full presentation - RWTH Aachen University

Mitglied der Helmholtz-Gemeinschaft 

Investigations of Degradation Phenomena in High Temperature 

Discharge Lamps using Thermo chemical Modelling 

Torsten Markus and Sarah Fischer 

Institute for Energy Research (IEF-2) 

Forschungszentrum Jülich GmbH 

52425 Jülich 

T.Markus@fz-juelich.de 

T. Markus RWTH-Aachen Jan 07 

1

Content 

• Insight into Modern Light Sources 

• Corrosion and Chemical Transport 

• Calculation Design 

• Results 

• Conclusion 

16. Juni 2010 Institute of Energy Research (IEF-2) 2

PCA-Lamps 

Lamp-burner 

(PCA) 

PCA = Poly Crystalline Alumina 


Aplications for High Intensity Discharge Lamps 



DyI 3 


Schematic of a High Pressure Discharge Lamp 

Wall Material Al 2O 3 

Tungsten Electrode 

Wall Corrosion 

Alumina 

Deposition 

1380K 

1340K 1340K 

1300K 

5000K 

1310K 

Metal Halide Melt 

e.g. NaI / TlI / DyI 3 

Interior with Gaseous Species 

e.g. Na, I, Hg, Tl, Dy, Ar, Xe, ... 

6000K 

3800K 3000K 

1300K 

Tungsten 

Deposition 


Chemical Transport in Ceramic Discharge Metalhalide 

Lamps (CDM) 

Cross Section of a Corroded Discharge Vessel 

in Horizontal Burning Position (after 9000 h of operation) 

Source 

Sink 

500µm 500µm 

PCA deposition 


Explaination of Transport Phenomena 

AlI 3(g)+Al 2O 3(s)= 3AlOI(g) 

hot 

Al O (s) 

2 3 Al O (s) 

2 3 

Al 

Al 

O 

AlO 

Al2O 

AlOI 

AlI 3 

Salt salt Melt melt 

3 DyI 3(l) + 4 Al 2O 3(s) = Dy 3Al 5O 12(s) + 3 AlI 3(g) 

DyI 3(g) + Al 2O 3(s) = DyAlO 3(s) + AlI 3(g) 

3AlOI(g)= AlI 3(g)+Al 2O 3(s) 

16. Juni 2010 Institute of Energy Research (IEF-2) 9 

Al 

O

From data assessment to an application 

calculation 

Augmented by 

Estimation techniques 

a) experimental trends 

b) ab initio calculations 

Experimental raw data 

(thermodynamic properties and phase equilibria) 

Assessment 

Programs 

Gibbs Energy Database 

Application 

Programs 

Calculations of: 

Thermodynamic Data 

Phase Equilibria 

Process Simulation 


Computation 

T, V = const 

Source 

(Condensed Phase) 

„Cooperative Transport Model“ 

R. Gruehn, H.J. Schweitzer, Angew. Chem. 95, 80 (1983) 

Thermochemical Database 

c p = A + B·T + C·T -1 + D·T -2 

ΔH 

f 

S 

T 

0 

f 

ΔH 0 T c (T)dT 

p 

0 

T 

T c ( T) 

0 

p 

S 0 dT 

T 

T T 

T 0 

p 

Z, 

gas 

ni Z, 

solid 

n 

i 

Komponents i = 1,...N 

Iterations Z= 1,...n 

T 

Computation 

T, p = const 

Sink 

(Condensed Phase) 


SimuSage Modelling 


Equilibrium reactors 


Scheme of the transport program 

Frist Step: 

Input of the salt melt 

Phase splitter 

Iteration 

Cycle 

Sink Source 

Phase splitter 

First Step: 

Input of Al 2O 3 


Example: NaI – CeI 3 – Mixture 

T Sink = 1200°C 

T Source = 1400°C 

V = 0,0285 dm³ 


Comparison between experiments and simulations 

NaI – CaI 2 - Mixture 

Composition NaI, CaI 2 

75% NaI, 25% CaI 2 50% NaI, 50% CaI 2 25% NaI, 75% CaI 2 

Simulation [mol Al 2O 3]: 1,22 · 10 -9 < 2,55 · 10 -9 < 7,16 · 10 -9 

Experimental certification: 

→ Agreement for NaI – CaI 2 – Mixture 


Comparison between experiments and simulations 

CaI 2 – CeI 3 - Mixture 

Composition CaI 2, CeI 3 

90,48%, 9,52% 87,1%, 12,9% 

Simulation [mol Al 2O 3]: 5,08 · 10 -8 < 5,58 · 10 -8 

Experimental certification: ≈ 

What causes this difference? 


SEM – Analysis with 90,48% CaI 2 and 9,52% CeI 3 

1200 °C 1400 °C 


Formation of secondary phases 

Calculated 

Activity 

Al 2O 3_ca(s) 1.0000E+00 

CaI 2_l(liq) 9.4561E-01 

Al 2O 3_cc(s2) 4.9718E-01 

CaI 2_ci(s) 2.4571E-01 

CeOI_ci(s) 1.1649E-01 

CeI 3_l(liq) 5.4007E-02 

Al 2O 3_l(liq) 3.5536E-02 

CeAlO 3_ci(s) 2.0447E-02 

CeAl 11O 18_ci(s) 1.3446E-02 

CeI 3_ci(s) 1.1285E-02 

CeAlO 3_l(liq) 5.4999E-03 

Ce 2Al 2O 6_ci(s) 4.1199E-03 

CeAl 11O 18_l(liq) 3.4167E-03 

CaAl 4O 7_ci(s) 3.0990E-03 

CaAl 2O 4_ci(s) 7.1260E-04 

Al_l(liq) 6.7438E-04 

Al_ci(s) 4.0705E-04 

CaO_ci(s) 1.4573E-05 

AlI 3_l(liq) 2.7872E-06 

CeO 2_ci(s) 1.6025E-06 

Al 9.3713E-07 

Ce 2O 3_ci(s) 6.0181E-07 


Example: NaI – CaI 2 – CeI 3 – Mixture 

T Sink = 1200°C 

T Source = 1400°C 

V = 0,0285 dm³ 

n(CaI 2)=0,0008mol 


Experimental Determination of 

Thermodynamic Data 


Differential Thermal Analysis (DTA) 

Simultaneous DTA with 

Thermogravimetry (TG) 

STA 429, Netzsch 

Principle of DTA 

Measuring of Phase 

Transition 

Temperatures 

Determination of the 

Quantity of Heat 

Studies in different 

Atmospheres 

Thermal Analysis from 

RT to 2800 K 


P T 

T 

T

Principle of Knudsen Effusion Mass 

Spectrometry (KEMS) 

Vaporisation studies up to 2800 K 

Identification of gaseous species 

Determination of partial pressures 

(10 -8 ... 10 Pa) 

Evaluation of thermodynamic data of 

gaseous species 

condensed phases 

Elucidation of corrosion processes 


Mass Spectrometer Knudsen Cell System (CH 5) 


Potential of 

Knudsen Effusion Mass Spectrometry 

Experiment 

Appearance-Potential 

 

 

r T 

G 

2 nd law 

 

 

 

r T 

H 

 

r 298 

H 

 

Identifiction of 

Gaseous Gpecies P i – partial 

pressure 

 

S 

r T 

 

r 298 

S 

 

k ° p – 

Equilibrium Constant 

II 

r 298 

H 

 

3 rd law 

r 298 

S 

 

 

a i – 

Thermod. Activities 


II 

 

 

m ix 298 

H 

G 

E 

m 

 

i

Temperature and Composition dependency 

of activity for the NaI – CeI 3 system 

a 

a 

1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

550 °C 0,8 600 °C 

0,0 0,2 0,4 0,6 0,8 1,0 

625 °C 

0,0 0,2 0,4 

XNaI 0,6 0,8 1,0 

1,0 

0,6 

0,4 

0,2 

0,0 

1,0 

0,8 

0,6 

0,4 

0,2 

0,0 

0,0 0,2 0,4 0,6 0,8 1,0 

650 °C 

0,0 0,2 0,4 

XNaI 0,6 0,8 1,0 


Phase Diagram of the System NaI–CeI 3 (DTA) 

Binary Excess Polynomial: 

G m E =(xNaI i )(xCeI3 j )(A + B*T + C*T*ln(T) + D*T 2 + E*T 3 + F*T -1 ) 

(calculated) 


Overview of the data to be optimized in the 

NaI-CeI 3 system 

Various experimental data on the binary NaI-CeI 3 system have been measured: 

– phase diagram data (liquidus points, eutectic points) 

– liquid-liquid enthalpy of mixing 

– activity of NaI(liq) at different temperatures 

OptiSage will be used to optimize the parameters for the liquid Gibbs energy model 

(XS terms). All other data (G° of the pure stoichiometric solids, as well as the pure 

liquid components) will be taken from the FACT database (i.e. remain fixed). 

A polynomial model for the Gibbs energy of the liquid will be used: 

G = (X 1 G° 1 + X 2 G° 2) + RT(X 1 ln X 1 + X 2 ln X 2) + G E 

where G E = H – TS E 

Using the binary excess terms: 

H = X 1X 2 (A 1) + X 1 2 X2 (B 1) 

S E = X 1X 2 (A 3) + X 1 2 X2 (B 3) 

Hence: 

G E = X 1X 2 (A 1 - A 3T) + X 1 2 X2 (B 1 - B 3T) 

Where A 1, A 3, B 1 and B 3 are the 4 

parameters to be optimized. 


G 

G° 1 

NaI 

x CeI3 

CeI 3 

G° 2

Summary 

Corrosion- and rearrangement – effects of the 

wall material limit the life time of High-Energy- 

Discharge-Lamps 

Cooperative Transport Model was programmed 

with SimuSage 

Simulations of the corrosion speed of lamp 

relevant salt mixtures 

Comparison of the experiments and simulations 

show remarcable agreement

full presentation - RWTH Aachen University

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?