College of Earth andMineral Sciences

,V,PENNSTATE.. .- :_-...

I. ! . . .... .. . .-- ----.. . - .- - .. . . . Z .- . -," -.- " ' -

ELECTE " - " . . . .

- - ,... .. a'.-'; J METALLURGY PROGRAM o=.4 -'- -

TECHNICA E -

" 7 .;"'.-"~~~~~-"*1 ....ECR OR ,.. . °" ' "'

.. . .. . . . . . .. ..-. ...- • ':

SDecember 1989 - -

SContract No. NOO14-4-k-0201HYDROGEN SULFIDE EFFECT ON HYDROGEN ENTRY I '

-- T-- ECIANISTIC STUDY . -* . ~ . Contrac .N. Tikrn001 '8kj2, ~

- D partientof MarteriaIs- Scienc -an-dEngineering -"4-.. % 7" .

.- The Pennsylvania State University _ -,.', - - "

. .. . .. ..... . * . .. .. -_ . .. -'. - -,- , . 5

I" DLi: UTI-.'--- A T- .A. "- N-, - -A - --

Approved c pu-. '-- . .. ...." h " i-'o l-Dut .zuon U-,., tLrtod 71 _ -. .. 4z

--- 21 019_ '

* ~*89 12 1

PENN STATECollege of Earth andMineral SciencesUndergraduate MajorsCeramic Science and Engineering, F%;€l Science, Metals Science and Engiccnng, Polymer Sccncc,Mineral Economics; Mining Engineering, Petroleum and Natural Gas Engineering;Earth Sciences, Geosciences: Geography; Meteorology.

Graduate Prbgrams and Fields of Research . - ..... - --

- Ceramic Szicncc and Engineering, Fuel Science, Metals Scicace and Engineering, Polymer Science, -... .: Mineral Economics, Mining Engineering, Mineral Processing, Petroleum and Natual Gas

* Engineering;-,-4-**Geochemistry and Mineralogy, Geology, Geophysics; Ccography; Meteorology.

Universitywide Interdisciplinary Graduate Programs Involving EMS Facultyand Students ' -.. .: .. ' . . -Zarth Sciences, Ecology, Environmental Pollution Control En.inccang. Mineral EngineeringManagement, Solid State Science.

Associate Degree Programs - -, ..

Metallurgical Engineering Technology (Shenango Valley Campus). -

Interdisciplinary Research Groups Centered in the CollegeC. Drew Stahl Center for Advanced Oil Recovery, Center for Advanced Materials, Coal Research , -- ..

Section, Earth System Science Center, Mining and Mineral Resources Research Institute, Ore . ..... •Depowsis Research Group. ,: "'

..... Analytical and Characterization Laboratories (Mineral Consttuton, - -Laboratories) K '4 .- *---*

Servi,.es available include, classical chemical analysis of metals and silicate and carbonate rocks, " -.

X-ray diffraction and fluorescence, electron microscopy and diffraction, electron microprobeanalysis, atomic absorption analysis, spectrochemical analysis, surface analysis by secondary ion .. - " -mass spectrometry (SIMlS); and scanning electron microscopy (SEM). - -. "-

The Penryivan.a State Univcrt3 . il ornpisance wid fedcsr2 ald sae laws. is corm ed to die poi0t1 t | ail penuns 1a i have equalace to programs. admission. aM employment wittho eg=d o) race. reigioa. s A. naon onpn. handa.:p. age. ix status a Jisabled of

-- n wma-ran-evetran. inm ail 4rfmavlactoio inquwim to dic Affirmative Acton 0fri B. Suznne Bwooa. 101 9.iiad Oiding. .

. Univcrnty Park. PA 16802; (14) 863-0471.- , . _ . -

- Ul Ud.s.i 027 -.

Produced by the Penn State Depa ien:t Publicanoo - ""

'REPORT DOCUMEHAfTION PAGE

la. AraotUr cuI .LAu1Wty, WW3am Ith. ALSIIE M.llvJl

NONE _JA. MCUJMY 0-4ICA110 AJIWY O.*e1)QaUIx~rIAvAj.&.ju1Y OA A(,OM

ib. D(CLA1S11ICA)IOM10Ow, Cj' YJW.I4CX..L

4. ?tDJOAMNC OAG.A't0ZAIOt4 A.LOAT ktjMjSS. MOMI0ALWA OAG .Aft r4 APA NuMsA4S)

La. KAMEJ Of ? AJQ'MiY-ft O.AC^LLA1)i 4b. 0uFCE SY*1 7,% NAU~ Of M01opitrua OgA-jA.j1oM

Howard W.. Pickering I i p-office of Naval Research

U. AZO At SS (0 rr. SUj I. Ana 2.pCoaql j Lb AOO&Utom l Sr. W .£JC004)

The Ptnnsylvania State University Arlington, VA 22217

University Park, PA 168021

U~. ftAMI OF IUliOINjft ?ej1OA'h a b. F FLS 1r-" 9. rxojwPAujr IN$1AU1AiHr DotflFicAtIOm Numsco

k. AOOXA(C.11416 a.4 Jnc, Poo. 10. SOURCE Of fUNOlhG humatis

PlAOGLAJ-% PiROJECT rAssc YiA4 ut-4r

tLLMLIhr too NO. NO . A4 asso S , O

11.___11 __ME______________________bon__ 84-K-0201 14315098f"Hydrogen Sulfide Effect on Hydrogen Entry Into Iron - A Mechanistic Study"

LL. ?CtOA. AuTIIoP4s)

Pickering, Howard W.

19. SUMt.MENT~AjA N~OTATION4

17. C-C)"1I CONSI IS L~ suj aS (Cnr.6 tc on ",env~ it rwwd sA .0vnojy by D'OCX ,-~

1C GAOUP SU"-AOUF Hydrogen charging in iron; HiS eftfect in acids on: surfaceH coverage, entry and overpotentUal; h-e.r., absorption and

I adsorption rate constants; modified I-P-Z model.

T1hc recently developed I-P-Z model is moldified in orderto analyze the observed enhanced permeation ofhydrogen th~at occurs in the presece~ of hydrogen sulfide during cathodic hydrogen charging of iron. Themodification accounts for the fact that the energy of adsorption becomes coverage dependent at the highercoverages and affects the hydrogcn evolution reaction (h.e.r.) in the pIesence of H2S. Charging experiments wereperformed on Ferrovac E-lron membranes 0.5mm thick using a Devanathan-Stachursi cell in deacrated, prc-electrolyzed solutions made from 0.lM HClOYt and 0.IM NaCIO 4 with pH values of 1 and 2. The transfercoefficienit, Xi, exchange current density, io, thickness-dependent absorption-adsorption raje constait, V.~recombination rate constant, k4, surfacc hydrogen coverage, Qjj and discharge rate constant, k .. we- e obtaamncdby application of the model to the. experimental results. As a result, the role of HiS has been ;- 1-fied. WYhilr QH-is increased in the presence of I-1S, the overpoteidal, d , is decreased consistent with an Li, -veuincreat .. in a,and the increased H entry is found to be the result of a decreased kj as well as the increased cr.. In addition, a ve-ryimportant relationship hias been derived that will enable the caldulation of the absorption, kgbs, and adsorption,kads,rate constants from the electrochemical permeation results for different membrane thicknmsses. d:

:0. 01$rAIS,,IO14jAVA1L.SjUrY Of A.ITLA.CT 1i. ,..OTIU.CT $LCUjryfl OL.SI1CArIONIUGJCLASfaohur 0Jh1ATE : $AAA(E AS Are 0 or< US r15 I

I. UAh Of RP£UohsiaLJ. a.oivouA.L. 11. t LIKhOOnKt oolO Ota o.s'j.j,.J '2r.I O1K ET

A. t. Sedrikr (202) 696-4401 1j -- A

.ZFn(r.n1473 JUN 8 r iouni t,0,0ru ,,rw ,. 000.i. C 0A; '.'C..I O Ac ,,r

841021 359

"Hdoe u-d fc nHdoe ntyIt rn-ALehnsi cdv

HYDROGEN SULFIDE EFFECT ON HYDROGEN ENTRY INTO IRON- A MECHANISTIC STUDY

Rajan N. Iyor, hizumi Takeuchi*, iMelirooz Zainanzadehi** andl-Howard W. Pickering

Department of Materials Science and EngineeringThe Pennsylvania State University

University Park, PA 16802

ABSTRACT

The recently developed I-P-Z model is modified in order to analyze the observed enhanced

permeation of hydrogen that occurs in the presence of hydrogen sulfide during cathodic hydrogen

charging of iron. The modification accounts for the fact that the energy of adsorption becomes

coverage dependent at the higher covei-ages and affects the hydrogen evolution reaction (h.e.r.) in the

presence of H2S. Charging experiments wvere performed on Ferrovac E-Iron membranes 0.5 mil

thikkusing a Devanathan-Stachurski cell in deaerated, pre-electrolyzed solutioas made from 0I

HCIO4 and 0.lM NaC1O 4 with pH values of 1 and 2. The transfer coefficient, ...,exchange current

density. ix,, thickness-dependent absorption -adsorption rate constant, k", recombination rate constant,

k3,stirfice hydrogen coverage, 0&~ and ditcharge rate constant, , 0 , were obtained by application of

the model to the experimental results. As a result, the role of H-2S has been clarified. While OH is

increased in the presence of H-2S, the overpotential, il, is decreased consistent with an observed

increase in a, and the increased H entry is found to be the result of a &e.reased k3 as well as the

increased (4. In addition, a very important reladonship has been derived that N% ill mnable the clculation

of the absorption, ka1bs, and adsorption, kLads, rate constants from flic electrochemical permeation results

for different membrane thicknesses.

SPresent address: Kashima Steel Works, Sumitomo Metials Industries, Ltd., Kashimia, Japan.

Present address:, Professional Services Industries - P T L Division, 850 Poplar Street, Pittsburgh,

PA 15220.

€ ," 2

INTRODUCTION

Since tile failure of steels in wet hydrogen sulfide (H2{S) is well known to occur in many

technological situations, especially in oil and gas field applications .1). many studies have been carried

out in order to better define and understand this problem. It is well known that in the presence of

H2S, the concentration of h,,drogen in steels is increased sharply (1-6). It is also now generally

accepted that when the concentration of hydrogen in a metal reaches a critical level, cracking .an occur

(7).

It has often been suggested (6), batsed on a measured increase in hydrogen uptake, that H2S has

a "poisoning" effect on the hydrogen evolution rmaction(n.c.r.). Contrary arguments based on

polarization and permeation experiments have provided different mechanisms 3,4,5,8) for the role of

H21S in the entry of hydrogen into iron and steel. There is also a mechanistic dilemma on tile

significant effect that H21S has on the kinetics of ihe h.e.r. and hence on hydrogen absorption.

There are at least four main proposals of the mechanism of the enhancement of hydrogen

absorption by iron due to H21S. The first of these, which was developed by Iofa and Kain (8)

considers the following sequence of reactions:

Step 1: HS- = (HS')ads

Step 2: (HS)ads + H30+ = (H-S-H)ads + H20

Step 3: (I1-S-H)ads + e" > (HS')ads + Hads

In this model, steps (1) and (2) are quasireversible and step (3) is considered to be rate limiting. On

the basis of their experimental data in 0. IN H2 SO4 , Iofa and Kam concluded that the HS- anion

accelerates the discharge of H+ ion. However, their interpretations were incomplete in the se that

the reaction parameters remained unknown and in fact, there is a far higher concentration of H2 S

compared to HS in acid solutions (9) as shown in Figure 1. So, Kawashima et al (3) proposed the

following reaction sequence: -.

Step 1: HI2Sads +e - H2Sads- ;

Step 2: H2Sads- + Hads+ - H2S ...... Hads

Step3: H2S...... Hads --> H2Sads + Hads . i .,L

f 1 3

They showed that 11S did not change the Tafel slope, but did decrease the cathodic overpotential in

acid solutions. In their model, H2 S works as a bridge to transfer the electron from the iron surfacc to

the hydrogen ion.

The poisoning model of Bockris et al 10) attributes the increasc in hydrogcn absorption to a

lowering of the IM-Hads bond energy, w. Lowering of w increases the activation cnergy for the

discharge process of the h.e.r. Thus, in systems where this model is applicable, the hydrogen

overvoltage would increase with increasing concentration of the poison. However, it is now well

established that H2S decreases the hydrogen overpotential (3,4,5,8) and, thus, the poisoning model is

not applicable in this case.

Bolmer (11) proposed an overall reaction model:

2 H2S+2e- - H,+2HS"

and rationalized that at currents below the diffusion limited current, the hydrogen evolution mechanis,n

is a function of the HS- concentration and/or the H2 S/HS" ratio. However, his

analysis was incomplete in that the reaction parameters for the h.e.r. and hydrogen absorption

remained unknown.

In the studies cited above, the experimental data were not sufficient and the interpretations were

based mostly on experiments which were not well defined. Also, up to the present time and because

of its complexity there seems to be no general model which can rationalize the significant effect of H2 S

on the h.e.r.

This paper reports the results of experimental studies directed at evaluating the effect of H2 S on

the hydrogen permeation and evolution reactions. It analyzes these results with a newly developed

general model of hydrogen entry into metals (12,13) in order to gain an insight into the permeation

mechanism and to arrive at otherwise difficult-to-obtain parameters such as the surface coverage and

the individual rate constants.

4

EXPERIMENTAL

A Devanathan-Stachurski cell 14 %%-as used to measure the permeability of hydrogcn through

iron membranes as a function of the charging current, pH aad I.IS concentration in the charging

solution. The apparatus shown in Figure 2 consists of two identical electrolytic cells separated by the

iron membrane. One of the membrane surfaccs (input surface) was galvanostatically charged with

hydrogen, using a Pt counter electrode. The potential of the input surface was, mcasurcd with a

saturated calomel reference electrode. The other membrane surface (exit surface) was coated wich Pd

and held in 0.1M NaOh at a potential of 150 mV (versus a HgO/Hg reference electrode) which

oxidized all of the hydrogen diffusing through he membrane; this oxidation current provided a

measure of the hydrogen pcrrneation flux.

The membrane was Ferrovac E-Iron, annealed in evacuated capsules at 9300C for two hours and

firmace cooled. The final preparation involved polishing with 600 emery paper and tcgrcasing with a

mixture of benzene and acetone. The final thickness of the iron specimen was

0.5mm and this iron membrane was fixed between the two cells, tightly sealing them with nbber 0-

rings. Details of the cell design, circuitry, edge effects, and special procedures can be seen elsewhere

(15)

Mixtures of .1. rICO 4 and 0.1 MI NaCIO 4 were used to make up the charging solutions with

pH values of I and 2. The solutions were pre-electrolyzed for 72 hours under an atmosphere of

prepurified nitrogen. To make up solutions containing H2S, 'irst a master solution of H-2S was

prepared by passing H2 S gas of 99.6% purity through the deaerated, pre-electrolyze 0. iM HC104

solution. The concentration of H2 S in this master solution was measured with an ion selective

electrode and found to reach 5.2x10"2 M, after two hours of H2 S flow through the solution. The

master solution of H2 S was mixed with the charging solution in order to give H2 S concentrations of

10-5 to 10-3 M, while maintaining the ionic strength constant.

In order to obtain reproduciblity of the permeation data, several different precharging procedures

were evaluated. The procedure that gave good reproducibility consisted of the following steps. The

charging solution was pre-electrolyzed in a pre-electrolysis cell with Pt electrodes at 2 mA for more

5

than 72 hours prior to starting the experiment. llen, the iron membrane was precharged at I mA for

30 minutes and then at 0.5 mA for 3 hours (during this time, the charging potential moved in the

negative direction s the permeation current increa.scd). Following this step, the cell % %as drained and

filled again with fresh pre-electroiyzed solution. Subsequently,, the polariztion and perncation data

werc recorded proceeding from a staring higher current (7- ImA cm"2) to lower currents. To perfomi

experiments with H2S , the charging solution was changed to the H2S pre-mixed solution and the

same sequence was repeated.

RESULTS

The cathodic polarization curves for different H2S conctntrations (ci1S) in the charging solution

at twt. different (acidic) pH values are given in Figures 3 (a) and (b). It is easily observed from these

plots that the hydrogen overvoltage (1) decreases asymptotically with increasing H2,S concentration.

Figures 4 (a) and (b) display the steady state permeation current (i.) as a function of the hydrogen

overvltage (ij), for the pHil and pH2 charging solutions. From these figures, it is obvious that i.

increases substantially in the presence of H2S. They also show that while all of the In ic vs ri and In

i. vs rj plots in the non H2S-containing solutions are linear, not all are linear in the presence of H2S.

In order to understand the mechanism and quantitatively analyze the partitioning between the

hydrogen absorption and hydrogen evolution reactions, the recently developed I-P-Z model (13) is

applied here. According to this model, the i., vs "ir plot (where ir = ic-i. is the hydrogen

recombination current density) and fi'j (=icea°l) vs i. plot (where c. is the h.e.r. transfer coefficient

and a = F!RT = 38.94 V-1) will be linear if the mechanism of the h.e.r. is a non-activated (i.e., the

energy of adsorption is not coverage dependent) discharge-chemical recombination process. Figure

5(a) shows that the relationship between i. and 4ir is linear at pH=1, for solutions with and without

H2S. Figure 5(b) shows that the i. vs 4ir plot is linear at pH=2 in the absence of H2S and also when

cH2S=2.5xl0"5 M whereas it is nonlineai for the higher H2S concentrations. The corresponding

vs i. plots are given in Fictures 6 (a) & (b). Although there is a fair amount of scatter in Figure 6(a),

mainly due to the lack of knowledge of the exact a values, the fij vs. i. plot is found to be linear

6

for the lower H2S containing pHI solutions. On the other hand, for the pH2 solutions (Figure 6(b)),

the fi vs. i. plot is found to be lincar for c1js=0, but nonlinear for c,1 .s=2.5:x|0" 5 M. *hus, for the

higher pH and higher HI2S containing solutions, deviations occar in the square-root rclationship

(Figure 5(b)) or the fin-i.. plot becomes nonlinear (Figure 6,b)), or both.

ANALYSIS AND DISCUSSION

1. Dwz Arnalsis Using the Basic I-P-Z Model

Since the i,. vs 4i. and fill vs i. plots are linear for the pHI solutions except at the highest H2S

concentration, including cl,,s=O (Figures 5 (a) and 6 (a)) zs well as for the pH2 solution at c1js = 0

(Figures 5 (b) and 6 (b)), the basic I-P-Z model (13) can bc applied for these cases. The following

equations represent the fluxes or currents of the discharge, recombination and pemieation reactions

(symbols are defined at the end of the text) (13):

ic = Fkj (l-0H)e-aotl = io' (1-0H) e-aa (1)

where io ' = F kl 3H+ C-aaE q

ir F k3 OH 2 (2)

io= F (DI/L) cs = cs/b (3)

OH= cs/k" (3a)

k" = kabs/(kads + (DI/L)) (3b)

And from these equations, the following relationships have been obtained(1 3):

i.= (k"/b) (F k3)' 0 .5 qi r (4)

ic eaa ---(b io'/k") io + io (5)

Thus, from Figures 5 and 6 (in essence, by regression analysis of Equations (4) and (5)) and Eq.

(1), all of the following constants are computed: the exchange current density, io (=io ' (1-0 ,)); the

disciharge reaction rate constant, ko; the recombination reaction rate constant, k3, and the thickness

dependent absorption-adsorption constant, k". For details of the computation procedure, see

elsewhere (13).

.K 7

2 Uodifiel I.P.Z Modd Incorporating Frwnkin-Teinkin Corrccrion for Activared Reactions

For pl-I = 2, the presence of I-I.,S causes a deviation from a linear i. ",ir tbchavior i Figure

5b) or linear fil c4 i,., behavior (Figure 6b). One possibility is the adsorption of HS' but this

would requirc that the pH in the interface region rise above -4.5 since it is clear from Figurc I that the

dissociation of H2S is negligible in solutions of pH < 4.5. Another is that H2S" can be formed (16,17)

by the following reaction:

H,S +e" 4 H-Sads (6)

having a high reaction rate constant. This fast side reaction occurring at the cathodic metal surface can

have three different effects:

(1) II2S can function as a bridge for hydrogen discharge through the 12 S" species, similar to the

proposal by Kawashima et al (3):

H2Sads + H30+O (H .... HS) + 1-10 (7)

(H .... H2 S) + 1420 - (- .... H2S)ad s + '120 -- Hads + 112S + 1"120 (8)

where H2Sads discharges the proton (Eq. (7)) and becomes the stable 112S molecule (Eq. (8)). This

step can decrease the overvoltage for the discharge reaction, in accord with the expcrimnental

observations (Figures 3 (a) and (b)).

(2) Besides the above effect, the recombination reaction can also be restricted effectively by the

(H...H2S) intemiediate of this side reaction slowing thc diffusion of hydrogen adatoms and/or

blocking the hydrogen recombination sites on the mletallic surface.

(3) At higher pH, in view of the higher [H2S]/[H+].ratio, chemical desorption is less strongly favored;

i.e., effect 2 could lead to a change in the mechanism from a primarily chemical desorption to an

electrochemical desorption mechanism.

The adsorption of HS" on t,'ie surface can produce the same result as the (H .... H2S) species in

effects (2) and (3) but HS" cannot match H2S" in Ittilittting the proton discharge step (effect (1))

because HS, unlike H2S, is not stable so the bridge mechanism is the more likely one. The above

effects (2) and (3) indicate activation of the hydrogen evolution reaction (h.e.r.) and under these

, 0 8

conditions, the Frumnin-Tcmkin (F-1) corrections have to be applied to the discharge and

recombination currents (18); thus, Equations (1) and (2) will be modificd as follows:

ic = io° (1-01i C e"a cf e'FOH (9)

ir = F k3 OH 2 Ccf OH (10)

where f = */?.r zind y is the gradient of the apparent standard free energy of adsorption with coverage.

Although not considered in Eqs. (8) and (9), the coverage by H2S" may also be important as far as site

availability and activation of the reactions are concerned. In order to determine the effect of H2S, the

relationship between OH and OH2S needs to be quantified by considering the change in OH by

additions of H2S to the solution. The implicit assumption in the present model is that this H2S effect

%;an be neglected as if the H2S', in acting as a bridge to discharge thc proton, will not occupy a surface

site for a prolonged duration. A typical value of f is 4 to 5 for hydrogen coverages (18); f = 4.5 will be

used here.

From Equations (3) and (3a) we obtain,

OH = b iJk" (11)

Inserting Equation (11) into Equation (10), it can be shown that

In (eir/i.) = (afb/k") i. + In (bjFk3)0 '5/k") (12)

Substituting OH from Equation (11) into Equation (9), one obtains

In [i e(i-k")/(l-bij.k")) = - o a 11 + In (io ' ) (13)

or

In (fiei .) -ar + In (io' ) (14)

wherefie~i" = ice e(afbi j k" )/(I -bi ,,/ l") (14a)

Thus, Equation (12) tells us that i. will not be linearly related to 4ir, when the h.e.r. is activated (f>

0), but that i.* will be linearly related to In ('Ii/i*). Also, according to Equation (14), In (fiei*) will

be linearly related to "l when the h.e.r. is activated.

9

3. Data Analysis by the Application of F-T Corrected I.P-Z Model

Figure 7 shows that the plots of In ei,..) vs i.. at pH = 2 are linear in accord with Equation

S12) for c=lOA 4 M and 1tls=3xl04 M. But for the lowest HS concentration, c1lzs-2.5x10"5 NI,

the plot may not be completely linear and is consistent with the h.c.r. being only partially activated, as

is already indicated above in the i, vs 4i, plot (Figure 5(b)) which is linear and in the fil vs i.. plot

(Figure 6(b)) which is nonlinear. Extended analysis showed that the data for ctnis=2.5x10 "5 M are

better analyzed with the F-T corrected I-P-Z model.

The value of a is determined from Equations (12) and (14) by an iterative procedure. Using an

initial value of ac = 0.5 and the k" obtained from the slope of the (-,r /i.) vs. i. plot (Eq. (12)), In

tfi.) is known for various values of ie and i. and as a function ofri. Thus, a regression analysis

of Eq. (14) can be done to obtain a new value of ox. Using this new value of cc, a new value of k is

obtained (from Eq. (12)) and a regressio:. analysis is again performed on Eq. (14). This procedure is

repeated until the a obtained from the slope of the In (fici) vs ii plot (Eq. (14)) converges to a fixed

value. Then, the In (fii..) vs il functions are plotted (Figure 8), and, besides a, k3, k" and io' are

determined from the regression analysis of Equations (12) and (14).

Having determined k" for all combinations of pH and c 12s (ic. with and without the F-T

correction to the I-P-Z model), OH (calculated using Equation (11)) vs 71 can be plotted and these are

shown in Figures 9 (a) and (b). These plots demonstrate that H2,S increases the hydrogen surface

coverage (OH) quite significantly and that the increase is asymptotic with increasing cH2s. In fact, at

pH = 2 (Figure 9 (b)), the 01i values increase significantly with small additions of H2S, but do not

increase any further with higher cjjs within the expeimental error. One explanation of this

asympototic behavior is that the adsorption sites, which could be a small fraction of the total surface

sites, e.g., kinks, become occupied at low H2S concentrations. Thus, the effectiveness of H2S in

slowing down the diffusion of H atoms on the surface and/or blocking the recombination sites with

side reaction (6), thereby increasing OH, reaches a saturation limit. Another explanation is that H2S"

adsorbs readily on the surface and at rather low H2S concentrations in the solution begins to interfere

with adsorption sites for hydrogen on the surface. The limit of this effect could be a decrease in GH

with increasing H2S concentration, which is not observed in Figs. 8 and 9. The equilibrium hydrogen

4 , 9

Figure 7 shows that the plots of In Nit/i,.,) vs i. at pH = 2 are linear in accord with Equation

(12) for cts=10"4 M and c1i1s=3xl0'4 M. But for the lowest H2,S concentltion, c11,s=2.5xl0"5 ,

the plot mnvy not be completely linear and is consistent with the h.e.r. being only partially activated, as

is already indicated above in the i, VS 4ir plot (Figure 5(b)) which is linear and in the fil vs i,. plot

(Figure 6(b)) which is nonlinear. Extended analysis showed that the data for clits=2.5xl0"5 N1 are

better analyzed with the F-T corrected I-P-Z model.

The value of a is determined from Equations (12) and (14) by an iterative procedure. Using an

initial value of c = 0.5 and the k" obtained from the slope of the (r /i.) vs. i. plot (Eq. (12)), In

(fi,i . ) is known for various values of ic and i. and as a function ofri. Thus, a regression analysis

of Eq. (14) can be done to obtain a new value of a. Using this new value ot o, a new value of k" is

obtained (from Eq. (12)) and a regression analysis is again performed on Eq. (14). This procedure isrepeated until the cc obtained from the slope of the In (fie,i) vs il plot (Eq. (14)) converges to a fixed

value. Then, the In (fieci) vs il functions are plotted (Figure 8), and, besides a, k3 , k" and io' are

determined from the regression analysis of Equations (12) and (14).Having determined k" for all combinations of pH and c11 s (ie. with and without the F-T

correction to the I-P-Z model), 6H (calculated using Equation (11)) vs il can be plotted and these are

shown in Figures 9 (a) and (b). These nlots demonstrate that HS increases the hydrogen surface

coverage (OH) quite significantly and that the increase is asymptotic with increasing clbS. In fact, at

pH = 2 (Figure 9 (b)), the OH values increase significantly with small additions of H2S, but do not

increase any further with higher c, 2s within the experimental error. One explanation of this

asympototic behavior in that the adsorption sites, which could be a small fraction of the total surface

sites, e.g., kinks, become occupied at low H2S concentrations. Thus, the effectiveness of H2-S in

slowing down the diffusion of H atoms on the surface and/or blocking the recombination sites with

side reaction (6), thereby increasing OH, reaches a saturation limit. Another explanation is that H2S"

adsorbs readily on the surface and at rather low H2S concentrations in the solution begins to interfere

with adsorption sites for hydrogen on the surface. The limit of this effect could be a decrease in OH

with increasing H2S concentration, which is not observed in Figs. 8 and 9. The equilibrium hydrogen

10

coverage (0.) can be obtained by extrapolating the OH vs i1 plots to 11 = 0. and these plots are in

Figure 9c. Then, io is obtained as: io= I (1-0), and k is obtained from Eq. 1.

4. H2 S Effects on the Various Kinedc Paraneters and the h.e.r. Transfer Coefficient

In order to thoroughly and quantitatively understand the mechanism by which H1-2S enhances

hydrogen entry into iron, al. of the important quantities are plotted versus etjS (Figures 10 to 15).

Figure 10 shows a plot of io as a function of CHs , indicating that io is unaffected by H2S. On the

other hand, cc is increased significantly by small additions of H2S as shown in Figure 11. Figure 11

also shows within the experimental scatter that ot asymptotically reaches a concentration independent

value at higher concentrations of H-S. The general increase in (x can be explained by the Kawashima,

et. al., bridge mechanism (described earlier) that facilitates easier proton discharge, and the asymptotic

behavior of ct with Cl12S may be a consequence of the H2S" saturation discussed above.

A plot of the discharge rate cunstant (klo) as a function of c112s , Figure 12, shows that kl1 is

unaffected by H2S. The large scatter in Figure 12 may be due to the actual desorption mechanism

being partly chemical recombination and partly electrochemical, especially when the c142S is high since,

as is reported below, chemical recombination less readily occurs with increasing H2S concentration.

Due to the extreme complexity of this problem and a lack of the pertinent kinetic data, an analysis

considering both of these desorption mechanisms simultaneously is not being pursued at present.

Another important parameter is the thickness dependent absorption-adsorption constant (k")

plotted as a function of c 2s (Figure 13), showing that k" is not altered significantly by H2S. The

slight increase that may be indicated in Figure 13 means that absorption is favored over adsorption

with increasing H2S concentration. Hence along with the increasing o and decreasing k3 (reported

below), an increasing k" may also contribute to the increase in steady state permeation with increasing

H2S concentration. An increase in k" could be caused by any of the factors associated with the energy

of adsorption becoming coverage dependent. The data in Figure 13 include those for pH = 1 (without

F-T-c) and also those for pH = 2 (with F-T-c) and they seem to merge quite well, further supporting

the proposed mechanism.

From tile foregoing discussion, the H2S bridging during hydrogei discharge (reactions (7) &

S)) and the activation of the h.e.r. by HI2S (ic. the slowing down of the d.ffusion of hydrogen

adatoms and/or the blocking of the hydrogen recembination sites) are considered to be the two major

effects of H2S. A further confirmation of the importance of these effects of H,S would be a

significant decrease in the rccombination rate constant with increasing C11 5s. Indeed, that this occurs,

is clear from Figure 14, which shows k3 decreases with addition of H2S to the solution and as with

the other parameters, then levels off with further additions of H2S.

5. Hydrogen Evolution Reaction Mechanism

Finally, the potential range in which the coupled discharge-recombination mechanism operates

can also be determined from the I-P-Z model (13) as follows:

ticI = [In (10 kl/k3 )]/(aa) (15)

• cu [in (0.1 kl/k3 )]/(aco) (16)

where ieI is the lower limit and TIe u is the upper limit of the potential range. The calculated values of

TcI and "e11 are plotted in Figure 15, as a function of et12s and cH+. The two solid slanted lines are

drawn, one through the lower points and one through the upper points, to indicate the lower and upper

bounds of the hydrogen overvoltages calculated for the coupled discharge-recombination mechanism

using the ki, k3 and a values obtained by application of the model to the data for the different H2S

concentrations, while the horizontal lines at "l =-0.35 and 11 =-0.45V indicate the lower and upper

bounds of the experimentally measured hydrogen overvoltages. For example, in the absence of H-,S

(which corresponds to the abscissa value of zero), the experimental overvoltage range is at less

negative electrode potentials than the overvoltage range for operation of the coupled discharge-

recombination mechanism. So, the discharge reaction may be rate controlling in this case. But with

increasing cH2s, the overpotential range for operation of the coupled discharge-recombination

mechanism shifts to lower overpotentials so that above a certain combination of the concentrations of

H2 S and H+ (i.e., [H2 Sj0 "2 5/[H+] 0 " 12 5 :; 0.2), this range overlaps the overpotential range of the

experiments (shaded region in Figure 15). Thus, with increasing H2S concentration, the h.e.r.

mechanism shifts from proton discharge controlled to coupled discharge-recombination controlled.

12

At higher pH values, one would expect that the effect of HIS in slowing tile "."ffusionl of

hydrogen adatomni will be even larger because of the higher (1'I2S]/[HI- ratio. What this would mean

is ihat the mechanism of H2 fornation tie. hydrogen lion) m11y, have to change into an

electrochemical desorption mechanism at higher p11 values. TI'lus prediction is full, consistent with the

findings of Murayama etal (19) in that the electrochemical combination route (ic. Had + HI+ + c" --4

H2 ) dominates over the chemical kombination route for 99.9% iron in acetate solutions of pH > 3 in

the presence of H2 S.

One final comment is to be made with regard to the importancc of the Equation (3b) in

enlightening our understanding of the surface and near surface reaction kinetics. Rewriting Equation

(3b), we obtain

(k')" = (kads/kabs) + (DI/kabs) (L) (17)

Equation (17) has been discussed elsewhcre( 1 3). The essence of Equation (17) is that if the plot of

1/k" vs /i. is linear, the values of kabs and kads are obtained (knowing DI). Thus, from the complete

polarization and permeation data for different memhrane thicknesses (L), one can systematically

investigate the important surface properties relating to the hydrogen absorption and adsorption rate

constants.

CONCLUSIONS

1. Cathodic hydrogen charging experiments were performed on Ferrovac E-Iron membranes in

perchloric acid solutions of pH 1 and 2 with and without H2 S. The steady-state, hydrogen permeation

current density (i0..,) was greatly enhanced in the presence of H'12S, while he hydrogen overvoltage (TI)

was decreased.

2. The recently developed I-P-Z model (13) is applicable to the h.e.r,/hydrogen permeation

process in the absence of H2S and for the pH 1 solutions of lower H2 S concentrations. From the

analysis the various rate constants, exchange current density, transfer coefficient and hydrogen

surface coverage were obtained.

13

3. For the higher pH and higher I'H2S concentrations, the I-P-Z model was modified by utilizing

the Frumkin-Temkin correction for ihe proton discharge and hydrogen recombination reactions. This

correction was necessary due to tih energy of adsorption becoming coverage dclldcnt, probably a

result of the product of the fast side reaction, H2S + a' -- I., restricting tle diffusion of

hydrogen adatoms and/or blocking the hydrogem recombination sites on the iron surface.

4. The hydrogen surface coverage (0H) and the transfer coefficient (t) are enhanced while the

recombination rate constant (k3) is decreased in the presence of H2S. The increase in cc explains tile

observed decrease in ovcrpotcntial (ij) on tie basis of the Kawashima cE al (3) bridging effect of H,S

(i., the overall reaction: HI-2S'=,+ H30+ - - + H2,S + H20).

5. On the basis of tile bridging effect and the slowing of tih recombination reaction, all of the

observations and calculated parametric changes by H2S have been rationally explained. These

parameuic changes are the increase in the transfer coefficient (a) and the decrease in tile recombination

reaction rate constant (k3). In addition, the overvoltage range for operation of the coupled discharge.

recombination mechanism is predicted, by application of the I-P-Z model to the data, to be at lower

overpotentials in the presence of 1-12S and to overlap the experimentally measured ow, 1rpotentials above

a certain H2S concentration, i.e., the h.c.r. mechanism shifts from discharge controlled to coupled

discharge-recombination controlled with increasing H2S concentration.

6. It is noteworthy that one can now calculate the important hydrogen absorption and adsorption

rate constants, by applying the I-P-Z model to the electrochcmical perneation data obtained as a

function of membrane thickness.

ACKNOWLEDGMIENT

The encouragement and financial support by A. John Sedriks, Office of Naval Research,

Contract No. N00014-84K-0201 are gramtefilly acknowledged.

LIST OF SYMBOLS

a a constant, F/RT, (volts)-1

al+ activity of hydrogen ions in the electrolyte, dimensionless

14

A amperesb a constant, LFD 1, mol (A cm)"ell+ hydrogen ion concentration, 10"Ph , mol liter I

:1s 1.H2S concentration in the .-Iectrolytc, mol liter-

cs suface hydrogeCn concentration, mol cm3

D1 bulk hydrogen diffusion coefficient, cm 2 s- I

Eeq equilibrium potential for the h.e.r., mV vs SHE

F Faraday constant, 96500 C g-eq"f a constant = y/RT, dimensionless

fie,i. a variable, f k -biosk"), A cm-2

fit, a variable, iceaO,, Acn "2

ic charging flux or current density, A cm"2

i. steady state pcmleation flux, A cm "2

io cxchangc current density. A cm 2

io0 io/(l-0e), A cm 2

ir steady state evolution flux, A cm 2

k 1° discharge reaction rate constant, mol (cm 2 s) 1

k I discharge reaction rate coefficient (= io'/F), mol (cm 2 sy

k.. recombination reaction rate constant, tool (cm2 s)-13

k-tb asrtorate constant, mol (cm,2 s)1I

kads adsorption rate constant, cm s i

k" thickness dependent absorption-adsorption constant, mol cnr 3

L membrane thickness, cm

M mol liter IR gas constant, 8.314 J (g-mol k)-

T temperature, K

Greek Symbols

a; transfer coefficient, dimensionless

TI overvoltage, Vtie cathodic overvoltage, V

TICeI lower limit of Ti0, V

711tu pper limit of Tlc, V

OH surface hydrogen coverage, dimensionless

OC tquilibrium surfacce hydrogen coveraige, dimecnsionless

-radient. of the apparent standard free energy of adsorption with hydrogen eomCe, J -m*IfO

REFERENCES

1. H-2S Corrosion in Oil and Ga% Production, A Compilation of Classic Papers, R. D. Tuttle and R.S. Kane, eds., NACE (198 1).

2. J. F. Newman and L. L. Shircir, Corrosion Science, Vol. 9, p. 631 (1969).

3. A. Kawashimna, K. Hashimoto and S. Shimodaim, Corrosion, Vol. 32, p. 321 (1976).

4. P. Sury, Corrosion Science, Vol. 16, p. 879 (1976).

5. Mly. Hashimoto, E. Sato and T. M'vurata, Hydrog~en in M'vetals (Conf. Proc.), JIMS-2 (1980), p.209

6. B. 1. Bcrkowirz and H. 1-. Horowitz, J. Elccrochcni. Soc., Vol. 129, p. 468 (1982).

7. J. OMly. Bockris, Inteniational Conference or, Stress Corrosion Cr;Icking!and HydroI~nEmhrittigment of Tron Base Alloys, Firminy, France, 1973, R. W. Stachle ci al cds., INACE-5(1977), p. 286.

8. Z. A. lofa and F. L. Kamn, Zashchita Mctallov, Vol. 10, p. 17 (1974); Protection of Mevitals, Vol.10, p. 12 (1974).

9. B. E. Boucher, Proceedings of the Fourdh 1nre 1atonal CongresoMelicCroi, NACE(1969), p. 550.

10. J. OM. Bockris, J. IMcBreen and L. Nanis, J. Electrocheni. Soc., Vol. 112, p. 1025 (1965).

11. P. WV. Bolmer, Corrosion, Vol. 21, p. 69 (1965).

12. R. N. Iycr, H. WV. Pickering and I. Zanianzadeh, Scripta MVet., Vol. 22, p. 911 (1988).

13. R. N. Iyer, H. W. Pickering and I. Zamanzadeh, J. Electrochem. Soc., 136 7463 (1989).

14. M. A. Devanathar. and L. Stachurski, J. Electrochemn. Soc., Vol. 1 11, p. 619 (1964).

15. S. S. Chatcrjce, B. G. Ateya and H. W. Pickering, M'vet. Trans., Vol. 9A, p. 3189 (1978).

16. Private Discussions with Konrad WVeil, Technische Hochschule Darmistadt, Darnistadt, FederalRepublic of Germany.

17. E. Hart and M. Anbar, The Hydrated Electron, John Wiley & Sons, Inc., Newv York (1970), p.113.

18. E. Gileadi and B. E. Conway, Modem Aspects of Electrochemistry, No. 3,J. O1M. Bockris andB. E. Conway, eds., Butterworths, Washington (1964), pp. 347-442.

19. H. IMurayamna, AI Sakashita and N. Sato, Hydrogen in IMetals (Conf. Proc.), JUMS-2 (1980), pr.297.

16

FIGURE CAPTIONS

Figure 1: Stability plots of H2S, HS- and S2 -as a function of pH. (9)

Figure 2: Schematic of the apparatus for the hydrogen permeationexperiment.

Figure 3: Hydrogen overvoltage, il, vs. hydrogen charging current density ic,for different H2S concentrations at (a) pH=l and (b) pH=2.

Figure 4: Hydrogen overvoltage, i1, vs. steady state hydrogen permeation currentdensity, io,, for different H2S concentrations at (a) pH=1 and (b) pH1=2.

Figure 5: Steady state hydrogen permeation current density, io, vs. square root ofhydrogen recombination current density, 4Ii for different 1-i2Sconcentrations at (a) pH=1 and (b) pH=2.

Figure 6: Relationship between fil and i., for different H2S concentrations at

(a) pH=1 and (b) pH=2.

Figure 7: Relationship between ",ir/io and io for different H2S concentrations at pH=2

Figure 8: Hydrogen coverage-corrected ic vs hydrogen overvoltage, i1, fordifferent H2S concentrations at pH=2.

Figure 9: Hydrogen coverage, 014, vs hydrogen overvoltage, r, for different H2Sconcentrations at (a) pH=1 and (b) pH=2.

Figure 10: H2S effect on the h.e.r. exchange current density, io.

Figure 11: H2S effect on de h.e.r. transfer coefficient, a.

Figure 12: -,3 effect on the discharge rate constant, ko.

Figure 13: H2S effect on the absorption-adsorption constant, k.

Figure 14: H2 S effect on the hydrogen recombination reaction rate constant, k3.

Figure 15: H2S effect on the hydrogen overvoltage range (between sloped lines) ofthe coupled discharge-recombination mechanism. Points: Calculated

17

(Eqs. 15 and 16) for different ki and k3 values obtained by applicationof die I-P-Z model to the permeation data. Cross hatched area:potential range of experimentation overlaps potential range of couplcddischarge-recombination mechanism.

-- --

I,0

IIR

-OD

o 0 0 0 00 C) (D0 r CMj

(9/) apij )ns JDlol )t) uQjpDjzj

EL

SL

0 wi

4- Uj

L.i

0~

C.'j

U-)

CC)

0.-0

LO LocoL

C)z

p p o

(11QA) IL 'afteH0AJaAQ UabOJPfiH L

CV,

Ln 0 0

&0oo0

O+E4c

QQ

Ii 0

r)

L

0 U) 0 )co ce) I

6

(110) U'abPIIO~aAOUa6JP0

CV,

700

o +EE

-S.

cr) c2

(110) LLla~llOAaAOUa6O~fl

00

OCu4

E

CL

=- 0

cr) Cr)

#U

(1I0A) 11 'abejj0AJaAO UaB0JPMH

____ ____ ___ ____ ____ ___ ____ e'

0-

0o

6

Co1n C>CII C C

c~cz---. 0 ~ 0 0

(=wE 6 0 w) 0

0o

(0

0

01

6 0

Qwobs/Vw) '~!

41

4 ' 00*

0 0

LOE

QwLuobs/vrf) (IL~oe)dxa li

0+

Q0

00

0)0

0 U) 0 1

N~ T 00

Q-wDbs/Vrf) (1.v'oe)dxa 31

00

E CV,

00

*/E 66

+3ML

CV,

040

0+30

0,

N -

00

LL

*0)

(-wo-bs/v

In 8 1C!

400

C;

C

L.

L1L

U) 0 LO17 1 0 00 0 0 0

E) 'a§eJ3AOo Ua6OJPflH az~eJJflG

00

0-

043

00

LO.

17 0

5e'3eaA:)U6OPR e+n

0 00

- - 0t

(fI -

~N. ~'1*CO

a.z)bs v 01

CV)00

o 2

c; 0, Z

o ~0 0

0 LO 0) LO) 0 1

)OV 50 U! aSeaXoUI

cte)0 0

0C E

%E2

N

0 00

00 C'j

C) C)T- )0

* S

00

LO 0

EL

0 0CV

0o0-

0 0

11 co C\J EC=; C; CE

((SOZW)IOW(F'

0

*C

0 C4

N

0 0

LIO(110) IL'Obe[OA13AOU350~rN

BASIC DISTRIBUTION IST

Technical and Suinnary Reports 1988

Oroanization Cooies Organization Conies

Defense Documentation Center Naval Air Prop. Test Ctr.Camerson Station Trenton, NY 08628Alexandria, VA 22314 12 ATTN: Library

Office of Naval Research Naval Contruction BattallionDept. of the Navy Civil Engineering Laboratory800 N. Quincy Street Port Hueneme, CA 93043Arlington, VA 22217 3 ATTN: Materials Div.Attn: Code 1131

Naval Research Laboratory Naval Electronics LaboratoryWashington, DC 20375 ' San Diego, CA 92152ATITN: Codes 6000 ATTN: Electron Materials

6300 Sciences Division262?

Naval Air Development Center Naval Missile CenterCode 606 Materials ConsultantWarminister, PA 18974 Code 3312-1ATTN: Dr. J. DeLuccia Point Mugu, CA 92041

Commanding Officer CommanderNaval Surface Weapons Center David Taylor ResearchWhite Oak Laboratory CenterSilver Spring, MD 20910 Bethesda, MD 20084ATTUI: Library

Naval Oceans Systems Center Naval Underwater System Ctr.San Diego, CA 92132 Newport, RI 02840AT-N: Library 1 ATTN: Library I

Naval Postgraduate School Naval Weapons CenterMonterey, CA 93940 China Lake, CA 93555ATTN: Mechanical Engineering ATTN: Library

Department 1

Naval Air Systems Command NASAWashington, DC 20360 Lewis Research CenterAttn: Code 310A i 21000 Brookpark Road

Code 5304B I Cleveland, OH 44135Code 931A 1 AT-N: Library

Naval Sea System Command National Bureau of StandardsWashington, DC 20362 Gaithersburg, MD 20899ATTN: Code 0514 1 Attn: Metallurgy Division I

Code 05R 1 Ceramics Division 1Fracture & DeformationDivision

gdval Facilities Engineering Defense Metals & CeramicsCommand Information Center

Alcxandria, VA 22331 Battelle Memorial Inst.ATTN: Code 03 1 505 King Avenue

Columbus, OH 43201 1

Scientific Advisor Metals and Ceramics Div.Commandant of the Marine Corps Oak Ridge National LaboratoryWashington, DC 20380 P.O. Box XATTN: Code AX 1 Oak Ridge, 1h 37380 1

Army Research Office Los Alamos Scientific Lab.P.O. Box 12211 P.O. Box 1663Research Triangle Park, NC 27709 Los Alamos, NM 87544ATTN: Metallurgy & Ceramics ATTN: Report Librarian

Program 1

Army Materials and Mechinics Argonne.National LaboratoryResearch Center Metallurgy Division

Watertown, MA 02172 P.O. Box 229ATTN: Research. Programs Office 1 Lemont, IL 60439

Air Force Office of Scientific Brookhaven National LaboratoryResearch/NE Technical Information Division

Building 410 Upton, Long islandBoiling Air Force Base New York 11973Washington, DC 20332 Attn: Research Library 1ATTN: Electronics & Materials

Science Directorate 1

Lawrence Radiation Lab.LibraryBuilding 50, Room 134Berkely, CA 1

NASA Headquarters David Taylor Research CtrWashington, DC 20546 Annapolis, MD 21402-5067Attn: Code RM 1 ATTN: Code 281 1

Code 2813 1Code 0115 1

RE/1131/88/75

4315 (036)

Supplemental Distribution List Feb 1988

Prof. I.M. Bernstein Profs. G.H. Meier and F.S.PettitIllinois Institute of Technology Dept. of Metallurgical andlIT Center Materials Eng.Chicago, Ill 60615 University of PittsburghPittsburgh, PA 15261

Prof. H.K. Birnbaunh

Dept. of Metallurgy & Mining Eng. vDr. W. C. MoshierUniversity of Lllinois Martin Marietta LaboratoriesUrbana, Ill 61801 1450 South Rolling Rd.

Baltimore, 1O 21227-3898

Prof. H.W. Pickering Prof. P.J. MoranDept. of Materials Science and Eng. Dept. of Materials Science & Eng.The Pennsylvania State University The Johns Hopkins UniversityUniversity Park, PA 16802 Baltimore, MD 21218

Prof. J. KrugerProf. D.J. Duquette Dept. of Materials Science & Eng.Dept. of Metallurgical Eng. The Johns Hopkins UniversityRensselaer Polytechnic Inst. Baltimore, MD 21218Troy, NY 12181

Prof. R.P. WeiDept. of Mechanical Engineering

Prof. J.P. Hirth and MechanicsDept. of Metallurgical Eng. Lehigh UniversityThe Ohio State University Bethlehem, PA 18015116 West 19th AvenueColumbus, OH 43210-1179

Prof. H. Leidheiser, Jr. Prof. W.H. HarttCenter for Coatings and Surface Research Department of Ocean EngineeringSinclair Laboratory, Bld. No. 7 Florida Atlantic UniversityLehigh University Boca Raton, Florida 33431Bethlehem, PA 18015

Dr. M. Kendig Dr. B.G. PoundRockwell International Science Center SRI International1049 Camino Dos Rios 333 Ravenswood Ave.P.O. Box 1095 Menlo Park, CA 94025Thousand Oaks, CA 91360

Prof. C.R. ClaytonProf. R. A. Rapp Department of Materials ScienceDept. of Metallurgical Eng. & EngineeringThe Ohio State University State University of New York116 West 19th Avenue Stony BrookColumbus, OH 43210-1179 Long Island, New York 11794

o Boris D. Cahan Prof. G.R. St. Pierrecept. of Chemistry Dept. of Metallurgical Eng.case Western Reserve Univ. The Ohio State UniversityCleveland, Ohio 44106 116 West 19th Avenue

Columbus, OH 43210-1179

Dr. K. Sadananda Prof. G. SimkovichCode 6323 Dept. of Materials Science & Eng.Naval Research Laboratory The Pennsylvania State UniversityWashington, D.C. 20375 University Park, PA 16802

Prof. M.E. Orazem Dr. E. McCaffertyDept. of Chemical Engineering Code 6322University of Virginia Naval Research LaboratoryCharlottesville, VA 22901 Washington, D. C. 20375

Dr. G.R. Yoder Dr. J.A. SpragueCode 6384 Code 4672Naval Research Laboratory Naval Research LaboratoryWashington, D.C. 20375 Washington, D.C. 20375

Dr. C.M. GilmoreDr. N. S. Bornstein The George Washington UniversityUnited Technologies Research Center School of Engineering& AppliedEast Hartford, CT 06108 Science

Washington, D.C. 20052

Dr. A.L. Moran Dr. F.B. MansfeldCode 2812 Dept. of Materials ScienceDavid Taylor Research Center University of Southern CaliforniaAnnapolis, MD 21402-5067 University Park

Los Angeles, CA 90089

Dr. B.E. Wilde Dr. Ulrich StimmingDept. of Metallurgical Engineering Dept. of Chemical Eng. & AppliedThe Ohio State University . Chemistry116 West 19th Avenue Columbia UniversityColumbus, OH 43210-1179 Ndw York, N.Y. 10027

Prof. J. O'M. BockrisDept. of ChemistryTexas A & M UniversityCollege Station, TX 77843