffin, G. W.; Horning, E. C.; Wentworth, 

W. E.: J. Chem. Phys. 60 (1974) 4994. 

[9] Karpas, Z.; Berant, Z.; Shahal, O.: Int. J. Mass 

Spectrom. Ion Processes 96 (1990) 291. 

[10] Su, T.; Bowers, M. T.: J. Chem. Phys. 55 (1973) 

3027. 

[11] Sennhauser, E. S.; Armstrong, D. A.: Can. J. Chem. 

58 (1980) 231. 

[12] Karpas, Z.; Cohen, M. J.; Stimac, R. M.; Wernlund, R. 

F.: Int. J. Mass Spectrom. Ion Processes 74 (1986) 

153. 

[13] Karasek, F. W.; Kim, S. H.: Anal. Chem. 47 (1975) 

1166. 

[14] Carr, T. W.: J. Chrom. Sci. 15 (1977) 85. 

[15] Karasek, F. W.; Kane, D. M.: Anal. Chem. 46 (1974) 

780. 

[16] Karpas, Z.: Anal. Chem. 61 (1989) 684. 

[17] Karpas, Z.; Berant, Z.; Stimac, R. M.: Struct. Chem. 1 

(1990) 201. 

[18] Karpas, Z.; Stimac, R. M.; Rappoport, Z.: Int. J. Mass 

Spectrom. Ion Processes 83 (1988) 163. 

[19] Karpas, Z.: Int. J. Mass Spectrom. Ion Processes 93 

(1989) 237. 

[20] Hagen, D. F.: Anal. Chem. 51 (1979) 870. 

[21] Sprangler, G. E.: Anal. Chem. 65 (1993) 3010. 

[22] http://webbook.nist.gov/ 

[23] Einax, J. W.; Zwanziger, H. W.; Geiß, S.: 

Chemometrics in Environmental Analysis, 

VCH-Verlag Weinheim 1997, ISBN 3-527-28772-8, p. 

153-163. 

[24] The ISAS Database 

http://ims.isas-dortmund.de/database 

[25] Karasek, F. W.; Kim, S.H.: Anal. Chem. 50 (1978) 

2013. 

[26] Lubman, D. M.: Anal. Chem. 56 (1984) 1298. 

[27] Karpas, Z.: Anal. Chem. 61 (1989) 684. 

[28] Karpas, Z.; Berant, Z.; Stimac; R. M.: Struct. Chem. 1 

(1990) 201. 


Detection of chlorinated and fluorinated substances using 

partial discharge ion mobility spectrometry 

H. Schmidt 1,2 , J.I. Baumbach 1 , P. Pilzecker 3 , D. Klockow 1 

1 

Institut für Spektrochemie und Angewandte Spektroskopie (ISAS), 

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

2 

Universität Dortmund, Fachbereich Chemie, Anorganische / Analytische 

Chemie, Otto-Hahn-Str. 6, D-44221 Dortmund, Germany 

3 

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

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

ABSTRACT 

Radioactive nickel foils ( 63 Ni), so far most 

frequently used in ion mobility spectrometry, 

are expected to be partly replaced by 

nonradioactive alternatives for ionization of 

analyte molecules, for instance partial 

discharge sources. 

A partial discharge ion mobility spectrometer 

was used for the detection of selected volatile 

halogenated compounds containing chlorine as 

well as fluorine atoms. Spectra of 

trans-1,2-dichloroethene, trichloroethene, 

tetrachloroethene and perfluorohexane 

measured with the developed ion mobility 

spectrometer (IMS) will be presented and 

discussed. 

KEYWORDS 

Ion Mobility Spectrometry, Partial Discharges, 

Chlorohydrocarbons, Fluorohydrocarbons 

INTRODUCTION 

The sensitive detection of halogenated 

compounds has been recognized as a 

significant challenge and important task for a 

number of years [1-3]. Chlorohydrocarbons, 

chlorofluorohydrocarbons 

and 

fluorohydrocarbons, i.e. aliphatic or 

cycloaliphatic hydrocarbons of low molecular 

mass with the hydrogen partly or entirely 

substituted by chlorine and / or fluorine are of 

considerable importance with respect to their 

applications in industry, in chemistry and even 

in medicine: Amongst other applications they 

are used as insulants, inert solvents and 

coolants as well as fire-extinguishing agents or 

propellant gases for generation of aerosols 

[4-6]. Some of these compounds were shown 

to be photolysed in the stratosphere thus 

leading to chlorine radicals which destroy the 

protecting ozone layer [7-8]. 

Known as a very sensitive and inexpensive 

method for rapidly determining gaseous 

analytes, ion mobility spectrometry was chosen 

for the detection of halogenated hydrocarbons 

[9-10]. 

In ion mobility spectrometry the way most 

frequently applied to ionize the analyte 

molecules is the use of radioactive sources. 

The β-radiation of such a source like 63 Ni, 241 Am 

or 3 H generates so-called reaction ions from the 

drift gas which in turn execute chemical 

reactions with the analyte molecules. However, 

despite their excellent long term stability as well 

as their simplicity ion mobility spectrometers 

(IMS) using radioactive sources suffer from 

several flaws: Their linear working range is 

limited and the practical use of IMS becomes 

difficult because of a continuously increasing 

number of regulatory requirements as to the 

use of radioactive material. Therefore, those 

IMS are expected to be partly replaced by 

non-radioactive alternatives. UV-lamps as well 

as lasers have been successfully adapted to 

put an end to some of the shortcomings 

mentioned [11-13]. Unfortunately, chemical 

substances having ionization energies which 

exceed 11.8 eV cannot be ionized using 

UV-lamps because of limited transmission of 

MgF 2 windows for VUV light of less than 106 

nm. As most of the saturated hydrocarbons 

containing fluorine as well as most of those 

containing both chlorine and fluorine have 

ionization energies beyond 11.8 eV, the 

ionization by partial discharges [14-17] was 

adapted in order to combine the availability of 

both positive and negative ions with the high 

flexibility offered by almost no limitations with 

regard to the ionization energy of the analyte 


H. Schmidt et al.: „Detection of chlorinated and fluorinated...”, IJIMS 3(2000)1,8-14, p. 9 

Discharge 

Grid 

Sample 

(Inlet) 

Reaction 

Chamber 

Grid- 

Bradbury- 

Nielsen-Grid 

0 – 10 kV 

Faraday- 

Circuitry 

+ 5 0 V 

- 5 0 V 

P 

10 kV – Module (negative) 

Sample and Drift Gas 

SU 

SU 

SU 

SU 

(Outlet) 

Drift Tube 

Aperture 

Grid 

SU Supply Unit 

P Pulse 

Plate 

Drift Gas 

Input 

Measuring 

Amplifier 

SU 

SU 

Figure 1: 

Scheme of a Partial Discharge Ion Mobility Spectrometer 

molecules. This means that similar conditions 

were established as with the use of radioactive 

material. 

In this article a Partial Discharge Ion Mobility 

Spectrometer (PD-IMS) is introduced for the 

sensitive detection of halogenated substances. 

Some of the characteristic features are 

described and a possible application is outlined. 

EXPERIMENTAL 

The measurements were carried out using an 

ISAS partial discharge ion mobility 

spectrometer with a drift tube of 6 cm and a 

reaction chamber of 3 cm length. The 

construction of the PD-IMS is shown 

schematically in Fig. 1, the most relevant 

experimental parameters are summarized in 

Table 1. Materials used for the IMS are Teflon ® , 



Carrier Gas + Analyte 

Carrier Gas 

Membrane 

Temperature 

Control 

Permeation 

Tube 

Cap 

Analyte (g) 

Analyte (l) 

Figure 2: 

Scheme of a Permeation 

System for the Generation of 

Test Gases 

nickel, stainless steel and aluminum, thus 

guaranteeing that only inert materials are in 

contact with the analytes. For generating the 

discharges a point-to-plain geometry was 

chosen. A needle made from stainless steel as 

well as a nickel filament served as electrodes in 

comparable experiments, the tip of the needle 

and the filament both having diameters of about 

50 µm and a gap of about 4.7 mm between the 

electrodes. Because of its mechanical lability a 

stabilizing tube was required for the accurate 

adjustment of the nickel filament. 

Due to chemical and physical properties of the 

chemical substances to be ionized, negative 

ions were preferred for detection throughout all 

the measurements. The IMS was kept at 

ambient temperature. By using temperature 

controlled permeation tubes which were filled 

with the respective analytes (as liquids) to be 

detected and capped by a polydimethylsiloxane 

membrane, it was possible to introduce test gas 

samples into the reaction chamber of the IMS 

through 2 mm (i.d.) Teflon ® tubings. Different 

concentrations of analytes which were in the 

range between 100 ppb v and 50 ppm v were 

Table 1: Experimental Parameter 

Electric Field Strength 

Length of Drift Tube 

Length of Reaction Chamber 

Discharge Voltage 

Electrode Geometry 

Gap between Electrodes 

Shutter Grid Opening Time 

Drift Gas 

Drift Gas Flow 

Carrier Gas 

Carrier Gas Flow 

Moisture 

Copyright © 2000 by International Society for Ion Mobility Spectrometry 

330 V/cm 

6 cm 

3 cm 

4 kV (if not further specified) 

Point-to-Plane 

4.7 mm 

100 µs (if not further specified) 

(10 to 1000 µs adjustable) 

N 2, Air 

150 mL/min 

N 2 

170 mL/min 

< 5 ppm v 

prepared by varying the temperature of the 

permeation tubes (Fig. 2). Nitrogen (99.999 %, 

Messer-Griesheim, Dortmund, Germany) was 

used both as carrier gas (150 - 200 mL/min) 

and drift gas (100 – 300 mL/min). With the help 

of a mass flow controller (Model 1479, MKS, 

Munich, Germany) the flux could be controlled 

precisely. Alternatively, synthetic air 

(Messer-Griesheim, Dortmund, Germany) was 

used as a drift gas. During all the 

measurements the moisture was continuously 

controlled (Moisture Monitor Series 35, 

Panametrics, Waltham, Ma, USA) and never 

exceeded a value of 5 ppm v. 

Results of the detection of the following 

substances are presented in this article: 

trans-1,2-dichloroethene (C 2H 2Cl 2, > 95 %), 

trichloroethene (C 2HCl 3, >99.5 %), 

tetrachloroethene (C 2Cl 4, for IR) (all 

Sigma-Aldrich Chemie GmbH, Deisenhofen, 

Germany) and n-perfluorohexane (C 6F 14, 85 % 

n-isomer, 99 %, ABCR, Karlsruhe, Germany). 

Drift voltage and discharge voltage can be 

varied with the help of the supply unit, which 

also allows to reduce the opening time of the 

shutter grid down to 10 µs. The 

current resulting from the ions 

reaching the aperture-grid shielded 

Faraday-plate is converted to a 

voltage and amplified by a custom 

built amplifier (10 V/nA). The 

measurements are controlled by a 

software custom developed with Test 

Point for Windows ® which also 

collects and stores the spectra using 

a 32 bit ADC board.


Drift Time (Center of Peak) / ms 

Peak Area / a.u. 

Peak Width / ms 

3 

2 

1 

10 

9 

1 

Peak Area 

0 

0 50 100 150 200 250 300 350 400 

Time of Experiment / min 

Figure 3: 

Stability of Peak Width, Position of Peak Center and Peak 

Area of a RIP using a Needle made of Stainless Steel 

RESULTS AND DISCUSSION 

Any ionization principle which is intended to 

replace radioactive sources needs to be 

investigated with respect to its merits. This can 

be done advantageously by making use of a 

reaction ion peak (RIP) of a drift gas because 

in this case experimental parameters 

determining the stability of the ion source can 

be adjusted very precisely without any influence 

of analyte ions. 

Apart from free electrons nitrogen, in contrast 

to oxygen, does not form negative reaction ions 

when applying negative partial discharges so 

that no reaction ion peak serving as an 

Drift Time (Center of Peak) / ms 

Peak Width / ms 

Peak Area / a.u. 

2 

1 

0 

9,0 

8,5 

8,0 

1 

Peak Width 

Position of 

Peak Center 

Peak Width 

Position of 

Peak Center 

Peak Area 


UNTITLED AGraph1 13.02.01 10:44 

Total 

Peak 1 

Peak 2 

PD-IMS 

Needle: Stainless Steel 

UNTITLED - Graph3 13.02.01 11:09 

Needle: Nickel Filament 

0 

0 50 100 150 200 250 

Time of Experiment / min 

Figure 4: 

Stability of Peak Width, Position of Peak Center 

and Peak Area of a RIP using a Nickel Filament 

PD-IMS 

indicator of stability could be 

observed in pure nitrogen. For 

this reason air instead of 

nitrogen was used as a drift gas 

for the investigation of the 

stability of PD-IMS 

measurements. It is assumed, 

however, that the results 

obtained can be transferred to 

measurements in nitrogen as a 

drift gas. Since in that case no 

reaction ion peak overlaps with 

product ion peaks, any ion 

mobility spectrum can entirely be 

assigned to the analytes 

introduced. Therefore ion 

mobility spectra taken with 

nitrogen as a drift gas are 

generally more simple than 

those with air. 

As figures of merit were taken 

the full width of half maximum 

(FWHM) of a reaction ion peak in air, its area, 

and the position of the peak center in 

dependence of the time after starting partial 

discharges for the first time. These experiments 

were executed using different needle-like 

electrodes made of stainless steel as well as of 

nickel filaments of comparable diameter. 

When applying a stainless steel needle, a 

run-in-time of up to 100 min was observed at 

the beginning of the experiment in which initially 

two separated peaks grew together to form one 

peak (Fig. 3). While running the discharge, a 

small amount of material may be sputtered 

from the tip such changing its 

diameter and giving rise to 

discontinuities which are visible in 

the shape, the position and the 

area of the peak. A nickel 

filament of cylindrical shape, 

however, does not vary in 

diameter, thus instabilities are 

drastically reduced. This can be 

seen from Fig. 4. Therefore a 

nickel filament was the preferred 

electrode in the experiments 

presented in this article. 

After having verified the 

satisfactory operation of a 

PD-IMS using the nickel filament 

as an electrode, experiments with 

selected analytes were carried 

out. In Fig. 5 ion mobility spectra 

of the negative ions of

UNTITLED - Trichloroethene(1) 13.02.01 11:15 

UNTITLED - Graph10 13.02.01 11:19 


Signal / a.u. 

1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

Signal / a.u. 

6 8 10 12 14 16 18 

1,00 

0,75 

0,50 

0,25 

0,00 

Trichloroethene 

C 2 

Cl 4 

Shutter Grid 

Opening Time / µs 

1000 

100 

C 2 

H 2 

Cl 2 

C 2 

HCl 3 

Drift Time / ms 

trichloroethene as a typical example of an 

analyte of the group of unsaturated 

chlorohydrocarbons are shown for two different 

shutter opening times. Obviously, several 

peaks are visible in the spectra. Reducing the 

shutter grid opening times from 1 ms to 100 µs, 

these peaks resulting from different ions are 

resolved more clearly, as expected. 

Spectra obtained from three 

different 

unsaturated 

chlorohydrocarbons which only 

0,3 

differ in the hydrogen / chlorine 

ratio, i.e. spectra of the negative 

ions of trans-1,2-dichloroethene, 

trichloroethene 

and 

tetrachloroethene, are compared 

in Fig. 6. Although being of 

comparable chemical structure, 

the mobility spectra of the 

compounds are clearly different. It 

is remarkable that for C 2Cl 4 almost 

exclusively small ions with a high 

mobility are formed, whereas 

larger ions with smaller mobilities 

preferably result from C 2H 2Cl 2 and 

C 2HCl 3. As can be seen from 

Table 2, in which only ions having 

relative intensities larger than 30 

PD-IMS 

Negative Ions 

6 8 10 12 14 16 18 


PD-IMS 

Trans-1,2-Dichloroethene 


Tetrachloroethene 

Negative Ions 

Figure 6: 

Comparison of Ion Mobility Spectra of Di-, Triand 

Tetrachloroethene in Nitrogen 


Signal / V 

0,2 

0,1 

0,0 

Figure 5: 

Ion Mobility 

Spectra of the 

negative Ions 

formed from 


in Nitrogen at 

two different 

Shutter Opening 

Times 

PD-IMS 

% are taken into 

consideration, the latter two 

substances form several ions 

which have the same mobility, 

e.g. K 0(C 2H 2Cl 2)=1.87 cm 2 /Vs, 

K 0(C 2HCl 3)=1.86 cm 2 /Vs. 

However, the intensity 

distributions of the ions with 

regard to their drift time show 

significantly different 

variations characteristic of 

each substance. 

Another parameter which is 

decisive for the intensity 

distribution of the ions formed by partial 

discharges is the discharge voltage applied 

between the filament and the discharge grid. 

This dependence is shown in Fig. 7 for 

perfluorohexane (C 6F 14) as an example of a 

saturated fluorocarbon, the class of substances 

showing the highest ionization energy of the 

halogenated hydrocarbons considered. Here 

the discharge voltage was varied between 4.0 

and 4.8 kV, all the other parameters such as 

the gap of 4.7 mm between the electrodes or 

the carrier gas flow were kept constant. 

Throughout the whole range of discharge 

voltages ions of at least 11 different mobilities 

were formed. The mobilities of the negative 

ions formed in perfluorohexane as well as their 

intensities as a function of the discharge 

voltage are summarized in Table 3. The total 

relative uncertainty of the mobilities was 

calculated to be ± 1.5 %. At low values of the 

discharge voltage small ions with high 

Perfluorohexane 

UNTITLED - C6F14,Entlspan,select,2D 13.02.01 11:26 

Discharge 

Voltage / kV 

4.0 

4.2 

4.4 

4.6 

4.8 

Negative Ions 

10 15 20 25 


Figure 7: 

Perfluorohexane: Dependence of the Ion Distribution in 

Nitrogen on the Discharge Voltage in the Ionization Region 

of the IMS at constant Drift Voltage


Table 2: 

Reduced Mobilities K 0 of the Negative Ions of C 2H 2Cl 2, C 2HCl 3 and C 2Cl 4 

C 2H 2Cl 2 

K 0 / cm 2 /Vs 

2.03 

1.87 

1.59 

1.48 

Intensity / % 

31 

100 

57 

70 

C 2HCl 3 

K 0 / cm 2 /Vs 

1.86 

1.72 

1.60 

1.39 

Intensity / % 

46 

46 

100 

70 

C 2Cl 4 

K 0 / cm 2 /Vs 

2.29 

2.11 

1.99 

Intensity / % 

100 

48 

34 

mobilities dominate whereas with increasing 

discharge voltages more and more larger ions 

having lower mobilities are generated. The 

occurrence of ions with medium mobilities is 

less significantly influenced. The intensity of an 

ion with K 0 ≈ 1.60 cm 2 /Vs, for example, varies 

between 100 % rel. intensity at a discharge 

voltage of -4.0 kV and 11 % at -4.8 kV, 

whereas the intensity of the ion with 

K 0 ≈ 1.33 cm 2 /Vs stays with rel. intensities 

between 37 and 47 %. Within the total 

experimental uncertainty calculated, 12 of the 

ions of the same K 0-value occur at more than 

three discharge voltages applied, only differing 

in their relative intensities. 

CONCLUSION 

A partial discharge ion mobility spectrometer 

can be advantageously used for the sensitive 

(ppb v-range) detection of fluorinated and 

chlorinated hydrocarbons. Even substances 

having ionization energies higher than 10.6 eV 

can be measured. Thus not only unsaturated 

but also saturated halogenated hydrocarbons 

can be detected. Using a filament made from 

nickel, the ionization source has been shown to 

work with good stability over a period of 4 hours 

and more. Some of the chlorohydrocarbons 

examined form ions of the same mobility but of 

different ion intensity distributions. Other ions 

formed are of unequivocal mobility and 

characteristic of the analyte introduced. When 

applying partial discharges to ionization of 

analytes the discharge voltage between the 

needle and the discharge grid has to be 

controlled very carefully as the ion intensity 

distribution strongly depends on this parameter. 

Acknowledgments 

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] Bruner, F., Quantitative analysis of trace organic 

compounds in the atmosphere: determination of 

halocarbons Comm. Eur. Communities, [Rep.] EUR 

(1982), EUR 7137, Appl. Mass Spectrom. Trace 

Anal., 203-14 CODEN: CECED9 

[2] Lepine, L. and Archambault, J.F., Parts-per-Trillion 

Determination of Trihalomethanes in Water by 

Purge-and-Trap Gas Chromatography with Electron 

Capture Detection, Anal. Chem. 64, 810-814, (1992) 

[3] Galdiga, C.U. and Greibokk, T., Simultanous 

Determination of Trace Amounts of Sulphur 

Hexafluoride and Cyclic Perfluorocarbons in 

Reservoir Samples by Gas Chromatography, 

Chromatographia, 46, 440-443, 1997 

[4] Elkins, J.W., Chlorofluorocarbons, in: The Chapman 

& Hall Encyclopedia of Environmental Science, 

Alexander, D.E. and Fairbridge, R.W. (Ed.), pp. 

78-80, Kluwer Academic, Boston, MA, 1999 

[5] Encyclopédie des gaz, L‘ air Liquide (Ed.), p. 60, 

Amsterdam, Elsevier, 1976 

[6] Riess, J.G. and Le Blanc, M., Perfluor-Verbindungen 

als Blutersatzmittel, Angew. Chem. 90, 654-668, 

1978 

[7] Molina, M.J. and Rowland, F.S., Stratospheric sink for 

chlorofluoromethanes: Chlorine atom catalyzed 

destruction of ozone, Nature 249, 810-814, 1974 

[8] Singh, O.N., The Hole in the Ozone Layer – a Much 

Discussed Phenomenon of this Decade, 

Naturwissenschaften 75, 191-193, 1988 

[9] Baumbach, J.I. and Eiceman, G.A.: Ion Mobility 

Spectrometry: Arriving On Site and Moving beyond a 

Low Profile, Appl. Spectr. 53, 338A-355A, 1999 

[10] Sielemann S.; Baumbach J.I. ; Soppart O. ; Klockow 

D.: UV-Ionenbeweglichkeits-spektrometer - 

Alternative zu den herkömmlichen 



Table 3: 

Reduced Mobilities K 0 of the Negative Ions of C 6F 14 as a Function of the Discharge Voltage 

Discharge 

Voltage / kV 

-4.0 

K 0 / 

cm 2 /Vs 

1.93 

1.79 

1.60 

1.45 

1.33 

1.24 

1.14 

1.09 

1.05 

1.00 

0.95 

Intensity 

/ % 

42 

74 

100 

47 

46 

67 

73 

21 

22 

11 

9 

-4.2 

K 0 / 

cm 2 /Vs 

1.94 

1.78 

1.60 

1.46 

1.33 

1.24 

1.13 

1.05 

1.01 

0.95 

0.90 

Intensity 

/ % 

7 

28 

85 

39 

40 

57 

100 

45 

42 

43 

39 

-4.4 

K 0 / 

cm 2 /Vs 

1.76 

1.58 

1.44 

1.33 

1.24 

1.17 

1.11 

1.04 

1.00 

0.95 

0.91 

0.86 

Intensity 

/ % 

10 

42 

31 

47 

59 

49 

100 

81 

81 

78 

72 

26 

-4.6 

K 0 / 

cm 2 /Vs 

1.77 

1.60 

1.46 

1.34 

1.25 

1.17 

1.11 

1.05 

1.00 

0.95 

0.91 

0.86 

Intensity 

/ % 

5 

16 

17 

43 

45 

55 

83 

88 

100 

100 

78 

19 

-4.8 

K 0 / 

cm 2 /Vs 

1.61 

1.47 

1.39 

1.32 

1.25 

1.18 

1.12 

1.06 

1.01 

0.97 

0.93 

0.90 

0.86 

Intensity 

/ % 

11 

15 

48 

37 

54 

61 

80 

100 

99 

98 

86 

55 

28 

63Ni-Ionenbeweglichkeitsspektrometern für die 

kontinuierliche Detektion halogenierter Alkene im 

ppbv-Bereich. GDCh-Umwelttagung, Karlsruhe, 

27.9.-1.10.1998 

[11] Leasure, C.S.; Fleischer, M.E.; Anderson, G.K. and 

Eiceman, G.A., Photoionization in air with ion mobility 

spectrometry using a hydrogen discharge lamp, Anal. 

Chem. 58, 2142-2147, 1986 

[12] Lubman, D.M. and Kronick, M.N., 

Resonance-enhanced two-photon ionization 

spectroscopy in plasma chromatography, Anal. 

Chem. 55, 1486-1492, 1983 

[13] Roch, Th. and Baumbach, J.I., Laser-based ion 

mobility spectrometry as analytical tool for soil 

analysis, Intern. J. Ion Mobility Spectrometry 1, 

43-47, 1998 

[14] Bradshaw, R.F.D., UK Patent 1,606,926, 1978 

[15] Shumate, Chr.B. and Hill, H.H., Coronaspray 

nebulization and ionization of liquid samples for ion 

mobility spectrometry, Anal. Chem. 61, 601-606, 

1989 

[16] Baumbach, J.I.; Irmer, A. v.; Klockow, D.; Alberti 

Segundo, S.M.; Sielemann, St.; Soppart, O. and 

Trindade, E., Characterisation of SF6 Decomposition 

Products Caused by Discharges in Switchgears using 

Ion Mobility Spectrometry, 4th International 

Workshop on Ion Mobility Spectrometry, Cambridge, 

England, 1995 

[17] Adler, J.; Arnold, G.; Döring, H.-R.; Starrock, V. and 

Wülfing, E., First Results with the Bruker Saxonia 

Corona Discharge IMS in „Recent Developments in 

Ion Mobility Spectrometry: Proc. of the 6th Intern. 

Workshop on Ion Mobility Spectrometry, Baumbach, 

J.I. and Stach, J., (Ed.), Bastei, Germany, 1997 


A micro-machined ion mobility spectrometer-mass spectrometer 

G.A. Eiceman* 1 , E.G. Nazarov 1 , and R.A. Miller 2 

1 

Department of Chemistry and Biochemistry, New Mexico State University, Las Cruces, NM 88003, USA 

2 

The Charles Stark Draper Laboratory, Inc.Cambridge, MA 02139-3563, USA 

ABSTRACT 

An ion filter for APCI mass spectrometry was 

constructed using a micro-machined radio 

frequency (RF) ion mobility spectrometer (IMS) 

operating at ambient pressure. The ion filter 

was positioned between the ion source and the 

flange of the mass spectrometer and allowed 

pre-separation of ions before mass 

spectrometry measurements. The 

micro-machined RF-IMS drift tube was 

fabricated from two glass plates separated by 

0.5 mm thick silicon strips providing a drift tube 

with dimensions of 30 mm long X 10 mm wide 

X 2 mm thick. Ions are swept, using a clean 

gas flow of about 2 liters/min nitrogen or air, 

through the drift tube containing metal 

electrodes deposited on the glass plates. 

During passage in the drift region, ions enter an 

RF field created using a 2 MHz asymmetric 

waveform with high (20,000 V/cm) and low 

(-1000 V/cm) electric fields; a superimposed 

DC electric field of –400 to +400 V/cm can be 

adjusted to select and pass ions through the 

filter and into the mass spectrometer. In this 

miniature design, ions that pass through the 

filter are deflected into a pinhole inlet on the 

vacuum flange of a mass spectrometer. Binary 

mixtures of volatile organic compounds were 

used to demonstrated continuous monitoring 

with a photo-discharge lamp ion source and ion 

pre-separation before mass analysis. 

KEY WORDS 

Ion filter, ambient pressure, radio-frequency ion 

mobility spectrometry, high field asymmetric 

waveform 

INTRODUCTION 

The revolution in biological mass spectrometry 

has required pre-fractionation of complex 

mixtures so mass spectra can be reliably 

interpreted and liquid chromatography has 

been successful as an inlet for mass 

spectrometry [1]. Despite the value of preseparation, 

liquid chromatographic methods 

can be time consuming and can limit sample 

throughput. An alternative could be preseparation 

of ions and a field asymmetric 

mobility spectrometer was successfully used as 

an ion filter for mass spectrometry [2-5]. This 

combination of mobility spectrometry with mass 

spectrometery (IMS/MS) decreased background 

chemical noise in mass spectra and 

simplified spectral interpretation for peptide 

measurements [6]. Since ion separation in IMS 

can occur in milliseconds to seconds, IMS/MS 

could improve sample throughput liquid 

chromatography. 

Over the past decade, analytical devices based 

on RF-IMS methods have undergone practical 

and technical advances [6-10]. In these 

analyzers, ions are carried between parallel 

plate electrodes using a flow of gas. The ions 

are exposed to an oscillating electric field of low 

and high strength so differences in mobilities 

result in a displacement of ions toward one of 

the electrodes. Another comparatively weak 

DC electric field can be superimposed on the 

RF field and can, at an appropriate magnitude, 

draw an ion into the center of the drift region to 

pass to the detector. The magnitude of the 

electric fields needed to control ion motion are 

dependent upon drift tube dimensions but 

representative values are ~25,000 V/cm for the 

high field and about -1000 V/cm for the low 

field. The DC field, also termed the 

compensation voltage is typically –400 to +400 

V/cm. The drift tube needed for these 

measurements is simple in design and 

operation, in contrast to the more conventional 

Received for review December 2, 2000, Accepted December 15, 2000 


G.A. Eiceman et al.: „A micro-machined ion...”, IJIMS 3(2000)1,15-27, p. 16 

time-of-flight drift tubes used in IMS. The 

design of high field IMS (also known as RF IMS 

for fields at radio frequencies) lacks ion 

shutters, an aperture grid or an assembly of 

rings to establish an electric field as found in 

conventional ion mobility spectrometer [11]. 

Consequently, an IMS/MS based upon this 

technology could be rugged, low-maintaniance, 

and economical. One improvement might be 

futher reduction in the size of the drift tube. 

Until recently, miniaturization of IMS drift tubes 

of any design had been considered 

unpromising since conventional ion injection 

schemes lead to losses in resolution with small 

drift tubes and unacceptably high noise created 

through micro-phonics in the aperture 

grid-detector. This noise obscures the ultra low 

signal levels found in drift tubes of reduced 

dimensions. In spite of these barriers, a few 

small drift tubes of conventional design have 

been successfully demonstrated on the 

miniature scale by several teams (US Army 

[12], Oak Ridge National Laboratory [13], and a 

team in Germany [14]). Recently, a small 

micro-machined drift tube for ion mobility 

spectrometry was fabricated using methods 

that are amenable to large scale and 

inexpensive manufacture [15]. In addition, the 

micro-machined drift tube is compatible with a 

range of atmospheric pressure ion sources 

providing an ion pre-filter of general utility for 

APCI-mass spectrometry. A particular 

attraction of the micro-machined drift tube was 

the planar configuration that would easily match 

the flange of a mass spectrometer. Moreover, 

the size and convenience of the small drift tube 

would reduce both cost and complexity of an 

IMS/MS instrument over conventional scale 

IMS analyzers. 

In this work, the combination of a micro-planar 

RF IMS with a mass spectrometer is described 

and some properties of ion filtering are 

demonstrated using binary mixtures of volatile 

organic compounds. The objective of this work 

was to merge the concept of ion filtering with 

comparatively inexpensive technology and 

these findings provide a first measure of the 

feasiblity and utility of the concept of micro 

IMS/MS. 

PRINCIPLES OF ION FILTERING 

The mobility coefficient of an ion, K, is 

independent of electric field until E/N exceeds 

40 Towsend (T d) and above this limit, the 

mobility becomes field dependent in a 

non-linear fashion [16], i.e. K can be 

represented as K E. The field dependence of K E 

is found to depend on even powers of E/N as 

shown in Figure 1A and given in Equation 1: 

K E=K*[1 + α 1(E/N) 2 + α 2(E/N) 4 +……] (1) 

and this may be simplified or approximated to 

Equation 2: 

K E ≈ K*[1 + α(E)] (2) 

where α(E) is a function of the electric field 

dependence of K versus E. According to this 

expression, K E increases with E when α(E)>0 

and decreases with increasing E when α(E)


ELECTRIC FIELD (V/cm) 

Mobility (arbitrary units) 

A 

1.8 

α>0 

α≈0 

1.7 

α0 

Gas Flow 

α0, 

α(E)


Ionization 

Source 

Ion Filter 

Plates 

Deflector 

Electrode 

+ 

+ + + + 

+ 

+ 

+ + 

+ + + 

+ 

+ 

Faraday Plate 

+ 

+ 

+ 

Plenum Gas 

Mass 

Spectrometer 

Figure 2: 

Schematic of micro-machined drift tube as an ion filter for atmospheric pressure chemical 

ionization mass spectrometer. Not to scale. 

be attained by applying a constant specific 

compensation voltage to the drift tube so that 

only ions with certain ∆K arrive at the end of the 

drift region. All other ions undergo annihilation 

through wall collisions. In another method of 

operation, this compensation voltage can be 

continuously changed through a range to 

create a profile of all ions in the analyzer. 

At the center of operation of an RF-IMS is ∆K 

and the interaction between the high field and 

ion structure. In many instances, K is 

increased with increased E and some rationale 

must exist to describe this dependence. 

Momentum and energy balance of ions in 

electric fields demonstrates that mobility 

depends [17] on energy ε=3/2 kT eff per 

Equation 3 where T eff is the effective 

temperature of ions 

ν 

K = = 

E 

q 

1 

( 

N 

3µ 

kT 

eff 

) 

1 / 

2 

1 

Ω ( T 

(3) 

If the effective cross-section for an ion-neutral 

collision (Ω ) is fixed for rigid sphere 

interactions and reduced mass µ is constant, 

the value of K should decreased with increasing 

T eff. Thus, increasing the electric field (i.e. 

eff 

) 

heating the ion so T eff increases) should result 

only in a decrease in K and only α(E)0 is commonly exhibited by 

poly-atomic organic ions. Though Ω is the only 

term available to explain high field dependence, 

there is no comprehensive model, at this 

writing, for the high field dependence where 

α(E)>0. Nonetheless, the formulas suggest 

that ion cross sections undergo decreases 

under high fields and this might be attributed to 

conformational changes or more likely 

ion-solvent declustering. 

EXPERIMENTAL 

Instrumentation 

Micro-machined Drift tube/Tandem Mass 

Spectrometer- A schematic of the 

micromachined drift tube is shown in Figure 2 

and consists of an photo-ionization source and 

the ion filter with a detector electrode and a 

deflector electrode. The drift tube is fabricated 

from two Pyrex wafers and one heavily boron 

doped silicon wafer as already described in 

detail [15]. The drift tube was attached to a 

Teflon base that could be attached to the 

flange of a TAGA 6000 APCI-tandem mass 

spectrometer (MS/MS) from Sciex, Inc. 

(Toronto, Ontario, Canada) as shown in Figure 



Figure 3: 

Photograph of filter in Teflon support on flange of mass spectrometer. 

3. The MS/MS was equipped with an Apple 

PowerMac 7100/66 computer and API 

Standard Software, Ver 2.5.1 (PE SCIEX). A 

detailed description of the TAGA triple 

quadrupole mass spectrometer has been given 

[18] and operating conditions are given in Table 

1. Ions were injected from the drift tube, 

through a hole in the detector electrode and the 

mass spectrometer flange into the pinhole of 

the interface plate of the MS/MS. In these 

studies, sample was ionized using 

photo-ionization with a discharge lamp. A 

miniature 10.6 eV (λ=116.5 nm, EG&G) 

photo-discharge lamp was positioned above the 

hole in the top Pyrex wafer of the spectrometer 

to permit the photons to enter the drift tube. 

The bottom ion filter electrode was grounded 

while the high voltage asymmetric field and 

compensation voltages were applied to the 

L5 (V) 

Table 1: 

Dimensions and Operating Parameters of APCI-Triple 

Quadrupole Mass Spectrometer 

Discharge Plate Potential (kV) 

Orifice (OR) Potential (V) 

L2 (V) 

L3 (V) 

L4 (V) 

0.6 

-60 

-53 

-45 

-40 

250 


L6 (V) 

R1 or Q1(V) 

R2 or Q2(V) 

R3 or Q3(V) 

Base vacuum (torr) 

upper filter plate. An 

amplifier was used to 

detect the charge 

collected on the 

detector plate. A 

polynomial waveform 

synthesizer model 

2920 was used to 

sweep 

the 

compensation voltage 

and a Tektronics 

model TDS340A 

storage oscilloscope 

was used to record the 

signal from the 

detector amplifier and 

produce an RF-IMS 

spectrum for the 

sample gas. 

A radio frequency (2 

MHz) asymmetric field 

is required to filter the 

ions in the 0.5 mm gap 

drift region and is 

produced by a 

soft-switched resonant 

circuit that incorporates a fly back transformer 

to generate the high voltage pulses. The circuit 

provides a peak-to-peak voltage of about 1400 

volts at a frequency of 2 MHz with a duty cycle 

of ~ 30%. The electronics included an isolation 

transformer in to permit all the analyzer 

electronics to float at +100V dc. 

A clean air generator (Model 737, Addco Corp, 

Miami, FL) was used to generate sample 

vapors and provide a flow to the drift tube. The 

gas was scrubbing through beds of activated 

charcoal and 13X molecular sieve providing 

moisture of ~0.5 to 1 ppm. The purified air was 

split into three independently controlled parts 

with 1 to 4 l/min used as drift tube flow, 10-300 

ml/min used to operate a diffusion-based vapor 

generator, and 0-300 ml/min used as a dilution 

gas flow with the sample flow. In experiments 

with mixtures of chemicals, 

-40 

-40 

0 

-40 

3 X 10 -6 

the dilution gas line was used 

also for preparing and 

introducing an additional 

vapor sample. All supply 

lines after the vapor 

generator were kept at 

~50°C to minimize 

wall-adsorption. This flow 

system allowed a constant 

flow to be delivered to the


drift tube while permitting changes in vapor 

concentration by adjusting the ratio of sample 

to dilution flows. Chemicals- Chemicals used in 

this work were acetone, benzene, toluene, 

xylenes and lutidines. All chemicals were 

purchased from Aldrich Chemical Co. (St. 

Louis, MO) and used without further 

purification. 

Procedures 

General Procedures- In preliminary studies, 

vapors of individual compounds in air were 

presented to the drift tube and scans were 

made for the compensation voltage (V dc) with 

the asymmetric waveform (V rf ) supplied to the 

drift tube. Concentrations were 0.1-1 ppm and 

scans of V dc provided a measure of ion 

behavior in a region where ion neutral 

interactions did not appreciably affect peak 

maxima with variations in concentration. Scans 

were obtained for each chemical using a 

Faraday plate at virtual ground potential as the 

detector. A scan occurred at a rate of 0.5 V per 

second between +10 to –10 V. When a 

characteristic compensation voltage for a 

chemical was determined, electronic control 

was switched to allow confirmation of the 

optimum value for V dc by manually tuning. 

During this step, the effectiveness of the filter in 

blocking ions at other settings of V dc was 

measured. In order to check ion yield in the 

filter, V dc and V rf were set to zero and the ion 

current was monitored with the Faraday plate. 

Obtaining Mass Spectra using the IMS Ion 

Filter- The ion filter was operated in three 

electrical configurations with the mass 

spectrometer. In all of these, the entire 

apparatus for the micro-machined drift tube 

including control and Faraday plate detector 

electronics were floated to +100 V as the flange 

of the mass spectrometer was +80 V. In one 

configuration, V rf and V dc were switched off and 

all ions were passed to the mass spectrometer 

via the drift gas. In a second configuration, the 

V rf was applied to the drift tube while V dc was 0. 

These were controls to ascertain both the 

presence of ions and the total filtering of ions 

by the micro-machined drift tube. Finally, V dc 

was tuned for certain peaks, established earlier 

using a Faraday plate detector on the drift tube, 

with the V rf on the drift tube. In all instances, 

the mass spectrometer was scanned from 15 to 

300 amu and 500 scans were averaged for a 

single mass spectrum. 

RESULTS AND DISCUSSION 

Ion characterization and resolution in a 

micro-machined drift tube 

The micro-machined drift tube exhibited 

behavior consistent with findings from drift tube 

of dimensions larger than those used and 

operating under comparable electric fields 

[4-10]. When ions for acetone are carried 

through the drift tube by the gas flow alone, in 

the absence of a RF high field, there is no need 

for a compensation voltage to maintain ion 

passage and ion losses through collisions on 

walls are minor. A peak for these ions appears 

at ~0 V dc as shown in Figure 4 where a the 

peak profile was obtained by sweeping the 

super-imposed DC voltage from –1 to +1 V dc (or 

fields of –20 to +20 V/cm). Ion passage in this 

drift tube at 2-3 liters/min occurred in ~3 ms so 

the time for response to the acetone vapor was 

comparable to that for conventional or 

time-of-flight mobility spectrometers [11]. 

Response is rapid by comparison to 

chromatographic methods [1]. When an 

asymmetric RF electric field of 2 MHz with V rf 

alternating between +1000 V (20,000 V/cm) 

and –300 V (-6000 V/cm) was applied to the 

drift tube, a peak for ions from acetone 

appeared at –4.8 V dc. The magnitude of V dc 

(also known as the compensation voltage) is 

governed by differences in mobility coefficients 

for the ion between low and high electric fields 

and the position can be altered by V rf and ion 

residence time in the drift tube. The peak width 

under the above conditions was ~0.5 V dc at full 

width half-maximum. 

So long as vapor levels were below 1 ppm, the 

peak for acetone appeared at characteristic 

values for V dc and the peak position was 

independent of vapor levels for acetone in the 

micro-drift tube. This was true also for other 

volatile organic compounds (VOCs) such as 

benzene and alkylated benzenes as shown in 

Figure 5. However, when vapor levels were 

greater than 1 ppm, ion molecule clustering in 

the drift region influenced both peak shape and 

peak position. This is clearly an unwelcome 

feature and has been seen by others with 

conventional sized drift tubes. A solution to this 

phenomenon has been proven earlier and the 

effect can be reduced or removed by isolating 

ion separation from any residual sample vapors 

from ion formation. In a next generation design 

of the drift tube, ions will be deflected into the 

drift flow from an ionization region. 

Nonetheless, instrument behavior was 



ION INTENSITY, Volts, a.u 

-7 -5 -3 -1 1 3 5 7 9 

COMPENSATION VOLTAGE 

Figure 4: 

Scans of V dc for peak from acetone vapors in 

micro-machined drift tube with and without V rf applied 

to the drift tube. The peak appears at V dc = 0 when 

V rf is off. When V rf is applied to the filter, a V dc of –4 V 

(under these conditions for V rf) are needed to pass 

ions through the filter. The value for V dc can be 

changed through variations in asymmetric waveform 

and drift tube parameters. In this experiment, V rf was 

1.1 kV. Flow rate was 2 liters/min. 

predictable and reproducible so long as 

concentrations were kept below 1 ppm with this 

design. 

The peak for acetone was used as a reference 

point in these studies and ions for acetone 

were separated from those of benzene and 

alkylated benzenes in binary mixtures. In the 

spectra for Figure 5, the V rf was 900 V and the 

peak for ions of acetone appeared at –2.5 V dc. 

Other VOCs and compensation potentials (V dc) 

at V rf of 900 V were benzene, -7.0; toluene, 

-5.0; m-xylene –3.2; o-xylene, -4.0; p-xylene, 

-4.1; 2,6-lutidine, +0.2; and 2,4-lutidine, +0.4. 

The peak shape for all these spectra was 

nearly symmetrical and with FWHM of ~0.5 V dc. 

The peak shape or ratio of V dc/∆V dc in this unit 

was calculated as 5-20 and is comparable to 

that of larger drift tubes [2-5]. Consequently, 

the reduction in dimensions and use of parallel 

plates has provided performance comparable 

to larger instruments build by conventional 

methods. In summary, the separation of peaks 

for acetone and other VOCs is complete for 

benzene, toluene, and two isomers of xylene 

(the identities of the ions will be described 

below). As noted above, little may be stated 

about the relationship between V dc and ion 

structure pending the development of a 

comprehensive model for operation of these 

RF-IMS analyzers. Nonetheless, RF-IMS has 

been successfully applied in biomedical and 

environmental uses. In the studies below, the 

separation between acetone and toluene 

(second spectrum from bottom in Figure 5) will 

be used as a measure of the micromachined 

unit to serve as a pre-filter to a mass 

spectrometer. 

Micro-machined drift tube with mass 

spectrometer 

Mass spectra obtained from the 

micro-machined drift tube are shown in Figure 

6A for ions from acetone moving through the 

drift tube at V dc=0 with RF voltage off. The 

mass spectrum exhibited a protonated 

monomer (m/z 59), a proton bound dimer (m/z 

117) and a proton bound trimer (m/z 176). In 

addition, hydrates of these ions are apparent at 

low intensity, e.g. MH + *H 2O at m/z 87. The 

presence of MH + , M 2H + and M 3H + in the gas 

stream was not surprising considering the 

vapor levels and ambient temperature. 

Moreover, no special efforts were made to 

scrub the air below ~1 ppm moisture. The 

findings are consistent with this design, where 

ions and neutrals of sample pass through the 

filter together and where a rapid equilibrium as 

shown in Equation 4 occurs in the drift tube. 

k f 

MH + + M M 2H + 

k r 

(4) 

As demonstrated for mobilities of ion-molecule 

clusters in conventional IMS, when k f and k r are 

fast compared to ion drift through a drift tube, 

the ion equilibrium occurs in an ion swarm 

without peak broadening or distortion that might 

reveal an ion mixture [19]. As a result, the 

peak corresponding to acetone shown in Figure 

4 is not a single ion but rather an ion mixture. 

An alternate interpretation of this spectrum is 

that a single ion, presumably M 3H + , is present 

in the micromachined mobility spectrometer 

and that this ion undergoes fragmentation in 

the atmospheric pressure to high vacuum 

region (the reverse reaction is given in 

Equation 5). Fragmentation reactions in the 

k f 

M 2H + + M M 3H + 

k r 

(5) 

expansion zone of MS/MS instruments have 

been observed and lens potentials have been 

used to add an early stage for collision induced 

dissociation in comparable instruments [20] . 

This explanation is consistent with a very easily 

dissociated proton bound trimer. Finally, as 



ION INTENSITY, Volts, a.u 

Acetone 

-10 -8 -6 -4 -2 0 2 

COMPENSATION VOLTAGE 

Figure 5: 

Scans of V dc for individual and mixtures of volatile 

organic compounds. Vapors included: A. 2,4 lutidine, B. 

2,6 lutidine, C. acetone with p-xylene, D. acetone with 

o-xylene, E. acetone with m-xylene, F. acetone, G. 

acetone with toluene, and H. acetone with benzene. V rf 

was 0.9 kV and gas flow in the drift tube was 2.5 

liters/min. 

acetone vapor concentrations is raised above 1 

ppm, the peak position gradually drifts in V dc as 

might be expected when ion clusters are 

formed. In summary, the bulk of evidence 

suggests but does not conclusively 

demonstrate that the peak for acetone is a 

mixture of protonated monomer, proton bound 

dimer and proton bound trimer (at trace levels). 

The somewhat unexpected facet to this work 

was the appearance of a proton based ion 

chemistry with a photoionization detector. 

Photoionization of acetone should have 

resulted in an M + . and instead the product ion 

was MH + . This suggests that ionization of 

acetone is occurring in air at ambient pressure 

through a yet unidentified intermediate reaction 

involving proton transfer. 

When high voltage is applied to the drift tube 

and V dc maintained at 0V, all the ions for 

acetone were blocked by the filter; that is, the 

ions were annihilated on the walls of the drift 

tube. Consequently, no ions were passed to 

the mass spectrometer as shown in Figure 6B 

and only two noise spikes were recorded (note 

abundance values). The position of the peak 

with the RF field applied was located at –2.5 V dc 

and when the compensation voltage was tuned 

to this voltage, ions from acetone passed the 

filter and entered the mass spectrometer as 

shown in Figure 6C. Under these conditions, 

ions were continuously passed to the mass 

spectrometer. At other values of V dc (including 

–15V and -3.5V), ions were not observed. The 


A 

B 

C 

D 

E 

F 

G 

H 

increase in the relative intensity of the MH + 

ion in Figure 6C versus that without the RF 

field (Figure 6A) suggests that the 

equilibrium in Equation 4 was altered slightly 

by the strength of the high RF field 

environment or that some ion speciation was 

occurring. For example, in another region of 

the spectrum at –0.22 V dc (not shown), the 

proton bound dimer was the dominant ion 

species and the protonated monomer had 

comparatively low intensity. 

The mass spectrum for ions of toluene from 

the micro-machined ion filter without RF 

fields and V dc=0 is shown in Figure 7A. The 

mass spectrum showed that the bulk of 

intensity was found with a monomer ion (M +. 

at m/z=92) and that there were small but 

measurable amounts of dimer ions (m/z 

184). A trimer ion was not observed. There 

is a small amount of m/z 106 (M + *H 2O) ions. 

When RF fields are applied to the drift tube 

(with V dc=0), the micromachined drift tube 

behaves as an ion filter and the ions 

annihilated on the drift tube walls (Figure 7B). 

As the compensation voltage is swept negative, 

ions do not appear until V dc reaches –4.0 V. 

This corresponded to the V dc necessary to pass 

ions for toluene through the filter to the mass 

spectrometer where the M +. is readily apparent. 

As the voltage is taken further negative, ions 

are once again neutralized in the drift tube and 

do not reach the mass spectrometer. This 

example and that for acetone demonstrate that 

the micromachined drift tube function as ion 

filters at ambient pressure and that the analyzer 

operated successfully in conjunction with the 

mass spectrometer. 

Ion pre-filter for APCI mass spectrometry 

The test of an ion filter is its ability to separate 

ions in a mixture on the basis of differences in 

mobility (at high and low electric fields) and to 

isolate and pass selectively the chosen ion to 

the mass spectrometer. The curve in Figure 5 

for toluene and acetone exhibits baseline 

separation for the peaks for each in a mixture 

and was chosen as the test case to 

demonstrate RF-IMS ion filtering in a mixture. 

The mass spectrum for ions in a mixture of 

acetone and toluene vapors without an RF field 

and with V dc=0 is shown in Figure 8A. As 

expected, there is no ion filtering under these 

conditions and ions of acetone and toluene 

both are seen in the mass spectrum as MH + 

and M 2H + for acetone and M +. for toluene.


A 

PERCENT RELATIVE ABUNDANCE, 0-100% 

B 

C 

50 100 150 200 250 

MASS/z 

Figure 6: 

Mass spectra of vapors of acetone A: with V rf = V dc = 0, B. with V rf = on and V dc = 0; and C. 

V rf = on and V dc = -2.5 V. Abundance values were: A: 6.2 x 10 6 , B. 2 x 10 3 , and C. 2.2 x 10 6 . 

V rf was 0.9 kV and gas flow in the drift tube was 2.5 liters/min. 



A 


B 

C 

50 100 150 200 250 

MASS/z 

Figure 7. 

Mass spectra of vapors of toluene: A. with V rf = V dc = 0, B. with V rf = on and V dc = 0; 

and C. V rf = on and V dc = -2.5 V. Abundance values were: A: 3.4 x 10 6 , B. 2 x 10 3 , 

and C. 0.5 x 10 6 . V rf was 0.9 kV and gas flow in the drift tube was 2.5 liters/min. 



A 


B 

C 

50 100 150 200 250 

MASS/z 

Figure 8: 

Mass spectra of mixture of acetone and toluene vapors: A. with V rf = V dc = 0, B. 

with V rf = on and V dc = -2.5; and C. V rf = on and V dc = -6.0 V. Abundance values 

were: A: 5.1 x 10 6 , B. 0.5 x 10 6 , and C. 0.2 x 10 6 . V rf was 0.9 kV and gas flow 

in the drift tube was 2.5 liters/min. 



When the RF field was applied at 0 V dc ions did 

not reach the mass spectrometer as expected 

from results with individual vapors (Figures 6B 

and 7B). When RF fields were applied and the 

compensation voltage was set to that for 

acetone, the mass spectrum showed the ions 

for acetone (MH + and M 2H + ) and no ion for 

toluene. Thus, the toluene was effectively 

filtered by the micromachined drift tube and the 

only ions for acetone were passed to the mass 

spectrometer as shown in Figure 8B. 

In the Figure 8C, the compensation voltage 

was set to that for the peak attributed to 

toluene but the resultant mass spectrum shows 

the presence of peaks for toluene as well as for 

acetone. The presence of these acetone ions 

suggests that the peak for toluene seen in the 

spectrum of Figure 5 is a cluster ions or a 

heterogeneous trimer ion of the kind 

(C 3H 6O) n*H + *C 7H 8 (where n = 1 and 2) as 

shown in Equation 6: 

selectively pass the ions for this substance to a 

mass spectrometer. The drift tube design used 

in these studies allowed ions and neutrals to 

flow together through the drift region and the 

formation of heterogeneous ion clusters was 

observed. Isolation of ion formation and ion 

characterization can be readily accomplished in 

a re-designed generation of drift tubes. The 

size and particularly the cost for this ion filter 

might favor the development of comparable ion 

filters for other ion sources such as 

electrospray ionization. 

ACKNOWLEDGEMENTS 

Support is gratefully acknowledged from the 

university cooperation grants from Charles 

Stark Draper laboratory (award No. 

DL-H-516600) and the FBI through contract no. 

J-FBI-98-111. Technical assistance from John 

Carr in electronics for floating the drift tube 

electronics was welcome. 

(C 3H 6O) n*H + 

+ 

C 7H 8 

===== 

(C 3H 6O) n*H + *C 7H 8 

(6) 

proton bound 

acetone species 

toluene 

heterogeneous ion 

This can be understood through the 

appearance of ions for acetone and may 

suggest that the proton bound dimer is in a 

secondary equilibrium with toluene neutrals. 

This derives from the mixture of ions and 

neutrals in the drift region where close 

proximity of neutrals and ions favors cluster 

formation. Such reactions will be prone to 

matrix effects and unreliable response in 

circumstances were chemically complex vapors 

are passed through the analyzer. In the next 

generation ion filter, ion formation and 

characterization will be isolated so that ion 

characterization can occur in a clean gas 

environment. This has already been 

discovered with conventional sized mobility 

spectrometers of the RF-IMS design and is 

equally valid in the microscale analyzer. 

CONCLUSION 

A micromachined drift tube for an RF based 

method of characterizing ions using differential 

mobilities alternating between low and high 

electric fields has allowed continuous 

monitoring of ions at ambient pressure. 

Moreover, this drift tube can be used to isolate 

the spectral peak of a specific substance, 

volatile organic chemicals in these studies, and 

REFERENCES 

[1] Niessen WMA. J. Chromatogr. A. 1999; 856: 

179. 

[2] Buryakov IA, Krylov EV, Makas AL, Nazarov EG, 

Pervukhin VV, Rasulev UKh. Sov. Tech. Phys. 

Lett. 1991;17: 446. 

[3] Buryakov IA, Krylov EV, Nazarov EG, Rasulev 

UKh. Int. J. Mass Spectrometry and Ion 

Processes 1993; 48: 114. 

[4] Carnahan B, Day S, Kouznetsov V, Tarasov A. 

“Development and Applications of a Transverse 

Field Compensation Ion Mobility Spectrometer” 

3 rd International Workshop on Ion Mobility 

Spectrometry, Cambridge, UK, August 6-9,1995. 

[5] Guevremont R, Purves RW, Day S, Pipich CW, 

Matyjaszczyk MS. Review of Scientific 

Instruments 1998; 69: 4094. 

[6] Carnahan BL, Tarasov AS. United States Patent 

5,420,424,May 30,1995 

[7] Ells B, Froese K, Hrudey SE, Purves RW, 

Guevremont R, Barnett DA. Rapid Comm. Mass 

Spectrom. 2000; 14: 1538. 

[8] Purves RW, Guevremont R. Anal.Chem. 1999; 

71: 2346. 

[9] Guevremont R, Purves RW, J Am Sos Mass 

Spectrom 1999;10: 492. 



[10] Guevremont R, Purves RW, Barnett DA, Ding L. 

Int. J. Mass Spectrometry and Ion Processes 

1999; 193: 45. 

[11] Eiceman GA, Karpas Z. “Ion Mobility 

Spectrometery”, CRC Press, 1994. 

[12] Harden CS. “Relative Performance 

Characteristics of Handheld Ion Mobility 

Spectrometers-The Chemical Agent Monitor and 

a New Miniaturized Instrument”, 5 th International 

Workshop on Ion Mobility Spectrometry, 

Cambridge, UK August 1995 

[13] Xu J, Whitten WB, Ramsey JM. “ Miniature Ion 

Mobility Spectrometry”, 8 th International 

Workshop on Ion Mobility Spectrometry, Buxton, 

UK, Aug. 8-12, 1999. 

[14] Baumbach JI, Berger D, Leonhardt JW, Klockow 

D. Intl. J. Environ. Analy. Chem. 1993; 52: 189. 

[15] Miller RA, Eiceman GA, Nazarov EG. Sensor 

and Actuators B. Chemical, 2000; 67:300. 

[16] McDaniel EW. Mason EA. “The mobility and 

diffusion of ions in gases”, John Wiley & 

Sons,1973. 

[17] Revercomb HE, Mason EA. Anal. Chem. 1975; 

47: 970. 

[18] Eiceman GA, Kremer JH, Snyder AP, Tofferi JK, 

International Journal of Environmental Analytical 

Chemistry 1988; 33: 161. 

[19] Preston JM, Rajadhyax L. Anal. Chem. 1988; 60: 

31. 

[20] Wensing MW, Snyder AP, Harden CS. J. Mass 

Spectrometry 1995; 30: 1539. 


Coupling of Multi-Capillary Columns 

with two Different Types of Ion Mobility Spectrometer 

J.I. Baumbach 1 , S. Sielemann 1 , P. Pilzecker 2 

1 

Institut für Spektrochemie und Angewandte Spektroskopie, Bunsen-Kirchhoff-Str.11, 

D-44139 Dortmund, Germany 

2 

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


Abstract 

Pre-separation using gas chromatographic 

Multi-Capillary Columns is used to overcome 

resolution problems if complex mixtures of 

volatile organic compounds are analyzed by ion 

mobility spectrometry. It is shown that 

inter-molecular charge transfer reactions take 

place and the interpretation of the acquired 

spectra is improved. 

Mixtures of selected volatile organic 

compounds (butanol, ethylmethylketone, 

pentane, propanol, pyridine, tetrachloroethene, 

toluene and trichloroethene) with masses up to 

1 ng are separated in a few seconds at ambient 

temperature. 

Introduction 

Ion mobility spectrometry (IMS) appears to 

have become increasingly popular as an 

analytical instrument during the past two 

decades because of better understanding of 

constructional parameters (drift tube design, 

electric field conditions, shutter opening time), 

instrumentation (membrane inlet system, spray 

injection, coupling to gas chromatographic 

columns) and methods of data handling 

(collection, interpretations, neural networks, 

curve fitting). Nevertheless, ion mobility 

spectrometry still suffers from limitations that 

are especially relevant for field analysis of 

mixtures. Among others concentrationdependent 

response characteristic and limited 

dynamic range should be mentioned. Methods 

to overcome these difficulties are described in 

the literature [1-9] and include, for instance, the 

combination of a gas chromatograph with an 

IMS system [10-20]. 

The operational principle of IMS has already 

been described frequently [3,6-9], therefore only 

a few essentials are briefly outlined here. In 

general, a continuous stream of a gas (air or 

nitrogen), carrying the analytes, passes through 

an ionization source ( 63 Ni ß-radiation, UV-light, 

discharges). The analyte molecules are ionized 

directly, or via ion-molecule reactions (proton 

transfer, electron attachment) with ionized 

carrier gas molecules. Ions formed in the 

ionization region of the IMS are periodically 

introduced via a shutter grid into the drift tube, 

where during their drift under the influence of 

an uniform electric field they collide with neutral 

molecules and separate into ion clouds. 

Through these collisions, ions attain drift 

velocities inversely related to their mass 

(among other properties - for details see 1-9 ), so 

that their collection on a Faraday plate at the 

end of the drift tube delivers time-dependent 

signals corresponding to the mobilities of the 

arriving ions. This ion mobility spectrum 

contains information on the nature of the 

different trace components present in the 

carrier gas. 

Depending on the number of different analytes 

in the carrier gas the IMS spectra may become 

quite complex. For instance using ß-radiation 

for ionization and air containing a small amount 

of water vapor, reactant ions such as (H 2O) nH + 

and (H 2O) nO 2 

- 

will appear, which react with the 

analyte molecules M to form ions like MH + or 

MO 2 

- 

(monomer ions), M 2H + (dimer ions) and 

even mixed species like MH + M´. 

This problem can be overcome by preseparation 

of the different analytes. One way of 

achieving this is the coupling of a gas 

chromatograph with IMS, a concept under 

consideration for the past 20 years, in which 

packed GC columns were soon replaced by 

Received for review January 25, 2000, Accepted July 10, 2000 


J.I. Baumbach et al.: „Coupling of Multi-Capillary Columns with two Different Types ...”, IJIMS 3(2000)1,28-37, p. 29 

capillary columns. IMS with 

63 

Ni-ionization 

sources [10,12-20] and UV-lamps [15] have 

been used. The operation temperature of the 

columns varied between 35 and 100 o 

C 

[10,13-16,18], and 100 and 200 o C [11,14,20], 

and even column temperatures higher than 200 

o 

C are reported [12]. The range of retention 

times was between half a minute [16,18-19], 

and 30 minutes [10-13,15,20] and occasionally 

expanded up to 2 h [17]. The 35 o C seems to be 

the lowest temperature above room 

temperature that can be regulated. 

Uncontrolled, room temperature would subject 

the GC column to undesirable temperature 

fluctuations due to the cycles of the climate 

control system or outdoor environment. A 

uniques report to operate a GC/IMS system at 

room temperature was reported by Leonhardt 

et al. 21 using UV-ionization of benzene, toluene 

and the isomers of xylene and extanding the 

"low" temperature approaches of IMS. 

The goal of this paper is to demonstrate that 

fast gas chromatographic pre-separation with 

Multi-Capillary Columns (MCCs) is ideally 

suited for combination with IMS and can 

effectively reduce the interaction of different 

analytes in a mixture within the ionization 

region. In addition to the elimination of 

competing elements of the analyte mixtures via 

pre-separation a simplification of a complex 

reagent chemistry strategy is reachable, which 

is necessary for any succesful IMS-alone based 

analytical application. Furthermore, the MCC 

delivers analytes directly to the most efficient 

ionization region of the IMS, in the case of 

UV-IMS near the surface of the UV-lamp, which 

is also full established in the literature as well. 

In the case of AVM, a commercial available 

IMS, the coupling was realized without 

changing anything at the instrument. Therefore 

it would be shown, what could be reached by 

adding a pre-separation unit in front of the 

unchanges inlet system of the IMS and whether 

an attachment of MCC should be considered. 

Because of the high carrier gas flow rates MCC 

are more suitable for combintation to IMS than 

for other detectors, used for example in gas 

chromatography. Details of characterisation 

of MCC with respect to chromatographic 

parameters are summurized in [22]. The 

deep minimum of MCC in the Van-Deemter 

curve makes the MCC unique compared 

with standard capillary columns. Therefore, 

especially for field applications, small 

changes in gas flow rates will have a effect 

neclectable. Thus they intend to be more 

compatible to IMS than other columns. 

Experimental 

Ion mobility spectrometer 

For the experiments two different ion mobility 

spectrometer were used, each one connected 

to a MCC and a simple, self-made inlet system. 

One custom made IMS was equipped with an 

10.6 eV UV-lamp, the other one (airborne vapor 

monitor, AVM) includes a 63 Ni-ionization source 

and was borrowed from Graseby Dynamics, 

Watford, UK, for the purpose of inter 

comparison. The main parameters of the 

different IMS systems used are summarized in 

Table 1. 

Table 1: 

Main parameters of the two different IMS systems used 

ISAS custom Graseby AVM 

made IMS 

Ionization source: 

10.6 eV UV-lamp 

63 

Ni (360 MBq) 

Electric field strength: 

375 V/cm 244 V/cm 

Length of the ionization region: 15 mm 

12 mm 

Length of the drift region: 

61.5 mm 

39 mm 

Diameter of the drift region: 

15 mm 

12 mm 

Shutter opening times: 30, 100, 300, 1000 

µs, adjustable 

180 µs 

Shutter delay: 

100 ms 

100 ms 

Drift gas: 

Nitrogen 

(99.999%) 

Air 

Drift gas flow: 

100 - 800 mL/min about 500 mL/min 

Temperature: 

24 o C 

24 o C 

Pressure: 

101 kPa 

101 kPa 

Materials 

All reagents were used as 

received from the 

suppliers: pentane (99 %), 

heptane („UV-Spectroscopy 

grade”), octane 

(puriss. p.a., 99,5 %), 

butanol (puriss. p.a., 99,5 

%), toluene (puriss. 

absolut), o-xylene (99 %), 

pyridine (puriss., p.a., 99,8 

%) all from Fluka, 

Deisenhofen; hexane from 

Fisons, MainzKastel; 

benzene (p.A., 99.7 %) and 

carbon tetrachloride (p.A., 

99.8 %) both from Merck, 



Darmstadt; propanol 

(purris. p.a., 99 %), 

m-xylene („water free”, 

99 %), p-xylene (99 %), 

chloroform (99 %), 1,1,1- 

trichloroethane (99 %) 

and 1,1,2-trichloroethane 

from 

Aldrich, 

Deisenhofen. 

A 1 ml gastight syringe 

was used for the 

injection to obtain 

qualitative impressions. 

Standard mixtures were 

prepared as vapors in a 

dilution apparatus. Micro 

liter amounts of neat 

liquid were injected in a 5 

l glass bottle equipped with a cap that had been 

modified with a Swagelok union. After adding a 

known sample amount of liquid with syringes 

the mixture was left for one hour. 

A sample was injected into the AVM using a 

syringe. For the UV-IMS a continuous and 

constant gas stream passed through the glass 

Ionisation Region 

Sample and MCC Inlet 

Drift Region 

Table 1: 

Main parameters of the two different IMS used 

ISAS custom 

made IMS 

Ionization source: 

10.6 eV UV-lamp 

Electric field strength: 

375 V/cm 

Length of the ionization region: 15 mm 

Length of the drift region: 

61.5 mm 

Diameter of the drift region: 

15 mm 

Shutter opening times: 

30, 100, 300, 

1000 µs 

adjustable 

Shutter delay: 

100 ms 

Drift gas: 

Nitrogen 

(99.999%) 

Drift gas flow: 

100 - 800 mL/min 

Temperature: 

24 o C 

Pressure: 

101 kPa 

Drift-Ring 

Apperture-Grid 

Graseby AVM 

63 

Ni (360 MBq) 

244 V/cm 

12 mm 

39 mm 

12 mm 

180 µs 

100 ms 

Air 

about 500 mL/min 

24 o C 

101 kPa 

bottle and the concentration decrease 

exponentially. The glass bottle was connected 

with the 6-way-valve by a Teflon tube and the 

sample was injected using a 250 µl gas-tight 

loop. 

Multi-Capillary Columns 

The MCCs were obtained from the Institute of 

Applied Physics, Novosibirsk, Russia 

(5%-Methyl-Phenyl-Silicone) and 

Alltech, Wallbronn, Germany (SE-30). 

Their main features are presented in 

Table 2, detailed characteristics of the 

MCC were reported earlier [22]. 

Faraday-Plate 

Gas-Outlet 

Drift Gas-Inlet 

Data-Acquisition 

Connection of MCC and IMS 

The outlet of the MCC was inserted 

directly to the ionization region of the 

IMS. In the case of the AVM the 

Electrode 

UV-Lamp 

Bradbury-Nielsen-Shutter 

Figure 1: 

Overall construction of the ISAS ion mobility spectrometer 

Carrier gas 

inlet 

Einlaß 

Injection 

Multi-capillary 

column (MCC) 

Graseby 63 Ni-IMS (AVM) 

Gas outlet 

Data acquisition 

Figure 2: 

Connection of the Multi-Capillary column to the AVM 



for each substance: 

positive ions: above 

negative ions: below 

Reactant Ion (Air) 

Chloroform 

Carbon Tetrachloride 

1,1,2-Trichloroethane 

0 5 10 15 20 25 


Signal / a.u. 

0.6 

0.4 

0.2 

Figure 3: 

Ion mobility spectra of positive and negative ions 

formed from chloroform, carbon tetrachloride and 

1,1,2-trichloroethane using a AVM (Graseby, 

Watford, UK) 

Reactant Ion Peak 

Chloroform 


1,1,2 Trichloroethane 

Signal / a.u. 

0.6 

0.4 

0.2 

Chloroform 


1,1,2 Trichloroethane 


Superposition of 

single components 

signal divided by 3! 

Mixture 

5.0 5.5 6.0 6.5 


Figure 4: 

Ion mobility spectra of negative ions formed in a mixture of chloroform, carbon tetrachloride and 

1,1,2-trichloroethane (reaction ions (RIP), single components, mixture, superposition of single spectra) 



CHCl 3 

-0,45 

-0,60 

CCl 4 

10 

0 

-0,15 

-0,30 

Signal / V 

6,4 

6,2 

6,0 

5,8 


5,6 

5,4 

5,2 

C 2 

H 3 

Cl 3 30 

50 

60 

70 

40 

20 

Retention Time / s 

6,4 

6,2 

CHCl 3 

205 ng 

RIP 

CHCl 3 

CCl 4 

C 2 

H 3 

Cl 3 


6,0 

5,8 

5,6 



5,4 

5,2 

CCl 4 

109 ng 

C 2 

H 3 

Cl 3 

6 ng 

5,0 

0 10 20 30 40 50 60 70 


Figure 5: 

Ion mobiltity spectra of negative ions formed in chloroform, carbon tetrachloride and 1,1,2-trichlorethane 

using Multi-Capillary column coupled to an AVM ion mobility spectrometer (3D-plot above, peak height 

diagramm below ) 



connection was made to the membrane inlet of 

the instrument to let the instrument unchanged 

themselves. The main construction of the ISAS 

IMS is presented in figure 1. The sample inlet 

or the MCC are fixed near the surface of the 

UV-lamp for highest ionization efficiency. All 

other parts of the IMS are as conventional used 

in ion mobility spectrometry. The connection to 

the AVM was made directly to the nose of the 

membrane inlet system (see figure 2). No total 

sample transfer could be realized, but the effect 

of the use of the pre-separation unit will come 

out clearly. Thus, these detail arising from the 

construction of the AVM as a warning device, is 

not so important for the following investigations. 

The figure is added to make clear, why the total 

amount of the sample would not be able to 

enter the IMS. 

Results and Discussion 

In figure 3 the response of the AVM in the form 

of spectra of chlorofom, carbon tetrachloride 

and 1,1,2-tichloroethane as single components 

within the carrier gas air (reactant ion peak, 

RIP) is shown. Spectra of both positive and 

negative ions can be detected, depending on 

the polarity of the electric field within the drift 

tube. Drift times of the ion swarms between 5 

and 9 ms are typical if a drift tube length of 39 

mm and electric field strength of 240 V/cm is 

used. The drift times for negative ions are 

shorter than for positive ions. 

The response of the same instrument to a 

mixture of chloroform, carbon tetrachloride and 

1,1,2-trichloroethane (205 ng, 109 ng and 6 ng 

respectively) is shown in figure 4. The single 

responses of each component and the reactant 

ion peak arising from air are plotted for the 

case of negative ions observed within the IMS. 

Additionally, the sum of all spectra divided by 3 

is presented to demonstrate, that the response 

of the IMS investigating a gas mixture is not a 

undisturbed superposition of the single spectra. 

The trichloroethane dominates the shape of the 

Table 3: 

Mobility values of different alkanes 

Analyte 

Mobilitiy / cm 2 /Vs 

Pentane 1.45 1.67 

Hexane 1.31 1.51 

Heptane 1.38 1.81 

Octane 1.32 1.67 

spectrum although the amount was 18 times 

lower than the amount of carbon tetrachloride. 

Figure 4 indicates also, that with mobilities of 

the negative ion of 2.53 cm 2 /Vs (reactant ions, 

RIP), 2.72 cm 2 /Vs and 2.44 cm 2 /Vs (chlorofom), 

2.82 cm 2 /Vs (carbon tetrachloride) and 2.77 

cm 2 /Vs (1,1,2-trichloroethane), respectively no 

adequate separation can be achieved for 

mixture analysis. 

Figure 5 (upper part) gives the results of the 

simple coupling of a 21 cm long MCC to the 

AVM applying a carrier gas flow through the 

MCC of 20 ml/min. This is in contrast to 600 

ml/min gas flow within the AVM. The retention 

times are 16.3 s for chloroform, 20.3 s for 

carbon tetrachloride and 56 s for 

1,1,2-trichloroethane, respectively. The 

procedure of production of negative ions via 

ion-molecules reactions from air leads to a 

decrease of the reactant ion peak (RIP) and a 

shift of the peak position corresponding to the 

mobilities of the sample ions produced, 

followed by relaxation of the reactant ion peak. 

Chloroform produces ions with mobilities less 

than the reactant ions and both other molecules 

ions with higher mobilities. The CHCl 3 is not 

readily identificable as unique because of the 

overwhelming concentration used to show the 

total shift of the peak from the reaction ions to 

analyte ions. The identification of the Cl-related 

cluster signals should be realised by use of 

IMS-Time-of-flight-Mass-Spectrometer, which is 

under development for the UV-IMS actually. 

To demonstrate the peak shift in Figure 5 

(below) a peak-height-diagram is presented. 

On the right side the single spectra are shown 

for accentuation of the advantage of 

pre-separation. The RIP and the peaks of the 

substances under consideration are separated 

clearly. Thus, a simple and not optimized 

connection of a MCC to the AVM extends the 

application. 

In the case of UV-ionization the mobilities of the 

ions produced varies often only slightly. Single 

spectra for n-pentane, n-hexane and n-heptane 

ionized with a 10.6 eV UV-lamp are shown in 

figure 6. In all cases two main ions are 

produced. The mobilities are of similar size 

(see table 3). Furthermore, the UV-IMS-spectra 

are broad and the overlap of the single 

UV-spectra would make a deconvolution of the 



Signal / a.u. 

0.10 

0.08 

0.06 

0.04 

0.02 

0.00 

Pentane 

Hexane 

Heptane 

Octane 

5.0 7.5 10.0 12.5 15.0 17.5 20.0 


Figure 6: 

Ion mobility spectra of pentane, hexane, heptane and octane for the single 

substances using the ISAS UV-IMS 

spectrum very difficult. 

At ambient 

temperature the 

application of the 

MCC leads to a total 

separation of the 

alkanes during 3 

minutes (see figure 7). 

The carrier gas flow 

through the 20 cm 

long MCC was 20 

mL/min (in this case). 

Retention times of 22 

s for n-pentane, 34 s 

for n-hexane, 66 s for 

n-heptane and 157 s 

for octane are 

reached. This is 

acceptable for field 

investigations. The 

cluster formation 

because of 

16 

Hexane 

14 

Pentane 

Heptane 

Octane 

12 


10 

8 

6 

1.06 

1.18 

1.15 Sample Amount / µg 1.11 

0 50 100 150 


Figure 7: 

Peak-height-diagramm of a mixture of pentane, hexane, heptane and octane using the ISAS UV-IMS 



Signal / a.u. 

Benzene 

Toluene 

m-Xylene 

6 7 8 9 


Figure 8: 

Spectra of benzene, toluene and m-xylene obtained with an ISAS 

UV-IMS 

concentration as analyte chemistry should not 

be considered here. 

A typical application field for UV-ionization is 

the detection of benzene, toluene and xylenes 

in the environment. Figure 8 shows the spectra 

of the single substances. The mobility values 

are 2.17 cm 2 /Vs, 2.13 cm 2 /Vs and 2.06 cm 2 /Vs, 

respectively. Thus, an overlap of 

the peaks still occurs also in this 

case. Smaller shutter opening 

times should be reachable if 

ionization efficiency could be 

enhanced by further optimization of 

the design of the ionization region 

of the IMS. Because of the narrow 

diameter of the UV-light beam the 

loss of analyte not ionized is to 

high. Further progress is 

attainnable at this subject. The 

coupling of the MCC leads to the 

results presented in figure 9. The 

3-D-plot of these analytes shows 

the total separation within about 

one minute. The retention times at 

ambient temperature are 10 s, 24 s 

and 60 s for benzene, toluene and 

m-xylene, respectively. The 

amounts of analytes are 1.57 µg 

(benzene), 1.58 µg (toluene) and 

1.62 µg (xylene). Figure 10 shows the linear 

range and the minimum detection limit of the 

IMS design 10.6 eV - UV-IMS for Benzene. In 

this case the linear range is about 2 orders of 

magnitude. The detection limits are in the range 

of 60 ng for benzene, toluene and the xylenes. 

0,30 

0,25 

0,20 

Signal / a.u. 

0,15 

0,10 

0,05 

0,00 

Benzene 

10 20 

30 

40 

50 

60 

70 

Figure 9: 

3D-Plot of the IMS-chromatogram for benzene, toluene and m-xylene using an ISAS MCC-UV-IMS 


Toluene 


m-Xylene 

10 

9 


8 

7 

6


Signal Area / a.u. 

10 0 

Benzene 

substance are nearly the same (between 8.9 

ms for ethylmethylketone and 9.34 ms for 

toluene) the MCC coupling allows the 

separation. The sample amounts are 787 ng 

(butanol), 910 ng (ethylmethylketone), 648 ng 

(pentane), 858 ng (propanol), 398 ng (pyrdine), 

504 ng (tetrachloroethene), 261 ng (toluene) 

and 530 ng (trichloroethene). 

10 -1 

100 1000 

Substance Amount / ng 

Figure 10: 

Linear range and minimum detectable limit of the 

response of the ISAS UV-IMS for benzene and 

spectrum of the lowest amount detected 

The possibility to achieve efficient separation of 

mixtures of butanol, ethylmethylketone, 

pentane, propanol, pyridine, tetrachloroethene, 

trichloroethene and toluene by MCC-UV-IMS 

during time intervals of about one minute at 

ambient temperature is shown in figure 11. 

Although the drift times of the ions of the single 

Conclusions and Further Applications 

Coupling of Multi-Capillary Columns with IMS 

leads to an increase of the scope of 

applications. It was shown for two different 

types of ion mobility spectrometers, that simple 

coupling methods can lead to efficient 

separation and operation at ambient pressure 

with retention times acceptable for field 

analysis. Heating of the column and the 

detection system could further reduce the 

retention time but increase the power applied. 

Detailed studies of the linear range and the 

detection limit by 63 Ni and different ionization 

energies of the UV-lamps will support the 

acceptance of ion mobility spectrometry for field 

Signal / a.u. 

0,2 

0,1 

0,0 

Propanol 

Pentane 

Ethylmethylketone 

Butanol 


Pyridine 

Pyridine 

Toluene 



0 25 50 75 100 

Toluene 

Ethylmethylketone 


Butanol 

MCC-UV-IMS 


Propanol 

Pentane 

25 

20 

15 

10 


5 

0 

0 

25 

50 


75 

100 

125 

Figure 11: 

IMS-chromatogram of a mixture of selected VOC´s (butanol, ethylmethylketone, pentane, popanol, 

pyridine, tetrachloroethene, toluene and trichloroethene) 



investigations and risk assessment. The 

inclusion of other types of phases on the 

Multi-Capillary columns and of different length 

would improve the analysis of more complex 

mixtures. 


Some results presented were obtained with the 

financial support of the European Union 

(contract EV5V-CT94-0546). The financial 

support of the Bundesministerium für Bildung, 

Wissenschaft, Forschung und Technologie and 

the Ministerium für Wissenschaft und 

Forschung des Landes Nordrhein-Westfalen 

and the co-operation with the company 

GRASEBY Dynamics, Watford, UK, the 

Institute of Applied Physics, the Institute of 

Catalysis and the company Sibertech, 

Novosibirsk, Russia, are gratefully 

acknowledged. 

References 

[1] H.E. Revercomb, E.A. Mason, „Theory of plasma 

chromatography/ gaseous electrophoresis - a review” 

Anal. Chem. 48, 970-983, (1975). 

[2] E.W. McDaniel, E.A. Mason, „The mobility and 

diffusion of ions in gases”, John Wiley, New York, 

(1973). 

[3] T. Carr, „Plasma Chromatography”, Plenum, New 

York, (1984). 

[4] H.H. Hill, W.F. Siems, R.H. St.Louis, D.G. McKinn, 

„Ion mobility spectrometry” Anal.Chem. 62, 

1201A-1209A, (1990). 

[5] R.H. St.Louis, H.H. Hill, „Ion Mobility Spectrometry in 

Analytical Chemistry” Crit. Rev. Anal. Chem. 21, 

321-355, (1990) 

[6] H.M. Widmer, M.A. Morrissey, „Neochromatographic 

technologies - ion mobility spectrometry” Chimia 43, 

268-277, (1989). 

[7] J.E. Roehl, „Environmental and Process Applications 

for Ion Mobility Spectrometry” Appl.Spectrosc.Rev. 

26, 1-57, (1991). 

[8] G.A. Eiceman, „Advances in Ion Mobility 

Spectrometry” Crit.Rev.Anal.Chem. 22, 17-36, 

(1991). 

[9] J.I. Baumbach, G.A. Eiceman, "Ion mobility 

spectrometry: Arriving on-site and moving beyond a 

low profile" Applied Spectroscopy 53, 338A-355A, 

(1999) 

[10] F.W. Karasek, R.A. Keller, „Gas chromatograph / 

plasma chromatograph interface and ist 

performancein the detection of musk ambrette” J. 

Chrom. Sci. 10, 626-628, (1972). 

[11] S.P. Cram, S.N. Chesler, „Coupling of high speed 

plasma chromatography with gas chromatography” J. 

Chrom. Sci. 11, 391-401, (1973). 

[12] F.W. Karasek, D.W. Denney, „Evaluatiopn of the 

plasma chromatograph as a qualitative detector for 

liquid chromatography” Anal. Letters 6, 993-1004, 

(1973). 

[13] M.A. Baim, H.H. J. Hill, „Tunable selective detection 

for capillary gas chromatography by ion mobility 

monitoring” Anal. Chem. 54, 38-43, (1982). 

[14] M.A. Baim, F.J. Schuetze, J.M Frame, H.H. Hill, „A 

microprocessor-controlled ion mobility spectrometer 

for selective and nonselective detection following gas 

chromatography” Am. Lab. 14, 59-70, (1982). 

[15] M.A. Baim, R.L. Eatherton, H.H. Hill, „Ion mobility 

detector for gas chromatography with direct 

photoionization source” Anal.Chem. 55, 1761-1766, 

(1983). 

[16] A.P. Snyder, Ch.S. Harden, A.H. Brittain, M. Kim, 

N.S. Arnold, H.L.C.Meuzelaar, „Portable hand-held 

gas chromatography / ion mobility spectrometry 

device” Anal. Chem. 65, 299-306, (1993). 

[17] H.H. Hill, W.F. Siems, „Ion mobility detection after 

capillary gas chromatography”, LC-GC 6, 810-814, 

(1988). 

[18] A.P. Snyder, C.S. Harden, A.H. Brittain, M.G. Kim, 

N.S. Arnold, H.L.C. Meuzelaar, „A portable, 

hand-held gas chromatography - ion mobility 

spectrometer” Am. Lab. 24, 32B-32H, (1992). 

[19] H.L.C. Meuzelaar, R. Sip, E. Smaragdis, W.H. 

McClennen, J.P. Dworzanski, „On-line monitoring and 

evaluation of high speed GC/IMS data flows” ERDEC 

Conference on Defense Research (1994). 

[20] G. Simpson, M. Klasmeier, H. Hill, D. Atkinson, G. 

Radolovich, V. Lopez-Avilla, T.L. Jones, „Evaluation 

of gas chromatography coupled with ion mobility 

spectrometry for monitoring vinyl chloride and other 

chlorinated and aromatic compounds in air samples” 

J. High Resol. Chromatogr. 19, 301-312 (1996). 

[21] J.W. Leonhardt, H. Bensch, D. Berger, M. Nolting, J.I. 

Baumbach, „Determination of Benzene, Toluene, and 

Xylene by means of an Ion Mobility Spectrometer 

Device using Photoionization” Proceedings Third 

International Workshop on Ion Mobility Spectrometry, 

Galveston, TX, October 16-19, 1994, J.Cross (Ed.): 

Third International Workshop on Ion Mobility 

Spectrometry, Lyndon B. Johnson Space Center, 

Houston, Report S-799, 49-56, (1995). 

[22] J.I. Baumbach, G.A. Eiceman, D. Klockow, S. 

Sielemann, A. v.Irmer: Exploration of a multicapillary 

column for use in elevated speed chromatography. - 

Int. J. Env. Anal. Chem. 66, 225-240, (1997) 


Detecting heroin in the presence of cocaine using ion mobility spectrometry 

A.M. DeTulleo, P.B. Galat, M.E. Gay 

United States Department of Justice, Drug Enforcement Administration, South Central Laboratory, 1880 Regal Row, 

Dallas, Texas, USA. 

ABSTRACT 

The South Central Laboratory has been using 

the Barringer field portable Ion Mobility 

Spectrometer (IMS) to support federal, state, 

and local drug cases for over seven years. In 

the course of hundreds of field operations, 

difficulty has been encountered in detecting 

heroin when cocaine is present. Cocaine and 

heroin, when present individually, are easily 

detected on the Barringer IMS at 

concentrations as low as 1 nanogram. 

However, depending on relative concentrations 

of mixtures of cocaine and heroin, the heroin 

may not be detected at levels as high as 25 

nanograms. In addition, heroin samples 

routinely contain other opium derivatives, such 

as acetylcodeine and 6-monoacetylmorphine. 

These two compounds exhibit peaks that have 

similar reduced mobilities to cocaine. The 

South Central Laboratory has encountered 

numerous situations where cocaine is initially 

detected. However, when using Gas 

Chromatography-Mass Spectrometry (GC/MS) 

for confirmation, both cocaine and heroin are 

detected. Therefore, the operator must be 

able to recognize this potential problem in law 

enforcement field situations. 

INTRODUCTION 

The Barringer Ion Mobility Spectrometer is a 

valuable tool used in the law enforcement field 

for detecting trace amounts of controlled 

substances. However, this instrumentation is 

merely a field test, does not produce conclusive 

results, and has certain limitations. Ionization 

by IMS depends on factors such as the carrier 

gas matrix, the concentration of the sample, the 

method of sample introduction, and the 

temperature in the drift tube.1 More 

specifically, when there are multiple reactant 

ions, such as a combination of cocaine and 

heroin, there is a competition between the two 

ions. This competitive nature of 

charge-exchange ionization prevents IMS from 

being a suitable technique for some multiple 

compound analyses.2 For example, a previous 

study determined the difficulty IMS has in 

separating and identifying mixtures of 

methamphetamine HCl and nicotine from 

cigarette use.3 In addition, Karasek, Hill, and 

Kim conducted research published in 1976 

identifying the major peaks detected on an IMS 

for heroin and cocaine. However, their IMS 

instrumentation is different than the portable 

instruments today and they did not address the 

IMS identification of a mixture of cocaine and 

heroin.4 Mixtures of heroin and cocaine 

posses a slightly different problem. The two 

compounds have adequate baseline 

separation, however, the response on the IMS 

is altered depending on their concentrations. In 

addition, even though the reduced mobilities 

(Ko) for heroin HCl (1.04 and 1.14 cm2/Vs) and 

cocaine HCl (1.16 cm2/Vs) are different, two 

other compounds, acetylcodeine (1.09 and 1.21 

cm2/Vs) and 6-monoacetylmorphine (1.13 and 

1.26cm2/Vs) posses Ko values that are similar 

to cocaine.5 Acetylcodeine and 

6-monoacetylmorphine are produced when 

acetic anhydride is added to crude morphine, 

which can contain other opium alkaloids such 

as codeine, to produce heroin. Therefore, 

when cocaine and heroin are present in the 

unpredictable law enforcement search for 

controlled substances, this can cause abnormal 

IMS response. 

EXPERIMENTAL 

The South Central Laboratory conducted 

studies to address the IMS detection of 

Received for review December 2, 1999, Accepted July 15, 2000 


A.M. DeTulleo et al.: „Detecting heroin in the presence...”, IJIMS 3(2000)1,38-42, p. 39 

Table I: Ion Mobility Spectrometer parameters 

Ion Mobility Spectrometer Barringer IONSCAN 400 

Drift Tube Temperature 237EC 

Inlet Temperature 

279EC 

Desorber Temperature 288EC 

Drift Gas (air) Flow 

300cc/minute 

Sample Gas(air) Flow 200cc/minute 

Exhaust Gas Flow 

502cc/minute 

Desorber Time 

8 seconds 

Drift Tube Length 

6.9 centimeters 

Scan Period 

20 microseconds 

Calibrant 

Nicotinamide 

mixtures of cocaine and heroin in law 

enforcement situations. The studies include 

recording IMS linearity and response for 

various concentrations of cocaine and heroin, 

individually and in a mixture. The IMS used 

was a Barringer Corporation 

Model 400 IMS using the 

parameters listed in Table 1. The 

IMS was calibrated routinely using 

authenticated cocaine HCl (Merck 

#V4511) and heroin HCl (DEA 

Special Testing Laboratory 

N10-P75A) standards, as well as 

the Barringer supplied calibration 

mixture called the verific. This 

study does not reflect results for 

cocaine base, heroin base, or 

black tar heroin HCl. All 

authenticated standards in 

laboratory grade methanol were 

introduced onto a teflon filter card 

using an eppendorf syringe. The 

methanol was then allowed to evaporate before 

analyzing the card. All filters were analyzed by 

the IMS prior to any sample placement to 

insure no contamination. In addition, all 

IMS work on this project was conducted 

in a clean, contaminant-free room 

dedicated to IMS work only. 

Linearity and Response for Cocaine 

IMS has limited linearity capabilities, 

therefore, the linear concentration 

response values are an approximation.6 

Three repetitions of cocaine ranging in 

concentrations from 1ng to 80ng were 

analyzed. The average IMS response 

was recorded in digital units (du) and is 

also called amplitude counts. Cocaine 

exhibited a linear response between 

approximately 1ng and 8ng. Above 

approximately 8ng, the amplitude response was 

relatively constant. Figures I and II illustrate 

Figure I: Cocaine HCl response on the IMS 400 

the response for cocaine. 

Linearity and Response for Heroin 

Three repetitions of heroin ranging 

in concentrations from 1ng to 300ng 

were also analyzed and their 

response recorded. Heroin 

exhibited a linear response between 

approximately 1ng and 25ng. 

Above approximately 25ng, the 

amplitude response was relatively 

constant. Figures III and IV 

illustrate the response for heroin. 

Mixtures of Cocaine and Heroin 

Figure II: 

Cocaine HCl linear response on the IMS 400 


Various concentrations of mixtures 

of cocaine and heroin were tested 

to determine the IMS response. 

There was no difficulty in detecting


Figure III: Heroin HCl response on the IMS 400 

the cocaine at any concentration of heroin. 

Therefore, the majority of the mixtures tested 

used the maximum response 

concentration for heroin at 25ng. 

Figure V illustrates the plasmagram 

obtained for the maximum response 

of cocaine (8ng) and heroin (25ng). 

As Table 2 clearly outlines, at 

heroin’s maximum response of 25ng, 

the cocaine concentration needs to 

be under approximately 8-9ng for the 

heroin to be detected by the IMS. 

lists the published and/or some 

experimental Ko values and drift 

times for cocaine, heroin, 

6-monoacetylmorphine, and 

acetylcodeine.7 Figures VI and VII 

show plasmagrams of 

6-monoacetylmorphine and 

acetylcodeine respectively. When 

these compounds are present in 

heroin, the similar peaks seem to 

add to the amplitude response of 

the cocaine. For example, 8ng of 

cocaine has an approximate 

average amplitude count of 

324.3du. 25ng of heroin has an 

approximate average amplitude of 

326.7du. However, in a mixture of 

8ng of cocaine and 25ng of heroin, the 

responses are 849du/1107du 

6-Monoacetylmorphine and 

Acetylcodeine Interference 

Heroin samples routinely contain 

other opium derivatives, such as 


Figure IV: Heroin HCl linear response on the IMS 400 

acetylcodeine. These compounds 

have peaks that have similar 

(Cocaine/Cocaine high) and 166du 

reduced mobilities and drift times to that of 

respectively. On the Barriner IMS 400, there 

cocaine. In fact, they posses small peaks to 

are two detection channels for cocaine. 

the left and right of the cocaine peak. Table 3 

Cocaine high is the channel that 

detects larger, more concentrated 

cocaine peaks. Figure 5 illustrates 

the peak pattern seen throughout the 

mixtures and Table II lists the 

individual and combined amplitude 

responses. Cocaine’s response is 

significantly larger when in a mixture 

with heroin. 

Figure V: 

Plasmagram for 8 ng of cocaine and 25 ng of heroin 

DISCUSSION 

Once the linearity and amplitude 

response is measured for cocaine 

and heroin individually, the IMS 

operator can judge the approximate 

concentrations of the drugs identified. 



Table 2: IMS response data for various concentrations of cocaine and heroin. 

*(PNA stands for Peak, but no alarm) 

Cocaine 

Response 

ng (du) 

4 (110) 

6 (213) 

7 (N/A) 

8 (324) 

9 (646) 

10 (747) 

15 (842) 

20 (964) 

10(747) 

15 (842) 

20 (964) 

40(1266) 

Heroin 

Response 

ng (du) 

25 (327) 

25 

25 

25 

25 

25 

25 

25 

10 (176) 

15 (216) 

20 (270) 

40 (443) 

IMS 

Response 

Cocaine 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Alarm 

Heroin will likely be detected if the mixture 

contains under approximately 8-9ng of cocaine 

and over approximately 25ng of heroin. 

Cocaine was always detected. However, 

recognizing the distinct plasmagram peak 

pattern allows a better judgement call for the 

IMS operator. The 6-monoacetylmorphine, due 

to similar reduced mobility and drift time 

(1.13cm2/Vs and 15.744 ms) to cocaine 

(1.16cm2/Vs and 15.221ms), can cause 

non-baseline separation. The IMS interprets 

this peak pattern as a wider, larger cocaine 

peak (Cocaine high) than is actually present. In 

addition, acetylcodeine, due to similar reduced 

mobility and drift time (1.21cm2/Vs and 

14.564ms) to cocaine can also exhibit the same 

patterns. The IMS operator could mistake this 

pattern for a larger cocaine response. 

Furthermore, the IMS may not detect the 


IMS 

Response 

Heroin* 

Alarm 

Alarm 

PNA 

Alarm 

Alarm 

PNA 

PNA 

PNA 

Alarm 

Alarm 

PNA 

PNA 

Figure VI: 

Plasmagram for 40 ng of 6-monoacetylmorphine 

IMS 

Amplitude (du) 

Cocaine/Cocaine 

High 

(mixture) 

603/748 

712/948/1536 

893 

849/1107 

952/1537 

746/933 

819/1031 

860/1096 

975/1472 

928/1485 

935/1568 

828/1659 

IMS 

Amplitude (du) 

Heroin 

(mixture) 

308 

209 

N/A 

166 

190 

N/A 

N/A 

N/A 

111 

163 

N/A 

N/A 

heroin, and therefore the operator could 

possible conclude that the sample only contains 

cocaine. In no instance did the IMS false 

alarm on any compound. Hence, all vacuum 

sweep samples collected in the field should be 

confirmed by GC/MS analysis, which will 

positively identify the cocaine, heroin, 

acetylcodeine, 6-monoacetylmorphine, and 

possibly other cocaine and heroin derivatives 

and alkaloids not mentioned. 

CONCLUSION 

The South Central Laboratory’s 

experimentation with the field portable IMS in 

the laboratory and in the law enforcement field 

has proven that IMS is a valuable tool in 

detecting the presence of trace amounts of 

cocaine and heroin. However, there are 

concentration limitations due to competing ions 

and interferences from 


acetylcodeine found in heroin 

residue. The two compounds make 

the IMS response for cocaine larger 

than it actually is. Concentrations of 

cocaine higher that 8-9ng may mask 

the IMS detection of heroin. Hence, 

the plasmagram peak pattern during 

these scenarios should be 

evaluated. Finally, a GC/MS 

confirmation of the sample should 

always be conducted on each 

sample.


Table 3: 

Published and/or experimental values of reduced mobilities (Ko) and Drift Times for 

Cocaine, Heroin, 6-Monoacetylmorphine, and Acetylcodeine 

Cocaine 

Heroin 

6-Monoacetylmorphine 

Acetylcodeine 

K o Published 

(cm 2 /Vs) 

1.16 

1.50 

1.84 

1.04 

1.14 

1.13 

1.26 

1.09 

1.21 

K o Experimental 

(cm 2 /Vs) 

1.1600 

1.0423 

1.1424 

1.1197 

1.2684 

1.0912 

1.2103 

Drift Time 

Experimental 

(milliseconds) 

15.221 

17.024 

15.538 

15.744 

13.897 

16.154 

14.564 

Figure VII: Plasmagram for 40 ng of acetylcodeine 

REFERENCES 

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

CRC Press, Boca Raton, FL, 1994, p. 153. 

[2] Eiceman, p. 49. 

[3] DeTulleo, A. M., “Methamphetamine Versus Nicotine 

Detection on the Barringer Ion Mobility 

Spectrometer”, 5th International 

Workshop on Ion Mobility 

Spectrometry proceedings, 1996, pp. 

215-223. 

[4] Karasak, F. W., Hill H. H., and Kim, S. 

H., “Plasma Chromatography of 

Heroin and Cocaine With Mass 

identified Mobility Spectra”, Journal of 

Chromatography, No. 117, Elsevier 

Scientific Publishing Company, 

Amsterdam, 1976, pp. 327-336. 

[5] Eiceman, p.154. 

[6] Eiceman, pp. 129-132. 

[7] Eiceman, p. 154. 

[8] Karasek, p. 332. 


A high resolution IMS for environmental studies 

J.W. Leonhardt, W. Rohrbeck, H. Bensch 

IUT Institute for Environmental Technologies Ltd., Rudower Chaussee 5, 12489 Berlin, Germany 

ABSTRACT 

IUT Ltd. develops various IMS devices for 

environmental purposes. The analysis of 

mixtures of aromatics has shown that there are 

problems to separate benzene and toluene ions 

properly by means of a low resolution cell 

(R=25). Similar problems exist in the negative 

mode for various halocarbons, for example 

trichloroethylene and dibromomethane. 

Therefore a new basic equipment was 

developed in order to improve resolution and 

transmission. New detector cells have a 50 or 

100 mm long drift tube with diameters of 20 or 

30 mm respectively. Ionisation is produced by 

tritium b-sources or by UV-lamps. The trigger 

pulse can be varied in the range of 10 - 350 

microseconds, the drift field is 400 - 700 V/cm. 

A resolution better than 100 was achieved. 

This value could be improved up to 200 by use 

of a deconvolution program. The simultaneous 

detection of chlorine and bromine ions 

produced in a sample of bromochloromethane 

is demonstrated. Further applications are 

discussed for benzene, toluene, xylene, 

halothan, isofluorene, formaldehyde etc. 

INTRODUCTION 

Small concentrations of toxic compounds in 

atmospheric air, in exhaust gases, in air at 

workplaces have to be measured selectively by 

a portable equipment. Ion mobility 

spectrometers were used to solve problems like 

the monitoring of phosphor organic agents, 

explosives and drugs [1-3]. 

The IUT Ltd. has developed some types of 

such spectrometers with high resolution and 

sensitivity. Monitors for benzene, toluene, 

xylene, formaldehyde, ethylenoxide, phosgene 

and halocarbons are available now [4]. 

In the developed systems tritium sources as 

well as UV-lamps are used as ionisation 

sources. Due to the high ionisation efficiency of 

tritium - all b-energy is transferred to gas 

molecules within a thin layer (thickness of 

about 1,5 mm) along the electrode - the 

sensitivity could be improved by one order of 

magnitude. 

The basic interaction of b-particles in a gas 

mixture - ambient air, dried air or carrier gases - 

is the production of ion pairs and excited states: 

+ − 

' 

A + β → A + e + β 

+ 

' 

A + β → A + β 

At normal air pressure ions like 

+ + 

N 

2 

and O2 

are not stable and get changed into cluster ions 

due to the water content. It can be assumed 

that the following positive ions exist in air [8]: 

+ + + 

NH 4 

, NO , ( H 2 

O ) n 

H 

These three cluster ions are called reaction 

ions. They produce the reaction ion peak RIP, 

which is a visible triplet. Charge transfer is 

realised in the presence of a molecule M with a 

higher proton affinity as the RIP-ions have: 

+ + 

( H O ) H + M → MH + nH O 

2 n 

2 

In the negative mode the negative cluster ions - 

the reaction ions - are 

O2 

− 

associated with some water. Their charge 

transfer goes to the molecule with a higher 

electron affinity. 

METHODS 

1. The Ion Mobility Cell 

The ion mobility detector is designed with 

cylindrical geometry. As shown in fig. 1 a tritium 

loaded disc electrode with 10 mm diameter in a 


J.W. Leonhardt et al.: „A high resolution IMS...”, IJIMS 3(2000)1,43-49, p. 44 

d r i f t 

gas 

shutter g r i d potential 

rings 

aperture grid 

insulator 

c o l l e c t o r 

emitter 

Betasource 

S 

a 

exit 

L d 

distance of 1..10 mm to a shutter grid electrode 

is used as ionisation source. Between these 

electrodes are the gas inlet and outlet. The 

ionisation volume may vary from 50 µl to 1,5 

ml. The drift tube is behind the shutter grid with 

ring electrodes. The collector electrode is also 

a disc protected by an aperture grid at the last 

ring electrode. The b-source can be substituted 

by an UV-lamp. 

2. Ionisation Sources 

Beta 3 H-sources. 

Beta-tritium-ionisation sources were developed 

in a joint venture with the Radium Institut in St. 

Petersburg. This special type of a ionisation 

source is used as emitter electrode, its energy 

spectrum is given in fig. 2. 

The spectrum demonstrates that tritium is 

gettered in a thin titanium layer only. The mean 

energy of emitted Beta particles is equal to 

=3.6 keV. The absolute activity may vary 

from a c = 0.42 MBq (free limit) up to 10 GBq. 

The ion production rate q(z) at a given distance 

from the emitter z can be evaluated by the 

formula (1). 

ac 

q ( z 

) = ⋅ ∆ε 

( z 

) 

( 1 ) 

w 

Figure 1: Scheme of an IUT - IM cell 

where w is the mean energy necessary to 

produce one ion pair. De(z) can be evaluated 

by Bethe’s formula for b-absorption in the non 

relativistic case as follows: 

( 2 ) 

ε o 

= 0. 

5 1 1MeV 

2,0x10 6 

1,5x1 

0 

1,0x10 6 

counts 

6 

5,0x10 5 

∆ε 

N εo 

1 16ε 

z 

− = 0. 

3ϕ 

⋅ 

∆z 

A ε 

( z 

) ln , ( ) 

I 

; 

0,0 

0 20 

keV 


Figure 2: 

Beta spectrum of a tritium source with 

low activity


s h u t t e r 

g r i d 

insulator 

potential 

rings 

aperture g r i d 

grid 

drift g a s 

protection g r i d 

g r i d g r i d 

collector 

lamp 

w i n d o w 

S 

a 

exit 

L d 

N, A: atomic and mass numbers; I=94.5eV, j = 

density. The mean number of ion pairs 

produced by one b-tritium particle is 

=102.8. About 30 % of the tritium in the 

emitter will contribute to the ionisation of the 

gas. That kind of ionisation source is available 

at IUT and Radium 

Institute in the range of 0.5 

MBq up to 0.5 GBq. 

Figure 3: Scheme of the IUT - IMS with a photoionisation discharge tube 

UV-photoionisation 

lamps. 

Hydrogen plasma discharge lamps were 

successfully checked for IMS-application too. 

The scheme of an ionisation source using 

photoionisation is given in fig. 3 - the exit 

window of the lamp is placed at the same 

position as the b-source. A special grid acts as 

an „emitting“ electrode, which avoids windows 

charging. There is the problem that photons 

should not enter the drift space. The geometry 

of the ionisation source has to be optimised 

with respect to the „shining“ into the drift space 

and to the carrier production rate. Unfortunately 

the ion distribution along z is much smoother as 

in the case of a b-source. This reduces the 

efficiency of such an arrangement. Photons flux 

of the lamp is typically 10 12 cm -2 s -1. . 

3. IMS-Resolution 

As discussed by R. St. Louis and H. H. Hill jr. 

[5] the defined resolution R is dominated by 1) 

the pulse width, t Pulse, 2) the diffusion 


R 

= 

1 6 l 

n 

2 

U 

U 

T 

d 

broadening, t diff., 3) the capacitive coupling 

between aperture and collector, t ap, 4) the gate 

depletion, t g, 5) temperature changes, 6) 

pressure changes, 7) coulombic repulsion and 

8) the amplifiers’ rise time, t R. The modified 

formula is equal to 

2 2 

K U ⎛ 

S 

d 

+ 4 

t 

Pulse 

− 

L ⎜ 

d ⎝ KU 

1 

Pulse 

⎞ 

⎟ 

⎠ 

2 2 

+ ⎛ ⎜ 

⎝ 

2 

a U 

2 

L U 

t K U 

where K = mobility, U T = temperature voltage, 

U ap = voltage at aperture grid, U d = drift voltage, 

a - distance between aperture grid and 

collector, S = distance between space charge 

and gate, L d = drift length. The discussion of (3) 

shows that R can reach 150 and even more at 

L d = 10 cm, t R < 10 µs and E = 1.5 kV/cm. 

These high field strengths may produce some 

problems in routine devices. The cells designed 

have the following specific parameters: 

System Drift U d 

length [kV] 

[cm] 

IUT-25 2,5 0,8 

ß/UV 

IUT-50 5,0 2,0 

ß/UV 

IUT-100 

ß/UV 

10,0 5 - 7 

t Pulse 

[µs] 

350 

30 

10 

In fig. 4 the resolutions of the reaction ion 

peaks for these three cell types are shown. 

d 

d 

ap 

U pulse 

[kV] 

0,2-2 

0,2-2 

0,2-2 

2 

⎞ 

⎟ + 

⎠ 

t R 

[µs] 

30 

30 

30 

2 2 2 

Pulse d 

4 

L 

d 

R 

25 

50 

120 

(3) 

T 

0,03 

0,3 

0,3


1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

0,000 0,002 0,004 0,006 0,008 0,010 0,012 0,014 

1,0 

0,8 

0,6 

amu0,4 

0,2 

0,0 

0,000 0,002 0,004 0,006 0,008 0,010 0,012 0,014 

1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

0,000 0,002 0,004 0,006 0,008 0,010 0,012 0,014 

Figure 4: 

Reaction ion peaks of 3 IM - detectors 

demonstrating the progress in the 

resolution 

4. Transmission 

The total charge picked off by the collector 

electrode with reference to the DC-ionisation 

current at saturation, I o, multiplied by t FWHM may 

characterise the transmission, that means the 

quality of the system: 

∫ i ( t ) dt 

o 

T = 

I t 

( 4 ) o 

⋅ 

FWHM 

. 

∞ 

The transmission of a photoionisation IMS can 

be determined by a given compound at a 

suitable concentration. The determination of I o 

may demand a special calibration arrangement. 

The used amplifier has a sensitivity of 5 × 10 9 

V/A and a rise time of 30 µs. Pulses with 

voltages between 0,2 - 2 kV were applied to the 

b-source. The IUT-50 system can be used as 

portable, hand held device or as a stationary 

unit. The data processing is carried out by 

means of a 32-bit processor. There is an 

alpha-numerical display for certain compounds 

in the mixture. The spectrum can be transferred 

to a PC. A special output is prepared for data 

transfer. The printed circuit board is designed in 

SMD-technology. 

RESULTS 

The photoionisation IUT-50 is designed for the 

sensitive detection of benzene down to the 

10 ppb level and in presence of other aromatics 

like toluene, xylene, cumene - often in much 

higher concentrations. The first version of a 

hand held system was checked in a chemical 


t[s] 

factory. It works with a membrane inlet system 

(fig. 5). Benzene could be measured down to 

30 ppb in ambient air. Benzene ions get 

quenched in presence of toluene, xylene and 

also cumene. As long as the device is used in 

0,8 

0,6 

0,4 

amu 

0,2 

0,0 

Figure 5: 

Peaks of benzene, toluene and xylene 

measured by means of an IUT - IMS - 25, 

photoionisation 

0,008 0,010 0,012 0,014 0,016 0,018 0,020 

t [s] 

Figure 6: 

Mixture of toluene (1.4 ppm ) plus different 

concentration of benzene 

the trace level c < 1 ppm the benzene content 

still can be evaluated from the remaining 

benzene peak in the spectrum as shown in fig. 

6. 

The probability of the quenching reaction, like 

between benzene and toluene, j, is rather high: 

+ K B T 

+ 

, 

B + T ⎯ ⎯ → B + T 

Toluol 1.4 

+ Benzol 62.4 

+ Benzol 33 

+ Benzol 17,5 

+ Benzol 9,2 

+ Benzol 4,9 

+ Benzol 2,6 

+ Benzol 1,37 

+ Benzol 0.72 

+ Benzol 0,38 

+ Benzol 0,2 

+ Benzol 0,11 

At benzene concentration of 1 ppm 1,4 ppm 

toluene cause a charge transfer of 80 % of the 

benzene ions. 

The region of response is up to 100 ppm. The 

situation could remarkably improved by the 

pre-separation of the aromatic compounds after 

sampling using an integrated column inside the


0,00 

-0,05 

-0,10 

-0,15 

current [amu] 

RIP 

-0,20 

-0,25 

Cl - 

-0,30 

0,002 0,003 0,004 0,005 0,006 

t [s] 

Figure 7: 

3D spectrum of 1 ppm benzene in toluene, 

xylene 

inner loop of the system. No additional carrier 

gas is needed. 

A 3-dimensional spectrum of about 1 ppm 

benzene in toluene, xylene is shown in fig. 7. 

Similar applications are thinkable for all 

compounds with a suitable photoionisation 

cross section at 10,2 or 10,6 eV. A typical 

1,6 

1,4 

1,2 

1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

cell (R » 100) gives 2 peaks in the spectrum: 

Cl - and P(-Cl) - . This situation can be used to 

identify the mentioned compound. 

The detection of anaesthetic gases is described 

by Eiceman [7] and has been carried out by 

means of the 25/50 devices. Spectra are given 

in fig. 10 for halothane, isoflurane and 

enflurane. 

This system can be applied excellently for the 

halocarbon determination in the GC-IMS mode. 

0,00 

Figure 9a: 

5 ppm phosgene in air, resolution 25 

-0,02 

10 

phosgene ion 

retention time 

-0,04 

20 

0,000 0,005 0,010 

drift time 

-0,06 

Cl - 

Figure 8: 

IM - Spectrum of a gasoline sample 

(Super Plus, BP) 

current [amu] 

-0,08 

application is the 3D-spectrum of gasoline 

(British Petrol, Super plus) shown in fig. 8, in 

which aromatics and alcanes produce a 

fingerprint picture. The identification of gasoline 

is possible. 

The determination of phosgene in ambient air is 

a classical industrial application. The typical 

spectra are given in fig. 9 (a) and (b). While (a) 

mainly resolves the Cl-ion, the high resolution 

-0,10 

-0,12 

0,0045 0,0050 0,0055 0,0060 

t [s] 

RIP 

Figure 9b: 

Phosgene ion peaks and reaction ion peak, 

resolution 100 



current [amu] 

0,05 

0,00 

-0,05 

-0,10 

-0,15 

-0,20 

monomer ion 

dimer ion 

current [amu] 

350 

300 

250 

200 

150 

100 

50 

0 

(a) Cl + Br in air, normal spectrum 

Cl - 

RIP 

(b) Cl + Br in air, deconvoluted spectrum 

Cl - 

Br - 

Br - 

4,00 4,25 4,50 4,75 5,00 5,25 5,50 5,75 6,00 

RIP 

Figure 11: 

Simultaneous 

detection of 

chloride and 

bromide ions in a 

sample of 

bromochlorometha 

ne a: normal 

spectrum b: 

deconvoluted 

spectrum 

-0,25 

drift time t [ms] 

The resolution power of 100 is demonstrated in 

fig. 11. (1) shows the Cl - - and Br - -Peaks being 

produced from bromochloromethane. Using the 

mathematical deconvolution [6] the spectrum 

could be improved again by a factor 2. 

0,7 

0,6 

-0,30 

-0,35 

-0,40 

Figure 10: 

Ion mobility spectrum of isofluran (narcotic) 

in the negative mode with membrane - 

inletsystem, 2.4 ppm) 

RIP 

RIP 

0,010 0,015 0,020 0,025 

monomer ion M + 

t [s] 

+ 

dimer ion M 2 

In fig. 12 the spectrum of malonacidester as 

somane simulance is shown. At a concentration 

of 100 ppb the monomer and dimer peaks can 

be seen clearly. Other examples for the positive 

mode are acrolein in fig. 13 and formaldehyde 

in fig. 14. In case of formaldehyde the gas inlet 

system is rather sophisticated. 

DISCUSSION 

The application of IMS-equipment is limited 

obviously by the still incomplete basic 

knowledge about reactions and mechanisms 

and furthermore by missing data bases. 

Therefore the suggestion of Karpas et al [3] to 

use calibration standards of the mobility scale 

is a very useful method to make comparable 

various systems. The GC-IMS is an interesting 

feature to improve the acceptance of this 

method [11]. An interesting contribution to the 

IMS application activities may be the system 

with an integrated column in the gas loop. This 

0,5 

0,4 

0,4 

RIP 

current [amu] 

0,3 

RIP (air) 

0,3 

0,2 

0,1 

current [amu] 

0,2 

0,0 

0,002 0,004 0,006 0,008 0,010 0,012 0,014 

drift time t [s] 

0,1 

Figure 12: 

0.1 mg/m 3 soman-simulance 

(malonacidester) and reaction 

ion peak (RIP) - positive mode 


0,0 

0,008 0,010 0,012 0,014 0,016 0,018 

t [s] 

Figure 13: Acrolein in air - positive mode


0,014 

0,012 

RIP 

REFERENCES 

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

CRG Press, Boca Raton, FL (1994) 

[2] Eiceman, G.A. Crit. Rev. Anal. Chem., 22, 471 (1991) 

current [amu] 

0,010 

0,008 

0,006 

0,004 

0,002 

0,000 

0,008 0,010 0,012 0,014 0,016 

Figure 14: 

Formaldehyde spectrum - positive mode 

system equipped with a photoionisation source 

is able to avoid and toluene. Fingerprints of 

flammable and other organic mixtures can be 

evaluated by a portable system. The 

PI-GC-IMS has good prospects to reach the 

sub-ppb range for many compounds. So far we 

are convinced that this method will have a good 

future also in emission determination. New 

application fields of IMS-devices are coming up 

in electrotechnical engineering [9] and 

microelectronics. The control of gas purity but 

also the characterisation of outgassings of 

polymers are topics of high interest [10]. 


The basic principles of this work were 

sponsored by the Federal Ministry of Science 

and Technology.No. 01 VQ 916B/O, 1991. 

t [s] 

[3] Karpas, Z., Wang, Y.F., Eiceman, G.A., Harden, C.S. 

Chemical Standards for Calibration of the Mobility 

Scale in Ion Mobility Spectrometry - in press - 

[4] PREVAC GmbH, Mikro-Ionisations-Gassensor 

(MIGA) zur Schadstoffbestimmung, speziell für 

halogenierte Kohlenwasserstoffe, BMFT. 

Förderkennzeichen 01 Q 916B/O. (1991 - 1994) 

[5] St. Louis, R.M., Hill, H.H. Jr. Ion Mobility 

Spectrometry in Analytical Chemistry. Anal. Chem., 

21, 321 (1985) 

[6] Ehart Bell, S., Wang, Y.F., Walsh, M.K., Qishi Du, 

Ewing, R.G., Eicemann, G.A. Qualitative and 

quantitative Evaluation of Deconvolution for Ion 

Mobility Spectrometry. An. Chim. Acta 303, 163 - 174 

(1995) 

[7] Eiceman, G.A., Shoff, D.B., Harden, C.S., nyder, A.P. 

Ion Mobility Spectrometry of Halothane, Enflurane 

and Isoflurane/Anesthetics in Air and Respired 

Gases. Anal. Chem. 61, 1093 (1989) 

[8] Caroll, D.I., Dzidic, I., Stillwell, R.N. and Horning, E.C. 

Identification of positive Reactions Observed for 

Nitrogen Carrier Gas in Plasma Chromatography 

Mobility Studies. Anal. Chem. 47, 12 (1975) 

[9] Baumbach, J.I. IMS Application in SF 6-switchers, 

private communication 

[10] Budde, K.I., Holzapfel W.I., Beyer, M.M. Application 

of Ion Mobility Spectrometry to Semiconductor 

Technology: Outgassings of Advenced Polymers 

under Thermal Stress, J. Electrochem. Soc., 142, 3 

(1995) 

[11] Snyder, A.P., Harden, C.S., Brittain, A.H., Man Goo 

Kim, Arnold, N.S., Menzelaev, H.L.C. Portable 

Hand-Held Gas Chromatographie/Ion Mobility 

Spectrometry Device. Anal. Chem 65, 299 - 306 

(1993) 


9 th International Conference on 

Ion Mobility Spectrometry 

ISIMS 2000 

SOCIETY 


for 

ION 

SP ECTROMETRY 

MOBILITY 

Halifax, Nova Scotia, Canada 

August 13-16, 2000 

Programme 

Fundamentals 1 

Effects of electric fields on mobility spectra in high temperature drift tubes at ambient 

pressure: Models and precision measurements 

Erkin Nazarov, G. A. Eiceman, J. E. Rodriquez, and J.A. Stone 

The effects of temperature on the detection of volatile vapors emitted from explosives 

using ion mobility spectrometry 

R.G. Ewing and C.J. Miller 

The Peak Shape Provided By IMS Instruments As Determined By Various Parameters 

Stefan Morley, Jürgen Landgraf, Rainer Lippe, Ulrich Gräfenhain, and Inka Schilde 

Thermal Stability of Nitrated Organics in IMS Analysis 

C. J. Miller and R. G. Ewing 

Instrumentation 1 

Miniaturized Ion Mobility Spectrometer 

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

Miniature 2nd grade Aspiration Type Ion Mobility Spectrometer with IMCell Technology in 

Detecting of Chemical Vapors 

Timo Jaakkola 

VIP Sources for Ion Mobility Spectrometry 

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

External Exit Gate Fourier Transform Ion Mobility Spectrometry 

E. Tarver 


ISIMS 2000 Programme, IJIMS 3(2000)1, p. 51 

Applications 1 

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

monitoring monoethanolamine on board submarines 

H R Bollan, and J L Brokenshire 

The Use of IMS and GC-IMS for analysis of saliva 

C. Fuche 

Rapid analysis of pesticides on imported fruits by GC-IONSCAN 

R. Debono, T. Le, S. Yin, A. Grigoriev, R. Jackson, R. James, F. Kuja, A. Loveless, S. 

Nacsono 

Detection and Identification of Toxic substances and their percursors using ion mobility 

spectrometers (IMS) in an unmanned aerial vehicle 

Vincent M. McHugh, Charles S. Harden, Dennis M. Davis and Donald S. Shoff, Gretchen 

Eyet, Tony O’Connor and Simon Pavitt 


TLC-Pyrolysis-IMS Coupling - A New Selective Detection Method for Thin Layer 

Chromatography Using the Chromarod Technique 

P. V. Sivers, H. Matschiner, M. Brodacki, and J. Stach 

Performance of a Micromachined Radio Frequency – Ion Mobility Spectrometer (RF-IMS) 

with Non-rad and Low-rad Ionization Sources 

Raanan A. Miller, Gary A. Eiceman and Erkinjon G. Nazarov 

Discussion of the technical aspects and test results of Barringer Sabre 2000. 

L. Fricano, T. Gabowicz, R. Jackson, F.. Kuja, L. May, S. Nacson, M. Uffe, S. Wiles R. 

Debono 

Evaluation of quantitative analysis by corona discharge ion mobility spectrometry 

Khayamian, T. and Tabrizchi, M. 

Fundamentals 2 

Use of neutral networks for the identification of class specific features in ion mobility 

spectra 

Suzanne Bell, Erkin Nazarov, G. A. Eiceman, J. E. Rodriquez, and Y. F. Yang 

Relationships for ion dispersion in ion mobility spectrometers 

Glenn E. Spangler 

Temperature corrections in ion mobility spectrometers 

Dennis Davis and Donald B. Shoff 

Neural Network Analysis of Ion Mobility Spectra for Chemical Class Identification. 

G.A. Eiceman and E.G. Nazarov 




A novel portal design for rapid real time detection of explosives’ vapors and particles 

L. Fricano, M. Goledzinowski, F.. Kuja, L. May, S. Nacson, M. Uffe 

Solid Phase Microextraction coupled to Ion Mobility Spectrometry 

J.I. Baumbach, G. Walendzik, S. Sielemann, D. Klockow 

The use of GC-IMS to Analyze High Volume Vapor Samples from Cargo Containers 

P. Lafontaine 


A Miniature Ion Mobility Spectrometry with a Pulsed Corona-Discharge Ion Source 

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

Electrospray Ionization Ion Mobility Spectrometry for Cationic Species 

Herbert Hill 

Poster session 

IMS in the undergraduate chemistry curriculum 

S.P. Sibley, B.L. Houseman, W.C. Blanchard * 

Manipulating Reaction Ion Chemistry To Enhance Detection In Ion Mobility Spectrometry 

Keith A. Daum, Robert G. Ewing, and David A.Atkinson 

Ion non-linearity drift spectrometer - a selective detector for high-speed gas 

chromatography 

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

Ion mobility spectrometer linearity studies of cocaine, heroin, methamphetamine and 

MDEA 

T. Dallabetta-Keller, A.M. Detulleo, P. B. Galat, and M.E. Grey 

Principles & Applications of Solid Phase Desorption Coupled to GC-IONSCAN® System 

A. Grigoriev, R. Jackson, R. James, F. Kuja, A. Loveless, S. Nacson 

Rapid Characterization of Bacteria by Ion Mobility Spectrometry 

R. F. DeBono, J. R. Jadamec, R. T. Vinopal 

Ion Mobility Spectrometry Analysis of LAMPA, GHB, and GBL 

R. DeBono, T. Le 


The Use of Handheld Ion Mobility Spectrometry Instrument in the Marine Environment – 

Overview of an Operational Assessment 

Chih-Wu Su, Steve Rigdon, Tim Noble, Mike Donahue 



Side-by-side comparison of small hand-held ion mobility spectrometers and surface 

acoustic wave devices 

Vincent M. McHugh, Charles S. Harden, Dennis M. Davis and Donald S. Shoff 

Gretchen Eyet 

Evaluation of materials as sample collectors for the Barringer 400 ion mobility 

spectrometer 

Nairmen Mina, Samuel P. Hernandez, Felix. Roman, and Luis .A. Rivera 

Instrumentation 3 / Applications 4 

Recent Developments of IMS Tube Configurations 

Rainer Lippe, Rüdiger Döring, Stefan Morley, Jürgen Landgraf, Ralf Burgartz 

Ion Mobility Spectrometry for Laser Desorption-Ionization Analysis 

G.A. Eiceman, D. Young, and D.A. Lake, M.V. Johnston, 

Use of IMS for a Study on CS Decontamination 

T. Donnelly 

Near-real-time analysis of toxicologically important compounds using the volatile organic 

analyzer for the international space station. Part II 

E.S. Reese, T.F. Limero, 


Development of a Detector for Outbound Currency 

Keith A. Daum, Garold L. Gresham and David E. Hoglund 

In-Situ Methylation of Methamphetamine during Ionscan Analysis 

Chih-Wu Su, Kim Babcock, Steve Rigdon 

Evolution of IMS technology within the Australian Customs service 

John Kerlin and Ginna Webster 

The Future of Ion Mobility Spectrometry in the NASA Manned Space Program 

Thomas Limero, Eric Reese, John Trowbridge 




ISIMS 2000 

SOCIETY 


for 

ION 

SP ECTROMETRY 

MOBILITY 


August 13-16, 2000 

Abstracts of Papers 


Abstracts of Papers presented at ISIMS 2000, IJIMS 3(2000)1, p. 55 

Effects of electric fields on mobility spectra in high temperature drift tubes at ambient 

pressure: Models and precision measurements 

Erkin Nazarov 1 , G. A. Eiceman 1 , J. E. Rodriquez 1 , J.A. Stone 2 

1 

New Mexico State University, Department of Chemistry and Biochemistry 

2 

Queens University, Kingston, Ontario, Canada 

ABSTRACT 

Advances in drift tube designs for ion mobility spectrometers have been impeded with suitable 

models and supporting precise measurements to describe the influence on ion behavior from 

electric fields between components. Electric fields in specific regions of a drift tube at 250°C were 

independently varied and finding were interpreted using a model for the transport of thermalized 

ions in a homogeneous electric fields at ambient pressure. Response was measured using peak 

shape, signal intensity, drift times and reduced mobilities( K o) for positive polarity reagent ions in 

purified air at 660 torr pressure with 0.15 ppm moisture. These findings provided several 

surprising conclusions that may govern the design of future drift tubes. Ion neutralization on the 

aperature grid with electric fields accounts for losses in signal intensity reaching ~100% below 


The effects of temperature on the detection of volatile vapors emitted from explosives 

using ion mobility spectrometry. 

R.G. Ewing and C.J. Miller 

Idaho National Engineering and Environmental Laboratory 

P.O. Box 1625 

Idaho Falls, ID 83415-2208 

ABSTRACT 

Vapor detection of explosive compounds (e.g. PETN and RDX) present in plastic explosives is 

relatively difficult due to the low vapor pressures. Although particulate from these compounds can 

be detected by ion mobility spectrometers operated at elevated temperatures, vapor detection 

provides a less intrusive way of detecting explosives. In order to facilitate the detection of plastic 

explosives, the uses of taggants has been proposed. Some potential taggants include: ethylene 

glycol dinitrate (EGDN), 2,3-dimethyl-2,3-dinitrobutane (DMNB), para- and ortho-mononitrotoluene 

(p-MNT, o-MNT) with minimum concentrations of 0.2, 0.1, 0.5 and 0.5 percent by mass 

respectively. DMNB and EGDN yield only fragment ions of NO 2 

- 

and NO 3 

- 

in IMS at elevated 

temperatures, which presents a problem for selective detection of these compounds. Lowering 

the temperature below 75 °C allows for the appearance of molecular ions or ion-molecule adducts 

which improves the identification of these compounds. A discussion on the role of temperature in 

ion creation and stability for these compounds will be provided. In addition, a comparison of vapor 

detection for EGDN and DMNB in pure form and in bulk explosives will be presented. 



The Peak Shape Provided By IMS Instruments As Determined By Various Parameters 

Stefan Morley, Jürgen Landgraf, Rainer Lippe, Ulrich Gräfenhain, and Inka Schilde 

Bruker Saxonia Analytik GmbH, Permoserstrasse 15, D-04318 Leipzig, Germany 

ABSTRACT 

It is widely accepted that gating time as well as the electrical field at the shutter grid have a strong 

impact on the peak shape of Ion Mobility Spectrometers (IMS). In order to discriminate 

neighbouring peaks from each other a high resolution is desirable. This fact does frequently lead 

to a solution that only reduces the gating time (i.e. the shutter grid pulse width) and effects merely 

in a deteriorated signal to noise ratio. Therefore, a compromise has to be found that satisfies 

both, the need for resolution as well as the need for intensity for a given tube design. 

Furthermore, it is a matter of fact that not only the gating time but also the electrical field nearby 

the grid contributes to peak shape. These effects are worked out into a mathematical formalism, 

which provides developers with an easily applicable tool to verify the effectiveness of their design 

and shows its stretch potential. The formalism itself is also compared to experimental results 

gained by various Bruker IMS instruments. 



Thermal Stability of Nitrated Organics in IMS Analysis 

C. J. Miller and R. G. Ewing 


P. O. Box 1625 

Idaho Falls, ID 83415-2076 

ABSTRACT 

Nitrated organics have been used extensively in explosives, double based powders, and explosive 

taggants. Detection of these chemicals (via vapor analysis) is a stronger indicator of an explosive 

device than particulate samples. Some of these chemicals are readily detected by ion mobility 

spectrometry (IMS); however, they are known to be thermally unstable. In IMS, the nitrated 

organics tend to decompose at elevated temperatures yielding a nitrate ion (NO 3- ). The 

temperature at which decomposition to a nitrate ion occurs is chemically dependent. Ideally, ion 

mobility spectrometers are operated at elevated temperatures to improve response, to reduce 

clustering and to maintain a clean instrument. For the detection of these nitrated organics, the 

temperature must be reduced in order to obtain molecular ion information. Although this nitrate 

may be an indicator of the presence of explosives, it is less specific than the molecular ion or a 

base fragment from a polymeric chain. A discussion of temperature dependence on the detection 

of various nitrated organics will be presented. 



Miniaturized Ion Mobility Spectrometer 

J.I. Baumbach 1 , M. Teepe 2 , A. Neyer 2 , H. Schmidt 1,3 , P. Pilzecker 4 

1 

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


2 

Universität Dortmund, Fachbereich Elektrotechnik, Arbeitsgebiet Mikrostrukturtechnik 

Otto-Hahn-Str. 6, D-44221 Dortmund, Germany 

3 

Universität Dortmund, Fachbereich Chemie, Anorganische / Analytische Chemie, 

Otto-Hahn-Str. 6, D-44221 Dortmund, Germany 

4 

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

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

ABSTRACT 

A novel micro machined ion mobility spectrometer (IMS) containing a 16 MBq- 63 Ni-ionisation 

source (3 mm diameter) and a small drift tube is described. The spectrometer was used for the 

sensitive detection of various volatile halogenated compounds down to the ppb v-range (pg). 

Results of investigations of different volatile organic compounds in air and nitrogen measured 

using the IMS developed will be presented and discussed with respect to detection limits and 

selectivity of the IMS. Examples of spectra of different mixtures of analytes are also presented, 

including ethanol, tetrachloroethene, benzene, toluene and xylenes. The results will be compared 

with those obtained using traditional ion mobility spectrometers using 63 Ni- ß-radiation sources, 

UV-lamps or partial discharge to ionise the carrier gas or the analytes. Advantages and 

disadvantages of the different spectrometers used will be summarised. 



Miniature 2 nd grade Aspiration Type Ion Mobility Spectrometer 

with IMCell Technology in Detecting of Chemical Vapors 

Timo Jaakkola 

Environics Oy, P.O.Box 349, 50101 Mikkeli, Finland 

ABSTRACT 

Environics Oy has utilized an Aspiration Type Ion Mobility Spectrometry (IMCell Technology) in its 

gas detectors since 1988. In an Aspiration Type IMS the ions are introduced into several electric 

fields arranged perpendicular to the flow stream in the measurement cell. The ions, generated by 

an Am241 source, are collected into the electrodes of the electric fields and the produced current 

is measured in each electrode. The currents forms a signal pattern that is spesific to each gas. 

Neural pattern recognization methods are used in the identification. The sum of the currents refer 

to the total concentration of the gas. One of the advantages of the measurement principle is an 

extremely simple and small structure of the measurement cell. Typically electrodes are made into 

a normal printed circuit board. 

Environics Oy has now developed a 2 nd grade IMCell measurement cell. Earlier, the ions are 

introduced into the IMCell from the total height of the flow channel. Therefore, the ions with similar 

ion mobilities can have an initial phase of being either in the upper or lower part of the flow 

channel. In the 2 nd grade IMCell, all the ions are introduced into the cell only from the upper part of 

the flow channel. The method increases the selectivity. Furthermore, the number of electrodes 

has been increased from 6 to 16 with a novel method of dual polarity. Dynamic electric fields 

allows tuning of the measurement cell to different gases. 

The IMCell technology with the 2 nd grade principle has made possible to develop a miniature IMS 

sensor unit. The developed sensor unit, mathematical modelling of the 2 nd grade Aspiration Type 

IMS and test results of several chemical vapors will be presented. 



Detection and Identification of Toxic substances and their percursors using ion mobility 

spectrometers (IMS) in an unmanned aerial vehicle. 

Vincent M. McHugh 1 , Charles S. Harden 1 , Dennis M. Davis 1 , Donald S. Shoff 1 , Gretchen Eyet 2 , 

Tony O’Connor 3 and Simaon Pavitt 3 

1 

US Army Edgewood Chemical and Biological Center, Aberdeen Proving Ground, Maryland 

2 

GEO-Centers, Incorporated, Aberdeen Proving Ground, Maryland 

3 

Graseby Dynamics Limited, Watford, Herts, UK 

ABSTRACT 

The United States Army (Edgewood Chemical Biological Center, APG, MD) and the United States 

Navy (Naval Research Laboratory, Washington, D.C.) have initiated a joint program to integrate a 

modified, hand-held Ion Mobility Spectrometer (IMS) into an Unmanned Aerial Vehicle (UAV) for 

real-time detection and identification of chemical clouds. The use of IMS technology provides an 

ideal approach to rapid, real-time chemical vapor detection, identification and quantification. 

The IMS point detection payload will incorporate the capability for acquiring a sample for 

subsequent retrieval and analysis, as well as for control of/communication with the payload 

components and external devices. The IMS point detection payload will possess the capability to 

store data from two IMS units, the UAV flight controller, a Global Positioning System and a 

meteorological sensor. The IMS point detection payload will be fully compatible with the small 

UAV, consume a minimum of electrical power and occupy a minimum volume. The IMS point 

detection payload will be integrated with the UAV to share a gas sampling system, electrical 

power source, and digital Input/Output (I/O) interfaces. 

Data will be presented, using surrogates for toxic, target chemicals, which demonstrate the 

effectiveness of an IMS when contained in a UAV nosecone and tested in a wind tunnel. 

(sponsored by the Defense Threat Reduction Agency) 



TLC-Pyrolysis-IMS Coupling - A New Selective Detection Method for Thin Layer 

Chromatography Using the Chromarod Technique 

P. V. Sivers 1 , H. Matschiner 1 , M. Brodacki 2 , and J. Stach 2 

1 

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

2 

Elektrochemie Halle, Weinbergweg 23, 06120 Halle, Germany 

ABSTRACT 

About 25 years ago Okumura and Kadono [1] suggested a new thin layer chromatography 

technique: The use of so-called chromarods, thin quartz rods coated with a suitable phase like 

aluminium oxide or silica gel instead of usual thin-layer plates. The chromarod technique allows 

the application of different types of detection methods after chromatographic separation and 

thermal desorption or pyrolysis. Most interesting are selective detection methods like NDIR or 

IMS. Especially IMS provides high selectivity for a lot of pyrolysis products like sulfur dioxide, 

nitrogen oxides, hydrogen halogenides and others. In addition functionalized hydrocarbons can be 

detected with high sensitivity after thermal desorption as well. In combination with NDIR carbon 

dioxide detection especially in connection with pyrolysis of hydrocarbons the whole range of 

organic compounds can be covered. 

The paper gives a detailed description of the developed thin layer chromatography method [2] 

which allows the investigation of up to 10 chromarods in one run. The detection capabilities using 

NDIR and IMS detection are compared. The experimental conditions for detection of pesticides, 

surfactants, fatty acids etc. are described in detail. Special attention will be paid to the influence of 

used carrier gases (nitrogen or oxygen) and pyrolysis temperatures. 

[1] Okumura, T. and Kadono, T.: U.S. Patent 3,829,205 (1974). 

[2] P. v. Sivers, M. Hahn, Chemie in Labor und Biotechnik 49 (1998) 424 



Performance of a Micromachined Radio Frequency – Ion Mobility Spectrometer (RF-IMS) 

with Non-rad and Low-rad Ionization Sources 

Raanan A. Miller 1 , Gary A. Eiceman 2 , Erkinjon G. Nazarov 2 

1 

Charles Stark Draper Laboratory, 555 Technology Square, Cambridge, MA 02139 

2 

Department of Chemistry and Biochemistry, New Mexico State University, Las Cruces, NM 88003 

ABSTRACT 

Radio frequency ion mobility spectrometry (RF-IMS) is a recently developed IMS-based technique 

that allows miniaturization of IMS drift tubes through micromachining, while preserving sensitivity 

and resolution. The micromachined device fabricated at Draper Laboratory exhibits excellent 

performance characteristics with a drift tube 3 × 1 × 0.2 cm 3 in size. 

The RF-IMS consists of a micromachined drift-tube containing an ionization region, a tunable ion 

filter, and a detector. A carrier gas (flow rate of 1 - 4 liters/min) transports the ions through the drift 

region where they are filtered by applying a dc bias-voltage and a radio frequency waveform to ion 

filter electrodes. Adjusting the ratio of dc bias voltage to RF voltage selects the ions that can pass 

through the filter and be collected at a detector. 

This paper compares performance results (e.g., detection limits, reproducibility and resolution) of 

the RF-IMS for a range of volatile organic compound using several non-radioactive and low 

activity ionization sources. These ion sources include a UV photo discharge lamp and a 100 

µcurie americium source. 



Discussion of the technical aspects and Test results of Barringer Sabre 2000. 

L. Fricano 1 , T. Gabowicz 1 , R. Jackson 1 , F.. Kuja 1 , L. May 1 , S. Nacson 1 , M. Uffe 1 , S. Wiles 1 , R. 

Debono 2 

1 

Barringer Research Limited, Mississauga, Ont. 

2 

Barringer Instruments Inc., Warren, NJ 

ABSTRACT 

Technical information on Barringer’s newest IMS-based analyser, the SABRE 2000 ® , will be 

presented. This versatile, handheld instrument provides capabilities for both vapour and particle 

analysis, and operation at high temperatures. From the initial conceptualization stage, this 

instrument was designed and engineered to accommodate both particle and vapour samples, 

integrated into a functional, compact platform. 

Both particle and vapour results for narcotics and explosives will be presented and discussed. A 

discussion of other potential analytes which can be detected using the SABRE 2000 ® will also be 

presented. 



Evaluation of quantitative analysis by corona discharge ion mobility spectrometry. 

Khayamian, T. and Tabrizchi, M. 

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

ABSTRACT 

In this work the capability of corona discharge ion mobility spectrometry (CD-IMS) in quantitative 

determination of some volatile organic compounds has been evaluated. Generally, in IMS the 

signal intensity of a product ion is not a linear function of the sample concentration due to the 

reaction between the reactant ion and sample molecule. However, a linear calibration curve can 

be obtained if the kinetics 1 of the ion molecule reaction (R + + P Õ R + P + ) is considered. As the 

concentration of the sample (P) is constant the reaction become pseudo first order, hence, the 

quantity Ln(R 0+ /R 0+ -P + ) should be a linear function of the sample concentration. In this expression, 

+ 

R 0 is the original reactant ion density and P + is the sum of all the product ion densities. Acetone 

and dimethylmethylphosphonate (DMMP) as two examples of volatile organic compounds were 

examined in this experiment. The exponential dilution flask method was used for introducing the 

standard samples into IMS. In both cases a linear calibration curve was obtained. Detection 

limits of 25 ppt (parts per trillion) and 1.5 ppb were obtained for acetone and DMMP respectively. 

The dynamic range for acetone was more than three orders of magnitude while for DMMP it was 

two orders of magnitude. These results are indicative of the capability of CD-IMS as a 

quantitative analytical technique at ultra trace levels. 

In order to evaluate the capability of the method for real sample analysis, the acetone 

concentration in human breath was directly determined. The acetone concentration in breath 

could be related to sugar concentration in blood. In this work, breath spectra of the diabetes 

people were determined and artificial neural network was used to construct a model between the 

breath spectra and the blood sugar. 

1 

Spangler, G.E., Lawless, P.A., Anal. Chem., 1976, 48, 1352. 



Use of Neural Networks for the Identification of Class Specific Features in Ion Mobility 

Spectra 

Suzanne Bell 1 , Erkin Nazarov 2 , G. A. Eiceman 2 , J. E. Rodriquez 2 , and Y. F. Yang 2 

1 

Eastern Washington University, Dept. of Chemistry and Biochemistry 

2 

New Mexico State University, Department of Chemistry and Biochemistry 

ABSTRACT 

Ion mobility spectra have traditionally been considered low resolution and information poor 

compared to techniques such as mass and infrared spectrophotometry. However, recent 

applications of neural networks in the analysis and classification of ion mobility spectra have 

demonstrated that spectra contain significant amounts of class-specific chemical information. An 

exhaustive study of a large spectral library, obtained at low moisture and across a wide range of 

temperatures and concentrations demonstrated that much of this class-specific information 

resides near the reactant ions. This region was not previously regarded as particularly informative 

or useful for spectral classification. These features are strongly suggestive of small ions that 

presumably arise from fragmentation. Further advances in neural network applications holds 

promise for developing neural network-based spectral identification procedures that will enhance 

the utility of ion mobility spectrometry both in the laboratory and in the field. 



Temperature corrections in ion mobility spectrometers 

Dennis Davis and Donald B. Shoff 


ABSTRACT 

One of the major trends in IMS research today is to extend the utility of IMS in field environments. 

This is especially important for IMS devices that require a high degree of portability (small size, 

low power consumption, e.g., near ambient temperature of operation is implied). In order to 

address extended utility issues, IMS devices must be subjected to increasingly more hostile 

conditions where temperature, pressure, and relative humidity cannot be controlled or predicted. 

The temperature, pressure and humidity all affect the reduced mobility of the chemical species in 

the IMS. The temperature has perhaps the largest effect, because it affects the kinetics of the 

formation and dissociation of the ions and the density of the drift gas. Methods have previously 

been developed for correcting these variables over small changes, e.g. five to ten degrees C. 

This paper will detail investigations over wider temperature ranges, ambient temperatures ranging 

from 10 to 45 C, and will discuss efforts to increase the efficiency of detection algorithms by 

applying temperature corrections to the identification algorithms used in IMS spectrometers. 

Organophosphorus and organosulfur compounds were used as test substances. 



Neural Network Analysis of Ion Mobility Spectra for Chemical Class Identification. 

G.A. Eiceman and E.G. Nazarov 

Dept. Chemistry and Biochemistry, New Mexico State University. 

ABSTRACT 

In this study we investigated the possibility of identification of chemical class from ion mobility 

spectra, the location of such information in the spectra, and the influence of temperature on it. 

It is well known that for identification of chemicals from IMS spectra the drift time of product ions is 

usually used. This approach can be applied successfully when the identification is from within a 

small group of known chemicals. However, when the reference database includes the IMS spectra 

for hundreds of chemicals belonging to around a dozen classes, with 8-10 spectra for different 

concentrations of each, the chemical identification of an unknown becomes problematic because 

such a database will contain multiple peaks with the same drift time. 

It has been discovered that neural networks can be used to determine chemical class of an 

unknown from its spectrum with a success rate of more then 90 %, using a large database as 

described above (total number of spectra exceeding 1000). This demonstrates that chemical class 

information is contained within IMS spectra. 

By neural network analysis of different sections of IMS spectra from the database, and by 

comparison of results of from a range of temperatures (between 50 and 250 o C), it was shown 

that the chemical class information in IMS spectra is provided by fragment peaks located close to 

trailing edge of the RIP at elevated temperatures, while at low temperatures they are distributed 

between RIP and product peaks. 

In the presentation the chemical basis for these results will be discussed. Additional proof of these 

conclusions will be provided by deconvolution treatment of spectra. 



A novel portal design for rapid real time detection of explosives’ vapors and particles 

L. Fricano, M. Goledzinowski, F.. Kuja, L. May, S. Nacson, M. Uffe 

Barringer Research Limited, Mississauga, Ont. 

ABSTRACT 

Barringer has designed and engineered a walk-through portal, with advanced pre-concentration 

and detection systems, based on a technology transfer agreement with Sandia National 

Laboratories. Barringer’s design provides speed, sensitivity, robustness and ease of servicing, 

while minimizing chemical interferences. 

The portal screens passengers for explosives using a non-invasive technique to collect both 

particle and vapour samples. The person being screened enters the portal, where a combination 

of a top-blowing fan, and wall-mounted, directional, air jets, dislodge any surface particulates, and 

transport them, together with any vapours present, to a 2-stage pre-concentrator. During 

sampling, portal air is drawn into the pre-concentrator at >15,000 litres/minute. 

Total analysis time, from initial portal entry to final result, is under 10 seconds per person. The 

new engineering features of the portal, and results obtained, will be presented. 



Solid Phase Microextraction coupled to Ion Mobility Spectrometry 

J.I. Baumbach 1 , G. Walendzik 2 , S. Sielemann 1 , D. Klockow 1 

1 

Institut für Spektrochemie und Angewandte Spektroskopie, Bunsen-Kirchhoff-Str.11, 


2 

Institut für Entsorgung und Umwelttechnik gGmbH, Kalkofen 6, D-58638 Iserlohn, Germany 

ABSTRACT 

Normally, ion mobility spectrometry is a useful analytical method for the sensitive monitoring of 

gaseous substances. To investigate water samples contaminated with organic components an ion 

mobility spectrometer (IMS) was coupled to a solid phase microextraction unit (SPME). For this 

purpose, a septum-sealed SPME unit was directly connected to an ISAS custom designed 10.6 

eV-UV-IMS. 

The experimental arrangement and results of investigations using an interface system for the 

transfer of volatile organic components (especially Benzene, Toluene, Xylenes) from the aqueous 

matrix into the IMS are presented and discussed with respect to reproducibility, detection limit and 

selectivity. 



Side-by-side comparison of small hand-held ion mobility spectrometers and surface 

acoustic wave devices 

Vincent M. McHugh 1 , Charles S. Harden 1 , Dennis M. Davis 1 , Donald S. Shoff 1 , Gretchen Eyet 2 

1 


2 

GEO-Centers, Incorporated, Aberdeen Proving Ground, Maryland 

ABSTRACT 

Within the U.S. Military, there has been a long-term and on-going debate with respect to the 

relative attributes of Ion Mobility Spectrometry (IMS) and Surface Acoustic Wave (SAW) 

technology for the detection and identification of chemical warfare agent. Almost all of the claims 

that have been made in favor of one technology or the other have been made on the bases of 

experimental data gathered in unrelated tests. Because of the on-going technology debate and 

because the U.S. Department of Defense has entered into a significant development program 

aimed at production of small, unobtrusive CW detectors, side-by-side evaluations of prototype 

IMS and SAW field instruments have been performed at the U.S. Army Edgewood Chemical 

Biological Center. Of particular interest in these evaluations were small field prototype 

instruments that will fit in the palm of a human hand -- instruments that are 40 in 3 (655 cm 3 ) in 

volume and weigh 2 pounds (900 gm). 

The purpose of the evaluation was to describe relative performance characteristics of devices 

based on the two technologies in experiments consisting of exposures to the same sample 

conditions at the same time — arguments resulting from possibilities of different analyte 

concentrations, water vapor concentrations and sample gas temperatures would be eliminated. 

Analytes included toxic organophosphorus analytes were studied (the so-called "nerve" agents), 

organosulfur analytes ("mustard gas"), and hydrogen cyanide (a "blood" agents). Analyte 

concentrations ranged from about 10 ppb to 10's of ppm, water vapor concentrations were 

intentionally kept high to stress the technologies, and sample gas temperatures ranged from 

about 25 to 36 C. The prototype devices were simultaneously exposed to analyte vapors in air 

and in mixtures of air and other substances that were known to cause undesirable responses in 

both IMS and SAW devices. IMS spectra were compared with SAW crystal frequency shifts for 

each of the analyte - interference - water vapor mixtures. Figures of relative merit to be reported 

include limits of detection, response time, ability to distinguish analytes by group and by type 

within a group, and ability to distinguish between analytes of interest and potentially interfering 

substances. 

The state-of-the-art in development of small, hand-held field detection devices based on IMS and 

SAW technologies will be reviewed. 



Evaluation of materials as sample collectors for the Barringer 400 ion mobility 

spectrometer 

Nairmen Mina, Samuel P. Hernandez, Felix. Roman, and Luis .A. Rivera 

Department of Chemistry, University of Puerto Rico at Mayaguez, Mayaguez, PR, 00681 

ABSTRACT 

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

affinity of various explosives (RDX, NG, TNT, PETN and DNT) for different materials and their 

adsorption/desorption thermal characteristics with a Barringer 400 IMS instrument. The explosive 

materials, dissolved in acetonitrile, were spiked with increasing amounts into the filters and 

membranes surfaces. The commercial filters and membranes, studied in this project, that 

withstood the high temperatures of the IMS injector/desorber were made of either fiberglass or 

cellulose, fine or coarse porosity of various pore size (0.5-40 mm), and medium to fast flow rate. 

The results from theses material were compare with the ones obtained from the currently used by 

Barringer which are a fiberglass filter and cotton swab. The data suggests that filter material 

properties such as pore size, surface roughness and porosity, flow rate and explosive vapor 

pressure 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 material 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. These experiments interrogate the adsorption/desorption thermal characteristics of the 

specific explosive material being evaluated with the different substrate materials 



Recent Developments of IMS Tube Configurations 

Rainer Lippe, Rüdiger Döring, Stefan Morley, Jürgen Landgraf, Ralf Burgartz 

Bruker Saxonia Analytik GmbH Leipzig 

ABSTRACT 

In order to design portable miniaturized IMS detectors with sufficiently high performance, various 

IMS tube concepts have been developed and evaluated. Bruker’s seven year old basic IMS tube 

as used in the RAID-1, served as the reference standard. Any new tube design was intended to 

have at least the same sensitivity and resolution, be smaller, lighter, and have lower power 

consumption. It should also be less expensive to produce. 

Four prototypes of IMS tube are discussed: 

• A miniaturized cylindrical tandem tube, each tube with 3mm internal diameter and a length of 

30-50 mm; 

• A tube with a rectangular cross-section and with multiple electrodes to exert more influence on 

the ion movement; 

• A disc-shaped IMS tube with an ionization region at the periphery and a steel rod ion collector 

in the center of the tube; 

• The “tube box”, effectively a conventional stack ring tube in a cuboidal housing, represents a 

very compact technology for the electrical and pneumatic parts that enables cost effective 

production. 

Three of these models were manufactured as initial prototypes and evaluated for their 

performance characteristics. The results of these investigations will be discussed. 



Ion Mobility Spectrometry for Laser Desorption-Ionization Analysis 

G.A. Eiceman 1 , D. Young 1 , D.A. Lake 2 , M.V. Johnston 2 

1 

Dept. of Chemistry & Biochemistry, New Mexico State University. 

2 

Dept. of Chemistry & Biochemistry, University of Delaware. 

ABSTRACT 

Laser desorption/ionization ion mobility spectrometry is being investigated as a technology for use 

in portable instruments for chemical analysis of adsorbates associted with aerosol particles. 

An ion mobility spectrometer has been designed to allow analysis of involatile analytes by laser 

desorption-ionization off a probe in the presence of an electric field and air at ambient pressure. 

The laser beam (266nm) enters the cell through a fused silica window and strikes a sample probe 

at an incident angle of 15°. Reflected radiation exits through another fused silica window. An ion 

shutter has also been included in this instrument, which allows the broad ion profiles produced by 

desorption off the rod to be modified to features comparable to conventional IMS peaks. Samples 

were prepared by depositing an analyte solution onto a glass probe and allowing solvent to 

evaporate. 

This configuration has been used to study a range of polycyclic aromatic hydrocarbons (PAH), 

including pyrene and coronene. Desorption-ionisation spectra have been obtained for these 

compounds, with detection limits in the sub-pg range in some cases. Effects of laser energy and 

amount of analyte on response magnitude have been studied, and evidence has been found of 

surface layering effects. Reduced mobilities are comparable to previous analyses of these 

compounds, and there was no evidence of analyte fragmentation but some indications of 

clustering. Use of the ion shutter has been found to significantly reduce peak widths whilst only 

reducing S:N by a factor of 2-3, which may allow application of the system to mixtures. 

Defining PAH response signatures is the first step to creating a database for compounds which 

may be encountered when an IMS is used for aerosol analysis. The high sensitivity and enhanced 

resolution of this method demonstrate the suitability for this application. 



Use of IMS for a Study on CS Decontamination 

T. Donnelly 

Police Scientific Development Branch (PSDB), Home Office, 

Sandridge, St. Albans, Hertfordshire, United Kingdom 

ABSTRACT 

CS (o-chlorobenzylidene malononitrile) is an incapacitant used by most police forces in the UK in 

the form of a hand-held spray. While water itself breaks down CS, other decontaminants are 

currently marketed which are potentially more effective than water. The overall aim of this study 

was to determine whether any of the decontaminants tested would be more suitable than the 

current recommendations for the decontamination of CS from vehicles and buildings. Itemiser ® , 

an explosives and drugs detector from Ion Track Instruments (ITI), was modified to detect CS for 

the purpose of this study. Itemiser was chosen for this study following a limited trial of five 

different trace drug and/or explosive detectors based on GC or IMS technology. Under standard 

operating conditions, Itemiser was found to have the best combination of sensitivity, sampling 

efficiency and speed of response. 

By utilising the speed of response of the instrument and the ease of sampling from surfaces, a 

study was carried out which looked at the efficiency of nineteen potential decontaminants when 

added to CS present on twelve different surfaces. Quantitative determination of CS was achieved 

over approximately three orders of magnitude using three peaks specific to CS. The instrument 

was operating in negative (explosives) mode. When used in this mode and under normal 

operating conditions, Itemiser is capable of detecting nanogram quantities of CS, applied directly 

to the paper trap used for sampling. Other instrumentation, based on GC technology, was used to 

verify that any reduction in CS peak strength was due to the neutralising effect of the 

decontaminant and not to any ‘false negative’ effects which might have occurred. 

The study showed that a number of the decontaminants which were tested were more effective at 

breaking down CS than water alone. However, there was also a tendency for these same 

decontaminants to produce permanent staining on the surfaces treated with them. In conclusion, 

therefore, the current recommendations for the decontamination of CS from vehicles and 

buildings, i.e. ventilation and washing with water, is still appropriate 



Near-real-time analysis of toxicologically important compounds using the volatile organic 

analyzer for the international space station Part II 

E.S. Reese, T.F. Limero, 

Wyle Laboratories, 1290 Hercules Dr., Suite 120, Houston TX. 77058 

ABSTRACT 

The management of health risks arising from air contaminants aboard the International Space 

Station (ISS) requires that the air be monitored frequently. Currently on the Space Shuttle, 

real-time monitoring does not occur; instead, in-flight archival sampling of the atmosphere is 

followed by analysis after the flight in the Toxicology Laboratory at the Johnson Space Center. 

Experiences during the Shuttle and Mir Programs demonstrated a need for near-real-time analysis 

of the atmosphere aboard the ISS. The International Space Station Volatile Organic Analyzer 

(ISS-VOA) will provide near-real-time information to aid in assessing the station’s air quality. 

The ISS-VOA is designed to collect a sample from the atmosphere and, using gas 

chromatography-ion mobility spectrometry, detect and quantify 25 toxicologically important 

compounds in near-real time. Compounds of interest include ketones, aldehydes, aromatics, 

alcohols, Freons ® , and others. The ISS-VOA will be installed in the International Space Station in 

2001 and used to help monitor the air quality on the space station. This discussion will cover the 

ongoing preparation procedures including calibration of the ISS-VOA for flight to the International 

Space Station. 



The Future of Ion Mobility Spectrometry in the NASA Manned Space Program 

Thomas Limero 1 , Eric Reese 1 , John Trowbridge 1 , John James 2 

1 

Wyle Laboratories, Life Sciences, Systems and Services 

2 

NASA/Johnson Space Center 

ABSTRACT 

The low cost, low maintenance features of ion mobility spectrometers (IMS) were the primary 

reasons that this technology was selected as the operating principle for the International Space 

Station’s (ISS) volatile organic analyzer (VOA). The VOA is scheduled to arrive on ISS in June 

2001 at which time it will begin the task of monitoring approximately 20 target compounds in the 

spacecraft air. Ground-based operation of the VOA has provided the opportunity to define the 

strengths and weaknesses associated with its operation. Most of these conclusions were 

presented at past IMS conferences. Although the VOA has yet to arrive at ISS, it is time to begin 

delineating the requirements for the second- and third-generation analyzers and to select 

technologies that could potentially meet the requirements. In the past 10 years, many 

technologies (e.g., mass spectrometry, Fourier transform infrared (FTIR)) have made large strides 

in reducing size and cost to become competitive with IMS. 

This presentation will discuss the requirements for the second generation VOA and for the 

following generation of VOA that might support a mission to Mars. Obviously, size, weight, and 

cost are very important parameters as NASA considers future VOA monitors, but other features 

such as performance, resource requirements, and mode of operation also need to be considered. 

A prototype VOA consisting of a small gas chromatograph and lightweight chemical detector, will 

be shown and data from the breadboard system will be presented. This system will provide an 

example, which illustrates the strengths and weaknesses of IMS technology for future generations 

of VOA. 




ISIMS 2000 

SOCIETY 


for 

ION 

SP ECTROMETRY 

MOBILITY 


August 13-16, 2000 

Abstracts of Posters 


Abstracts of Posters presented at ISIMS 2000, IJIMS 3(2000)1, p. 79 

IMS in the Undergraduate Chemistry Curriculum 

S.P. Sibley 1 , B.L. Houseman 1 , W.C. Blanchard 2 

1 

Goucher College, Towson, MD 21104 

2 

Blanchard & Co. Inc. 27 Glen Alpine, Phoenix, Md 21131 

ABSTRACT 

Current programs in chemistry at the undergraduate level provide good theoretical understanding 

of gas phase behavior. General chemistry courses include introduction to the concept of 

ionization energy, electron affinity, kinetic theory, and molecular mobility. Physical chemistry 

courses provide further theoretical background in the kinetics, thermodynamics, bonding and 

quantum mechanics of gaseous ions. Unfortunately, most curricula fail to provide much, if any, 

experimental instruction to verify any of the above gas phase phenomena. 

An effort to fill this significant instructional gap is being developed using IMS as the experimental 

medium. Comparisons, for a series of small molecules will be made, using simulation and 

experimental data. The simulation will use molecular modeling software such as CAChe. The 

experimental data will be obtained using IMS hardware. Basic IMS principles and equipment 

design will also be taught. 

The lecture materials, apparatus, and experiments dealing with ionization energy, electron affinity, 

ion mobility are being developed. A simplified IMS using a non-rad ion source and an ion well in 

place of a shutter grid is being used. 

This effort will be applicable to several areas in the undergraduate curriculum including courses in 

general, analytical., and physical chemistry. In the teaching of general chemistry the electron 

affinity, and the size and shape of molecules and the effect on mobility can be demonstrated. In 

analytical chemistry, the ion source, instrument design and potentials to move and position ion 

can be described. In physical chemistry, the ion energies, structure, kinetic energies, and the 

verification of computer simulation can be demonstrated. 



Manipulating Reaction Ion Chemistry To Enhance Detection In Ion Mobility Spectrometry 

Keith A. Daum, Robert G. Ewing, David A.Atkinson 


ABSTRACT 

In previous ion mobility spectrometry (IMS) studies, alternative reagent ions have been added to 

the reactant ions to enhance the sensitivity and selectivity of analyses. This study investigates the 

formation of product ions in the gas phase through proton abstraction, electron attachment, and 

adduct formation. In the negative mode, the gas phase acidity and electron affinity of the analyte 

have a role in product ion distribution as do the gas phase basicity, electron affinity, and 

concentration of the added alternative reagent ions. This study demonstrates that reactant ion 

chemistry can be manipulated to enhance detection and resolve interference problems during IMS 

analysis by promoting the formation of unique product ion distributions. It is also shown that the 

source chemical of the added reactant ions many have a role in the product ion distribution. 



Ion mobility spectrometer linearity studies of cocaine, heroin, methamphetamine and 

MDEA. 

T. Dallabetta-Keller, A. M. DeTulleo, P.B. Galat and M. E. Gay 

United States Department of Justice, Drug Enforcement Administration, South Central Laboratory, 

1880 regal Row, Dallas, Texas USA 

ABSTRACT 

The Drug Enforcement Administration’s (DEA) south Central Laboratory has been using the 

Barringer field portable Ion Mobility spectrometer ( IMS) to support federal, state and local drug 

cases for over eight years. In the course of hundreds of field operations and court testimony, the 

question of drug residue quantity has been a major concern. The vacuum sweep methodology for 

collecting trace drug residue is utilized by the South Central Laboratory. This collection technique 

focuses on a particular area or object and is strictly random. In addition, the IMS is only linear for 

a narrow range of concentrations and therefore is not a reliable technique for conclusive 

quantification. The linearity range of commonly encountered controlled substances, such as 

cocaine, heroin, methamphetamine, and methylenedioxyethylamphetamine (MDEA) were 

documented. All of the drugs were prepared using authenticated standards at concentration 

ranges from 1 nanogram to 300 nanograms. They were analyzed on the Barringer IMS model 

400. These data allow the forensic chemist to formulate a general concentration of controlled 

substance during field operations. In addition, these data are kept with the IMS log book 

information to be reviewed by the DEA’s instrument review committee for instrument integrity. All 

linearity data were compared to the figures published by the manufacturer. 



Ion Mobility Spectrometry Analysis of LAMPA, GHB, and GBL 

R. DeBono, T. Le 

Barringer Instruments Inc. 

30 Technology Drive 

Warrren, NJ, 07059. 

ABSTRACT 

Results obtained using Barrringer’s IONSCAN ® for the detection and characterization of LAMPA, 

Lysergic Acid N.N. MethylPropylamide; GHB, Gamma-Hydroxybutyrate; and GBL, 

Gamma-Butyrlactone are presented. 

LAMPA and LSD, Lysergic Acid Diethylamide, have the same molecular weight and overall 

general structure. The difference is in the alkyl groups attached to the amide; LAMPA has a 

methyl and propyl group, while LSD has two ethyl groups. The drift times of these two compounds 

were analyzed using the IONSCAN ® in standard narcotics mode, where LSD and LAMPA have 

unique drift times, which are within 100 µsec of each other. The current detection algorithm can 

correctly identify whether the sample is either LSD or LAMPA, since identification criteria requires 

a peak to be within ± 45 µsec of its expected drift time. Instrument sensitivity is of the order of 1 

ng. 

GHB and GBL have gained a loyal following among young people in recent years, becoming a 

staple at many nightclubs and at parties known as raves. GHB has been classified federally in the 

United States as Schedule I, effective March 2000, it is also known as a “date rape” drug. 

Positive and negative IONSCAN ® characterizations of these compounds will be presented. 



Rapid Characterization of Bacteria by Ion Mobility Spectrometry 

R. F. DeBono 1 , J. R. Jadamec 2 and R. T. Vinopal 2 

1 

Barringer Instruments, Inc., 

30 Technology Drive 

Warren, NJ, 07059 

2 

University of Connecticut at Avery Point, 

Groton, CT 

ABSTRACT 

Over the last 10 years Barringer has developed a family of Ion Mobility Spectrometry, IMS, 

instruments to detect explosive and narcotic residues at trace levels, without prior sample 

clean-up, or pre-concentration steps. Recent studies at the University of Connecticut have 

demonstrated that the same rapid technique can be used to detect specific components of 

bacterial cells, and generate unique fingerprints, (complex ion-peak spectra, in both positive and 

negative modes), for a broad spectrum of bacterial species grown on standard microbiological 

media. 



GC for industrial use 

H. Bensch, J.W. Leonhardt, 

IUT Berlin 

ABSTRACT 

The integration of GC separation columns into the gas loop of an IMS requires both a spectral 

sampling technique and data evaluation due to the 3 dimensional spectra. This technique opens 

up a new dimension in supersensitive trace analytics because of the separation of sophisticated 

mixtures before entering the IMS. 

Most impressive results are as follows: 

• the determination of nicotine in smokers breath air 

• the determination of NH 3 in clean room areas 

• the determination of COCl 2 in HCl/Cl mixtures 

• the determination of vinylchlorides in ambient air 

• the determination of CWA mixtures 

GC-IMS opens up new prospects in industrial applications, in research and in medicine. 



The Protection of civil facilities by means of a stationary IMS. 

H. Bensch 1 , J.W. Leonhardt 1 , G. Bensdisch 2 , E. Wolters 2 and K.M. Baether 3 , 

1 

IUT , Berlin 

2 

Drager Sicherheitstechnik GmbH 

3 

Dragerwerk AG 

ABSTRACT 

In the past, terrorist’s attacks were directed to various public facilities. Chemical warfare agents 

and industrial poisons were used frequently. Therefore, monitoring of fresh incoming air is an 

important means of protection of buildings. 

The specificity of such systems is given by following demands: 

• High reliability of the system. 

• No false alarms even at high concentration of organic compounds in the ambient air. 

• Ability to integrate into a more complex gas warning system. 

IMS stations developed and tested arrived the following technical parameters 

• Sensitivities in µg/m 3 : GB-0,7, GA- 0,4, GD- 4,2, VX- 0,4 L-15, HD- 2,1, HN3- 19. Most 

important industrial poisons are detected in the level of 1- 5 µg/m 3 . 

• Response: < 3 seconds at high concentration level 

The complex protection system in air channels of special facilities was tested under the 

consideration of solvents, window cleaners, and other organic compounds in ambient air. There 

were observed some effects due to cross sensitivities with the monomer ions of monitored 

species. By means of the data evaluation code, these effects could be excluded. A number of 

systems have been in operation successfully for 15 months.

Full issue - B & S Analytik GmbH

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?