Sardisco & Pitts - Corrosion of Iron in H2S-CO2-H2O & Protectiveness of the Sulfide Film as a Function of PH (1965)

li~_·>;, ..

. . . -·

'.t . " ' .,,~:~~,· " . ~'. . ' ·. ' . . . . . . . . .· ... •\: '

' .

Corrosion of Iron in an H 2s-co2-H20 System Composition and Protectiveness of the Sulfide Film as a Function of pH *

* * By J. B. SARD ISCO and R. E. PITTS

INTRODUCTION

Composition of the iron sulfide film, formed as a result of lfquid phase H2S-COrH 2 0 corrosion of Armco(1 iron, has been related to the H2 S concentration in the H2 S- C02 gas and to the time of exposure of the metal to the corrodent. 1 '

2 It has been shown also that the addition of FeS to the H2S-COrH2 0 solutiori alters the nature of the iron sulfide film. 2 It was not ascertained, however, which species of the bivalent sulfur (H 2 S, HS- or s=) was responsible for altering the nature of the sulfide film as the total bivalent sulfur concenrration was changed.

The purpose of the investigation reported herein w 2. s twofold: (l) to measure the composition and protectiveness of the iron sulfide film, formed during the corrosion of Armco iron by liquid p~rnse H2S-C02 -H2 0 at various pH values, and (::') to determine which, if any, of the bivalent species (H 2 S, HS- or s=) contributed to the relative non-protectiveness of the sulfide film. Protectiveness of the film is defined as the extent to which the film, after it is formed , protects the metal from further corrosion.

C:XP ERi MENTAL APPARATUS AND TECHNIQUES

'.V eight loss, weight of film remaining on the sp2cimens and film composition were obtained at V<G·ious initial pH values. In addition, the protective­ness of the sulfide film was measured at different pH values.

'Y~' ight Loss and Weight of Film

The apparatus in which the corrosion reaction wos carried out, the specimen preparation and the experimental procedure are described in detail elsewhere. 2

For this phase of the investigation Armco iron specimens, 19 mils thick, were corroded. Two corrosion cells were used, each containing three specimens with a surface area of 9.68 cm 2 and two of 1.61 cm2 each. From the six 9.68 cm2

s pecimens average weight loss and weight-of-film

,,:,· .-: c~ rr.itte<l for publication AprJI 5, 1965.

* L: n:z.:cd Gas Corporation, Shreveport, Louisiana.

<11 Armco Sceel Corporation, Middletown, OhJo.

ABSTRACT

Composition and protectiveness of the iron sulfide film, formed as a result of the liquid phase H 2S-C02-H2 0 corrosion of Armco iron, have been related to the initial pH of the solution. Composition of the film was determined by electron diffraction analysis. Protectiveness of the film was deter­mined by measuring the weight loss of iron specimens previously corroded and covered by sulfide films formed in H 2S-C02 -H2 0 solutions of different pH values .

Composition and the protectiveness of the iron sulfide film were affected by the initial pH of the H 2S-C02 -H 2 0 solution. The film was least protective in the pH region of about 6.5 to 8. 8. The film that was least protective contained essentially kansite, whereas the film that was most protective contained pyrite and troilite along with kansite.

Because the HS- anion is the pre­dominant species in the pH region of about 6. 5 to 8. 8, it is indicated that the HS- was at least partially responsible for the film being less protective.

data were obtained and on the l. 61 cm 2 specimens electron diffraction analyses were made. The corrosion reaction was carried out for 43 hours at 24 ::'.:: l C (75 F). Specimen preparation and corrosion reactions were carried out in a dry box.

The different pH values of the H2 S-C02 -H2 0 solution were obtained by adding NaOH or HAc to distilled water, removing the oxygen from the solution and then saturating the solution with the H2 S- C0 2 gas. Initial pH values were measured in a cell at the end of the rrain of corrosion cells. H2 S-C02 gas was bubbled continuously through the cells at approximately 10 cc/min for the duration of the test.

350

November 1965 PROTECTIVENESS OF SULFIDE FILMS 351

Protectiveness of the Sulfide Film

For this part of t.he investigation the apparatus and experimental procedure were identical to that discussed above with the following exceptions. Three 9.68 cm 2 Armco iron specimens were placed in each of four corrosion cells. At an adjusted pH the corrosion reaction was carried out for 41 hou:rs. From the specimens of two of the cells the weight loss and weight of the film remaining on the specimen were obtained at that pH. The specimens of the other two cells were rinsed with acetone, dried and then, with the film still on the specimen, recorroded for an additional 24 hours in a fresh oxygen-free H2 S-C0 2 -H 2 0 solution. The pH of the fresh oxygen- free solution for each test was about 4. 0. Finally, the weight lossE:s of these specimens were measured. The protectiveness (P) of the sulfide film formed at the various pH values was calculated from the following expression:

l / P = W L/g of film,

where

W L = weight loss at 65 hours - weight loss at 41 hours, and

g of film= \Veight of film remaining on the speci­mens after the initial 41 hours of corrosion.

Film Composition

The corrosion product of the small specimens (1.61 cm 2

), which were corroded for 43 hours, was analyzed by transmission and reflection dif­fraction. The experimental procedures employed

,.--... for both of these techniques have been discussed in detail elsewhere. 1

'3

For the transmission diffraction studies the film was stripped from tte Armco iron by a water-free I2 -CH3 0H solution in the dry box. 4

The stripped film was transferred to the electron microscope without the film coming in contact with air.

Specimens used for reflection diffraction studies were mounted on the diffraction manipulator of the microscope in the dry box and transferred to the microscope in a plastic bag. The specimens came in contact with air for 2 to S seconds.

Processing the Electron Diffraction Data

The radii of the rings of the electron diffraction plates were measured by a diffraction plate comparator. Ring intensities were visually estimated. A program for the electronic computer

.....---._ was written to aid in the identification of the iron sulfide films. The sulfide films consisted of one to four components. Input data for the computer were radii of the rings and values of the microscope constant. Interplanar distances were calculated from the equation:

dh k l = K/2R

where dh k 1 is interplanar distance, K is the microscope constant and R is the radius of the ring. The computer output contained the following:

l. Interplanar distances, radii diameters and ring intensities;

2. Suspected compounds along with the ratio of the number of rings matched to the number that should have matched;

3. Average error for all pairs of rings for each compound;

4. The compounds found in the corrosion product, each marked by an asterisk.

Figure 1 contains two examples of electron diffraction patterns interpreted by the computer. Arrows indicate components that were identified also by manual methods. It was decided arbitrarily that one minor ring of a component could be missing without negating the presence of that component. See, for example, pyrite, troilite and marcasite of plate S28R and troilite of plate 579R. Interpretation of the patterns by the computer has been verified by manual methods for at least 180 of 200 patterns analyzed. It is to be noted, however, that this program is being used as an aid in analyzing the diffraction patterns and not as an absolute method.

RESULTS AND DISCUSSION OF RESULTS

Table 1 contains results obtained by electron diffraction analysis of films formed at various initial pH values. Components of the films were troilite, kansite and pyrite or marcasite . Because the diffraction patterns of pyrite and marcasite are very similar, no distinction was made between them. The numerator of each fraction in the table is the number of diffraction plates in which each component was identified and the denominator is the total number of diffraction plates obtained at that pH.

From the data in Table 1, the following inferences can be made. In the pH region of6.6 to 8.4, kansite was the only component in the film. In the pH regions 4. 0 to 6. 3 and 8. 8 to 11. 0, kansite, FeS 2 (pyrite or marcasite) and troilite were present in the film with kansite being the major component.

Because kansite is responsible for the non­protectiveness of the iron sulfide film, 1 '

5 the film should be most protective in the pH regions where the film contains components in addition to kansite.

Figure 2 contains the non-protectiveness factor (l / P) of the sulfide film as a function of the initial pH of the H2S- C0 2 - H2 0 solution. The data are tabulated in Table 2. The non-protectiveness of the film formed at each pH was measured by the technique described in the experimental section. Figure 2 shows that the film was least protective

'~52 CORROSION - NATIONAL ASSOCIATION OF CORROSION ENGINEERS Vol. 21

PLATE '28R B 0, 0 KV, 2,094 TEST XB5 ~TD ,16

RJNG tNTENS!TY R DA

0,3 0.658 J. a 2•

60 0,352 0,70• 2, H•• 20 0. 4 05 0. 810 2,5,,2

20 0,42 Q,85b 2,H6J

10 0,45"4 Q,908 2,J062

10 0. 5 1 1.10 1,900

0,582 1. lb• 1 ,7990

60 O,b12 1.22• 1. HOB

0. 6

10 60

1. 1,2.,.

• B 1,23 l

13 50 0. 9"4 9 1.898 1.tOJJ

-+ KANSITE bib 99,2 ······ -+PYRITE 617 97,8 TRO!LITE 5/6 97,3

-+·•ARCAS!TE bl7 98,o ILP~A FE O/b 0.

PLATE 579R TE:!T XI IJ•

JO 0 '3• 2 0 . 68 • J,0 629

20 0,.01 0.802 2,6122

JO 0. •52 0. 90 . 2.3175

20 0,496 0. 992 2 . 1119

100 0,569 1 .1 38 1,8•09

10 0. 566 1.172 1, 7175

10 0,660 1.no 1.5171

10 0'8 01 1. b 0 2 1,3077

0,9CC 1,840 1 , 13 86

10 0 '9 7 8 1.956 1. 0 '1 1

Figure 1 - Computer Print- Out.

~ 4.0.----.---.,---~---,------,--------~

...J

3 :-J :: I .01f-----j---j----j-------C--------i-- - --<

3 4 5 6 8 9 10 11 pH

Figure 2 - Non-protectiveness (l/P) of the film vs pH at ::rn H2 S concentration of 2. 00 psi a. Tot:i.l specimen surface area 29. 04 cm".

in the pH region of about 6.5 to 8. 8 and becomes more protective at the pH value of about 9.2.

There is a slight maximum in the potential of the steel electrodes in the saturated hydrogen sulfide solution versus pH curves of Ewing6 in the pH region of about 6. 8 to 8. 8. This observation is another indication that the protectiveness of the film varies as the pH of the solution is changed. There is, however, no obvious correlation between weight loss at 43 hours of Figure 3 and non-protectiveness of the film.

Differences in the physical appearance of the sulfide films formed at different pH values are

~.

November 1965 PROTECTIVENESS OF SULFIDE FILMS 353

borne out by the replica electron micrographs of Figure 4. The relatively protective films formed at pH values of 3. 2 and 11. 0 appear to be more tightlv pao:ked than the less protective film formed at a pH of 7.2.

10:-T I I 8

1 1 -·-1----1· --+---+---+---1---~ Hil I I

' I I : "' 9 x

: ! 0 - We .iqht loS!: at i 6- We1~ht Loss , o!

I

43 hour< 5 hours

§ 5 ' uf [Experomentol err'or wos :+:·0.2 X 10-2grom~ (/) I . I I ' 0 i I

~ 4 ~' ----".'----l-----'----1----'------l.--~

,~~1 ! . ' i i

: !

' , . • I ' I ~' -----~.,..~--.:_,- --- =~- - - - ---:?

4.0 50 6.0 7.0 pH

8 .0 9.0 10.0

Figure 3 - \\.eigr.t loss and film weight vs pH at an H 2 S concenrratio" of 1. 12 psia in H2 - C02 gas. Total specimen surface area 20. 04 cm;;.

11.0

In the intermediate pH region (7. 0 to 11. 0), HS­is the predominant species. As was brought out in the preceding paragraphs, 6.6 to 8.8 is the pH region in which the sulfide film formed is least protective. It appears that the bi sulfide ion suppresses formation of FeS 2 and troilite. As a

T . .\BLE 1 - Electron Diffraction Results

Initial Component pH Kansite Pyrite Troilite

3.1 2/3 0 1/ 3 4. () 2/3 1/3 1/3 4.2 5/5 2/5 2; 5 6.3 5/6 1/ 6 1/6 6.6 4/ 4 l/4 0 7.2 3/3 0 0 8.4 4/4 0 0 8.8 4/ s l/S l/S

11. 0 4/4 3/4 1/4

pH 3.2 pH 7.2 pH 11.0

Figure 4 - Replica electron micrographs of iron sulfide film formed at various pH values. :.HOO X

T . .;BLE 2 - Data for Protectiveness Sh1dies

Concenrration of the H2S in the H2S-C02 gas was 2.0 ps ia.

AVERAGES OF THREE COUPONS

Test .fl Hr 41 Hr 65 Hr No. Wt Loss< 1l Film Wt Wt Loss< 1l

pH I/P Initial pH/ar Time

l . 0870 .0054 .0975 l. 94 3.2 3. 7 /30 Hours 2 .0365 .0108 .0567 l. 87 3.4 4. 8/ 24 Hours 3 .0482 .0068 . 0630 2.18 ·L3 4 .6/24 Hours 4 .0257 .0090 .0478 2.46 6.4 6. 4/ 23 Hours 5 .0239 0102 .0499 2.55 6.5 6 .0022 .0069 .0240 3. 16 7 .-1 7.0/22 Hours 7 .0031 .0069 .0265 3 . 39 S.4 8.6/22 Hours 8 .0035 .0104 .0201 1.60 9.1 7.8/24 Hours

(l) Corrected for Blank Weight Loss

C2 l Initi:i.l pH adjusted by adding acetic acid.

Normality Na OH

(0.093) 0

0.003<2)

0.015 O.Ol? 0. 15 2 l 0.32 0.5

(Normality Acetic Acid)

1 ;• 'f' : 1 ' • ~ ~ ' I ; , ' ~; , ' '

.. ;::- . -' . . . ' .. ·{;"' •,, . . .

I ·: .. ; • ..._ ~ ~ ', ' '

354 CORROSION - NATIONAL ASSOCIATION OF CORROSION ENGINEERS Vol. 21

result, in this intermediate region, the film contained essentially kansite and is the least protective.

CONCLUSIONS

It can be concluded that the composition and protectiveness of the iron sulfide film were affected by the change in the initial pH of the H2 S-C02 -H 2 0 solution and that the film was least protective in the pH region of about 6. 5 to 8. 8. The sulfide film that was the 1 east protective contained essentially kansite.

ACKNOWLEDGMENTS

The authors express their appreciation to Jack D. Wisterman of the Computer Section of the l ' nited Gas Research Laboratory for writing the

r--- computer program and to Joseph E. Rountree for the electron diffraction analyses.

REFERENCES

l. J. B. Sardisco, Wm. B. Wright and E. C. Greco. Corrosion of Iron in an H2 S-C0 2 -H2 0 System: Corrosion Fi L'll Properties on Pure Iron. Corrosion, Vol. 19, No. 10, 354t-359t (1963i October.

2. J. B. Sardisco and R. E. Pitts. Corrosion of Iron in an H2S-COrH 2 0 System: Mechanism of the Sulfide Film For­mation and Kinetics of the Corrosion Reaction. Corrosion, Vol. 21, 1\o. 8, 245-253 (1965) August.

3. C. J. Arceneaux. Techniques for Handling Air-Sensitive Materials in Preparations for Electron Microscopy. Norelco Reporter, 7, 16 ff (1960) January-April.

4. W. H. J. Vernon, F. Wormwell and T. J . Nurse. The Thickness of Air-Formed Oxide Films on Iron. J. Chem. Soc., 621-632 (1939).

5. F. H. Meyer, 0. L. Riggs, R. L . McGlasson and J. D. Sudbury. Corrosion Products of Mild Steel in Hydrogen Sulfide Environments. Corrosion, Vol. 14, No. 2, 109t-115t ( 1958) February.

6. S. P. Ewing. Electrochemical Studies of the Hydrogen Sulfide Corrosion Mechanism . Corrosion, Vol. 11, No. 10, 497c-50lc (1955) November.