INFORMATION TO USERS
This was produced from a copy of a document sent to us for microfilming. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the material submitted.
The following explanation of techniques is provided to help you understand markings or notations which may appear on this reproduction.
1.The sign or “target” for pages apparently tacking from the document photographed is “Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you of complete continuity.
2. When an image on the film is obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because of movement during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that shoutd not have been filmed, you will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., is part of the material being photo graphed the photographer has followed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete.
4. For any illustrations that cannot be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In all cases we have filmed the best available copy.
University Microfilms International 300 N /6EB ROAD. ANN ARBOR. Ml 48106 18 BEDFORD ROW, LONDON WC1R 4EJ, ENGLAND 7916045
VAU * TE-LIN THE PASSIVITY OF IRON—CHROMIUM ALLOYS.
THE OHIO STATE UNIVERSITY* PH.D.* 1979
Uni\wsitv Microfilms International joon ^“ droao . ann arsoh .mi4bio6 PLEASE NOTE: In all cases this material has been filmed 1n the best possible way from the available copy. Problems encountered with this document have been Identified here with a check mark .
1. Glossy photographs_ 2. Colored Illustrations 3. Photographs with dark background 4. Illustrations are poor copy 5. Print shows through as there Is text on both sides of page 6. Indistinct, broken or small print on several paqes throughout ____ 7. Tightly bound copy with print lost 1n spine _ 8. Computer printout pages with Indistinct print ______9. Page(s) _ lacking when material received, and not available from school or author______10. Paae(s) seem to be missing in numbering only as text follows 11. Poor carbon copy______12. Not original copy, several pages with blurred type ______13. Appendix pages are poor copy 14. Original copy with light type ______15. Curling and wrinkled pages 16. O ther______
Irtveraty Micrrinlms tiemational 10 N ZEES RD.. ANN ARBOR VII aSTOB <3131 761-4700 THE PASSIVITY OF IFfON-CHROMIUM ALLOYS
DISAEHTATT OH
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
*
By
Te-Lin Yau, B.S., M.S.
The Ohio State University
1979
Reading Committee: Approved By
Professor R. W. Staehle Professor R. A. Rapp Professor J. A. Begley
/Ad vi ser Department of Metallurgical DEDICATION
To my mother, Mra. Yang Chih Yau, whose wisdom and guidance gave me the patience and perseverance to pursue the doctoral program*
To my late father, Mr. Chiu Ho Yau, whose total commitment to the education of his children and the underprivileged made this dissertation possible.
ii ACKNOWLEDGMENTS
I would like to thank my adviser, Professor R. W.
Staehle, for his guidance and counsel during the course of this investigation. Thanks are also due Professors R. A,
Rapp and J. A. Begley for their interest.
I also wish to thank Dr. S. Smialowska for her interest, suggestions, and stimulation.
Financial support by the National Science Foundation is gratefully recognized.
Finally, the author extend a special vote of thanks to Miss D. M, Essock for the polish of this dissertation, and to his sister, Shiow Yun, and to Miss Jue Hua Tsai for all the varied tasks they have helped him to complete. VITA
April 9, 19ij-5>.... Born - Hunan, Republic of China
1969...... B.S.E.S., Cheng Kung University, Tainan, Taiwan
1970-1971 ...... Mechanical Engineer, Taiwan Shipbuild ing Corporation, Keelung, ^aiwan
1971-197 2...... Research Assistant, Department of Engineering Science, Tennessee Technological University, Cookeville, Tennessee
1973...... M.Sc., Tennessee Technological University, Cookeville, Tennessee
1972-197 9...... Research Associate, Department of Metallurgical Engineering, The Ohio Stat'e University, Columbus, Ohio
FIELDS OF STUDY
Major Field: Metallurgical Engineering
Studies in Metallic Corrosion. Professor M. G. Fontana, Professor F. H. Beck and Professor R. W. Staehle
Studies in Physical Metallurgy. Professor J. P. Hirth and Professor G. W. Powell
Studies in Mechanical Metallurgy. Professor J. W. Spretnak
Studies in Thermodynamics. Professor G. R. St. Pierre TABLE OF CONTENTS
i'age
ACKNOWLEDGEMENT...... 1 ii
VITA...... iv
LIST OF TABLES...... viii
LIST OF FIGURES......
INTRODUCTION......
1.1 Scope of Investigation...... 4 PASSIVITY......
2.0 Introduction...... 6 2.1 Definition...... 2.2 A General Description of Passive Metals and Alloys...... *...... 2.3 Cathodic Reactions...... 11 2.4 Anodic Passivation...... 12 2.5 Theories of Passivity...... 16 2.5-1 The Film Theory...... 16 2.5-2 The Adsorption Theory...... 20 2-5-3 Combined Film-Adsorption Theory...... 23 2.5-4 Other Theories...... 24 2.6 The Passivity of Alloys...... 26 2.6.1 Theories of Passivity in Alloys...... 29 2.6.2 The Passivity of Fe-Cr Alloys...... 30 2.6.2.1 Iron...... 30 2.6.2.2 Chromium...... 32 2.6.2. 3 Iron-Chromivim Alloys...... 34 2.7 Processes of Film Formation...... 67 2.7-1 Initial Fast Oxidation...... 67 2.7.2 Kinetics of Film Growth...... 66 2.7.3 The Transition Period...... 71 2.8 Solubility, Dissolution and Passivity...... 73 2.8.1 Solubility of Iron Oxides and Chromium Oxides...... 74 2.8.2 Formation and Dissolution of Anodic Oxide Films...... 75 2.8.3 Mechanisms of Oxide Dissolution...... 77
v TABLE OP CONTENTS (Continued)
AUGER ELECTRON SPECTROSCOPY...... 83
3.0 Introduction...... 83 3.1 Historical Note...... ' 81; 3.2 The Detected Volume and Auger Electron Mean Escape Depth...... 86 3.3 Quantitative Analysis...... 87 3.3*1 Absolute Caloul ations ...... 89 3.3*2 Quantitative Analysis with Standards.. 90 3.3*3 Quantitative Analysis without Standard 3*4 Chemical Effects in AES...... 9:V 3.5 Composition-vs-depth Profiles...... 101 3.6 Conclusion...,...... 10o
EXPERIMENTAL PROCEDURES...... 109
ij.,0 Introduction...... 109 i|_.l Electrochemical Measurements ...... 110 4.2 Electrolyte Solutions...... IIP I4..3 Electrochemical Test Cell...... Ilf Auger Unit...... 117 ij-.5 Procedure...... 123 I;.5.1 Specimen Preparation...... 123 4.5*2 Experimental Procedure...... 12 9' 4.5.3 APPH and NAPH...... 126 ^Auger Spectrum...... 128 i;.5*5 Preliminary Alignment...... 130 i;.5*6 Depth-Concentration Profile...... 132
EXPERIMENTAL RESULTS...... 133
5.1 Polarization Behavior of Alloys...... 133 5*2 Current Decay Curves...... 133 5.3 Auger Analysis...... lij-3
DISCUSSION...... 207
6.1 Polarization Behavior of Alloys...... ' 207 6.1.1 Rest Potential...... 210 6.1.2 Critical Current Density and Passivation Potential...... 21^ 6.1.3 Passive Current Density...... 221; 6.1.1; Transpassive Potential...... 227 6.1.5 Summary...... 728 6.1.6 Conclusions...... 230
vi TABLE OP CONTENTS (Contuiued)
6.2 Current Decay Curves ...... 2 32 6.2.1 Determination of Passive Properties of Pe-Cr Alloys...... 232 6.2.2 The Kinetics of Film Formation...... 239 6.2.3 Summary...... 2^6 6 . 2 . [|. Conclusion...... 2ipti 6.3 Auger Analysis...... 2ij.fi 6.3*1 The Maximum Chromium Zone in the Passive Films...... 230 6.3.2 The Average Chromium Content in the Passive Films...... 259 6.3-3 The Composition of Passive Filins...... 262 6.3.4 Summary...... 27 5 6.1|. The Sulfate Ion Effect...... 276 * FINAL CONCLUSIONS...... 278
REFERENCES...... 285
vii LIST OP TABLES
.Table Page
1. Some of the Properties of Pure Iron and Pure Chromium...... *...... 33
2. Standard EMF series of Metals . . 37
3. Galvanic Series of Metals...... 39
1;. Normal Potentials for the Redox Processes on Cr in Acid Solutions...... $2
5. Composition x, Lattice Parameter a, Degree of Inversion I, and Molecular Formula at the Boundaries of Oxygen-rich or Metal-rich..*... 61;
6. Solubility of Some Iron Oxides and Chromium Oxides in Water...... 76
7. Summary and Comparison of Techniques for Surface Chemistry Study...... 108
8. Chemical Composition of Materials Used in this Study...... 121;
9. Some Quantitative Auger Electron Spectrometry Experiments...... 12/
10. Electron Energies of Auger Peaks for Various Elements...... 129
11. The Quantitative Relationship between the Chromium Concentration and the Auger NAPH Ratio...... *..... 20$
12. The Electrochemical Characteristics of Fe-Cr Alloys...... 209
13* Film Thickness (8) Calculated from the Total Charge Passed during Current Decay Plotted vs at % Cr...... 231;
11;. The Current Densities of Fe-Cr Alloys after Twelve Hours of Passivation at 600 mVgce 236
v m LIST OP TABLES (Continued)
Table Page
15* The Chromium Concentration and the Required Sputtering Time for (-pfficrW x ...... 256 16. The Average Cr Content in the Passive Film of Fe-Cr Alloys...... 260
17. The Peak-to-peak Heights of Several Low Energy Peaks Observed in the Passive Films of Fe-Cr AlHoys after Exposure to IN H?S0. for Twelve Hours at 600 mV ...... 265 see 18. The Peak-to-peak Heights of Several Low Energy Peaks Observed in the Passive Films of Fe-Cr Alloys after Exposure to IN NapS0i (pH 3) for Twelve Hours at 600 mV ...7...... 266 ^ see 19. The Peak-to-peak Heights of Several Low Energy Peaks Observed in the Passive Films of Fe-Cr Alloys after Exposure to IN NapS0. (pH 6.5) for Twelve Hours at 600 mV.Jv .7..... 267 D L v 20. The Peak-to-peak Heights of Several Low Energy Peaks Observed in the Passive Films of Fe-Cr Alloys after Exposure to IN NapS0i (pH 10) for Twelve Hours at 600 mV .... 268 D U C 21. The Major Low Energy Spectra of Fe, Cr, and their Oxides...... 269 LIST OF FIGURES
Figure Page
1. Polarization Curve of a Normal Metal...... 14
2 . Polarization Curve of an Active-passive Metal. 19
3. Effect of Alloying Elements on Characteristic Points of an Anodic Folarization Curve for Fe in H2S0^ Solutions...... 27
4* Effect of Alloying Elements on Characteristic Points of an Anodic Polarization Curve for Ti in H^SOt Solutions...... 28 2 4 5. The Phase Diagram of Cr-Fe System...... 38 6 . The Corrosion Rates of Fe-Cr Alloys in Intermittent Water Spray, Room Temperature.... 41
7. The Corrosion Potentials of Fe-Cr Alloys in 1±% NaCl...... 42
8 . Standard Flade Potentials for Fe, Cr, and Fe-Cr Alloys...... 43
9. Critical Current Densities for Passivation of Fe-Cr Alloys...... U4 10. Potentiostatic Polarization Curves for Fe-Cr Alloys in Sulfuric Acid Solutions...... 44
11. Potentiodynamic Polarization Curves for Fe-Cr Alloys in IN H2S0^ at 25 C..(...... 4^
12. Effect of Chromium on the Corrosion Rate in Nitric Acid Containing 32% HNO^...... 47
13. A Model for the Passive Film...... 54
14. Potential-pH Diagram for the Fe-H^O System, at 25 C (Considering as Solid Substances Only Fe, Fe3^4 antl Pe2°3 >...... 1?. Potential-pH Diagram for the Fe-H-0 System, at 25 C (Considering as Solid Substances Only Fe, Fe (OH)^ and Fe(OH)-)...... 60 x LIST OP FIGURES (Continued)
Figure Page
16. Potential-pH Diagram for the Cr-IUO System, at 25 C (Considering as Solid Substances Only Cr and Anhydrous Cr20^)...... 61
17. Potential-pH Diagram for the Cr-H?0 System, at 25 C (Considering as Solid Substances Only Cr and Cr (OH) ^ )...... 62
18. Po£ential-pH Diagram for the Cr-H20 System, at 25 C (Considering as Solid Substances Only Cr and Cr(0H)3*nH20)...... 63
19. Film Thickness and Dissolution vs Log Time for Iron in Borate-Boric Acid Solution at pH 8.6.. 78
20. Model for Reductive Dissolution of Iron Cxides...... 80
21. Variation of the Mean Escape Depth with the Kinetic Energy of the Escaping Auger Electron. 88
22. Mo and W Auger Signals as a Function of Coated Mo on W Substrate 9 3
23. Most Prounced Auger Electron Energies...... 9f
2l|., Spectra of the Carbon KLL Auger Transition for Diamond, Graphite, Silicon Carbide, and Mo2C.. 99
25. Auger Spectra of Metal and Oxidized Metal Surface...... 100
26. Auger Spectra during the Gradual Removal of the Passive Film on Iron...... 102
27. Representative Auger Spectra from Various Depth of 150 Angstrom Nichrome Film on Silicon Substrate...... 10i|.
20. Various Auger Peak-to peak Heights from Nichrome Film on Silicon Substrate as Function of Film Thickness...... 105
29. Schematic Diagram of the Experimental Electrochemical Setup...... Ill
xi LIST OF FIGURES (Continued)
Figure PaFc
30. Close Up View of Actual Apparatus Including Instruments in Electrochemical Measurements... 112
31. Electrochemical Cell Apparatus Used in Electrochemical Measurement...... 113
32. Electrochemical Test Cell...... 113
33* PHI Model 5U5 SAM...... 118
3l|-. Noble Vaclon Pump Element...... 121
35* PHI Model 10-503 Specimen Manipulator and Holder...... 122
3 6 . Effect of Specimen Alignment...... 131
37* The Current Density-potential Curves of Fe-Cr Alloys in IN H2S0^...... 13U.
3 8 . The Current Density-potential Curves of P’e-Cr Alloys in IN Na^SO^ (pH 3)* ...... 135
39* The Current Density-potential Curves of Fe-Cr . Alloys in IN Ila^SO^ (pH 6.5)...... 136
I4.O, The Current Density-potential Curves of Fe-Cr Alloys in III Ha2S0^ (pH 10)...... 137
Ipl. The Current Decay Curve of Fe-Cr Alloys in IN H 2 S 0 ^ ...... 139
^2. The Current Decay Curve of Fe-Cr Alloys in IN N a 2S0^ (pH 3)...... 1
I4.3 . The Current Decay Curve of Fe-Cr Alloys in IN Na^SO^ (pH 6.5)...... I*!-!
Iptj.. The Current Decay Curve of Fe-Cr Alloys in IN Na2S0^ (pH 10)...... 1^2
lj.5. The Composition-depth Profile of the Fe-5Cr Alloy in IN H2 S0^ at 600 mVsce 12 Hours. . lipil
I4.6 . The Composition-depth Profile of the Fe-l5Cr Alloy in IN H^SO, at 600 mV for 12 Hours... 11^5 dL tj. S C 6 xii LIST OF FIGURES (Continued)
Figure
tl_7 - The Composition-depth Profile of the Fe-25.5Cr Alloy in IN H-SO, at 600 mV for 12 Hours... l|_ SC 6 48. The Composition-depth Profile of the Fe-5Cr Alloy in IN Na^SO. (pH 3) at 600 for 12 * u u. s e t ? IJours ..... 7 ......
49. The Composition-deptn Profile of the Fe-l5Cr AlJoy in IN Na_S0| (pH 3) at 600 mV for 12 Hourstt ...... T t-L ...... 3 C t?
50. The Composition-depth Profile of the Fe-25.5Cr Alloy in IN Na.jSOi (pH 3) at 600 mV for 12 Hours...... f. .'f...... ???......
51. The Composition-depth Profile of the Fe-5Cr Alloy in IN Na„S0i (pH 6.5) at 600 mV for 12 Hours...... 7. .7...... ??......
52. The Composition-depth Profile of the Fe-15Cr Alloy in IN Na„S0, (pH 6.5) at 600 m V ^ o for 12 Hours...... i . A ...... ???......
53. The Composition-depth Profile of the Fe-25*5Cr Alloy in IN NaoS0, (pH 6.5) at 600 mVQ_ for 12 Hours...... *+......
54* The Composition-depth Profile of the Fe-5Cr Alloy in IN Na-SO. (pH 10) at 600 mVQ„_ for 12 Hours..... 7 ...... ??!......
55. The Composition-depth Profile of the Fe-15Cr Alloy in IN Na^SO, (pH 10) at 600 mV for 12 Hours..... f . . ^ ...... !?......
56. The Composition-depth Profile of the Fe-25*5Cr Alloy in IN Na~S0i (pH 10) at 600 mV for 12 Hours...... i . A ......
57. Low Energy Spectra for theConditions in Figure 45......
58. Low Energy Spectra for the Conditions in Figure 48......
59* Low Energy Spectra for the Conditions in Figure 47...... LIST OF FIGURES (Continued)
Figure Page
60. Low Energy Spectra for the Conditions in Figure 40...... ' 169
bl. Low Energy Spectra for the Conditions in Figure 1+9...... 1?2
62. Low Energy Spectra for the Conditions in Figure %0...... 175
6 3 . Low Energy Spectra for the Conditions in Figure 51...... *...... 1/3
6 Low Energy Spectra for the Conditions in Figure 52...... *...... 150
65* Low Energy Spectra for the Conditions in Figure 53...... 1^3
66. Low Energy Spectra for the Conditions in Figure 54...... 13c
6 7 . Low Energy Spectra for the Conditions in Figure 55...... *...... 139
68. Low Energy Spectra for the Conditions in Figure 56...... 192
6 9 . The Low Energy Spectra of Cr^O^...... * 198
70. The Low Energy Spectra of Cr(OH)7(precipitated from CrCl^+NH^OH on Agfoil) i ...... 199
71. The Low Energy Spectra of Cr (OH)(pure Cr in aerated H^O for 30 minutes) ...... 200
72. The Low Energy Spectra of Fe^ ^CrQ ^0^...... 201
73. The L ovj Energy Spectra of Pure Chromium...... 203
74. The Low Energy Spectra of Pure Chromium...... 204
75* The Quantitative Relationship between the Chromium Concentration and the Auger HAPH Ratio...... 206
7 6 . Rest Potentialsof Fe-Cr Alloys...... 211 xiv LIST OF FIGURES (Continued)
Figure Page
77. Effect of pH on the Rest Potentials of Fe-Cr Alloys * * * t-n -L1 f—o
7 8 . The i . . of Fe-Cr Alloys...... 21b crit1 79. Effect of pH on the i .¥ of Fe-Cr Alloys.... 216 C ” -L w 1 80. The i .. of Fe-Cr Alloys...... 217 c n up 81. Effect ofpH on i_ *¥ of Fe-Cr Alloys..... 2j 8 c r j. l ^ 82. The Passivation Potentials of Fe-Cr Alloys.... 219
83* Effect of pH on the Passivation Potentials of Fe-Cr Alloys...... 220
8l|_. Supperposition of Fourbaix Diagrams for Chromium and Iron Showing the Insoluble Products and some of the Equilibria Defining the Phase Boundaries...... 222
85. The Passive Current Densities, i , of Fe-Cr Alloys...... ? ...... 22,6
86 . Effect of pH on the Passive Current Densities, i , of Fe-Cr Alloys...... 226 P 8 7 . The i of Fe-Cr Alloys...... 237 p12 88. Effect of pH on the i of Fe-Cr Alloys...... 238 p12 89. The Intercepts, a, and the Slopes, m, of the Early Stage of the Current Decay Curves of Fe-Cr Alloys...... 2J+2
90. Effect of pH on the Intercepts, a, and the Slopes, m, of the Early Stage of the Current Decay Curves of Fe-Cr Alloys...... 214.3
91. The Intercepts, a, and the Slopes, m, of the Later Stage of the Current Decay Curves of Fe-Cr Alloys...... 244 92. Effect of pH on the Intercepts, a, and the Slopes, m, of the Later Stage of the Current Decay Curves of Fe-Cr Alloys...... 2 4 5 xv LIST OF FIGURES (Continued)
Figure Page
93- The Chromium Profiles of the Passive Films of Fe-Cr Alloys...... ■ 29I
9i|. Required Sputtering Time to Reach ( ^ — ) vs Chromium Content...... 1 262
95- Required Sputtering Time to Reach )yi„v *■ T rcTUf niaA vs pH*...... 293
96. The Sulfur Profiles of the Passive Films of Fe-Cr Alloys......
97. The (at 75 Cr) in the Passive Film vs the Alloy Chromium Content after Twelve Hours of Passivation at 600 mV ...... see 25Y 98. Effect of pH on the (at % Cr) in the Passive Films of Fe-Cr Alloys111 “after Twelve Hours of Passivation at 600 mV „...... 258 see 99. The Average Cr Content in the Passive Film vs the Alloy Cr Content...... 261
100. The Average Cr Content in the Passive Film vs pH Values...... 263
101. The Composition of the Passive Films on Fe-Cr Alloys...... 272
102. The i vs (at % Cr) ...... 282 V>2_2 max 103. The 1^ vs (at % Cr)Qve...... 283
xvi INTRODUCTION
The phenomenon of passivity in metals has been recogniz
ed for a long time. The basic phenomenon was first recorded
by Bergman, who noticed that when iron was added to a solution of silver in nitrous acid no precipitation ensued (1). heir
(1), who investigated this action in the year 1790, conducted several superb experiments. He perceived that iron acquired an altered state in silver solution and then become insetico in nitric acid, and that ordinary iron remained visibly unattacked when put into strong nitric acid. Thereafter, the passivity of pure metals and alloys has been stud i e*d extensively.
Schonbein (2), in 1036, pointed out that iron could oe passivated by anodic polarization as well as by strong nitric acid. He had also made a series of tests upon silver, copper, tin, lead, cadmium, bismuth, zinc, mercury, but none of tnese metals indicated any resemblance to iron, for all of tnem were corroded when serving as anodes. He did not experiment with either cobalt or nickel, he suspected that these metals would act in the same manner as iron did. That is, he implied that only magnetic metals could be passivated. According to the present understanding of passivity, this is not true.
Nevertheless, the name "passive” for the altered iron, which he suggested originally, has been generally adopted.
The controversy on the mechanism of passivity started
in the days of Schonbein and Faraday. Schonbein (2)
explained it with the viewpoint that passivity occured due to
the altered state of the metal, e.g., its conversion into an
allotrope. Faraday (3) attributed the passive state to the formation of thin and insoluble oxide or similar compound on the metallic surface. From then on, many methods of investigation have been used to study the passivity of pure metals and alloys. Some of the methods used are as follows: electron diffraction, electrochemical, optical, and micro- chemical methods. However, little is yet known about passive films, in spite of the fact that the usefulness of many important alloys, e.g., stainless steels, depends on their ability to be passivated. The main reason is that the films are usually too thin to be detected by high energy x-ray and electron diffraction. There is still much work to be done in this field. Some of the unresolved problems to be considered in this study on iron-chromium alloys are as follows: (i) the composition of the passive film itself and the effect of the pH of the solution on the composition of this passive film, and (ii) the reason(s) for the critical influence of alloying Cr on the protective quality of the passive film.
The importance of research on the passivity of Fe-Cr alloys is clear when one considers that iron i3 the most widely distributed heavy metal in the earth and that iron is
the metal most commonly used by human beings. In fact, the
production and consumption of iron and steels can be used as
an index to indicate the strength of a country.
The use of iron can be traced back to the earliest his
torical times, and is recorded as early as is the use of
bronze. Notwithstanding, iron and steels rust in most
environments, even in the atmosphere and in water. Hence,
although they have an unusually versatile resistance to
alkalis and certain acid solutions, they still have a repu
tation for poor corrosion resistance. The prevention of
corrosion of iron and steels is, obviously, a very important
subject of study. Vernon (4) has presented four methods
preventing metallic corrosion, i.e. (i) modification of procedure, (ii) modification of environment, (iii) modifi
cation of metal, and (iv) protective coatings. Here we are
interested in the third method.
Iron forms many types of alloys with other elements for*
different service purposes. In the field of special steels,
stainless steels occupy a very important place. The term
stainless steel usually refers to those steels containing 1? wt % Cr or more.
As early as 1021, Berthier pointed out that iron alloyed with chromium turns out to be more resistant to acid attack and that its resistance increases with increasing chromium content {£) . The credit for discovering stainless steels is often given to the pioneering workers, Brearley, Strauss and Maurer,
However, Zapffe (6) named Guillet and Monnartz as the true discoverers of stainless steels because of their important investigation on basic properties, Guillet (7) studied the microstructures and mechanical properties of the majority of the current stainless steels of the martensitic , ferritic, and austenitic categories. The classifications which he laid down are still used today. The remarkable property of passivity in Pe-Cr alloys beginning at a minimum of i, ul.
Cr was first discribed by Monnartz (?) after his bro^d ; Lit.iy on the chemical properties of Pe-Cr alloys. He also demon strated the effect of carbon, the possibility of 'stabili zation1, and the beneficial effect of molybdenum.
1.1 Scope of Investigation
The corrosion resistance of a metallic material, in almost all cases of practical interest, depends on the surface film formed in the environment. In some cases, a protective film can reduce the corrosion rate to virtually zero; thus it is obvious that the surface film is very im portant in determining corrosion rate. The purposes and particular concerns of this investigation, then, are a more precise characterization of the passive films of Fe-Cr alloys and a rationalization of the role of alloying Cr with respect to its critical influence on the protective quality of the films. Important existing information will be used to ex plain the role of alloying elements in passivity.
This study does to survey the polarization behaviour of
Fe-Cr alloys in sulphate solutions, to examine the property alternations with changing alloy composition and solution pH, and to establish the composition profile within the passive film by using the Auger electron spectroscopy technique. The aim of this work is also to determine whether a correlation exists between electrochemical data and the results of Auger analysis which show the properties of the three studied Fe-Cr alloys, with different concentrations of Cr (5*15» and 25*5 wt %), to passivate and to correlate the electrochemical properties with the composition of the passive film. PASSIVITY
2.0 INTRODUCTION
Since the days of Faraday and Schonbein, the phenomenon of metallic passivity has been the subject of investigation by scientists and engineers. Hundreds of papers relating to this phenomenon have been published. Such intense interest in metallic passivity arose because of the complex nature of the phenomenon, the specific conditions under which it occurs, and the practical importance of increasing the corrosion resistance of metals and alloys.
2.1 DEFINITION
It is not easy to develop a universally acceptable definition of passivity, particularly since the phenomenon has been so extensively studied and discussed in the 20th century. Basically passivity refers to the loss of chemical reactivity ty certain metals and alloys under particular environmental conditions (9). Two definitions of passivity were given in "Corrosion
Handbook (l0)M.
"Definition 1. A metal active in the EMF serieu, or an alloy composed of such metals, is considered passive when its electrochemical behavior becomes that of an appreciably less active or noble metal. Definition 2. A metal or alloy is passive if it substantially resists corrosion in an environ- ment where thermodynamically there is a large free-energy decrease associated with its passage from the metallic state to appropriate corrosion products."
Metals and alloys, called passive by Def.l, usually also
conform to Def. 2 based on low corrosion rates. Examples of
metals are iron, chromium, nickel, titanium, stainless steels,
etc. These metals and alloys show a marked reduction in
reaction rates under pronounced anodic polarization. However,
corrosion inhibition, such as that experienced by iron in
acid to which organic compounds have been added, is also
included under this definition. A distinction should made
at this point between Passivity and Inhibition. The term
inhabition is used in cases where at a constant affinity the
corrosion rate is effectively decreased upon varying the
concentration or activity of a chemical substance which is
not involved in the considered corrosion reactions (ll).
Definition 2 is an even broader definition. Cases, such as those of lead in sulfuric acid or magnesium in HgO, are
called passive by Def. 2 but not by Def. 1. The corrosion potentials of these metals are too active to cause pronounced anodic polarization.
The most encompassing definition was probably that
offered by Wagner (11,12 ) as followsi A metal is passive if
the steady-state corrosion rate of the metal becomes lower as
its potential is shifted in the more noble direction. Alter natively, a metal is passive if the steady-state dissolution rate, in the absence of external current, is lower at a 8
higher concentration of an oxidizing agent than that at a louer
concentration of the oxidizing agent. However, the defini
tion could well include pitting corrosion in the passive
range where even the average corrosion rate is quite high.
A precise definition of passivity might also require that thi
dissolutin rate of the metal be extremely and unusually low
under condition where a large free energy decrease is
associated with the interaction of the metal and environment
(13).
Based on the previous disscussions, passivity may be
rationally defined in the following manner*
Passivity is a state of chemical inactivity or an
"inert state" of metals or alloys, under conditions where
there is a large free energy decrease associated with the
interaction of the metal and environment, caused by an
increased anodic potential or oxidizing power.
According to this definition not just any surface film can cause a passive state, only protective surface films.
Only these will prevent dissolution processes and can, therefore, be properly termed passive films.
2.2 A General Description of Passive Metals and Alloys
Passivity, unlike a metal's melting point or other physical properties, is not an absolute property. A metal may have different degrees of passivity depending on environmental conditions (llj.). For example, chromium, aluminum, iron, cobalt, nickel, titanium, tantalum, and
niobium are the easiest metals to passivate. An alloy with
a high degree of passivity can be obtained by alloying weakly
passivating metals with a strongly passivating metal. The
construction of stainless steels is a good example of such
an alloy.
Usually, oxidizing circumstances promote passivity while
reducing circumstances can destroy the passive state.
Examples of oxidizing conditions include anodic polarization,
and oxidizing solutions such as HNO^, AgNO^, KgCrgO^, H Cl Oy
KCIO^, HIO^, and so on. Oxygen itself, either in a solution
or in the air, is a strong passivating agent. Several metals,
such as chromium, tantalum, titanium, and aluminum, which can
spontaneously passivate in air and in many oxygen-containing
solutions, are called self-passivating. These metals have a
very stable passive state and can repassivate after mechanical destruction of the passive film. A metal can also
be passivated either by anodic polarization, or by contact with a more noble metal. As a matter of fact, oxidizing power is equivalent to electrode potential ( 9 )•
Decrease in temperature can assist the passivating process. For example, copper is reported to be passivated in concentrated nitric acid at -11°C but not at room temperature
(15 i 16). After copper is passivated at -11°C, it can maintain its passive state in nitric acid at room temperature.
On the other hand,, certain reducing circumstances can destroy the passive state. These circumstances include
cathodic polarization, contact with a less noble metal, and
environraentes containing hydrogen or carbon monoxide (par
ticularly when heated). Several other factors can also
promote the destruction of the passive state, i.e. increasing
the temperature and adding halogen ions. Magnetic fields are
liable to destroy passivity; however, there is no decisive
conclusion since the magnetic field often accompanies
secondary factors which are destructive, such as increased
temperature, mechanical shock, or physical effects due to
magnetostriction (10).
The effects on passivity of another factor, velocity,
are complex. The passive state of iron in concentrated ,
or lead in dilute H^SO^ can be destroyed if the velocity of
the medium is too high. On the contrary, self-passivating
materials such as titanium and stainless steels frequently
are easier to passivate when the velocity of the environment
is high (9). Since increase in the velocity of the medium
results in increase in the limiting diffusion current density,
the mixed potential is therefore in the passive range.
Passivity is not always connected with the action of the
oxidizing solution. Some metals can be passivated in non
oxidizing solution. For example, magnesium is passivated in
HF since magnesium fluoride is very insoluble in this acid
(17). Similarly, molybdenum has good resistance to HC1 and
HF (18, 19). Aeration of the above acids will reduce 1 ] molybdenum's corrosion resistance.
Finally, the passive state is persistent even after the
passivating process is ended. For instance, passivated iron
can maintain its passive state for a long time in dilute nitric acid, water, water vapor, and other environments.
Persistence of the passive state should be an important factor in developing an acceptable theory of passivity, which will be discussed in Section 2.5.
2.3 Cathodic Reactions
The corrosion of metals is a reaction accomplished by the simultaneously anodic and cathodic reactions. Either or both of the anodic and cathodic reactions may control the corrosion rate of metals. Cathodic reactions in corrosion is discussed in this section.
In industrial practice, many oxidizing species are used.
But for all of these the kinetics of reduction has not been studied sufficiently. However, the two most important cathodic processes, i.e., the hydrogen evolution reaction
(h.e.r.) and the oxygen reduction reaction, are frequently studied. The reason is due to the fact that hydrogen ions and water molecules are invariably present in aquous solution, and since most aqeous solutions are in contact with the atmosphere, dissolved oxygen is normally present.
In the absence of oxygen and any other oxidizing species, the h.e.r. will be the only cathodic process possible. 12
The principles of the h.e.r. are explained in Reference 20.
However, when dissolved oxygen is present both cathodic
reactions will be possible. The mechanism of the oxygen
reduction reaction is discussed in Reference 21.
In acid solutions, where the activity of the hydrogen
ions is sufficiently high, then, the h.e.r. is:
H^0+ + e --► + ^2° solutions) (2,'J) ■4" In neutral or basic solutions, the activity of H^O is
too low for it to participate in the h.e.r., and under these
circumstances water is reduced:
^2° + e — ^H2 + (neutral and basic solutions) (2. <■ )
Similarly, the oxygen reduction reactions are:
% 02 + 2H30+ + 2e — - 3H20 (acid solutions J (2 . .i)
+ H^O + 2e _». 20H~ (neutral and basic
solutions) (2.1].)
It should be noted that all the above reactions result in an
increase in pH in the diffusion layer.
Although the concurrent reduction of H^O and dissolved oxygen occurs frequently this does not exhaust the possi- +1 +2 bilities, and reactions such as M -*+e—*■M (metal ion re duction), M++e— (metal deposition), Cl^—►Cl” (chlorine reduction) may accompany either or both of the above reactions.
2*1]. Anodic Passivation
The phenomenon of passivity is one of the most important aspects in metallic corrosion. Although difficult to define,
passivity can be described by characterizing the behavior of
metals which show this phenomenon. Usually, the common
engineering and structural materials, such as iron, cobalt,
nickel, chromium, silicon, titanium, aluminum, and alloys
containing these metals, are more susceptible to this kind of
phenomenon. Also, some metals, e.g. zinc, tin, lead, cadmium
and uranium, exhibit passivity effects under certain conditions.
Figure 1 demonstrates the behavior of what can be called
a normal metal, that is, a metal which does not display passivity characteristics. The copper-sulfuric acid system is an example of this case. Here, if the electrode potential is increased, the current density of the electrode will increase rapidly. The reaction rate increases exponentially as described by the Tafel equation. A plot on a semilogari- thmic scale, then, will yield a straight line.
Figure 2 illustrates the typical corrosion behavior of an active-passive metal. Iron in sulfuric acid solution is an example. The behavior of this metal or alloy can be classified into three regions, active, passive, and trans- passive. In the active region, the metal acts just like a normal metal. The metal corrodes generally according to the
Tafel equation but is modified by corrections from concen tration overvoltage, resistance overvoltage, and the cathodic current density. However, when the current density reaches a critical point, 'the corrosion rate drops suddenly to ■J'
log i
Figure 1. Polarization curve of a normal metal* 1‘
(+)
Transpassive
E,
Passive
PP ! Active crit
(-) log i
Figure 2. Polarization curve of an active-passive metal. 16
i . the passive current density. The potential at i ..is p* c c n t called the critical passivation potential, E . The passive Ir r region starts at E^, the activation potential. The metal will
begin to activate at EA, if the electrode is polarized in the
cathodic direction. During the active-passive transition, a
■3 factor of 10 to 10 decrease in corrosion current can be
observed. In the passive region, the current density lias
little dependence on the potential. The passive current den
sity depends on the chemical composition and the environment.
Also, it should be noted that i is extremely time-dependent.
Finally, at very high potentials, say above ET, the metal
loses its passivity and corrodes at an increasingly rapid
rate. This region is called the transpassive region.
2.5 Theories of Passivity
Many theories have been proposed to explain the passive
state of metals. A survey of these theories can be found in
several pieces of literature (10, 16, 22-3)* However, the
definite mechanism for the cause of passivation is still not well understood (9, 2ij.). In general, the film theory and
the adsorption theory are considered to be the two major theories and will be discussed in detail. Other theories will also be mentioned.
2.5.1 The Film Theory
The earliest and most fully developed theory of passivity 17
seems to be the film theory, which explains the passive state
as the presence of a thin, insoluble and protective film on
the surface formed during the corrosion process. In many
cases this film is some kind of metallic oxide. Almost all
the properties mentioned in Sec. 2.2 confirm or at least do
not deny the film theory.
The following examples are the qualitative evidence for
the existence of a passive film on the surface of a passive
metal (25):
1. Mercury does not wet passive metal, but it amalga
mates with bare metal surfaces.
2. Thin, invisible films can be stripped off of
passivated iron, and then will become visible.
3. The use of an optical technique clearly shows the
differences in the optical properties of the "active" and
the "passive" states of a metal.
Faraday (3) was the first one to propose the film theory,
fie explainedthe passive state of iron in "fuming" nitric
acid to be attributable to the formation of a protective
layer of oxide. He states: "Why the superficial film of oxide, which I suppose to be formed when the iron is brought
Into the peculiar state by voltaic association, or occasion ally by Immersion alone into nitric acid, is not dissoived by the acid, is I presume dependent upon the peculiarities of this oxide and of nitric acid of the strength required for these experiment;...*" Faraday's oxide film theory has been generalized, and is
called film theory. Many workers, principally Evans (26),
Tronstad (27), Hedges (28), Glasstone (29), Mears (30),
Bonhoeffer (31), Franck (32), and Vetter (33), have supported
this theory of passivity. Evans (26) has described it
representatively as follows: "Most cases of passivity with
which the author is personally familiar appear to be
attributable-directly or indirectly-to a protective film,
although not always an oxide-film",
The passive films, those, for example, on chromium or
stainless steels, are usually too thin to be detected by high-energy electron diffraction. As a result, the existence of passive film has been doubted for many years. Evans (3^-1 ) was the first one able to analyze the oxide film of passive iron after stripping this film off in iodine-methanol solution. Iodine does not react with iron oxide, but only attacks preferentially along the interface between metal and metal oxide. It is possible, then, to strip the film from the base metal without dissolving the whole metal.
The optical method has been used to determine the thickness and the composition of the passive film. According to Tronstad's study (27), the thickness of passive film on iron, which is passivated in concentrated nitric acid, is about 25-30 8. A thicker film (about 90-110 8 ) is formed on carbon steel under the same conditions, but a thinner one is formed on austenitic stainless steel (about 10 8). IV
The kinetics of passive film growth seems to be logari
thmic in nature according to inverse logarithmic models
(35* 36) or logarithmic models (37* 38)- However, none of
these models were perfectly consistent with the temperature
and potential dependencies of the growth constants. There
is experimental evidence (39) showing that the growth
kinetics could be expressed with an equal degree of accuracy
by either logarithmic or inverse logarithmic kinetics.
Many investigators (33* 37 I4.O-I4.6) agree that the
composition of anodically passivated iron is duplex in nature,
the layer nearest the metal being magnetite Fe^O^ and the
outer layer being y-Fe^O^. These two iron oxides have
similar crystal structures: Fe^O^ is an inverse spinel with
lattice parameter ao=8.391l 8 (I4.7) * y-Fe^O^ has the cubic,
spinel-like structure with aQ=8.33 8 (I18). However, this model is not universally convincing everyone since the passive films rarely have a thickness of more than 50 8* and under some conditions the film thickness can be as thin as 10 8 (i(9).
It is not reasonable to have two overlying layers of oxide in which each layer's thickness is less than that of a unit cell
(13). There are other models for passive films such as chemisorbed oxygen or oxygen-containing species (14.9-51) * monomolecular (or less) oxide (52), two-dimensional oxide
(53)» three-dimensional oxide (5U)* nonstoichiometric oxide
(55)» and unknown oxide (56)* Clearly, there are disagree ments even among those supporting the film theory of 20
passivity.
One of the objections to the film theory of passivity
is that the formation of oxides or compounds does not always
result in passivation. This is true even for a passivating
metal like chromium. Oxide covered chromium, obrainod by
its being heated in air, can be more active than the
polished chromium. This fact was brought out by Hittorf as
long ago as 1899 (58). Thus, tnc overall properties of the
surface film are important to the passivity of a material.
Those properties include the composition, the thickness,
the defect structures, the adherence, the electric and ionic
conductivities, the solubility and so on. The passivity of
a metal or alloy is not a single-factor determined property.
2.5»2 The Adsorption Theory
There is no question as to the presence of surface film
on a passivated metal. As was pointed out in the last
section, the existence of this film can be established by stripping methods, optical methods and others. However, this still is not direct evidence that passivity is associated with the Isloating behavior of passive film on the corrosion process. Of course, the passive film can serve as a diffu sion harrier film, and does account for the slow-down of the corrosion process. Yet, there may be still other more fundamental mechanisms for the cause of passivity. One of them is the adsorption theory. Belck, in 188?, suggested that an adsorbed layer of
gaseous oxygon can cause the passivity of metal (58)*
Experimentally, Langmuir (59, 60) showed that tungsten with
a layer of adsorbed oxygon is inactive in hydrogen even at
1,200°C whereas the compound can be reduced by hydrogen
at 500°C. However, if gaps are created due to evaporation
or reduction reaction, hydrogen suddenly and quickly reacts
with the oxygen. Langmuir (61) also proposed that a
continuous layer of adsorbed oxygen atoms be necessary for
the establishment of passivity. Other researchers (62-65),
on the other hand, claim that the coverage of adsorbed
oxygen on more active sites, such as kinks and ledge3, can
significantly suppress the corrosion reaction.
An adsorbed species other than oxygen has also been
suggested for the cause of metal passivity. After studying
the effect of oxygen, an oxide film and different anions and
cations present in the solution on the passive state of metals, Gerasimov and lioskvichev proposed that metal passivity
is determined by the chemisorption of its surface by hydroxyl
ion (66).
According to the adsorption theory, a layer of oxygen adsorbed from the air or from aqueous solutions, rather than the oxide, is responsible for the passivity of iron, nickel, cobalt, chromium and stainless steels (67-69). The adsorbed oxygen functions to separate the metal and solution, to displace adsorbed H^O and anions, and thus to increase the 22
activation energy for hydration and dissolution of the metal
lattice (70).
The adsorption theory emphasizes also the observed shift
of the electrode potentials of passive metals. According to
Ershler's data for platinum in an HC1 solution, a 6 percent
coverage of the Pt surface with adsorbed oxygen will change
its potential in the noble direction by 0.12 V and reduce
the corrosion rate tenfold (71)- However, it is very
difficult to have controlled initial state of the metal
surface for the fundamental studies of passivity. It is
quite possible that the surface prior to passivation may be
covered with a layer of some adsorbed oxygen or even an oxiue.
Even in a completely deaerated solution, a metal, for example
titanium, may react with the oxygen of the water and be
covered with a film of adsorbed oxygen or even an oxide
film (72). Consequently, it is not clear whether the measured
amount of oxygen less than one monolayer is uniformly
distributed over a completely clean surface or whether the
oxygen merely severs to fill the openings of an already
existing adsorbed oxygen film or an oxide film.
Hoar (73), who favors the film theory of passivity,
argues that above the potential of zero charge the adsorbed
water dipoles will be orientated with oxygen attached to the metal. Furthermore, above the reversible potential of
M + Ho0 MO - . . + 2H+ + 2e“ it will be kinetically easy 2 solid for protons from the adsorbed dipoles to transfer to other 23
adjacent to them in solution. The adsorbed oxygen ions are
then taken up by any surface metal cations to produce a
monolayer of solid oxide.
2.5*3 Combined Film-Adsorption Theory
The film theory and the adsorption theory represent two
schools of thought regarding the passivation mechanism of
metals; however, they do not contradict, but rather supplement
one another. As the adsorbed film thickens, the free energy
of adsorption per mole of oxygen decreases. Thus, multilayer
adsorbed oxygen on metal has a tendency to change into an
oxide film ( 7U-) • anodic process will ^e retarded by
change in the double-layp'r structure, and the increased
difficulty ions will experience in passing through the protective film. Hence, a combined film-adsorption theory can more fully explain the mechanism of passivity of metals.
3ome investigators (75-78) support this intermediate theory of the passivity of metals.
Perhaps the most important mechanistic point of pa-ssivity to be established is the reason for the anodic current’s dropping so quickly between E and before reaching a stable passive state (see Figure 2). There is experimental evidence which suggests that complete coverage is unnecessary for the occurrence of passivity. However, in this case, a thicker film is needed for a change which will last after the film-forming conditions have been removed. This idea is 2k brought cut by Frankenthal (79) in his study of Fe-2lp5 9r
alloy in sulfuric acid. Ilia modal involves (i) a primary
film of only a small fraction of an oxygen monolayer, which
coull produce temporary passivation and cause the reduction
of she anodic current in the potential range from IS :;o E,, £9 l-J •/* and (Li) a secondary film cf three-dimensional oxide, which
is very stable and resistant to electrochemical reduction.
Here, it is net necessary that the composition of the oxide
film bo identical to that of the bulk ~ \.i des of :;he tnl.
2 .5.14. Other Theories
At present, the film and the adsorption theories are the
most accepted and validated. Several other theories have
been proposed to explain the passive state, most of them
widely differing. Some of these theories are the allotropy
theory (2, 80-83)> the hydrogen solution theory (81p— 36), the reaction velocity theory (87-89), and the electron con figu ration theory (90-97)* Only the electron configuration theory will he discussed briefly here.
According to the electron configuration theory, the ease of transition metals in forming a passive state (e.g. chromium* nickel, cobalt, iron, molybdenum, and tungsten) is related to incomplete inner shell (d electron) energy levels in the metallic state. This theory assumes that these incomplete energy levels tend, to fill with electrons. The passive state is ascribed to the unfilled d electrons in the metal or alloy and the active state, to the electron-filled condition of the
d band. Thus, adsorption of oxygen or oxidants accompanies
passivity because it leads to the acceptance, from the metal ,
of electrons; the d bands of surface atoms remain iiieoin plcto.
Adsorption of hydrogen or reducing agents, on the other hand,
supplies electrons and may convert the metal into the active
state.
The electron configuration theory does not deny film
formation, but rather considers it an adjunct which can offer
additional protection to the metal. However, this model is
not accepted by Horth and Pryor (98), Evans (99-100), and
others (16, 101), Pryor, et al. (98) argue that this theory
may not extend to other environments, namely chloride, and
that the protective film is really three dimensional; the
reaction rate may be interpreted in terms of the concentra
tion of defects. The chloride ion is not a reductant;
according to the electron configuration theory, it should not selectively destroy the passive state. Evans (100) disputes the view that the unfilled d-band would in the attachment of oxygen atoms; oxygen is a non-metal and its atoms tend to take up electrons rather than dispose of them.
It is also argued that this theory can not explain the following pertinent facts: some non-transition metals, e.g. aluminum, magnesium and beryllium, can be rendered passive under suitable conditions; several transition metals, (ex cepting chromium, nickel, iron, tungsten, cobalt, molybdenum, i?U
titanium, niobium, and tantalum,) do not possess high passive
properties. It is not clear why particular unfilled inner
shells, such as those found in chromium, are able to display
high passive properties.
Finally, regular properties of metals depend on the
internal structure of atoms and accordingly on the electron
configuration. Passivity, however, is complex in nature and
depends not only on the metal, but also on the environment.
It is obvious, then, that the proposal that the passive stele be attributed to the electron configuration only, is inadequate.
2 .6 The Passivity of Alloys
Alloys are practically and commercially important materials. They may have many advantages over pure metals.
By adjusting the alloying composition, it is possible to improve their mechanical and chemical (corrosion) properties.
The influence of alloying elements on the changes in the properties with regards to the passivation of alloys is rather complicated.
Tomashov (102) summarizes the effect of alloying elements on characteristic points of the anodic polarization curve for Fe and Ti in H^SO^ solution as shown in Figures 3 and I;.. An alloying element can make the passivating ability of an alloy more perfect by various means: by decreasing the critical current density (i . and the passive current * c n t 27
(+)
Cr, Mo, V -r___ JT Ni?Si E pit * Cr, Ni, Mo, Si, V, W
-Mo, V
;Cr, Ni, ^=Cr, Mo, V, Ti, Nb, Ni E Si, W
Cr, Si
J C$
Ni, Ti, Mo £ i Tr----' Tr>[Cr t«=sNit TiT' 1 Ni, Mo (-) ■*; • iv i X p 'crit log i
Figure 3- Effect of alloying elements on characteristic points of an anodic polarization curve for Fe in HpSO, solutions. ^ Arrows indicate favorable (^=) or unfavorable (— for passivation direction, (Reference 102), 28
-►Ta, A1
Al,Cr' ^=Cr,Mo,Zr
Cr,Mo,Ta
XI Cr 'PP
i log i
Figure ij., Effect of alloying elements on characteristic points of an anodic polarization curve for Ti in HpSO, solutions. ^ The meaning of the arrows is the same as in Figure 3* (Reference 102). 29
density (i ), by shifting the critical passivation potential P (E^) and the activation potential (EA ) to more negative
and potentials, and by shifting the pitting potential (E pi. ,)u the transpassive potential (E,p) to more positive potentials.
However, as we can see from Figures 3 and I4., no alloying
element can achieve all of these goals. The influence of an
alloying element on corrosion properties greatly depends on
the type of alloy also on conditions of corrosion. As an
example, chromium may cause a great reduction in icr^ t for
Fe-Cr alloys in H„SO, solutions, but will increase i .. for J 2 4. crit Ti-Cr alloys under the same conditions.
2.6.1 Theories of Passivity in Alloys
Corrosion processes involve the electrochemical reactions between an electrolyte and a, metal surface. Thus, modifi cations of the surface may have significant influence on the passivation of a corrosion system. Adding an alloying ele ment to a metal can certainly modify the surface of a corrosion system, and consequently change the corrosion processes.
It is well known that if a noble metal such as gold is alloyed with a less noble metal such as copper, the corrosion resistance of the alloy increases suddently beyond a certain threshold content (1 mole of gold to 7 moles of copper for this case). Beyond the threshold content, a layer of pure gold can be formed on the surface to prevent further corro sion. A similar occurrence is also observed in Fe-Cr alloys 30
and other alloys. However, Tammann (103) considered these
to be two different kinds of characteristics: parting limits
for systems like Au-Cu alloys, and passivity limit3 or
- critical alloy concentrations for systems like Fe-Cr alloy.
Theories of passivity used to describe pure metals arc
also proposed, for passive alloys; nominally, they are the
film theory, and the adsorption theory. Considering Fe-Cr
alloys, by the film theory, the passivity in high chromium
alloys ( 12 wt % Cr) is due to a layer of protective oxide
film similar to that suggested for pure chromium. Details
wi 11 be discussed in the next section. Tammann (10L|.) and
others proposed that the passivity of stainless steels is cr-e
to a layer of oxygen instead of oxide.
2.6.2 The Passivity of Fe-Cr Alloys
The properties of an alloy are closely related to the
properties of the components of the alloy in their individual
state. Therefore, the properties of pure iron and pure
chromium will also be reviewed in -this section in addition
to the properties of Fe-Cr alloys.
2.6.2.1 Iron (105-107)
Pure iron is a grey, rather soft metal with structures
in the cubic system. It melts, at 1536°C and boils at 3000°C;
its density is 7.071+ g/cm*^ at 20°C. One of the most
remarkable physical properties of this metal is its magnetism. 31
The magnetic susceptibility of iron is more than a million
times greater than of any other element except cobalt and
nickel. Only a-ferrite possesses ferromagnetism and 5-ferrtLc
is alight paramagetic. Iron has a strong affinity for oxygen and it r-ujJbly
oxidized. It burns brightly in oxygen, and will also burn
in air when in finely divided form, with magnetic Fe^O being
formed. Heated iron reacts very energetically with chlorine,
sulfur, phosphorus and steam.
Iron dissolves in dilute acids, such as sulfuric acid
and hydrochloric acid, and hydrogen is liberated. In cold
dilute nitric acid hydrogen is not evolves; instead, the acid
is changed to ammonia, and then reacts with excess nitric
acid to develop ammonium nitrate. With concentrated nitric
acid, iron, after a momentary reaction, does not appear to react and is said to have become passive. Passive iron does not dissolve in dilute nitric acid, precipitate copper from copper sulfate solution, lead from lead nitrate, or silver from silver nitrate, but it will dissolve in reducing acids, e.g. dilute hydrochloric acid. Other oxidizing agents, e.g. chromic acid and hydrogen peroxide, can make iron passive too.
The rusting of iron is a special case of corrosion because of its great economic importance. Rusting is not a simple process, and workers do not agree on the simple facts. It seems that the presence of water, an electrolyte and oxygen are essential for the occurrence of rusting. Rust generally contains ferrous oxide, ferric oxide, water, and
carbon dioxide. For details on rusting one can refer to
Evans' work (108) and others.
Iron does not appreciably corrode in alkalis, except
at high temperatures.
Some of the properties of pure iron and also pure
chromium are listed in Table 1.
2.6.2.2 Chromium (10£, 107, 109)
Chromium is a white and lustrous metal. Pure chromium
is relatively soft and can be rolled, forged, or extruded at
high temperatures (800-1250°C). Yet, commercial chromium
is among the hardest of the common metals because of the
presence of small amounts of carbon. It has a high melting
point, 1875°C, but its boiling point is not proportionately
high (2665°C at 760 mm Hg). Its density is 7*19 g/cm^ at 20
The common form of chromium is body-centered cubic and the
structure is changed to face-centered cubic at about l8L(.0oC.
Chromium is extraordinarily resistant to numerous
chemicals, including oxidizing acids, particularly at low
temperatures. It retains its bright finish in air, even in
the presence of moisture, and only tarnishes when it is warm
On the other hand, at elevated temperatures, it reacts with many substances. It burns in the oxy-blowpipe flames. It
also combines with sulfur, nitrogen, carbon, silicon, boron, halogens, and several metals, though only when it is hot. Table 1. Some of the properties of pure iron and pure chromium
Property Value of Fe Ref. Value of Cr Ref.
Atomic number 26 106 2k 109
Color Silvery-white 106 Gray 109
Atomic weight 55.8U0 106 51.996 109
Density 7.873 g/cm3 (20°C) 106 1,1k g/cm3 (20°C) . 109
Atomic volume 7.09U cm3/mole 106 7.29 cm3/mole 109
Crystal structure -*l81pO°C: body-centred 106 -910°C:body-centred 109 cubic cubic ~l81j.0 C : cubic 910°C*t-1390°C: face- centred cubic =-1390 C:body-centred cubic Melting point I535°c 106 1903 * 10 C 109
Heat of fusion 65 cal/g 106 67.3 cal/g 109
Boiling point 3000°C 106 26J|2°C 10 9 1 2^ 2„ 6,, 2 _ 6-,6. 2 , 2„ 20 6 _ 2. 6_,5] 1 Electron configura Is 2s 2p 3s 3p 3^ M-s 106 Is 2s 2p 3s 3p 3d l|.s 10 9 tion Oxidation states -2,-1,0,+1,+2,+3,+Ji, 106 -2,-l,0,+l,+2,+3,+!|, 1 0 9 +5,+6 + 5,+6 Chromium dissolves in nonoxidizing acids, for example,
hydrochloric and sulfuric acids, slowly when the2 are colc
and more rapidly when they are hot. It is not attacked by
cold aqua regia or nitric acid, either concentrated or dilut.
It is passivated by these acids and by other oxidizing solu
tions of chlorine and bromine. Passive chromium acts like a
noble metal, and does not dissolve in nonoxidizing acid.
Chromium is attacked by fused hydroxides at red heat,
and also by nitrates and chlorates, but not by alkali carbonates. p 2 6 The electron configuration of chromium is Is , 2s 2p ,
2 6 ^3 1 3s 3P 3d , Its . This structure of two outermost shells is favored over 3d^lj.s^ because of the increased stability of tbi half-filled 3d shell. Chromium has the following oxidation states as follows: -2, -1, 0, +1, +2, +3» +6. The +3 most stable and important state is Cr , which has a strong tendency to form complexes. Therefore, atoms or radicals bound to it are often not separated off in solution.
Some of the properties of pure chromium are also summarized in Table 1.
2.6.2.3 Iron-Chromium Alloys
Iron is capable of forming many valuable alloys. Iron- chromium alloys are undoubtedly among the most important of all alloys since chromium -is a fundamental ingredient of stainless steels. Alloying with chromium can prevent iron 31;
and steels from rusting under ordinary conditions. A steel
with more than 12 wt % Gr is called stainless steel.
The iron-chromium is basically a continuous solid
solution as shown in Figure 5 (110). This solid solution
is close to ideality (111). Yet, there are two important
zones, the gamma loop and the sigma phase, in this system.
Chromium is known as a ferrite stabilizer, which extends the
alpha phase field and narrows down the gamma phase field into
the so-called ’gamma loop*. This loop can be seen on the
left side of Figure 5 corresponding to high temperatures
(831-1394°^) and low chromium contents (<12.7 wt %),
There is another zone, occupied by a sigma phase, centered around the 45 wt % Cr axis at low temperatures. The sigma phase is an Fe-Cr compound with a tetragonal structure.
The alloy containing a sigma phase is very hard and extremely brittle. Above 821°C the sigma phase dissolves back into the chromium-rich alpha phase.
Chemically, chromium is a rather active element when not passivated. It can easily displace copper, tin, and nickel from their aqueous salt solutions. This can be seen from the standard EMF series of metals as listed in Table 2 (112).
The EMF series is a sequence of the standard oxidation- reduction (redox) potentials for all metals. The EMF series is determined by galvanic coupling between metals in equili brium with their ions at a concentration equal to unit activi ty. This series indicates the tendency of the metal to be 36
Cr-Fe Chromium-Iron
1600
1600 '<■
1400
(200 J>
(000 i
800 ! cr 600;
400 (____ io 20 30 40 50 60 70 80 90 Cr
iV p h j m ('ore^i'ijqe Cf'*■ .rmurn
Figure $. The phase diagram of Cr-Fe system. 37
Table 2. Standard EMF series of metals (Reference 112)
Metal-metal ion Electrode potential vs equilibrium normal hydrogen (unit activity) electrode at 25°C, volts______Au=Au -^+3e +1 ,1^98 0p+4H++4e=2Ho0 +1.229 +? Pt=Pt+2e + 1.2 +2 Pd=Pd +O.987 Ag=Ag++e +0.799 2Hg=Hg*2+2e +0.738 Fe ">+e=Fe +0.771 0?+2HpO+4e=40H +0.401 Cu=Cu +2e +0.337 Sn+^+2e=Sn+2 +0.15 2H++2e=H2 +0.000 Pb=Pb+2+2e -0.126
Sn=Sn+2+2e - 0.136 Ni=Ni+2+2e -0 .2^0 Co=Co+2+2e -0.277 Cd=Cd+2+2e -0.403 Fe=Fe+2+2e - 0 . 4 4 0 Cr=Cr+^+3e - 0 . 7 4 4 Zn=Zn+2+2e -0.763 Al=Al+3+3e -1.662 Mg=Mg+2+2e -2.363 Na=Na++e -2.714 K=K++e -2.925 38
ionized and go into solution. Thus, the more positive reclox
potentials correspond to the more noble metals, and vice
versa. Table 2 shows that chromium is near the bottom of the
sequence.
Since the equilibrium in metal-metal ion galvanic couple
rarely occurs in actual corrosion problems, and since alloys
are not included in this EMF series, the 'galvanic series'
has been elaborated. This series is an orderly arrangement
of metals and alloys according to their relative electrode
potentials in a given environment. This series gives a more
correct prediction of galvanic relations than the EMF series.
The galvanic series of metals and alloys in sea water, based
on galvanic corrosion tests, is given in Table 3 (113)-
Passivity has a great influence on galvanic corrosion t. ehavior.
For example, passive stainless steels occupy more noble
positions as compared with the lower positions of active
stainless steels.
A high degree of corrosion resistance can be achieved by
alloying chromium with iron. The most important influences
of alloying chromium on corrosion properties are the reducing
of the critical current density for passivating, which means
that Fe-Cr alloys are more easily passivated and the increas
ing of the stability of passivity once it has been established.
The true critical current density for iron, in the absence of p any corrosion product, is about 1? A/cm in IN , Tout -2 2 only about 3-2x10 A/cm for chromium under the same 39
Table 3* Galvanic aeries in seawater (Reference 113)
Platium Gold Noble Graphite Titanium Silver f Chlorimet 3 (62 Ni, 18 Cr, 18 Mo) LHastelloy C (62 Ni, 17 Cr, 15 Mo) '18-8 Mo stainless steel (passive) 18-8 stainless steel (passive) •Chromium stainless steel 11-30% Cr (passive) rlnconel (80 Ni, 13 Cr, 7 Fe) (passive) LNickel (passive) Silver solder 'Monel (70 Ni, 30 Cu) Cupronickels (60-90 Cu, ij.0-10 Ni) Bronzes (Cu-Sn) Copper •Brasses (Cu-Zn) rChlorimet 2 (66 Ni, 32 Mo, 1 Pe) LHastelloy B (60 Ni, 30 Mo, 6 Fe, 1 Mn) rInconel (active) LNickel (active) Tin Lead Lead-tin solders r18-8 Mo stainless steel (active) Ll8-8 stainless steel (active) Ni-Resist (high Ni cast iron) Chromium stainless steel, 13% Cr (active) rCast iron LSteel or iron 2 0 2 i*. aluminum (ij..5 Cu, 1,5 Mg, 0,6 Mn) Cadmium Active Commercially pure aluminum (1100) Zinc Magnesium and magnesium alloys conditions (llif). There is more than a factor of 50°
difference in passivating ability between iron and chromium.
For those Fe-Cr alloys or stainless steels, which contain
more than 12 wt % Cr, critical current density is reduced
to such small values that they are self-passivating in
aerated aqueous solutions. A chromium content of 12 wt % is
commonly called the critical composition for passivity of 1.1,
Fe-Cr systems. In addition, the passive state of high-Cr
alloys is much more stable than that of low-Cr alloys.
Aronowitz and Hackerman (115) claim that the passive film ot.
alloys with more than 12 wt % Cr can not be removed unless
there is severe cathodic treatment or repolishing of the
surface. On the contrary, it is very easy to take away the
oxide film of iron by cathodic reduction (1;3> 116).
The corrosion properties of Fe-Cr alloys were studied by investigators (117-127). Some of their results are shown in Figures 6-12. As we can see from these figures, chromium has a notable effect on the corrosion properties of Fe-Cr alloys, and in particular, there is a radical change around the critical composition. The low-Cr alloys, below the critical content, have properties similar to Fe, while above the critical Cr content, the alloys have behavior similar to
Cr. The actual critical composition based on corrosion data frequently depends on the media to which the alloys are exposed rather than on the chosen criterion. A glance at corrosion rate (Figure 6), corrosion potential (Figure 7), 009 008
007 006
& .005
° .004 .003 .002
001 0 0 2 4 6 8 10 12 14 16 18 20 Wt % Cr
Figure 6. Tne corrosion rates of Fe-Cr allots in intermittent water spray, room temperature (Reference 117). k?-
-p i—I •
Wt % Cr Figure 7* “The corrosion potentials of Fe-Cr alloys in NaCl (Reference 118). ue . tnadFaeptnil frF, Cr, Fe, for potentials Flade Standard 8. gure
Standard Flade potential, V (SHE) - 0.1 o.i 0.3 0.5 0.7 n eC aly {eeec 119-121). {Reference alloys Fe-Cr and 0 10 Wt % Cr % Wt oh ad Lennart and * Rocha * King and Uhlig and * King * Franck 20 100
10=s IIjSO, (0‘n.er)
kT r
pH - 3
pH = 7
o Direct method Qx* Indirect method
-3 10 0 5 10 Wt % chromium
Figure 9* Critical current densities for passivation, of Fe-Cr alloys in deaerated 3% NapSO, at pH 3 and 7, 25 C (Reference 121). Data4 in 10% HpSO. , room temperature, by R. Olivier (Reference 122). OJ Figure 10. Potentiostatic polarization curves Tor curves polarization Potentiostatic 10. Figure Current density (amp/cm 10 10 10 10 10' 10 10 10 10 ' - 0.3 : Cr.8: Fe-lZ+Cr,5: 7: Fe-l8Cr, 6: Fe-l6Cr, : e67r 3 Fe-9.5Cr,if: 3: Fe-6.7Cr,2: Fe-12Cr, uvs t 7 b Oiir (Reference Olivier Kolotyrkin M.Y. 8* by Curve 122). 7# by to 0 Curves solutions. acid sulfuric in alloys Fe-Cr oeta i ot (SHK) volta in Potential (Reference lllj.)# Fe-2.8Cr, 1: Fe, 0: o 03 06 09 12 +1.5 +1.2 +0.9 -0.6 +0.3 * ■ —
- -o
kS 0 10
-1 10
10
(M
-p •H
o
6 10 -l\.QO 0 +1*00 +800 +1200 +1600 Potential, m v (SCE)
Figure 11, Potentiodynamic polarization curves for Fe-Cr alloys in IN HpS0. at 25 C. 1; Fe, 2: Fe-lOCr, 3? F&-20Cr, 1*: Cr. (Reference 125). Figure 12. Effect of chromium on the corrosion rate corrosion theon chromium of Effect 12.Figure
Corrosion rate, mg/rim . 2ij. hr. I 2000 2500 3000 5000 1 3500 5500 6000 7000 4
.OOO 500 000 0 0 : 5C 2 60°C, 2:C, 151: 32%HN0-. containing acid innitric Rfrne 128).(Reference 5 Wt % Cr 3 10 : boiling. 15 J
20
k*
Flade potential (Figure 8), and critical current density
(Figure 9) indicates that they all yield almost the same
critical composition. However, in 32?£ HNO^ the critical Cr
content for the passivity of Fe-Cr alloys is at about 7 wt ^
Cr ( as shown in Figure 12) (128), whereas in FeSO^ solution
the critical value is increased from 15 to 20 wt% Cr (10).
In HC1 acid there is no critical chromium concentration since
Fe-Cr alloys are not passive in this solution. Only in
neutral aqueous solutions is the critical composition equal
to 12 wt % Cr (see Figure 9).
Figures 10 and 11 are anodic polarization curves for
Fe-Cr alloys in sulfuric acid solutions, by potentiostatic
and potentiodynamic techniques, respectively. There are
oscillations in the active-passive transition regions (shown between dotted lines in Figure 10) of Fe and low-Cr alloys when using the potentiostatic technique, but not when using the potentiodynamic technique. Keeping them at a constant potential in these regions would cause a violent oscillation of current to continue. By the potentiodynamic technique a quick enough scanning rate can suppress these oscillations, nonetheless Figures 10 and 11 indicate that a film of greater stability can be more easily formed on high Cr alloys than on iron and low-Cr alloys.
It is probable that the composition, the structure, the thickness, the dissolution rate in solution, and the electro nic and ionic conductivity of the passive film distinguish the stainless from the non-stainless alloys and steels. As
explained earlier (Section 2.1), it is difficult to define
passivity; however, it is clear that the passive film has the
following properties as the most complete passivation is
reached (2 5 )t
1. Extremely low ionic conductivity.
2. Noticeable electronic conductivity.
3. Extremely low chemical solubility and dissolution rut
1;. Thermodynamic stability in a wide range of potential.
Good adherence and high compressions! strength.
The dissolution rate of the passive film is a function
of the chemical composition at the passive film/solution
interface. At steady state the dissolution rate is equal to
the corrosion rate of the metal in the passive state. The
very low corrosion rate is one of the most significant
properties of the passive metal. The chemical composition
and thickness of the passive film are dependent on the
potential. The steady thickness of the film is controlled
by the ionic movement through the film, which is a function
of the electric field strength across the passive film. The
field strength and consequently the ionic current decrease as
the potential difference decreases and the film thickness
increases. The film thickness changes until the steady state is reached, when the ionic current density is equal to the corrosion rate. In this case the anodic rate of film for mation is equal to the dissolution rate in the solution (14.2 ). Studying the chemieal composition of passive films of
Fe-Cr alloys is of considerable value in understanding the
nature of their stability. However, because of experimental
uncertainties, still little is known in this field, although
many researchers have studied it. Evans (129) attempted to
isolate a film from unheated 18/8 stainless steel by using
the anodic treatment. In this method, the specimen is made
the anode in potassium chloride solution. The anodic dis
solution of the specimen starts at the bottom and enters up under the film, which is cleaned in hydrogen, and removed b;,
a water jet. But, in this attempt the collected skin contains too much residual metal to be transparent. The same method has been used to study 13 wt % Cr stainless steel (131>) and, in this case, a very thin transparent film can be obtained.
Vernon, et al. (131) used anhydrous methanol-potassium iodine as the stripping reagent in an apparatus designed to remove water and oxygen from the system, which is a modified
Evans' method (34)* ln this study, they found a marked enrichment of Cr in the film of 18/8 stainless steel, parti cularly when the specimen was highly polished. Later, a similar method (132-133)» known as the bromine-methanol technique, was used to analyze the composition of passive film from stainless steels. In these studies, chromium shows no significant preferential enrichment within the passive film; however, silicon and molybdenum are enriched notably. Many studies using other techniques show chromium enrichment
(in the form of chromium oxides, iron-chromium oxides, or
unknown oxides) within the passive film on stainless alloys
(122, 13^--lill) as well as on stainless steels (1I|_2—It*.7*) •
Cr203 has been suggested by Olivier (122) as the
substance that is formed on the surface of Fe-Cr alloys and
18/8 stainless steel. He recorded potential-time and
current-time curves of the passivation of Fe-Cr alloys in
10# ^ 2 ^ 4 s°lu ti°n. He observed an arrest of the potential
after the sharp increase, and before the oxygen evolution potential is reached, for those alloys with 6.7 wt % Cr or more. The reaction, as follows, is suggested for the cause of passivity:
2M2+ + 3H20 ___ M203 + 6H+ + 2e",
According to Sukhotin, et al. (12i|) the passive film on chromium in sulphuric acid solution should be mixed oxides of
Cr20^ and Cr02, its composition being close to CrC^ g. Their- reasons are as follows:
1. The passive film forms on the electrode at potentials _2 between -0.3 and -0.1 V, and can be oxidized to Cr2C^ at
EQ ^ 1.15 V. As shown in Table i(. (12lj.), all these properties are characteristic of mixed oxides of Cr20g and Cr02 ,
2. Cr20^ is a dielectric with high electronic resistivity; however, the passive film generally should have appreciable electron conductivity (25). Otherwise, it will be difficult to explain the ease of electrochemical reactions Table i|. Normal potentials for the redox processes on Cr in acid solutions (Reference 12i+)
Redox reaction Oxidation of Chromium Cr = Cr+2 + 2e -0.91 Cr = Cr+3 + 3e -0.7U 2Cr + 3H^0 - Cr203 + 6H+ + 6e -0.58 Cr + Cr+-"+ 3H20 = Cr203 + 6H+ + 3e -O.ij.1 Cr + 2H20 = Cr02 + i4-H+ + lj.e -0.15 Cr + 3H20 = CrO^ + 6H+ + 6e 0.29 2Cr + ?H20 = Cr207"2 + 1^H+ + 12e 0.29 Oxidation of Cr+2 +2 + 3 Cr = Cr J + e -0.i|.l 2Cr+2 + 3H2 0 = Cr203 + 6H+ + 2e O.O87 Cr+2 + 2H20 = Cr02 + Ij.H+ + 2e 0.62 2Cr+2 + 7H20 = Cr207~2 + lij.H+ + 8e 0.90 Oxidation of Cr+^ Cr J + 2H2 0 = Cr02 + If H + e 0.15 2Cr+3 + 7H20 = Cr207"2 + ll|.H+ + 6e I .33 Oxidation of C**203 Cr203 + H20 - 2Cr02 + 2H+ + 2e 1.11*. Cr2°3 + 3H20 = 2Cr03 + 6H+ + 6e 1.15 Cr2°3 + IjfUO = Cr207"2 + 8H+ + 6e 1.17 Oxidation of Cr02 j. u n _ j. ou+ j. o^ t to at a passive electrode. By the method of impedance measurement, Lovrecek and Sefaja (126) concluded that the passive film on chromium in sulphuric acid solutions is electrically non-homogeneous Cr^O^ with a thickness of 25 8. By using ESCA and AES for investigating the passive fii.ns formed on chromium electrodes in sulphuric acid baths, Romand, et al. (127) concluded that the composition is close to Cr^O^ with a 13-16 8 thick layer. Their results also indicate that hydrated species are located at least in the outermost part of the passive film. Okamoto (llt-5) has proposed that passive films are not very protective unless they enclose, in addition to iron anu oxygen, some hydrogen, probably as water. Another important parameter for the high corrosion resistance of stainless steels is the amorphous nature of the film. The role of water is expressed in Figure 13. Figure 13(a) shows that the passive film is a hydrated oxide film having amorphous structure. Metal ions can dissolve throughthe bared part of the film in the form of M0H+. Then MOH+ reacts with the neighboring H20 molecules, and precipitates as oxide in the bared part, as shown in Figure 13(b). Thus, Okamoto (1U5) concludes that the passive film is basically hydrated Cr20^ on stainless steels, and that the film changes to the less hydrated structure by the aging process. At any stage of aging, the film contains one of these three different bridges, 5k \s •0-M-0Ho -0-M-0H„ / \ 2 0 OH 0 OH \ / ■0-M-0Ho -0-M-0Ho / \ 2 / \ 2 OH OH OH OH H X*/ MOH (H20) -OH-M-OH, s ' t *= OH„OH- OH OH H n 2/ 2 'to ■0-M-0H„ -0-M-0Ho / \ 2 I \ 2 0 OH V 0 OH / \ / -0-M-0H„ ■> -0-M-0Ho / \ ^ / \ 2 (a) (a*) Figure 13. Metal ions dissolved through the undeveloped part in the film (a) are captured to form film (a') due to the bridging of OH surrounding the part (Reference lij.5) • - M - OUg, - HO - M - OH -, or -0 - M - 0 -, depending on the loss of a proton. Heumann and his co-workers (131+-135) have used an electrochemical method to identify the composition of a passive film on chromium surfaces. Their potential/current density curves of Cr, and CrOOH in 0.1 N HoS0^ in the transpassive region are parallel to each other. In reference to the Cr curve, there is a 35 niV shift of the CrOOH curve to a lower potential, and a R0 mV displacement of the Cr^O^ curve to a higher potential. The shift of the CrOOH curve is attributed to the reaction area of powdered CrOOH, which is larger than that of chromium. Tney conclude that CrOOH is present in the passive film of Cr and Fe-Cr alloys. Recent x-ray photo-electron spectroscopy (ESCA) studies (1^6-147) show that Cr0x (0H)^.*n 1^0 is the major constituent of stable passive films on 13/8 stainless steel, Fe-30 wt %Cr alloys (with and without 2 wt % Mo), and amorphous iron alloys. Cr(OH)^ formed on Fe-Cr alloys1 surfaces has been suggested by Weidinger and Lange (136). Tney have considered the possibility of many chemical and electrochemical reactions and proposed the following reactions for each state: 5£- (1) In the active state, H + (aq) ♦ e" * H 2° (* )l E o= 0 Cr+2 (s) ► Cr+ 2(aq); E°= -0.91 Cr+2 (s) + 2 JI20(lJ y Cr(OH)P + 2 Ii+(.'*q)i E°= -0.59 V (2) In the passive state, C r (O H )g + H 20 (1 ) --- > Cr(0H)3 + H+ (uq) + o". E - -O.36 o (3 ) In the transpassive state, Cr(0H)3 ► CrO^ + 3 li+ (aq) +3e"; where (s), (g), (1 ), and (aqj mean solid, gas, liquid, and dissolved respectively. Olefjord and Fischmeister (141), from their EECA studie.; of Cr steels in oxygenated water, state that the passive fil,.* generally consists of two layers* an outermost one of Cr(Oll),, and an underneath one of Fe-Cr oxide ( l’ e2+xt'r2 it t} the composition of the film alters with passivating Li ;.e* brief times give an Fe-Cr ratio close to that of the alloy; long times yield increasing Cr enrichment through the whole film, with the highest concentration at the outer layer. On the other hand, the low Cr steels ( 3«9 and 7.8 wt l,j Cr ) were passive for a short time, during which the Cr enrichment v/ac high ( Q times ); the chromium-rich layer, after it:. breakdown, was substituted by PeOOK. Prazak, et al. (137) have postulated that spinel iron- chromium oxide be the component within the passive film of Fe-Cr alloys on the basis of their electrochemical behavior in IN solution. They have noticed two limiting compositions in the transpassive region. Up to 16 wt , Cr content, there is no continuous corrosion in this region. Jteels with 18 to 30 wt Cr content are continuously corroded, and show the secondary passivity. Steels with rnoru than 35 wt ‘j Cr content are continuously attacked, but do no;. exhibit the secondary passivity. The corrosion behavior of these alloys are attributed to the formation of the crystalline structure of the spinel oxide. Qualitative changes in the properties of the oxide could occur in the care of a spinel oxide, when chromium ions occupy more than ’■ and I of the cation sites. These limiting values are equivalent to a theoretical composition of 15*5 and 30.7 wt Cr, respectively, in the Fe-Cr system, which are close to the Measured values. Okuyama (ldO), et al. have determined the compc.,i. lion h the passive films of iron-chromium alloys and 1 8 / 8 stainless steel from the chronopotential diagram., with the colorimetric analysis of dissolved metallic ions. They conclude that the passive films consist of the duplex Fe oxide layer of Fe_G. j r and Y -Fe„0„ and fine grains of Cr oxide ( probably Crp0~ ) 2 3 dispersed through the layer. They sug that corrcc: i o.' 5ft resistance does not depend on the thickness of the passive film but on the activity of Y-Fe^ i n the passive film. Corrosion resistance of Fe-Cr alloys is improved by the Cr enrichment in the passive film due to the decrease of pi I and cyclic anodic and cathodic polarization. Tammann (10*0 and others imputed the passivity of stainless steels to a layer of chemisorbed oxygen rather than oxides. All known iron oxides and chromium oxides are unstable in solutions of pH < 1 (see Figures 1^ to 18) (1A8). ;Jtill, iron, chromium and Fe-Cr alloys can be passivated in very acidic solutions. Frankenthal (l*f9) suggested that an unknown Cr*0 species or a compound of unknown structure and thermodynamic properties be responsible for the passivation of those materials in very acidic solutions. This possibili ty cannot be ruled out, although one of those known chromium oxides can be metastable and can exist within the passive film • There are not many studies on the structure of passive films of Fe-Cr alloys. Yearian, et al. (150) have measured the lattice parameters of Fe-Cr spinels. Their results are tabulated in Table 5* Spinels of the FeFe^2_x )^r „ C'l.» 0 FeDH** ■ Fe( 0 H l2 ta. ■'JFeBO a - - $ 3 )—-S -2-1 0 1 10 II 12 13 14 15 16 pH Figure li+.. Potontial-pH diagram for the Fo-HgO system, at 2$ C (considering as solid substances only Fe, Pe-0, and Fe20 ). (Reference lij.8). J ^ 60 - 2 - 1 0 1 10 II 12 13 14 15 15 ^ Fe(OH)3 F c( OH)2 HFeO 10 II 12 13 14 15 16 pH Figure 15. Potential-pH giagram for the Pe-H 0 system, at 25 C (considering as solid substances only Fe, FefOH)^ and Fe(OH)^), (Reference llj.8) * 6 L -2-1 0 1 3 4 5 7 8 9 10 11 12 13 14 15 16 1 i n irr- '7 iogC- 01 1 ■<> HaCrO*] HCrO, C r l O H I t Cr+ + --Q3------V 4 ® Cr( OJ b (J II 12 3 14 1.' figure 16. Potential-pH diagram for the Cr-H„0 system, at 25 C (considering as solid substances only Cr and anhydrous Cr203 ). (Reference II4.8 ). CIV) 0.0 0.8 0,0 0,6 0,2 - 0,2 Cr(OH) - 0,6 - 0,6 - 0,8 OO, 30) -1/t Cr - 1,6 Figure 17. Potential-pH giagram for the Cr-H 0 system, at 2 % C (considering as solid substances only Cr and Cr(0H)_). (Reference lij.8). - 2-1 0 1 6 10 II \2 13 14 15 16 2.2 — ' n-- rT T " 2,2 Io j C- 0 ! i E(V)2 J * i - (i 0 2 rs* 1.8 1 . 8 HjCrO^ CrjOj" 1,6 HCrO 1,6 14 W'J! 14 1.2 1,2 1 I 0,0 0,8 Cr(OH 06 0,6 0.4 0,4 0,2 0,2 0 CrOH ! CrlOHlJj ' * a G) w)(^Xj£i - 0,2 -v| f,r f i - 0,2 ; Cp(OH)3 . nH jO - a , 4 -0.4 - 0.6 - 0,6 - 0,6 - 0,8 -l -1 r(OHl - 1,2 - 1,2 ~14 -1,4 Cr°2 ^ > ^ 5 ^ - 1,6 - 1,6 - 1,8 i t__ i__ L-- 1 , 8 -2 -1 8 9 10 II 12 13 14 10 10 pH Figure 18. Potential-pH diagram for the Fe-f^O system, at 25 C (considering as solid substances only Cr and Cr (OHj^-nH^O), (Reference lij-8). Table 5* Composition x, lattice parameter a, degree of inversion I, and molecular formula at the boundaries of oxygen-rich (O.R.) or metal-rich (M.R,), (Reference 150) Boundary x a(S)±0.002 I Formula M.R. 0 8.396 1.0 Fe+3(Fe+2Fe+3>\ 0 Fe+3(Fe+2Fe O.R. 8.390 1.0 +H M.R. 0.30 8.38U 1.0 Fe+3(Fe+2Fe +3o.7Cr+3o . 3 )0l, O.R. 0.56 8.365 1.0 Fe+3 (Fe+2Fe f 30.!|4Cr 30 . 5 6 )\ M.R. 0.72 8.38U 0.73 Pe0.27Peo^73(Feo!73FeO?55CrO/ !)01 O.R. 0.82 O .83 Fe"*"2 Fe+3 (Fe"**2 Fe"*"3 Ct*^"3 ,)0, 8.365 0.17 0.831Feo .83 0.35 r o . M.R. 1.28 8 ,l(.06 0 Fe+2(Fe+3 u * 72Cr 31.28^°i^ O.R, 1.31 8.399 0 Fe+2(Fe+30 69Cr 31 . 3 1 )0ll M.R. 2.0 8.376 0 Fe+2(Cr+32 0)C\ O.R. 2.0 8.375 0 Fe+2(Cr+32^ o cQ = 13.584 A (151). Cr(OH)^ has the same structure as Cr?u^ but with different lattice parameters: aQ = 5*288 A and o c = 4.871 A (I52). Chromium oxide hydroxide, CrOOH, is is 0 o orthorhombic systems with lattice parameters: a = 4.861 A, o o 0 b = 4.292 A, a nd c = 2.960 A (153). 0 o electron diffraction is used to resolve the composition and structure of the passive film by means of electrons iron, or transmission through, a thin foil separated from the n.e L.d. surface, or passivated directly from a very thin foil. wc.00 and Kruger (154), in an electron diffraction study of a series of Fe-Cr alloys, conclude that the passive film of io,v Cr alloys are spinel-like, while the film on alloys with Cr-content 12 wt is amorphous in character. Thus, the5f imply that it is probably the difference between their stric tures which distinguishes stainless alloys from nonstain- less alloys. Hoar (155) has suggested that glassy oxides u.ay be more protective since ionic mobility is smaller in glasses than in crystalline structures which contain more defects. electron diffraction studies by other workers (132, 133, 156) show that the films give nondistinct crystalline diffraction patterns for stainless steels. Numerous ellipsometric studies (27, 157-162) have ascertained the thickness of the passive film on Cr, Fe-Cr alloys and stainless steels. They generally agree that the o thickness is in the range of 10-60 A, depending on the 66 potential, the composition, and the environment. The thick ness of the passive film itself does not seem to have great significance in the passivity of metals since a thicker film does not necessarily provide higher corrosion resistance. 67 2 .7 Processes of Film Formation The processes of the formation of the passive film involve the initial stages of anodic film formation on a b a re surface and the growth of the passive film. Usually, the formation of the passive film begins at a fast rate, but soon slows to the characteristic logarithmic process for those films in the thin-film range. i£ventually, the growth ul' tne film proceeds to some liming thickness. Details will be discussed in the following sections. 2 .7 .I Initial Fast Oxidation Two mechanisms have been proposed to describe the initial stages of film formation. The first mechanism, proposed by Muller (163)1 suggests that the film grows by a orecipitation process. Here it is assumed that the passive film which forms as a result of anodic processes is a precipitate. Upon application of a constant anodic current the metal starts to dissolve with n+ M --- > M + ne . (2.5) The dissolution is blocked by the precipitation of an insoluble film on the surface of the metal, Mn+ + Xn“ > MX. (2.6) This type of mechanism is characterized by an "induction time'; which is defined as the time required for the metal ion 66 to exceed the solubility product for the appropriate protective compound on the bare surface. Bockris et al. (16^-165) have interpreted their ellipsometric results to support this mechanism. The second mechanism is a nucleation and growth mechan ism. It postulates that the passive film results from two dimensional growth centers which are nucleated randomly on the metal surface. The species providing the growth centers absorb on the surface and diffuse to the growth centers by a surface diffusion process. Experimental and theoretical studies of this type of mechanism have been carried out by Fleischmann, Thirsk and co-worker (166-170). 2 .7.2 Kinetics of Film Growth The growth kinetics of thin passive films have been studied and models developed by Mott (171), Cabrera and Mott (35), Hoar and Mott (172), Fehlner and Mott (38), Haul i' Ilschner (173), Dewald (17*0» Pryor (56), Cohen and Sato (kS) Fromhold (175), and others. The growth laws in this area arc both direct and inverse logarithmic. However, it should bo pointed out that it is often possible to fit data to both direct and inverse logarithmic laws (176). The most commonly used model is set out in a series of papers by Mott (171) and Cabrera and Mott (35)* This model is based on a field assisted cation diffusion mechanism, according to which the potential drop across the film 69 remains constant while the field decreases as the film thickens. The rate controlling step is the field assisted movement of the cation across the metal-film interface. This model gives an ion current which.varies exponentially with the field; the film thickness, x, follows inverse- logarithmic kinetics l/x = A - Bln t (2.7) where A and B are constants depending on temperature and the potential drop across the film. A second model which rationalizes logarithmic kinetics has involved a place-exchange process. This modei was first proposed by Lanyon and Trapnell (1??) and developed by Bley and Wilkinson (178). The same mechanism is postulated for anodic oxidation by Sato and Cohen (45). According to their proposal, film growth advances by the field assisted place- exchange of metal-oxygen pairs. All such pairs in a given row normal to the surface are assumed to exchange places simultaneously. The ion current varies exponentially with the potential; the film thickens logarithmically with time; that is, x - A + Bln t (2.8) where A is a constant depending upon both temperature and 70 potential, and D is a temperature dependent but potential independent constant. Fehlner and Mott(38), by using a different model, have also obtained the logarithmic growth law. They assume that the structure of the film changes with thickness so that the potential drop across the film increases as the film grows in such a way that the field remains constant. With this assumption they obtained the same form of the equation as Equation (2.8). The parabolic and the cubic growth laws are usually not suitable for the formation of thin passive films. The parabolic law can be derived on the conditions that the diffusion of ions or migration of electrons through the onid . is the rate controlling step, and the rate is inversely proportional to oxide thickness. This law indicates a non- steady state diffusing controlled reaction (179)- Under specific conditions, some metals appear to react according to a cubic law or according to some other rate law with exponents different from 3 (e.g. , 2.5» etc. in xn = k £t + A ) (179)- Usually such behavior is restricted to short exposure periods. Cubic rate law has been explained as a combination of diffusion-limited oxide formation and oxygen dissolution into the metal. In other such casses as well, such irrational rate laws can probably be explained by the superposition of a morphological complication and ionic diffusion through the oxide. In general, then, such rate laws probably do not represent any new or significant mechanism. nevertheless, despite reasonably extensive experimental studies, detailed theoretical models have yet to be validated, furthermore, there is, as yet, no model here which can interpret the effects of alloying elements. 2.7*3 The Transition Period The detailed processes involved in the transition froi., a bare surface to a three-dimensional oxide film are still poorly understood. Fehlner and Mott (38) have put forward ' mechanism for the transition based on place-exchange model. According to their model, prior to the oxide formation, th<_ sticking coefficient for oxygen stays high as the surface atoms rearrange to accommodate chemisorption on neighboring cation sites. When the metal surface is completely saturated with oxygen, the sticking coefficient and the heat of adsorp tion have values characteristic of physical adsorption. Here film rearrangement is necessary for the formation of multi- layers of oxide. The place-exchange process is usee to describe this initial film growth. This process is described by Eley and Wilkinson (I78) to be one in which oxygen reacts with the underlying metal atom by a series of simultaneously activated exchanges between metal and oxygen atoms. Obviously, place-exchange will be more difficult after the formation of a monolayer. Oxygen ions will stay on the surface longer, and the movement of the metal ion to the surface requires activation energy. Some force on the mete] ion must lower this activation energy, so that place exolmn - can go on while the reaction rate decreases. 73 2.8 Solubility. Dissolution and Passivity Virtually all important engineering metals and alloys are thermodynamically unstable under service conditions. These materials are usable only because they form a more or less protective film on the surface which reduces the corro sion rate in many cases to effectively zero. The corrosion behavior of metals is controlled by the solubility and the dissolution rate of this protective film. The thickness of the film reaches, for a variety of metals, a limiting thick ness, As the outer surface of the film dissolves into the solution, it3 thickness is maintained by the electrice field. In cases where the formed film is not protective, this process many not be actually controlling but can still exert an ini'Iu-. ence on the corrosion behavior of the metal. This section discusses solubility, dissolution and passivity under the assumption that the dissolution behavior of the bulk oxide is similar to that of the corrosion product film. Thermodynamically, using the Gibbs free energy of the formation of oxides, one can plot a diagram describing the equilibrium conditions between the metal and its oxidation product in potential vs. pH coordinates. Such diagrams were plotted for all the Metal-Water systems by Pourbaix (1L0). Pourbaix has applied the "corrosion,11 "immunity," "passivation" designation to all the metal-water systems and has assumed that all solid corrosion products give passivity. Poteritial- pH diagrams are very useful in predicting the regions of pH and potential where corrosion can be expected. However, for example, iron can be passivated in IN H^SO^ (Figures 10-11). But according to Figures li|-l$, at pH< 1 no thermodynamically stable iron oxides can be formed. This example shows that passivity is difficult to explain as simply the formation of a stable oxide phase, without introducing any additional ideas. One of the additional ideas is considering the kinetic processes. Kinetically, the above difficulty is explained as follows: the existence of a film, regardless of its thermo dynamical instability, can take place if the rate of its formation is higher than the rate of its dissolution, or at least is equal to it. Therefore, the passivation of meta.Ls can still be found in some "corrosion" regions of Pourbaix diagrams. Obviously, the solubility and the dissolution ra.e of oxides or similar compounds are closely related to the phenomenon of passivity in metals. P.8.1 Solubility of Iron Oxides and Chromium Oxides The qualitative data on the solubility of iron oxides and chromium oxides can be found in"CRC Handbook of Chemistry and Physics" (l80): natural hematite (Fe20^) is soluble in HC1 and H2S0^ and slightly soluble in HNO^J natural magnetite (Ke30^) is soluble in concentrated acids and insoluble in alcohol and ether; Fe^O^-xH^O is soluble in acids and insolu ble in alcohol; natural goethite (FeOOH) is soluble in HC1; ^r2^3 ins°lub^e in acids, alkali and alcohol; C^O^'xHpO is soluble in acids and alkali and slightly soluble in HIT^OH FeCr^O^ is slightly soluble in acids; and all these oxides are insoluble in cold and hot water. It seems that Cr^O^ is the most stable one among the above oxides in the total pH range of aqueous solutions. The quantitative data on the solubility of some iron oxides and chromium oxides in water are given in Table 6. It can be seen that Cr(OH)^ has the lowest solubility in water and Fe(OH)^ has the second lowest solubility in water. 2.8.2 Formation and Dissolution of Anodic Oxide Films Generally, the thickness of a passive film at any time for an ideally flat surface is actually dependent on the difference in rates of film growth and film dissolution as in the following form d? = ^ f ~ 1D^ npF (2.9) where dx/dt is the film thickening rate, i^, the anodic current density for the formation of passive film (exclusive other reactions, e.g. __^.211 +2e ), i^ the dissolution current density of the passive film, M the molecular weight of the passive film, n the valency of the metal ions, the density of the passive film, and F the Faraday's constant. Lumsden and Staehle (182) have reported the film thick- ness-log t curves and the amount of dissolution for iron in borate-boric acid solution at pH 8.6. For example, these 76 'able 6. Solubility of Soine Iron Oxides and Ohroiui uiu Oxides in Water (181). Formula Dg%^y Temperature pH ^/JjfsoiuU O i l 2 0° C 5.25 Fe2°3 Fe„0, 5.18 50 7 0.926 3 8- F e (O H ) 3.25 18 7 l.k-2 25 9.2 0.732 Fe ( OH ) 3 . 1 2 18 7 Ij.. 8x10' 6 3 F e O O H 1+.28 25 7 0.037 F e O 5.75 -10 C r ( O H ) 25 0 . 3 8 x 1 0 5.21 25 7.9 0 . 0 6 Cr2°3 curves for a potential of 61^.0 raV^e are given in Figure 19. Figure 19 shows that the growth kinetics of the ellipsometric film thickness become logarithmic one to two seconds after polarization. The dissolution curve (LC-LE ) in Figure 19 indicated that a peak appears in the range of 0.1 sec which is followed by a decrease in dissolution. This shows that the total amount of dissolved oxide decreases until t=3 sec, and signifies that there is a precipitation which occurs after an initially rapid dissolution. Between t=3 and 100 sec, the film formation curve (L^,) and the dissolution curve (LC-LE) increase and the film thickness is close to the limiting value. 2.8.3 Mechanisms of Oxide Dissolution The study of oxide dissolution can be classified into two groups: (i) dissolution of anodic films and (ii) disso lution of ionic solids. These two classes have been identi fied as reductive dissolution and chemical dissolution respectively. The reactions for the case of oxides can be written as below: Reductive dissolution M30^+2e+i;H+— 3M+2+J+OH" (acid) (2.10) M^0^+2e+2H20__» 3M02“24^H+ (base) (2 .11) Chemical dissolution M0+H+__ M+2+0H" (acid) (2.12) M0+H20____M02"2+2H+ (base) (2.13) 78 -I 1 1 T I I I I | I------1-—I---TTTTTf “I I Ilf Film Thickness and Dissolution vs. Log Time for Fe Potential • 6 4 0 mVH Lc * Film Thickness Cole from Totol Chorqe , Le ■ Elhpsometnc Film Thickness Lc~Le « Amount of Dissolution i i i 1111___ i__i—1 i i i i 11 Log Time (sec) Figure 19. Fi1m thickness and dissolution vs. log time for iron in borate-boric acid solution at pH 8.6 (182). 79 +2 Thus the dissolution of oxides is favored to give M with _p increasing acidity and to give M02 with increasing basicity. The dissolution of iron oxide has been studied by Pryor and Evans (183-185)* Baram (186) and Azuma and Kametani (1H'7). Pryor and Evans have investigated bulk ferric oxide (X83) * ferric oxide films on iron (18L|.)» and bulk ferric oxide particles floating on a bed of mercury immersed in an acid electrolyte (185). The results of Pryor and Evans (183-185) show that ferric oxide dissolves more easily by a reductive process than by chemical dissolution. The schematic model for reductive dissolution is shown in Figure 20, The argument is that electrons are supplied by Fe p. Fe+2 + 2e (2.1!+) and adopted by Fe203 + 6H+ + 2e — ». 2Fe+2 + 3H20 (2.15) thereby the film is dissolving. The dissolution of ferric oxide in acid solutions in the absence of metals or other reducing agents will occur by a non-reductive mechanism. The apparent activation energy for dissolution is much greater (in the absence of complexing agents) for the chemical process than for the reductive process, Baram (106) studied the dissolution of reagent grade Fe^O^ in mixed sulfuric and hydrofluoric acids. His results indicate that ferric oxide dissolves more quickly in the mixed acids than in either acid alone. Baram explains tnis eo O.IN- HCI Electrolyte Fe_ ++ Fe_ ++ Thicker Oxide Film Typical Electron IRON Path Figure 20, Model for reductive dissolution of iron oxides (101;). 81 to the formation of an iron fluoro-sulfate complex. Azuma and Xametani (187) have investigated the dissolu tion kinetics of ferric oxide in hydrochloric, sulfuric, nitric, perchloric, hydrofluoric, phosphoric, and hjdrobromic acids from 0 to 100°C and 0,37 to 5*83 normality. Their data indicate that the rate of dissolution decreases in the order HF * HC1 > H2S0^_> HNO^> ITCIO^ (all li! at 95°C). The comparison was net extended to UPr and IJ^FO^. The dissolution of the passive films has beer; considered from several points of view. Gerischer and co-wurkers (188- 191)have studied the role of defect structure as it relates to electronic properties in the dissolution, processes; Vermilyea and co-workers (192-19)4-) have investigated the dissolution kinetics of Alo00 and Kg0 films in solutions containing various inhibitors; ITii (195)) has examined the effect of inhibition on the dissolution of NiO; Gilman, Johnston and Sears (196) and V/estwood and Rubin (197) nave considered effects of step poisons on the etchpitting of ionic crystals; Pryor and co-workers (98, 198) have suggested a defect-related model involving the replacement of 0 by monovalent ions. As a brief summary, the dissolution of passive film seems related to the combined efforts of defect interactions and step poison effects. Finally, on the effects of alloying elements, Abe and. StaehTe (199), using transiently strainer1 electrodes, demonstrated that both chromium and nickel additions to an iron base alloy yield a decrease in the amount of dissolution before the formation of a stable film. AUGER ELECTRON SPECTROSCOPY 3.0 Introduction Auger electron spectroscopy (AES) has recently developed into one of the most sensitive methods for chemical analysis of solid surfaces. The method, which with a particular small o depth resolution ( in the order of 10 A ), is essentially a surface probe as the data come from the top few atomic layers or so, depending on the material and ionizing energy. The spectra of AE3 can provide several kinds of information. Basically, AE3 is capable of uniquely identifying all elements above helium. It can also supply reliable quantit ative information and in many cases the character of chemicy L binding. While AES was originally used purely in research laboratories, it is now serving as a powerful tool for routine analyses. The applications of AES are widespread in numerous fields, such as thin film analysis, surface physics, chemistry, metallurgy, semiconductor technology, and. , dneru'i processing. It is suitable for all kinds of solids: metals, semiconductors, and insulators. These applications grow fanl as the fundamental mechanisms in AES are better understood* A few reviews on the technique and applications have apperaed lately (200-207). This chapter is primarily discussed 83 the applications of Auger Electron Spectroscopy in surface composition dertermination. 3*1 Historical note AES is based on the so-called 'Auger effect* , v.hich was first described by the French physicist P. Auger in 1925 (208, 209) while working with x-rays. Auger discovered the .iUger- electron tracks in a Wilson cloud chamber. He took the cloud chamber pictures of the photoelectric effect caused by a beam of monochromatic x-rays on atoms of neon, argon, and xenon. He observed that in most cases one to three electrons were released beside the photoelectron. These electrons result from an internal conversion involving the K,L, and ... levels, and have energies corresponding to the energy differences among these levels. They are considered to constitute a spectrum characteristic of elements. Lander (210) in 1953 Has discovered some small peaks in the secondary electron distribution curves, N(E), and suggest ed that this method can be used to identify surface impuriti es. The high sensitivity of this method was not realized until a further development was made by Harris (211) in 1968. Harris used a standard electrostatic velocity analyzer and, combined with electronic differentiation of the secondary electron spectrum, to study the Auger spectra of surface. Weber and Peria (212) in 196? adopted the readily available LEED optics into Auger electron spectrometers. How, this is very popular technique. Furthermore, the coaxial electron tat ic analyzer (also known as cylindrical mirror analyzer) was admitted into Auger electron spectrometers by Palmberg, horn and Tracy (213) in 19&9 to improve greatly the speed and sensitivity. The cylindrical mirror analyzer (CMA) has the advantage of a high transmission rate, along with the high versatility and wide applicability. It is more sensitive than any analyzer used before. Frequently, AES is used a I ■ .ng with sputtering of inert gas ion to obtain composition-vs- depth profiles (211;). Recently, AES employs finely focused electron beams with conventional deflection to do the two- dimensional compositional analysis of surfaces (215). The scanning system in conjunction with inert gas ion sputtering makes three-dimensional compositional visualizing possible. besides electron sources, Auger electrons can also be excited by other sources, such as x-rays (216) and ions (217,218). modern Auger units have capability of adding various accessories (219-220), such like Electron Jpectroscopy for Chemical Analysis (33CA), Secondary Ion Mass Spectroscopy (SIMS), Ultraviolet Photoelectron Spectroscopy (UPS), Low Energy Electron Diffraction (LEED), fracture attachment, sueci men heater, computer system, etc. 86 3.2 The Detected Volume and Auger Electron Mean Escape Depth The understanding of the detected volume is important for interpreting AES data. The detected volume is simply the volume analyzed under specific conditions, By altering the direction of the incident electron beam, the detected volume will be accordingly changed, and then the Auger signal v/il.1 also be affected. At first, the detected volume is determined mostly by the escape depth of the Auger electrons and therefore by the energy Y (Z) * It is almost independent of Ep. Second, when a few elements are present within the dui.ee- ted volume, the distribution of all the constituent is impor tant since atoms nearer to the surface yield more Auger electrons. In a quantitative approach, both the escape depth and depth distribution have to be learned, but none of them is given by the N(E) curve. A simple solution to this is to specify surface concentrations using a two dimensional termi- 2 nology, i.e., n atoms/cm with n proportional to peak height, and then to estimate the significance of n in relating to the detected volume. The disadvantage of this method is the possibility of being mislead if details of the detected volume are neglected. The escape depth is equal to the mean free path of inelastic scattering, \ , on the kinetic energy, E, of the ejected Auger electrons. The escape depth is generally 8? estimated empirically by noting the attenuation of a charac teristic substrate Auger signal as a coating layer of a particular material is made to increase in thickness. A compilation of escape depths from a variety of experiments is given in Figure 21 (220). The escape depth drops from a high value ate low energies to a minimum at about 70 eV. It also implies that the escape depth is not extremely dependent on the parent atom. This result is expected since the pro minent loss mechanisms involve valence band excitations and the valence electron density is not highly varying with the material. The insensitivity of the escape depth with material can much simplify quantitative analysis. The detected volume is less than the area excited by the incident electron beam* since this area is determined by the range of the primaries. The detected volume can be increased by using glancing bombardment since a bigger surface area can be excited. Generally, the optimum angle lies between 10° to 30° from the surface. The information of the depth dis tribution can be obtained by studying the dependence of the Auger current on incidence angle. 3*3 Quantitative Analysis Auger electron spectroscopy has increasingly important role as a surface study tool. In practical used of AES, an accurate determination of the composition is not often necessary. In failure analysis, as an example, a conclusion iue 1 Vraino tema sae et ih the depth with escape the of mean Variation 21.Figure iiCAPl 01HN i it n t i tM111 M it <3* ■: kinetic energy of the escaping Auger electron.Auger escaping theof energy kinetic • : eeec 2e :Reference A: 220e; Reference 220c; Reference Reference Reference 220a; 220g; ILCCTHMIMMV fcVl ILCCTHMIMMV +: ©: 0: Nt m Reference eeec 22 Oh,Reference Reference m 220d; 220f; 220b; I Ml /18 can be made by simply comparing the difference in the surface composition between good and bad parts. However, in m-jlj.c applications, it is important to have accurate quantitative determination. Reports on quantitative work have appeared in the literature (221-230)* In the following sections, three ;'proud :e s for the quantitative analysis are d i sens s< .-.i. 3,3.1 Absolute Calculations The Auger peak size is reasonably proportional to the surface concentration. Nonetheless, It is very difficult lu have a complete and accurate treatment at this time. The absolute calculation in ASS is of theoretical interest only. In the absolute quantitative analysis, it is essential to relate the Auger signal of an element to the surface concentration of that element. The Auger current 1^ produced by a WXY transition in a element is given by Palmberg (221) as exp (- H fa c o 8 0 ) d ftd E d H where Ip (E, H) is the excitation flux density, a (E) js I:,--, ionization cross-section of the core level V/ in the element by electrons with energy E, N(H) is the atomic concentration of the element at depth H, y is the WXY transition probabi lity factor, A is the mean free path for a WXY Auger electron, and exp (-H/Acos0) is the Auger electron probability for 9i) escape originating a distance H from the surface at an angle 8 with respect to surface normal. To simplify Eqn. (3.1)» it is assumed that the chemica3 composition is homogeneously distributed. It is convenient to separate the excitation flux density into two terms, Ip(E,H) = Ip + I b (E,0) (3.2) where Ig(E,0) is the excitation flux due to backseattered primary electrons. With these assumptions, the Auger current. i3 expressed as (221), IA - Ip TN Y o(E) X (1+Rb ) (3.3) wir re it is the backscattering factor and T is the trans mission of the analyzer. Obviously, for a complete calculation of Eqn. (3.3)» no have to know the ionization cross-section, the Auger yield, backscattering factor, and accurate Auger current. The problem becomes even more complicated if the surface is rough, and the composition is inhomogeneously distributed. With those difficulties, the quantitative analysis in this way is not possible for routine Auger analysis. Two methods which present more possibilities for quantitative analysis will be discussed in the following sections. 3.3*2 Quantitative Analysis with Standards A simple and convenient procedure may be applied in AEd to obtain quantitative analysis with an accuracy similar to that achieved in electron microprobe analysis (222), where 91 it 13 customary to compare the x-ray signal from the sample with that from pure element standards. Similarly, the Auger current from a sample may be compared with that from a pure element standard under the same experimental conditions. In the first approximation, the atomic concentration of element X in the sample is CX = GX, STD --- — X, STD = CX, STD ---- — — (3.1+) IH -1 X,STD where and 1^ are the Auger currents from the sample H H and standard respectively, and and 1^ slpD are the Auger peak-to-peak heights from the sample and standard respec tively. But this simple relationship can result in large errors when the electron transport properties are signifi cantly different. A more accurate result can be obtained by using Eqn. (3.3) to relate the concentration of element X in the test specimen (N^) to that in the standard (N^ * N . N ( X* ) ( *X.STD , NX X,STD < IX STD ‘“ XT' > * (1+R„ ) X .STD (1+R r ) (3.5) BX The advantage of Eqn, (3.5) t comparing to Eqn. (3*3)» 1^ tha:. ionization cross-section and Auger yield data are not needed, and also only the relative measurements of the Auger currents are required, If the composition of the test sample and the standard is similar, the escape depth and backscattering factor can be removed from Eqn. (3»5)> and this simplified equation is identical to Eqn. (3*4)* The backscattering factor can be determined by comparing Auger yield vs. Ep curves with theoretical ionization cross- section vs. Ep curve from gaseous targets where the back- scattering factor is negligible. The experimental results (231) show that the backscattering factor increases with the density of the specimen and the primary beam energy, referr ing to the binding energy of involved core electrons. The backscattering factor can be reduced by operating Ep near the ionization threshold, but this operation also reduces the sensitivity since there is a decrease in ionization cross- section. A more practical solution is to develop an empirical equation of Fig as a function of Ep, E^, and D (the density of the specimen). This could be sufficiently accurate for quantitative analysis. The influence of the backscattering factor on the Auger yield has been reported by Tarng and Wehner (232) as shown in Figure 22. They coated Mo uniformly on W while recording various Auger peaks from each element. As we can see from Figure 22, the 120 eV Mo Auger signal climbs to a value about 20/S higher than from pure Mo at about eight monolayers of Mo. So, the backscattering factor for the 120 eV Mo peak is 20% 9 1J l# • l Mo ON W ti si i IS woaoumnsof ftuoitotiiiD Figure 22. Mo and W Auger signals as a function of coated Mo on W substrate (Reference 232). greater for a W matrix than for a Mo matrix. The matrix effects on the ©3cape depth is not well stu died. Available Auger data seem to lie close to a 'universal escape depth curve (see Figure 21), but the data are not accurate enough to rule out the influence of the matrix on escape depth. Further studies on this direction are required for improving the quantitative capability of AES. 3.3.3 Quantitative analysis without Standards A less accurate but very convenient quantitative analysi can be achieved by using elemental sensitivity factors. The atomic concentration of element X is given in the following equation (221, 233)- i (3. 6) C* / I °XUX / I where is the relative sensitivity of element i, and d^ is the scale factor defined by di = Vm.i1?,! (3.7) where L. is the lock-in amplifier sensitivity, E . i.: t.n^- j. m j 1 modulation energy and Ip ^ is the primary beam current. Sensitivity factors can be found in Reference 233. Two advantages of this approach are the discardment of standards and insensitivity to surface roughness. The latter advantage is understood since all Auger signals are equally affected iy surface topography. 3 Chemical Effects in AES The Auger electron energy is correlated to atomic cor ■ levels. The individual levels are named by the x-ray nota tion K,L,M,N, etc. After the ionization of core levels under electron bombardment, the atom relaxs to the equilibrium state through the Auger process (208) in which an electron from a higher lever fills the vacancy and transfer the avail able energy to a third electron (the so-called Auger electron) If the energy is high enough, the Auger electron is ejected into the vacuum. The kinetic energy E of this Auger electron is approximately equal to ^ - Ex - Ey for a VXY Auger tran sition. The more accurate calculation of the kinetic energy E can be found in References 2 0 1 , 23i^ and 235. The Auger transition can occur in all elements except for hydrogen and helium which do not possess core electrons. Auger electron energy depends only on the energy levels of the parent atom, and does not depend on the energy of primary electrons which initiate the ionization. The most noticeable peaks of Auger transition have been plotted and are shown in Figure 23 (233)* Note that most elements can be uniquely identified, even when several coexist on a surface. Hence the qualitative aspects of AES are excellent. Generally, for purposes of identification, Auger peaks result from KLL transitions for elements from Z=3 to LMM transitions Figuro ATOMIC NUMBER 0 -t 1 * 23 200 1 - Mostprouncod Augerelectronenergies. . a - ^ U * i t t i - - f - - - t ■ If* * U 4 ^ - «..!a -- . j - * _ b:=t=i = t = : _b r_t T * each elem ent. All data w as reco id ed under similar c o n d itio n s using using s n itio d n o c similar under ed id reco as w data All ent. elem each h sme gr eto cr c ytm. System y p sco ectro p S lectron E uger A e sam the Dots in d icate the ele ctio n energies of the principal Auger peaks lor lor peaks Auger principal the of energies n ctio ele the icate d in Dots t - ■ V —t--- — * + ■ ■ t-- -V- — -t— * ■ ------ELECTRON ENERGY (eV) ENERGY ELECTRON ' — —i —i- ;— r i— J— i LMM: ------ 2000 —J 4 . . . .. —W 2200 2400 Cu Nh Flit Hn Tm Au Mu n* Th for elements from Z=ll|. to Z-Ij.0, MNN transitions for elements from Z=1|.0 to Z=79, and N00 transitions for heavier elements. For convenience, the Auger peak energy is commonly located at the minimum point in the dN(E)/dE peak. Auger electron spectroscopy can be used not only to ascertain the elemental composition of the solid surface, but also to measure changes in binding energy of inner shell and valence electrons which are strongly affected by a chemical environment. Therefore, this technique is potentially suitable for identifying surface compounds. A change in the environment of an atom in the surface region may change the Auger spectra in the following way3 (236-237): (a) Shifts in energy due to inner shell energy level shifts caused by changes of the valence elements; (b) Splitting and changes in spectrum shape generated by a redistribution of valence electrons; (c) Changes in the shape of low energy peaks resulting from variation in energy loss mechanisms; (d) Mo detectable changes. Chemical shifts in Auger spectra are not as easily analyzed in AES because three levels are involved. For a WXY transition, the measured chemical shift will bo AE - Ey ~ ~ -Ey ■* ( Ey + Ay - - A x ■ iiy ~ Ay ) = - A w + A x + Ay (3*8) where A^* A^» and Ay are "the respective shifts for the core levels W, X, and Y. Since all three A* s are not equal, tm. 98 measured AE is not easily interpreted. The situation becomes more complicated when the valence band is involved in the Auger transition. The redistribution of electrons within the valence band can cause not only Auger peak shifts but also the peak 3hape3. An interesting study (236) of characterizing surface chemical states has been carried out on carbon in various forms. The KLL carbon Auger spectra of diamond, graphite, silicon carbide, and a metal carbide M02C are displayed in Figure 2lj.. The spectrum change in this case is sufficiently distinctive to serve as a means of identifying the form of carbon at the surface in ideal cases. The Auger spectrum from carbon is particularly sensitive to bond hybridization. Hybridization of L shell electrons and, accordingly, changes in the transition probability function is proposed as the reason for the different peak shapes. With carbon, the differences in peak shapes (as shown in Figure 2 ) have been readily identifiable with the retarding field analyzer (RFA). This kind of information for some other elements can not be offered by RFA. However, AES with a cylindrical mirror analyzer (CMA) can disclose some fine structures associated with major Auger peaks for identifying the chemical state of surface atoms. Figure 25 illustrates Auger spectra of metal and oxidized metal surfaces (238). On a chromium surface, for example, two peaks appear at 3 2 . 0 e V a n d 14-5 -ij. eV for chromium oxide peak position whereas the pure 11- Mo„C SiC GRAPHITE DIAMOND 200 250 300 E (eV) Figure 21+. Spectra of* the carbon KLL Auger transition for diamond, graphite, silicon carbide, and I4o„C (Reference 23 ). METAL METAL METAL /OXIDE OXIDE OXIDE 30 40 50 60 40 50 ELECTRON ENERGY(eV) ELECTRON ENERGY (e-V) ELECTRON ENERGY(eV) a) CHROMIUM b) MANGANESE C) IRON Figure 25. Auger spectra of metal and oxidized metal surface (Reference 238). oo l oo lh! chromium peak is at 37*2 eV. Seo. et al. (239) have studied the chemical shifts in the Auger spectra of passivated iron. Changes in the low energy iron peak, which results from the ^VV transition reveal the chemical states of the passive film of iron. Two peaks are associated with this transition after 5 ft of film is removed as illustrated in Figure 26. These spectra are obtain ed by sequentially sputtering using argon ion bombardment followed by Auger analysis. The position changes of these Auger peaks (as shown in Figure 26) seem to imply that the passive film consists of two phases and the phases exist as two overlapping layers. Since Auger transitions often involve valence electrons, the Auger peaks of adsorbed species can be expected to display different bonding conditions. Yet, interpretation of Auger spectra in terms of chemical bonding is not always straight forward. The changes in the Auger spectrum is not always related to the valence band electron density distribution. Auger electrons leaving the surface may lose discrete amounts of energy through plasmon excitations, core level ionization, and interband excitations and appear as distinguishable peaks. Mularie and Peria (2l|.0) have demonstrated that some of the fine structure below the LVV Si Auger transition energy at 91 eV coincides energy loss peaks arising from plasmon excita tions. The 91 eV Si peak and the fine structure were compared 10. r- I rgn ru jfiHil CjO Mio tO fTkVpi |l.nuk MJ'v ■ ' t e » ttVI Figure 2b, Auger spectra during the gradual removal of the passive film on iron (Reference 239). 103 to plasmon loss peaks generated by elastically scattered electrons. 3.5 Composition-vs-depth Profiles Since the Auger signal originates from a layer only within a few Sngstrom of the exposed surface, the changes in the Auger signal as material is removed from the surface indicate the variations in the specimen's composition. The adding of sputter-etching gun on AES (2l|_l) has made possible the acquisition of data relating to specimen composition not only at the surface but as a function of depth into the specimen. This technique entailed the sequential removal of species of the specimen surface by AES analysis of the newly exposed layers. The ion bombardment gun provides a controllable high current density of inert ion beam to be projected on a speci men's surface for sputtering out the species from the surface. Figure 27 illustrates Auger spectra from a 150 8 nichrome film on a silicon substrate (203)* The spectra are represen tative of the film composition at the surface, 100 2 into the film, and when 200 2 of material have been removed. The com plete profile is plotted in Figure 28. This plot indicates clearly the 150 2 Cr nichrome film, the silicon substrate, and the oxygen at the surface and interface. Quick and accurate depth profile of surface structure is now possible with a powerful ion bombarment gun. Sputter- 101+ NidiroMii1 Mm *JS Ueloro ■V"' SlJUllCMfUJ UJ fOO A X) Ni (\j, ■O -V • -1" W Af c Cf 58 ;Cf - i— t - -1 t— <-----*--■ i— t-— ( 0 200 400 GOO 800 1000 Elccirun cnurgy (eV) Figure 27 . Representative Auger spectra from various depth of 150 angstrom nichrome film on silicon substrate (Reference 203). 10‘> Nichroino Iihn iii evaporated C d529eV ) * 1 \.V . . * • . ./ , NtlB'lOi'V) \ \ S.I92cVI y - — 0151 leVI *ij 5 Cl272c VI ( '(!-■ -*■< ?00 250 Material removed lin angstromt! Figure 2 0. Various Auger peak-to-peak heights from nichrome film on silicon substrate as function of film thickness (Reference 203), * ] U6 ing rates of greater than 200 2/min of tantalum oxide, for example, are easily achieved with this kind of gun. Removal rates with this gun can be controlled over orders of magnitude by varying sputtering potential from 300 eV to 3 keV, by varying the beam current density from 0 to over 200 pA/cm , and by scanning the beam. Scanning capability is an important feature for this kind of gun. When used in depth profile studies, scanning provides uniform removal rates that produce wide and flat-bottomed craters to remove the undesir able characteristics of crater walls from the point of the arialysi s . 3.6 Conclusion To detect the Auger electron excited by an electron beam provides a sensitive method to investigate surface atoms. AES is a nondestructive technique for surface chemical analy sis which has the attributes both of qualitative and quanti tative analysis. AES is well suited to the detection of light elements, except hydrogen and helium, and to the study of extremely thin layers on the solid surface. Therefore, the understanding of the passivity could be achieved by the careful use of AES. AES, when combined with noble gas 3putter ion bombard ment, can furnish valuable concentration depth profiles for surface-related phenomenon study. The Auger spectra may provide information regarding the chemical state of the atoms observed. But the ultimate capability of this approach is still not known at present. Theoretically, a fairly complete picture of bonding of atoms in surface regions could Jse developed, but much work remains to be done. Table 7 briefly summarizes and compares AES with some other methods for surface analysis (2l|_2). Nonetheless, none of these methods applied so far has enabled us to determine the chemical composition of the surface layer precisely. Table 7. Summary and comparison of techniques for surface chemistry study (Reference 2i|2) Sensitivity Elements AU'iinc Ljscts Approximate Major Technique Ever.ation , Emission Detectable Ideal Conditons Cost Aduni.ipe Radiography “ G u y s Lou eticigy £J-ijys Of or ^ Vets limited Lose Cost emitter Chemical -- All *=5 loss Cost Etvhinp dements Mu re-probe Electrons X-rays Ckwjd for =5 570.000 Versatility »nd Lnety’y 25 keV elements Dtspetsiun Z > 11 Can detect Z - 5 [on loni Ions Z > 2 —.001 550.000 Anah sis of Scattering 0.5 to 3.0 first keV Atomic layer Election X-rays Elections All = 01 550,000 Obtain Spectroscopy elements Chemical for Cf.imic jl State An j K ms Infcttr.ation Ion ions Ions Best fat = 01 '50.000? L s*c s\ ith Backscattenng 0.5 to 3000 Heavy atmosphere Spe-tromeiiy keV Elements contaminated Element Z < 10 samples probably not de’tec table Aucet Flections L lections Good tt’I =.(1111 550.000 Ana’s its of Spec: vv c p y 2-5 keV I idt! out;-: layers Elements Are!;- 1C' all Can detect f l-m.ents all dements simul ssul, Z > 2 taneously ROT EXPERIMENTAL PROCEDURES h. 0 Introduction The primary objective of this investigation is to study the characteristics of the passivity of alloys, and to develop information for achieving an understanding of the nature of this important phenomenon. Specifically, the study concerns the detailed influence of alloying elements to account for the achievement of passivity. In order to conduct this study, it is necessary to keep the environmental and continuum variables constant throughout the investigation. Thus all the experiments in this study are conducted on the specimens for which identical surface conditions are maintain ed and the environmental conditions are the same. Both electrochemical techniques and Auger analysis are used to characterisize the protective films on iron-chromium binary alloys. The electrochemical techniques are used to determine the potential-current density curve and current density-time curve at a selected potential. The Auger analysis will be alternated with sputtering so that the com position profile can be obtained. The Auger results will also be analyzed for chemical shifts in low energy range and com pare to the spectra of some chromium oxides and iron-chromium ox i d e • 109 110 This chapter describes the details of the experiments conducted in this study and also contains the experimental observations for investigating the phenomena of passivity. The next chapter discusses the results of the experiments des cribed in this chapter. This chapter contains three basic categories* (a) Materials (b) Equipment (c) General procedures followed in all of the experiments and specific procedures followed for specific experiments• ^.1 Electrochemical Measurements All electrochemical measurements were carried out under constant potential conditions. The potential is the independ ent variable, and the current is the dependent variable. A working electrode, a reference electrode, and an inert counter electrode were required to operate the cell under potentiosta- tic conditions. The electronic instruments used to measure the electrochemical data and control the experiments were* 1. Wenking potentiostat Model ?191 66TS10 2. Wenking Step Motor Potentiostat Model SMP 66 3. Keithley Electrometer Model 610B k. Esterline-Angus Stripchart Recorder Model E1101S. The full experimental setup is given schematically in Figure 29 and also in two photographs, Figures 30 and 31. To Solution Reservoires ♦ MRS f Rec. Rec.: Recorder P: Potentiostat Kater-Luggin Probe MRS! Multi Range SMP Step motor Counter Electrode Selector Switch Pctertiostat E: Electrometer Re: Reference Electrode SB: Jalt Bridge figure 29. Schematic diagram c; Figure 30. Close up view of actual apparatus including instruments in electrochemical measurements. I , I ! » I! ) ' Figure 31- Electrochemical cell apparatus used in electrochemical measurement. 11U ^.2 Electrolyte Solutions Solutions used were IN HgSO^ and IN NagSO^tpH 3, 6.3 and 10). They were prepared using double distilled water and analyzed reagent-grade NagSO^ and HgSO^. The pH value was adjusted by either NaOH or HgSO^. The solution was kept in a storage vessel and de-aerated by prolonged bubbling with perpurified argon, prior to the introduction of the solution into the cell. The argon leaves the storage vessel through a gas washing bottle which maintains a slight excess pressure within the vessel. Twelve hours of bubbling with argon were found to be sufficient to remove dissolved oxygen from the 2000-ml of electrolyte solution in the slorage vessel. o Measurements were made at 2^*1 C. 4.3 Electrochemical Test Cell All electrochemical tests were conducted using the two compartment pyrex glass cell shown in Figure 32. The 500 and 150 milliliter capacity compartments enclosing the working electrode and the counter electrode respectively, were separated by a finest grade fritted glass disc. This provided a good separation of anode and cathode reaction products and allowed control of the gaseous environment in each compart ment. Accommodations for a solution inlet and a thermometer were furnished in the working electrode compartment. Deaera tion was executed by passing perpurified argon into each compartment of the cell through coarse fritted glass disper- PLtUNUN ILECIROBt 116 sion tubes. A slight positive pressure was maintained in the cell by emitting the gas through water traps which were a part of the electrode assemblies. All accessories were made of Teflon or pyrex glass and no lubricant was used. A Haber- Luggin probe, which entered the cell at a degree angle, was constructed from a short length of pyrex glass tubing, drawn at one end to a fine capillary tip of about 1 millimeter in diameter and closed at the other end with a finest grade fritted glass disc. An inverted L-shaped piece of glass tub ing containing a stopcock at one end was coupled to the probe to complete the solution bridge. The probe, as well as the working electrode assembly, could be raised or lowered at will through Teflon seals so that the capillary tip could always be brought up to the surface of the specimen. Upon completion of the cleanness treatment the probe was filled with solution siphoned from the working electrode compartment of the cell. The other end of the solution bridge was attached to a bottle containing the same solution. The cell was interconnected by a solution bridge to a second bottle containing still the same solution and, then, was interconnected by an agar bridge to a third bottle containing a saturated potassium chloride solution. Two bottles of electrolytic solution and the fine fritted glass disc were used at the end of the primary solution bridge to prevent the flow of the saturated potassium chloride solution into the cell. A saturated calomel electrode was placed in the third bottle. The counter electrode was made of 0,005 inch platinum foil with a surface area of 2.50 square centimeters. When not in experimental use, the counter electrode was stored in a flask to prevent contaimination from atmospheric dust and unnecessary handling. Auger Unit The principles of AES have been briefly reviewed in the previous chapter. AES is accomplished by irradiating the surface of a solid with a primary electron beam where energy analysis is performed on the resultant secondary electrons. An energy enalysis of these secondaries enables one to ident ify the elements from which they originated. The spectrum obtained experimentally is the derivative of the electron distribution* i.e. dN(E)/dE, as a function of electron energy E. The Auger analysis was carried on in a stainless steel, ultra-high vacuum system using sorption* titanium sublimation, and ion pumps. These pumps gave a contaimination-free vacuum. The complete AES unit, which is a PHI Model 5^5 Scanning Auger Microprobe with Varian Vacuum Systems, is show in Figure 33. The sorption pumps in this unit are Varian VacSorb rough ing pumps which are designed to evacuate a leak-free vacuum system to a pressure of a few millitorr ( microns Hg ). Pumping operations are accomplished only through the process Figure 33. PHI Model S h S SAM. ri9 of physical adsorption. At liquid nitrogen temperature, gas is trapped and kept in the pump* When the pump returns to room temperature or is heated to higher temperatures, it is reactivated, i.e., the stored gas is released. The VacSorb pump differs from a mechanical roughing pump in that it requires only a finite amount of gas on each pump down cycle. A VacIon-pumped high vacuum system requires only a roughing pump which attains a vacuum in the low millitorr range, rather than a continuous operation backing pump (2^.3). The titanium sublimation pumps are Varian Titanium Filaments Model 916-002^ and are controlled by a Varian Control Unit Moder 922-0032. Titanium sublimates from a special filament when it is resistance-heated. The flesh titanium is deposited on surrounding surface and forms stable, solid compounds with the chemically active gas atoms which strike the surface. The gas atoms usually strike the film in a molecular form. The capture of gas atoms accounts for the pumping speed. The detailed theoretical basis is discussed in the Varian Vacuum Instructions (2i4i). The ion pumps in this system are Varian Noble Vaclon pumps which are developed to pump noble gases with a better stability and a greater speed than those attained with the conventional diode-type ion pumps. Each Noble-Vaclon pump element contains the cylindrical anode cells ( which are operated at a ground potential ) and a cathode grid { which consists of titanium strips operating at 5 kilovolts )• The IL'O pump functions within a field. Ions, produced by collisions of electrons with gas molecules in the anode cells, are accelerated towards the titanium grid. They strike this grid at a glancing angle, sputtering titanium onto the wall of the pump cell behide the grid. The pump wall serves as a large area for net titanium build-up, the condition necessary for secure argon burial. A noble pump element is illustrated in Figure 3U* General information concerning the features of this type of pump is also given in the Varian Vacuum Instruc tions (2l±$) . Auger spectra were obtained using a PHI 5^5 Scanning Auger Microprobe. The major optics part of the SAM is the PHI 15-110 optics which consists of a CMA with an internal scanning electron gun; the electron optical axis of the CMA and scanning electron gun are normal to the axis of the specimen manipulator. Mechanical X-Y-Z and rotation adjust ments on the specimen manipulator ( refer to Figure 35 ) allow the necessary motions to locate a specimen at the focus of the CMA for performing Auger analysis. The control cabinets of this SAM unit provide the necess ary power to operate (a) the components of the internal electron gun which consists of a tungsten triode source, two electrostatic lenses and electrostatic defletion plates; (b) the CMA and processing the Auger electron signal; (c) the sputter ion guns; (d) the ion pumps and the sublimation pumps. Two scanning displays are included in this unit for monitoring r ■ » / 7 \ * AAA * ANODE OBUOUE cIMPACT Auif s MAXIMUM Sl'UtU PING ARU.H PUMP WALL Tl SPUTTER SPU11f Nt D CATNOOt IlIANiuM AHOON ATOMS UuHit U HLPt Figure 3l|_. Noble Vaclon pump element (Reference 2^5). Figure 35. PHI Model 10-503 Specimen Manipulator and Holder 12.j the absorbed current and Auger signals. The X-Y recorder and oscilloscope display are needed for recording and displaying Auger spectra. A TC gauge and oven control unit and a digital ion gauge control are also included in this SAM unit. ^•5 Procedure if.5-1 Specimen Preparation Iron-chromium alloys were chosen as experimental specimens because they are very important materials in the commercial sense and they can develop a very stable passive state. The alloys were prepared by the International nickel Company. They were melted in a vacuum induction furnace, poured into 3-a'r graphite ingot molds and heated and extruded into cylinders. The compositions of the binary alloys, given by International Nickel, are listed in Table 8. Pure iron ai^; pure chromium were also used since they are the major elements for constituting the stainless steels. The iron specimens were cut from a piece of Armco ingot iron with the composition listed in Table 8, The very pure chromium was fabricated by Materials Research Corporation with the composition also shown in Table 8. All materials were shaped to a rod of f " diameter and about 10/16" long. Before bakelite mounting was applied, all the specimens were annealed at 8 2 5 ° C for four days and water quenched. The microstructure is ferritic with uniform grain size. A stainless steel wire was soldered on the back of Table 8. Chemical composition of materials used in this study. (in wt %) Elements Fe Cr Ni Si Cu Mn Mo Na s c P N 0 H Alloy Designation. F e (aJ bal .ooi+ .009 .003 . 02? .01+7 .001 .025 .035 .005 .002 Fe-5Cr 5 - - - - — ——— — — — — F e - l 5c i bi a l 15 — — - - - —— — — - — — Fe-255cibia l ? 5 .5 — — — — — — - - — — — — C r ^ .0 3 ' bal .001 .ois .001 — .001 .003 — .006 — .001 .02 .005 Chemical analysis supplied by: (a) Research and Technology Division of Armcc Steel, Middletown, Ohio; (b) The Internation Nickel Co., Suffern, N.Y.; (c) Materials Research Corp., Orangeburg, N.Y. 125' each specimen. Unwanted area exposed portions of the connec- (t tors were masked with glytal. A thread of 1/16 in diameter was rrade on the back of each specimen. The thread fit the Auger specimen holder. Afterwards, the specimens were ground to a 600-grit finish through a series of metallographic papers and then mechanically polished with 6 ftra and 1 pin diamond abrasives using kerosene as the lubricant. The specimen were cleaned and degreased with detergent. The final polishing was done with 0,05 Mm gamma alumina and washed with double distilled water and finally washed with spectrographic grade methanol, dried quickly in argon gas, and placed into the electro chemical cell immediately. *+.5*2 Experimental Procedure After mechanical polishing, the specimen was placed in the cell and deaerated solution was permitted to enter. Before each experimental run, the specimen was cathodically polarized at -900 mV (see) for five minutes, and then the potential was switched to the initial potential for that experiment immediately. First, the polarization curves were run to determine the passive region, if any. After these curves were accomplished, the specimen was repolished and the specimen was put into the cell again. The specimen was also cathodically polarized as before. The potential of the specimen was instantaneously changed to a potential in the 120 passive range, which was determined by polarization techniques; and current-time relationship was monitored. After the specimen was removed from the electrochemical cell, it was rinsed in double distilled water, dried in argon gas, and transferred to the Auger unit. The Auger analysis was alternated with sputtering so that the composition profile was obtained. The Auger results were also analyzed for chemical shifts. This provided information on the valence of the species as distributed through the film as well as the band structure of the film. 4.5.3 APPH and NAPH Auger peak heights involve quantitative information on the number of atoms of a certain type in the surface region. There appears to be a linear relationship between the surface coverage and the number of Auger electrons in the surface area of many elements (229-230). Some of these experiments have been summarized by Sickfaus (229) in Table 9* In the table the adsorbates and substrates are given along with the Auger electron information that is the monitor of coverage* e.g., dN/dE, peak-to-peak heights. Therefore, Auger peak-to-peak height, defined as the length between the top and bottom of a peak being commonly designated as APPH, is used as a quantity for Auger analysis. However, it is difficult to compare the results of the different investigations on an absolute basis since in many cases, under experimental conditions it is 127 Table 9, Some quantitative Auger electron spectrometry experiments ( Reference 229 ) i\< In 11 l>. 11 c S,lll>tl .it r MlllHllrl C.llllll .1 lldll I iivci,, i;i- ) ,i i-■ iii-ii- r . M .. 11 liij J jV /. Ill11 i in 1 111 - .11 Cx) Tv i!N/,!!■: inn cut 11 ,11 ' 1 JX III 1 i m * 1m .HIM It ll W ( lllll) ds';,u: irjll c II) ri-ill 1111' 7 V In" i in ’ 11-.- n I >) K ,ix/ \ . i U (1011) i/A’/i/y-: ii ml cm rent lu" In ' i in 1 I,,n ii I'll 1' Sia ,1 i f KiDJ itN/.tf: I’.ilinl.it'il- 11 , Si, il t f Hut) , ix/tti-: l f r >->',--.1 . III m 0 ,. i lit, il.i(i 11 S >nj il i tiuu) ,/r v/.//; i ni ,t ii i L ii in <1«', C.lll lll.l Ir l( ■ U Sil ,1 c< 100) r/.v/J/-; cri n-i-M -c l inn 17..1", C,ill III.1 K ll- i: Sij, ,1\-{IIMI) .ix/tti: cfii-,s-M , In m 1 O' ,. n \ \ ( ! 10) f N cxjii iiiiri- 0 < I ■\« N i ( IlHI) Knm lii ii cell 0 v 1 (, A 1 ii i.-.ir Jinx l< ~. .< (, A in ■ ■ 1,111 II K Ci-ii i n r(,v/«(/-: inn cm i cm I) 7 *VU- II H i -1 ii-.n Mr, u ./A’, nin cm r.-ni 1 0M.ii.-r 11.. , 1 r. ii -.i : u Mr. AN /dE inn i ill n-iil 1 In l i> cl i .,.],in ., i It.' An A'(/■.), JN/dE (J< 1 ) 1 17 l.i>. r. in ,i Mi in-.,t A*: It,- (~) d Ay,si-: d< < > 1 l i l . i y n x n..i. i,hi .i i t i, (-) ,/A', .//■: ’() 1 0 liy.it n- -ii!iHe ii it,- <11 (-) dX/ o Si(1 II) dN/dE ,Hi| isinntct f y II. 1 <0 < 1 5 l.iy.-ix Ii.., t 11 Sil 11)11) ,/A7'd-: cllipsn mu 11 y 0.1 < 0 < 1 . S l.iycrx 11 in'. ,1 1* sid n ) d.V/dE y e n SXlll" S/, in" cm ‘ inn.line.ir r, \ l\ S, ( I, l\ (1 <’•).-( MID) dN/dE rllijixiomi'lry ■'-■llJ" cm 1 1 inc.,)' (in iniiliTdli-s) 128 impossible to keep the same situations and therefore the APPH's are often presented in arbitrary units. The normalized Auger peak height, defined as the ratio of APPH's between the measured element ( e.g. carbon ) and base metals ( i.e. iron and chromium ), is designated as NAPH. The NAPH was used rather than the APPH in this study to eliminate sensitivity changes resulting from changes in operating conditions during the measurement. For comparison with the Auger spectrograms reported in this study, Table 10 lists the Auger peak positions for various elements. Those underlined electron energy values are the major peaks (233). Generally speaking, the sensitivity of AES is approximately 0.1 at.?£ and may be varied with the material elements. ^•5*^ Auger Spectrum Auger spectra were obtained using a PHI Model 5^5 Scann ing Auger Microprobe. A 2 kV and 2 mA electron beam was used having a diameter less than 5 The high current density of the electron beam, coupled with the scanning capacity, enable the system to generate the absorbed current micro graphs as well as surface composition data in several forms. The controls on the CMA control unit were set for observ ing the Auger peaks with the following conditionsi X-axis scale 10 eV/div. and 100 eV/div. (6 steps* 0, 10, 20, 50, 100, 200 eV/div.) 1 2 8 impossible to keep the same situations and therefore the APPH’s are often presented in arbitrary units. The normalized Auger peak height, defined as the ratio of APPH's between the measured element ( e.g. carbon ) and base metals ( i.e. iron and chromium ), is designated as NAPH. The NAPH was used rather than the APPH in this study to eliminate sensitivity changes resulting from changes in operating conditions during the measurement. For comparison with the Auger spectrograms reported in this study, Table 10 lists the Auger peak positions for various elements. Those underlined electron energy values are the major peaks (233). Generally speaking, the sensitivity of AES is approximately 0.1 a.t.% and may be varied with the material elements. 4.5*4 Auger Spectrum Auger spectra were obtained using a PHI Model 545 Scann ing Auger Microprobe. A 2 kV and 2 mA electron beam was used having a diameter less than 5 |im. The high current density of the electron beam, coupled with the scanning capacity, enable the system to generate the absorbed current micro graphs as well as surface composition data in several fo rm s. The controls on the CMA control unit were set for observ ing the Auger peaks with the following conditions* X-axis scale 10 eV/div. and 100 eV/div. (6 steps* 0, 10, 20, 50. 100, 200 eV/div.) Table 10, Electron Energies of Auger Peaks for various Elements ( Reference 233 ) Element Electron Energies (eV) C 272 S 11*0, 1£0 Fe )±Z, 86, 550, 562, 591, 598, 610, 6£L, 221* H i Cr 26, 1*19, W 7 , 1*59, 1*82, lj-89, 52% 571 0 1*75, W , 210 Electron Multiplier 1 kV (0-3 kV, as required) Modulation Voltsp_p 4 eV (10 steps* 0, 1, 2, 3, 4, 6, 8, 10, 15, 20 eV) Generally, the low x-axis scale gives an expanded energy scale and a higher resolution. This is very important for obtaining information on the chemical shifts. *1.5.5 Preliminary Alignment The performance of the CMA can be optimized by adjust ing the axial sample position and beam power supply focus and deflection controls. In the PHI Model 5^5 SAM system, the specimen position is very easily determined by calibrating the energy position of an electron peak of known energy e.g., an elastic peak. A focused electron beam of energy e'/g directed along the symmetry axis of the CMA is used to bombard a specimen mounted perpendicular to the CMA axis to maximize the peak-to-peak height of the elastic peak. The position in energy where the peak crossed the energy axis is equal to eV^ which is known exactly. At the completion of the calibration procedure the specimen is placed at the source point of the analyzer and the pass energy of the CMA is calibrated. The details are discussed in the PHI Model 5^5 Instruction Manual. The importance of specimen alignment is demonstrated in Figure hZ (214.6 ). The spectra are sulfur 150 eV peak in an energy expanded scale at various positions. It indicates that the intensity of the Auger peak is a function of the distance Segregated Sulfur on Nickel C 1 1 1 ) F a c e Un-Focused Cclose to gunl Focused VW CVOLTAGE WIDTHJ 200 eV SCCSPECTRUM CEMTFR] 100 eV Un-Focused Cfar from gun] Figure 36. Effect of specimen alignment ( Reference 2E6 ), between the specimen and the electron gun, and the detailed spectrum sturcture changes as the position varies. lienee, poor alignment may affect both qualitative and quantitative analysis. ij..$.6 Depth-concentration Profile After the composition on the surface had been recorded under ultra-high vacuum conditions (better than 10” ^ torr, 1 torr=l mm Hg), the system was backfilled with a high-purity argon for sputtering to study the width of the passive film. Active gases were removed, from the argon by titanium sublima tion pumping. The composition profile of the film was deter mined by repeated sputtering with argon ion bombardment at pressure of if. x 10 ^ torr followed by Auger analysis. The energy of the ion was 600 V. The sample was positioned appro ximately 6 millimeters in front of the analyzer. It is necessary, of course, to undertake a procedure for optimizing the focus and deflections of the ion bombardment gun for obtaining an effective and consistent sputtering rate. The focus and deflection are optimized by observing the peak on the oscilloscope and adjusting the focus and deflection controls to maximize the peak height. The details are also discussed in the PHI Model Sk-S Instruction Manual. EXPERIMENTAL RESULTS 5.1 Polarization Behavior of Allovs The characterization of the polarization behavior of a metal can be obtained by using the sample as a working electrode and automatically varying the voltage while record ing the change in current behavior. During anodic polarization, all experimental alloys showed active-passive behavior. The passive region is readily visible since the current is very low over a region of potent ial. The current density-potential curves of iron-chromium alloys exposed to the IN and IN Na2SO^ (pH 3, 6 .5 , and 1 0 ) are shown in Figures 37-UO. The curves were determined by step scanning rate at 15 mV/min. at 2^-l°C. 5•2 Current Decay Curves The potentiostatic and the potentiodynamic curves can practically indicate the actual behavior of passive metals. Those measurements can also lead to the studies of the react ion mechanism, the kinetics of corrosion phenomena, and metal deposition. Despite their broad applicability and extensive use, considerable uncertainty and insufficiency in the polarization curves still exist. One thing, that the polariz ation curve does not provide, is the dependence of passive current density with the time; and the passive current density 10 -2 e o a E aJ * !» -p •H -3 ra 10 C o T3 -P C O O 104 - 5 -TOO - 4 0 0 - 1 0 0 200 500 800 1100 1400 1700 2000 cure 37 Ihe current densi tv-uc tential curve? cf Fe-Cr alleys ir. 17 }1„Z0 1, J t-J* ■n & Current density, amp/cm tv CD o Oi O, 0 u> r-| - n *•< 0 1 ■ 'T! I - O M 'S o i-1 4 • • IU ._, {5 3 *T1 K OJ lit I H- £ rp vO Hi ur Current density, amp/cm' CD Oi Q w t; CD cH15 »i O & O CD n c/> wI o i \ ‘rl O r+ "is CD \ 15 rt f'-*• Ul CV' o t-* o O i; 1 rv3 '5 o* ll U o \ «J fit o * S '.1 ff ^ ! ■o o “Tj v ! fit CDI M 4 - n o S*. o p hi-j^ f> ■'j rn H* \ l\>p C/J CD 4 ’t1 n: M' ' H 9CI 2 s’ Current density, amp/cm i;*8 is extremely dependent on the time. The measurements of the current density-time curve can meet this purpose. Also, this study provides information on the performance of metals and alloys and the kinetics of the film growth in the long run. The current density-time curves, recorded by strip chart recorder were plotted on a log-log scale, and are shown in Figures i4.l-l4.J4.. These figures give the current decay curves for three test alloys in each electrolytic solution at a selected potential in the passive range. Generally, the current maximum had poor reproducibility since it depended on the original surface conditions which were not easily reprod uced. Emphasis was, therefore, placed only on the relative change of anodic current density. Each point on the current decay curve represents a statistical average. Sometimes, particularly for the Fe-5Cr alloy, oscillations or other anomalous behavior could accompany passivation even in the passive state. The effect of these sudden jumps in the current was that the current did not always return to a value as low as it had prior to the oscillation. This kind of anomalous behavior can obscure the true passive current and make it difficult to determine the true current value. From the current decay curves, increasing the chromium content can reduce the anomalous behavior. In other words, the alloy with higher chromium content will form more stable passive film. 1 o3 M l 0 « - 5 el 0 F e o C r Fe-1 5Cr Fe-255Cr 1 0*1---LJ..111llll I 1 m m l * i ■ milt i i i nm l i i n mil t i i mnl t i 139 1 1 -FeSCr Fe-15Cr Fe 25.5Cr 1 10 10i Figure lj.2. The current decay r'e-Cr all eye in IT r ^ S O ^ (pH 3)« 1 0 - 4 1 0 '«cl 0 -S' 1 0 1 0-7 F e 5 C r Fe-15Cr Fe-25 5Cr 1 0 i i ininl i i i mill i i imill x i i i m m l .iii m i l l .— i— u .j i.uil., i__l 1 0 1 o’ 1 0J 1 0 1 0 1 0 Time sesond Figure [|3. The current decay curves of Fe-Cr alloys in IF Lla-SO, (pF 6.5). d LL 1 0 ■== 1 0 Fe 1 0 Fe -5Cr Fe-1 5Cr -7 Fe-25-5Cr 1 0 , 08l— » 1 U I m l 1 I I l i a i l l i i it UiL 1 l m ml u L I I I x u iiL -F- 1 10 10 10 i o t o! 1 0 IV lime,second Figure J^. The current decay curves cf Fe-Cr alloys i:. ja„c»u- 13). 5.3 Auger Analysis The in-depth profiles of anodic oxide films formed on Fe-Cr alloys in sulfuric acid and sodium sulfate solutions were made by the simultaneous use of AES and sputtcring- etching with Ar+ ion to throw light into the thickness and composition of anodic oxide films and the experimental vari ables which affect them. The results were given as shown in Figures L|_5-68. The Auger peak-to-peak heights of the con stitutional elements, Fe(?03 eV), Cr(529 eV), 0(510 eV), S (150 eV), and C (272 eV), were plotted versus sput tering time in Figures I4.5 - 56. Each point on the composition- sputtering time profiles is an average value over the detected area including all crystallographic planes, grain boundaries, surface defects, etc. Nonuniform sputtering and mixing due to knock-on effects are not considered. These would affect compositional variation. The abrupt changes in the composition within a few angstroms m a y not b e detected. T h e significant l o w energy p e a k s of selected sputtering time at the expand scale were given in Figures 57“68. Figures \±$ - 56 show the Auger analysis of the passive film forms on the Fe-Cr alloys which had been exposed to electrolytic solutions at 600 raV (see) for 12 hours. Given plots are the normalized Auger peak heights, i.e., the ratio of the peak height of a measured element and the sum of the Fe(?03 eV) and the Cr(529 eV) peak heights vs. sputtering time. Usually, it is best to use peaks occurring above about 100 eV for quantitative analysis since the low energy Auger 4 3 Cr o + « X o re OS 0 2 0 . 30 Sputtering Time (min) Figure [4.5 * The composition-depth profile of the Fe-5Cr alloy in m li: H5S0, at 600 mV for 12 hours. £ L \. S C G i Fe-15 Cr IK HoS0, 600 mV, 12 hr 3 o — a - Fe u. X o ' 10 20 . 30 Sputtering Time (min) Figure I4.6. The composition-depth profile of the Fe-1 $ Cr alloy in IK H^SO^ at 600 ™Vsce for 12 hours. Fe-25,5 Cr IN H2S0^ 600 mV, 12 hr o. o a - Fe a> x rg Sputtering Time (min) Figure ^7. The composition-depth profile of the Fe-2505 Cr alloy in IN HoS0i at 600 irtf for 12 hours. 2 q. see Fe-£ Cr IN Na2S0^ pH3 600 mV, 12 hr o _ Fe X -Q_ Sputtering Time (min) Figure lj.8. The composition-depth profile of the Fe-5 Cr alloy in IN Ila^SO] (pH3) at 600 mV for '2 hours. 4~ Fe-15 Cr IN Na2S0^ pH3 600 mV, 12 hr Fe A - Cr a* -A. Sputtering Time (min) Figure h9. The composition-depth profile of the Fe-1$ Cr alloy in IN Na~S0( (pH3) at 600 mV for 12 hours, c see 4 Fe-25.5 Cr in Na2S0^f PH3 600 mV, 12 hr _ 3 W Fe O + a> LU x I © • mm n o s B 20 . 30 Sputtering Time (min) Figure 50. The composition-depth profile of the Fe-25.5 Cr alloy in ll'J IIaoS0, (pK3) at 600 mV a for 12 hours. ^ L}. S C G Fe-5 Cr IN Na2SC^, pH6.5 600 mV, 12 hr o- 0 □- Fe a - Cr ♦ - s x Sputtering Time (min) Figure 51* The composition-depth profile of the Fe-5 Cr alloy in 11* Na^SO^ (pH6.5 ) at 600 ^*sce For 12 hours* « 4 Fe-l5 Cr 600 mV, 12 hr 3 h* 0 - o- Fe Cr O + o> u. s. • mm X 2 e • mm 0 ID 21 . 30 Sputtering Time (min) Figure $2, The composition-depth profile of the Fe-l5 Cr alloy in Ill Ha^SO. (pH 6.5) at 600 :;Y for 12 hours, ^ o Sputtering Time (min) Figure 53* The composition-depth profile of the Fe-25.5 Cr alloy in Hi NapSO, (pH 6.5) at 600 r:V for- 12 hours, L iLL b U V Fe-5Cr IN Na2S0^_f pH 10 600 mV, 12 hr o * o Q - Fe Cr o> X Sputtering Time (min) Figure 54 • The composition-depth profile of the Fe-5 Cr alloy in IN Na?S0, (pH 10) at 600 mV for 12 hours. ^ see Fe-15 Cr IN Na2S0^, pH 10 600 mV, 12 hr IB ■ 20 . Sputtering Time (min) Figure 55, The composition-depth profile of the Fe-15 Cr alloy in II: Ha«S0. (pH 10 ) at 600 mV g for 12 hours. 2 4 ■’rii Fe-25.5 Cr IN Na2S0^, pH 10 600 mV, 12 hr 10 ' n . 30 40 Sputtering Time (min) Figure 56. The^composition-depth profile of the Fe-25.5 Cr alloy in IN IJa^SO^ (pH 10 Ij at 600 mV for 12 hours. i- O C V \jr V\ i<;6 t„ = 0 rain sp Sensitivity=0.2 t sp =0.5 min Sensitivity=0.2 t„ =i m m sp 40 E(e») Figure 57. Low energy spectra for the conditions in Figure 1+5 ■. 157 sp Sensitivity=0o5 dN( E) t =2 rain sp Sensitivity=0.5 t ap=2*5 min Sensitivity=0.5 I m i Ill i I In i il II II I 1Y1 I 11 m ill II I 20 30 4 0 50 GO E(ev) Figure 57*(continued) 158 t sp =5 min Sensitivity=1 dN( E) dE t =8 min sp Sensitivity=1 2D 30 40 SO GO E(ev> Figure 57.(continued) 15 9 t =12 min sp Sensitivity=1 dN(E) 20 30 40 60 GO E(ev) Figure Si* (continued) 160 tsp=0*5 min Sensitivity=0,5 t „ =1 min sp tsp=1 -5 mln E(e») Figure 58* Low energy spectra for the conditions in Figure 1*6 . 1 6 1 t sp= 2*5 rain Sensitivity=0.5 dN( E) 20 30 40 50 (0 E( ev) Figure 58*(continued) tsp=3*S min Sensitivity=0.5 ij_i iJ 1111 11111 111111 n 11 1111 i 11 111 11111 1 20 30 40 60 E(ev) Figure 58,(continued) t =Ll min sp Sensitivity=0.5 ii 1111 ii 11 ii 11111111 n 1111 E ( e v ) Figure 58*(continued) t sp=lS min Senaitivity=1 ill 11 m 11 ii 11 I n 11111 ii 1111 i 11 11 i 11 111 I 20 30 40 so 60 E( ev) Figure 58.(continued) 169 t =0 sp Sensitivity=0* 5 tap“ 0 *5 nlln Sensitivity=0.5 d N ( E) sp Sensitivity=0.5 ‘ sp'1 *5 min Sensitivity=0.2 Figure 59. Low Energy spectra for the conditions in Figure h i* t =2 min sp Sensitivity^O.t? dN( E) t =3 min P Sensitivity=0.5 m i I in i 1 h ii \ li i il n ii ll 11 i 11 11 i 1i.ii.iJ 20 30 4 0 50 6 0 E ( e v ) Figure 59. (continued) 167 t =4. min sp Sensitivity=0,5 tsp=5 min Sensitivity=0.5 t sp =7 min Sensitivity=0.5 l i 1111 i 111 i i I I n i I 11 11 I I i Vi 11 i i i I n i i I 40 t( e*) Figure 59.(continued) 16; t sp=25 min Sensitivity=0.5 dN( E) 30 40 SO E ( e v ) Figure 59.(continued) 16 v sp Sensitivity=0 JN( E) dE sp / Sensitivity^ 0.2 llllllllllli 111 I JJJLL 30 4 0 50 BO £( e») Figure 60. Low energy spectra for the conditions in Figure U8. 170 t =1 min sp Sensitivity=0.2 t =2 min sp Sensitivity=0.2 I 1 1 I L L_1 J HU 40 f ( e » ) Figure 60*(continued) 171 -ii. min Sensitivity=0.5 =to min Sensitivity=0,5 40 E( e») Figure 60 „(continued) 172 t =0 sp Sensitivity=0* 2 sp Sensitivity=0.2 dN( E) sp Sensitivity=0.2 ZD 30 40 50 fiO E( *») Figure 61* Low energy spectra for the conditions in Figure 49. 173 tap"1 *5 “ in Sensitivity=0.2 dN(E) t3p=2’5 min dE Sen3itivity=0.2 Sensitivity=0,2 40 E(*») Figure 61.(continued) t''sp =U.^ min iiensi tivi ty=Ci dN(t) t -10 m m sp Sens!tivity=0 t =lk min sp ^ Sensitivity=0.£ III I » HI I » I E(e») Figure 61. (continuted) 17! tsp=1'5 min -2 m m Senaitivity=0,5 t sp =3 m m Sensitivity=0.5 t =6 min sp Sensitivity=0.5 40 E( e») Figure 62. Low energy spectra for the conditions in Figure 5o. 1 7 6 rain Sensitivity=0,5 t = 2 min sp Sensitivity=0.5 t sp =3 min Sensitivity=0.5 min sp Sensitivity=0.5 40 E( e v ) Figure 62.(continued) 177 sp Sensitivity=0.5 dN( E) t =10 min sp Sensitivity=0,5 I m i In ii limlii iilm i Ii inli i ii I 20 30 40 SO 00 «•*) Figure 62.(continued) 173 t =0 sp Sensitivlty=0,5 t sp =0,5 min Sensitivity=0.5> tsp= 1 '5 rain Sensitivity=0.5 40 E(e») Figure 6 3 . Low energy spectra for the conditions in Figure 51. 1/9 t =2 min sp Sensitivity=0.5 dN( E) t =2.5 min dE tsp=3 min Sensitivity=0.£ i-llll 111 I 111 IllilU -l ti ii I i m Ii ii i In ii I 0 30 40 50 GO E( ev) Figure 63*(continued) l8o Sp Sensitivity=0.5 tsp=0‘S mln dN( E) Sensitivity=0.5 sp Sensitivity=0,5 t sp =1*5 min Sensitivity=0.5 n 30 40 GQ E(*») Figure 6/*. Low energy spectra for the conditions in Figure 52. 1 0 1 tsp= 2 '5 min Sensitivity=0.$ t =3 min sp Sensitivity=0.5 t =L min sp ^ Sensitivity=0.3 t sp =5 min Sensitivity=0,5 4 0 f i g u r e 6il , f ( « » ) C o n t i n u e d ; 1 8 ,.' sp Sensitivity=0.5 dN( E) t = 8 min sp Sensitivity=0.5 I I I I I I I I I I I I I I I I H I I I I I I ill I I I I II II I E( «*) Figure 61+. (continued) t = 0 . 5 rain dN( E) 3p Sensitivity=0.5 t„^-l min sp Sensitivity=0.5 20 30 40 60 «•») Figure 6£. Low energy spectra for the conditions in Figure t sp= 1 . 5 min Sensitivity=0.5 t a p = z ’^ m i n Sensitivity=0.5 dN(E) dE t =Ll min sp Sensitivity-0.5 30 40 50 (0 E(e») Figure 65*(continued) t =5 min sp Sensitivity=0«5 sp Sensitivity=0.5 11 n I in i 111 11 111 i 1111 ii I 111 111 11 t 11 111 I 20 30 40 SO SO E(«) Figure 65.(continued) l6o t =0 sP 5enaitivity=0.5 t =1 min sp Sensitivity=0.2 dN( E) dE t3p=1*5 min Sensitivity=0.2 II i i 1 n i i 1 it 11 111 i 11 n II I u i i 11 n i l i m 1 20 3D 40 50 60 E( ev) Figure 66. Low energy spectra Tor the conditions in Figure 5b- * 187 t =2 min 3P Sensitivity=0.2 t =2*5 roin 3p ^ Sensitivity=0.2 tsp=3*5 min Sensitivity=0.2 40 E(e*) Figure 66.(continued) Sp Sensitivity=0.2 dN( E) t =11 min sp Sensitivity=0.2 111111111 i 1111111111 Lii 11 UJJL ZQ 3D H 50 GO E(e») Figure 66.(continued) 189 t =0 sp Sensitivity=0.2 t 3p = 0 *5 m ln Sensitivity=0,2 d N t E) sp Sensitivety=0« 2 ^ p " * m i n Sensitivity=0.2 ZD 4030 SO GO E( **) Figure 67. Low energy spectra for the conditions in Figure 55. tsp=7 min Sensitivity=0.2 tsp=9 min Sensitivity=0*2 t =11 min sp Sensitivity=0.2 t =16 min sp Sensitivity=0.2 dW( E) t =22 min dE sp Sensitivity=0.2 1111 H 111 m 11 n 111111 11 30 40 SO GO E( ev) Figure 67.(continued) 19 9 t =0 sp Sensitivity=0,1 dN( E) * 3 5 = ° * s rain Sensitivity=0.2 JJULL 20 30 40 (0 E ( t * ) Figure 68* Low energy spectra for the conditions in Figure 56. 193 t =1 min sp Sensitivity=0.2 t =2 min sp Sensitivity=0.2 i i I m i I in i In ill ii 11 I i ii i 11 it i 11 i 11 I Figure 68* (continued.) 1 % t sp =7 ' min Sensitivity=0.2 ‘ s p 3 9 rain Sensitivity=0,2 t =11 m m sp Sensitivity=0.5 III I ll II II I in i 11 u i I n II I 40 E( ev) Figure 68,(continued) 199 t sp =12 min Sensitivety=0.2 t =17 min sp Sensitivity=0.5 40 E(e») Figure 68, (continued) 196 t =18 min sp dNf E) Sensitivity=0*5 dE I n i i I m i 1 ii ii 111 111 i l l l L l x j u I . ... I JJLU !0 30 40 50 SO E(«) Figure 68*(continued) 197 peaks are generally more susceptible to distortion by magnetic effects and localized specimen charging. Furthermore, Weber and Johnson C2I4.7 J showed that dN(E)/dE is a linear function of the elemental concentration, provided that there is no change in peak shape or shift in energy. No such changes were observed in the selected Auger lines. Also, the selected Auger peak should not be superimposed on another peak. The following Auger lines were used: >1(150 eV), C(?72 eV) ,and o 0(510 oV). The sputtering rate was o.. ..^.ated to be 1-2 A/min. Shown in Figures 57 - 68 are tlow energy spectra resulting from the IA0 «V7 transitions at various depths. Those spectra demonstrate the changes in the chemical states of the passive film from the surface to the matrix. -Those changes, by comparing them to standard oxides, can be used to detect surface compounds. Figures 69-72 display the low energy spectra of several chromium oxides and spinel iron-chromium oxide. The chromium oxide specimen was Fisher's certified grad chromium sesqui- oxide wh: ch v/as mounted by using silver print on stainless steel foil. The same means was also used for other powder samples. The spectrum of chromium sesquioxide shows two major low energy peaks at 31*5 and ^7 eV { as seen in Figure 69 ). The chromium hydroxide, Cr(OH)^, was prepared in two ways: (1) Pure chromium was passivated in oxygen-saturated double-distilled water for a few minutes at room temperature; (2) Precipitating Cr(OH)^ on Ag foil from a solution of I9t3 dN( E) L ll l i.lii I I III I llll II I I I I I I I II I I lLl_i_l_l 20 30 49 SO GO E( ev) Figure 69* The low energy spectra of Cr20^. 199 dN( E) 20 30 40 50 CO E (e v ) Figure 70. The low energy spectra of Cr(OH)~ (precipitated from CrCl^+ KH^OH on Ag' foil). J 2 0 0 dN(E) 20 30 46 SO GO £( ev) Figure 71, The low energy spectra of C r ( O H K (pure Cr in aerated H20 for 30 minutes). J 201 d N ( E) n I i llll i III II III i il II II11II I Ii II i li m 28 38 4 8 SO E( ev) Figure 72. The low energy spectra of Fen ^CrA ,--0-,. 202 0.1 M CrCl^ into which a few drops of NaOH was introduced. This foil was then rinsed in a large amount of double- distilled water. Both samples lead to the same major low energy peaks at 32 eV and ^ 6.5 eV as shown in Figures 70 and 71. The oxide Fe^ crucible with a gas burner until the NH^ and H 20 were expelled and then heating the residue at 1000°C for 4 hours. The structure was verified by x-ray diffraction using diffracto meter and powder techniques. Figure 72 indicates that the low energy peaks of this spinel iron-chromium oxide are at 33»5 » and 52.5 e V , The low energy Auger lines of pure chromium and pure iron are at 36.5 and ^8.8 eV, respectively, as presented in Figures 73 and 7k> Table 11 and Figure 75 give the quantitative relationship between the normalized Auger peak-topeak height of chromium and the atomic percentage of chromium of Fe-Cr alloys. It. 111 m 11.111 i n 111 ii ii 11 ii i fi 111 jjJ 20 30 40 SO 00 £ ( e v ) Figure 7 3 . The low energy spectra of pure chromium. dN(E) ULU 11 .! 11 111 i 1111 11 1111 . 11 11 i I 20 30 40 to E( ev) Figure 7^. The low energy spectra of pure iron. 205 Table 11. The quantitive relationship between the chromium concentration and tne Auger NAPH ratio Nominal wt % Cr at % Cr NAPH Ratio pfl^ r Fe 0 0 0 Fe-5Cr 5 5 . 4 0.07 Fe-l5Cr 15 15.9 0.21 Fe-25*5Cr 25.5 26.4 0.36 Fe-50Cr 49.9 51.7 o. 64 Fe-7QCr 78.2 79.4 0.84 Cr 100 100 1.0 IIA PH Ratio { Cr/Fe+Cr iue 75, Figure 0.2 G.I+ 0.6 1.0 0.6 100 Fe 0 AH ratio. NAPH hoim ocnrto ad h Auger the the and between concentration relationship chromium quantitative The 20 ho , ie fa C = A1 p^~Cr at ^ if NAP11 p = Line % Cr at at % Cr 60 80 Cr 206 DISCUSSION 6.1 Polarization Behavior of Alloys The passivity of metals is complicated by the introduc tion of alloying elements. The alloying additions may have the following effects upon the passivity of alloysi 1. changes in the defect structure, e.g. impurties or imperfections, which may have influence upon local action cells, 2. changes in grain boundary attack caused by segrega tion and heat treatment, 3. changes in the affinity of the ingredient metals for each other and for the non-metals, 4. the effects of the relative volumes of the various phases, 5* changes in the composition and/or the structure of the passive film, e.g. the possibility of forming ternary compounds, 6. changes in the physical, chemical, electrical, and mechanical properties of the passive film, 7- the effects of the mutual solubilities of the corro sion products present in the passive films, 8. changes in the diffusion rates of atoms in the alloy and of ions in the passive film, 9* changes in the kinetics of anodic and cathodic 207 208 reactions. Generally, it is difficult to interpret the influence of alloying elements on the polarization characteristics of iron-base alloys, particularly in the case of multiphase alloys where the structural changes may be important. Changes in the composition and structure will affect the kinetics of anodic and cathodic reactions which determine the rate of corrosion processes in aqueous environments - Con sequently, it is only possible to relate significant courses which alter the corrosion properties of single phase alloys. Several characteristic values from the polarization curves obtained in this study (Figures 37-40) have been summarized in Table 12. They are the rest potential first critical current density (icr^^ ) ar*d its corresponding critical potential for passivation (Epp )t second critical current density ^ corresPond^nS critical potential for passivation (E )» the passive current density PP2 (i ), the first transpassive potential (E„ )» and the second P i transpassive potential (E™ ). Since the polarization curve 12 of an alloy shows the characteristics of the components of the alloy in their individual state, therefore there may be more than one value for the same electrochemical characteristic in an alloy. The effects of Cr-content and pH value of the electrolyte on the electrochemical characteristics of Fe-Cr alloys will be discussed in the following sections. Table 12. The electrochemical characteristics of Fe-Cr alloys. E E Et Specimen e r ^crit^ ^crit^ e t Electrolyte PPa PP2 *P n 2 2 (mV) (mv) (amp/cm2 ) (mv) (amp/cm2 ) {amp/cm ) (mv) (mV) IN H2S0^ Fe -505 0 3 .1|jcl0"! 8.1x10"? 1500 Fe-5Cr -520 -100 1 .1x10",, 5 .2x10"? 900 1550 Fe-l5Cr -56o -250 8.6x10"? -100 2,2xl0"7[ 6.8x10"? 900 1500 Fe-25.5Cr -570 -I4.50 2 ,6x10** -100 1.6X10"4- U.i*xl0"b 900 1600 IN Na2S0^ Fe -620 +50 2.3x10”! 7 .0x10"? 1500 Fe-5Cr -614.0 -100 9.5x10"? 3 .2x10"? 800 II4.OO (pH 3) Fe-l5Cr -660 -500 1.7x10"? -100 i.5xio“? 3 .0x10"? 800 1380 Fe-25.SCr -665 -600 1.7xl0"J -100 3 .2x10"^ 2 ,8x10"° 750 IN Na2S0^ Fe -760 +200 2 .3xlO"p 9.5x10"? 1200 Fe-5Cr -770 0 9.7x10"? 5 .5x10"? 1200 (pH 6.5) Fe-15Gr -1*75 ------0 2 ,8x 10"? 2.5x10"? 900 1100 Fe-25.5Cr -U50 -100 2 ,l|xl0" 1 ,1x10"° 900 - IN Na2S0^ Fe -695 -150 1 .7x10“* 5.3x10"? 1200 Fe-5Cr -630 -100 8 .2x10"? S.ixio"? 1200 (pH 10) Fe-l5Cr -280 ------+1^00 2 .7x10"? 1 .9x10"? 950 1200 Fe-25.5Cr -214.0 +150 2 .9x10" 1.9x10"° 900 > 210 6.1.1 Rest Potential For a long while, the rest potential of a metal was used as a criterion of its corrosion resistance. Later it was understood that the measurement of rest potential alone is insufficient to predict the corrosion resistance of the metal. As we know, the rest potential is determined by both the cathodic reaction(s) with current I and the anodic reaction(s) with current Ia » Without knowledge of the course of these reactions, the rest potential measurements may be misleading. An increase in ER can be due to an increase in IR or a decrease. Similarly, a decrease of ER may mean an increase or decrease of IR . Furthermore, ER remaining constant could simply be the result of simultaneous increasing or decreasing of I_and I . With these uncertainties, the rest potential of a C a metal is still an important corrosion property. The effects of chromium content and pH on the rest potentials of Fe-Cr alloys are shown in Figures 76-77. Figure 76 shows that the effect of chromium content on the rest potentials of Fe-Cr alloys is diverse. With increas ing chromium content, the rest potential decreases gradually in acid solutions, but the rest potential increases in neutral and basic solutions, particularly, with a rapid change around the critical chromium composition. Figure 76 also indicates that high chromium alloys have much nobler rest potentials in neutral and basic solutions than in acid solutions. On the other hand, the rest potentials of low chromium alloys is > see - -700 -500 -J -300 -200 80 4 .OO c 0 L Figure Figure 5 76 10 et oetas f rF alloys. Cr-Fe of potentials Rest wt wt 15 % Cr H 3' pH H 0 pH H 10 pH 20 25 211 212 -200 -400 o / Fe-l5Cr > -5oo -600 -700 Fe* -800 0 2 6 8 10 Figure 77. Effect of pH on the rest potentials of Fe-Cr alloys. 213 nobler in acid solutions than in neutral and basic solutions. The nobler rest potentials of high chromium imply a better passive state of the alloys. A metal with nobler corrosion potential does not necessarily have higher corrosion resist ance, since its passive film may have relatively high or low dissolution rate. Further discussion will be given in Section 6.2. Figure 77 gives the effects of pH on the rest potentials of Fe-Cr alloys. We can see that the rest potentials of all specimens decrease first with increasing pH, and then the changes are reversed at pH 6.5 for iron and the low-Cr alloy and at pH3 for two high-Cr alloys. The decreasing se^nents of the rest potentials hint that the oxidation of the metallic atoms to their ions is the anodic reaction, and the hydrogen evolution is the cathodic reaction. The increasing segments of the rest potentials signify that other reactions are involved, e.g. the film formation reactions. In summary, the rest potential measurements show that the corrosion reactions can be changed by varying composition and varying pH. An iron base alloy with high chromium content or in high pH environment has noble rest potential and should be easily passivated. However, to determine the corrosion resistance of a metal, both rest potential and dissolution rate should be considered. 2 1 / + 6.1.2 Critical Current Density and Passivation Potential The appearance of an anodic current density maximum is one of the important characteristics of an active- passive metal. The anodic current density maximum is charact erized by the critical passivation potential ISpp and the critical current density The critical current densiti es and critical passivation potentials of Fe-Cr alloys are plotted in Figures 78 -83 . We can see from these electrochem ical measurements that in some cases, it is possible to have two critical current densities for high-Cr alloys. Since chromium is the more active element, then, as increasing the potential from £R to the noble direction, we observe first the critical current density (i,**. ) at more active potential 1 and second the critical current density (i_ *+. ) at more noble 2 potential. icr^^ and icr^t characterize the maximum chromium dissolution and the maximum iron dissolution, respectively. Only the two high-Cr alloys in the two acid solutions have i__.+ • The effects of chromium content and pH on cr 1 ^ c n are displayed in Figures 78 -79 , respectively. Despite the fact that there are only two points for each curve on those figures, it can be seen that icr^ decreases with increasing Cr content and pH. Also the slopes of curves become steeper with higher chromium content and higher pH: the slope changes from -5-7x10“^ a m p / c m ‘y'TSCr for IN HgSO^ solution to -1.5x10“-^ 2 — 2 amp/cm /jSCr for IN Na^SO^, pH3 solution and from -2*3x10“ Current density, amp/cm' 10 10 Figure Figure 78 Te cl^ f eC alloys. Fe-Cr of icrlt^ The . 10 * IN NaoSO, pK 3 pK NaoSO, IN * wt wt % Cr 20 25 30 215 Current density, amp/cm' 10 10 -1 -2 iue 9 Efc o p o the on pH of Effect 79. Figure f eC aly. 1 alloys. Fe-Cr of pH 6 8 10 216 OJ Current density, amp/cm' iue 0 The 80. Figure ^CT±t^ O Q A • fF-r alloys. F«-Cr of IN IN UN IN H 6.5 pH 2 % *23 a G0^, Na2 Na2S0^> Na2S0^, pH H 3 pH pH H 10 pH 6.5 10 217 Current density, amp/cm2 M* OQ c: ►j O' vn (D CD M •“!> (5 O cl- O H3 K O 3 o 4 H* c+ vn vn ro o 0 o H) (D CD 1 CJ P o » see ixr 8. h psiainptnil o F-r alloys. Fe-Cr of potentials passivation The Figxire 82. -700 -^00 -20C -100 600 500 100 200 300 0 H 6.5 pH o NapSO,IN , pH6.5 3 pH Na2S0^, 1N A H0SO, •IN NNpO pHIO NapSO, IN 5 10 wt wt % 15 Cr 20 H 3 pH 0 pH 25 PP 30 219 see -700 -600 -ij.00 -200 -300 1^00 200 100 300 0 0 5 Figure Figure o • Fe Fe-25.5Cr 83 Efc o p o te passivation the on pH of Effect , oetas f eC alloys. Fe-Cr of potentials Fe PP PP2 10 220 2 21 amp/cm 2/pH for the Fe-15Cr alloy to -8.1x10"^ amp/cm2/p}{ for the Fe-25*5cr alloy. This signifies that, with increasing chromium content and pH, i decreases more quickly. Also, crit, higher chromium content in the alloy can more effectively suppress the dissolution of the alloy in the active range. Two high-Cr alloys in the neutral solution and in the basic solution do not possess i„„*+ • This can be understood cr x from the superimposed Pourbaix diagrams of chromium and iron (Figure 8i|)i chromium oxide is stable from pH5 to pH12.5 and at low potential ranges, iron oxide is more stable in basic solutions and at high potential ranges. The effects of chromium content and pH on i r.„j + , which 2 characterizes the maximum iron dissolution, are shown in Figures 80-81, respectively. Figure 80 reveals that i .. 2 decreases from a high level for iron and the Fe-5Cr alloy to a much lower level for two high-Cr alloys. There are 3-5 orders of different magnitude. Alloying iron with 12 wt fo or more of chromium can very effectively reduce i .. • During cr 2 the running of the polarization curves, it was observed that much less black corrosion products ( probaoly sulfate or oxy- sulfates ) were formed on iron-chromium alloys, in the active range, than were formed on iron specimens. Furthermore, the amount of precipitated corrosion products formed is negligible for stainless Fe-Cr alloys. Hence, in contrast to the case of iron (21j.8), the predominate factor for the onset of passiva tion of two high-Cr alloys is pH in the acid ranges and not 222 e h (v ) 2.0 1.6 1.2 0.8 0.4 -0.4 Fe/Fe - 0.8 Cr/Cr - 1.6 -2 0 2 6 8 10 12 16 pH Figure 8l(.« Supperposition of Pourbaix diagrams for chromium and iron showing the insoluble products and some of the equilibria defining the phase boundaries. 223 the accumulation of cations nor the amount of primary sulfate which is formed. The effects of pH on i__j +. are given in Figure 81. It cri T*p can be seen that pH has little influence on i .. of iron and 2 Fe-5Cr alloy. i .. of these two specimens decreases only 17 c m t£ slightly with increasing pH. This implies that ferrous sulfate could be the major compound formed on the surface in the active range of iron and the Fe-£Cr alloy. of the two high-Cr 2 alloys is not just at lower levels, but also it is more strong ly influenced by changing pH. At pH 6.5 and pH 10, chromium oxide can be formed on the alloy even at low potentials as shown in Figure 6I4.. This can inhibit the formation of iron sul fate or oxysulfate and help the formation of iron oxides. The refore, irt„ - of high-Cr alloys is a stronger function of pH. cri 2 The corresponding characteristics with icrj_£ is designat ed as the critical passivation potential, E , which marks the P P begining of the passivation process. In acid solutions, E shifts to the active direction with increasing chromium content as shown in Figure Q2. In acid and neutral solutions, E PP2 drops from the value of pure iron to the value of the Fe-5Cr alloy and then it becomes insensitive to the chromium content of the alloy. However, on the other hand, in the basic solut ion, E increases first and then drops from 15 wt$ Cr. It PP^ seems that, in acid solutions and neutral solution, chromium can narrow the active region of iron-base alloys by shifting the passivation potential to more negative values. 22/j. Epp of two high-Cr alloys decreases with increasing pH in the acid range as show in Figure 8 3 . However their Epp 2 increases through the whole test pH range. Epp of iron and 2 the low-Cr alloy increases first and then decreases. Also, Figure 83 shows that, with increasing chromium content, the changes of Epp's with pH become less abrupt. 6.1*3 Passive Current Density Passive current density, i , is the current density in Jr the passive range, i^ remains reasonable constant until the potential is raised to the transpassive potential. Though i r is almost independent of potential, it is an extreme time dependent function. The dependence of ip with time will be discussed in Section 6.2. The passive current density from polarization curves (Figures 37—J+O) is plotted vs. chromium content and pH in Figures 85-86, respectively. Figure 85 demonstrates that ip decreases with increasing chromium content in all test solutions. However, the effects of pH on ip are diverse as shown in Figure 861 with increasing pH, ^ of iron and Fe-5Cr alloy decreases first following by an increase, and decreases againj i of the Fe-15Cr alloy in- Sr creases and then decreases; i of the Fe-25*5Cr alloy decreas- kr es and then increases. As we can find out from the current density-time measurements that i is extremely dependent on time. Therefore, it would be better not to make any decisive conclusion from i*s of polarization curves. However, it kr 225 10 A 1N TJapSOi , pH 3 IN NaP30. , pH 6.5 IN HapoOj t pH 10 oj oS pH 3 * -P a> pll 10 •a pH 0 -p 10 : 20 wt % Cr Figure 85* T*1© passive current densities,i ,of Fe-Cr alloys. r c\i ft o E Current density § 10 iue 6 Efc o p o te asv current passive the on pH of Effect 86. Figure 10 -6 este, p o F-r alloys. Fe-Cr of , ip densities, pH OCr 5Cr 22? seems fair to say that alloying iron with chromium can reduce i in all tested solutions. P 6.1,4 Transpassive Potential There are two transpassive potentials, E™, and Em » *1 LZ listed in Table 12. E„ refers to the potential for the 11 breakdown of chromium oxide. It is interesting to note that there is an E- for the Fe-5Cr alloy in acid solutions but not A1 in the neutral and the basic solutions. This means that there is enough chromium enrichment in the external layers of the alloy accumulated in the passive range of the Fe-5Cr alloy in the acid solutions to reveal the breakdown of chromium oxides, but not enough chromium oxide is formed in the neutral and the basic solutions. This supposition can be verified by AES's composition-depth profile studies. The close value of ’s A1 in all cases suggest that the same kind of chromium oxide is responsible for the passivity of Fe-Cr alloys. Above E„ , X1 chromium oxide dissolves as CrO^ ions ( in neutral and basic solutions ) or Cr20y ions ( in acid solutions ). Between ET and E-, , secondary passivity may be observed. L1 LZ The phenomenon of secondary passivity may be attributed to the formation of a new kind of surface compound (2ij.9) or the adsor ption of oxygen or anions from the solution on the surface of the metal (250-251). ET is not the potential for the breakdown of iron oxide 2 but the potential for the oxygen evolution, however the pro 228 perties of the oxide film can change at this potential. E™ x2 should be independent of the composition of the specimen but dependent on pH, and it decreases with increasing pH. The Fe-25*5Cr alloy does not show E^, in all the tested IN Na^SO^ solutions. The reason is because the electrochemical proper ties of the passive film of the Fe-25»5Cr alloy are very close to those of pure chromium which does not possess secondary passivity either. In the case of pure chromium, the dissolu tion of chromium oxide and the oxygen evolution are united together, and result in a concurrent current-potential behavi or at potentials above S* . 1 6 .I.5 Summary We can see from these previous sections that, by increas ing the chromium content in the iron-chromium alloys, there is a progressive reduction in the critical current density and the passive current density. All Fe-Cr alloys do not indicate the violent oscillation of current occurring at anodic potent ials in the active range which is a characteristic behavior of Fe in dilute sulfuric acids. Furthermore, chromium can effectively suppress the active corrosion of iron-base alloys by narrowing the active region of potentials in which the dissolution of alloys occures. In addition, the following statements can be madet 1. In acid solutions, two high-Cr alloys (15 and 25.5 wt Cr ) have two critical current densities, but in all other cases ( two high-Cr alloys in neutral and basic solu tions, and the Pe-^Cr alloys in all tested solutions ) there is only one critical current density. 2. All the critical and the passive current densities decrease with increasing chromium content. But, an alloy with higher chromium content will be more easily passivated, and it has better corrosion resistance in the passive state. 3. EDC decreases with increasing chromium content and ^*1 increasing pH in the acid pH ranges. However, all Fe-Cr alloys have the same Epp in the acid pH ranges. In neutral and basic solutions, the effects of chromium content and pH on E ™ are not straightforward. 2 4. The low chromium content in the Fe-5Cr alloy results in the absence of E^, in IN NaSO^ solutions ( pH=3, 6.5 and 10 ). On the other hand, the high chromium content in the Fe-25«5Cr alloy results in the only ET in IN NaSO^ solution 230 6.1.6 Conclusions The electrochemical properties of Fe-Cr alloys can par tially be rationalized by using the superimposed Fourbaix diagrams for chromium and iron (Figure 8i|). Upon increasing the potential from to the noble direction, it is expected to observe the following* (a) In acid solutions, first the alloy passes the chrom- ium dissolution line (Cr/Cr ) and then the iron dissolution line (Fe/Fe ). Then, it will be possible to have two current maxima in the active range. However, the formation of metast able chromium oxide ( from the extension of Cr/Cr^O^ line ) is also possible earlier than that of metastable iron oxide ( from the extension of Fe/Fe^0^ ). Therefore, the formation of chromium oxide can greatly reduce the iron dissolution rate ( or in Figures 37-ij-O ). After passing the extended Fe/Fe^0^ line, the formation of iron oxide together with chromium oxide is also possible. At high anodic potentials ( above Fe /Fe^O^ line ), at 2 ^ pH =s5, stable iron oxide (Fe^O^) can be formed on the surface of the alloy. (b) In neutral pH ranges, the formed chromium oxide on the surface of the alloy will be a stable one. The stable chromium oxide should be even more effectively retarding the iron dissolution reactions than the metastable chromium oxide. This expectation can be noti-ced by comparing Figure 39 to Figures 37-38. The formation of stable iron oxide is also possible at high potentials as discussed in the previous 231 paragraph. (c) In basic pH ranges but not greater than pH 12.6, by measuring the polarization curves starting from active potent ials, first the chromium in the alloy can directly be oxidized as the chromium oxide and, upon advanced polarization, the passive film on the alloy can also contain Fe^O^ or Fe^O^, It seems that the Fe-Cr alloys don’t have to pass a metal/ metal ion dissolution line and icri-t's sJl0Ul^ no^ present. However, pure iron and the Fe-5Cr alloy still possess high + as shown in Figure l+o. Those high i„„4+ 's can be attributed to the action of sulfate ions: probably the non- protective iron sulfate is forming in the active range. High chromium content in the two high-Cr alloys can reduce greatly. However, at higher potentials ( above Fe/Fe^O^ ), there will be a competition between chromium oxide and iron oxide, and therefore the passive film may contain chromium oxide or iron oxide or both. Taking into consideration the measurements of the polariz ation curves of the studied alloys, it seems that the bene ficial effects of alloying chromium on the corrosion properti es of iron base alloys are inherent in its more active elect rochemical nature. The active nature and the low critical current density of chromium make it possible to passivate the alloy even in a weak oxidizing solution. 232 6.2 Current Decay Curves 6.2.1 Determination of Passive Properties of Fe-Cr alloys Polarization curve measurements are an important tech nique in studying the electrochemical behavior of metals. Particularly, the potentiostatic polarization curves provide information which correspond closely to the actual behavior of passive metals. In spite of their wide applicability and extensive use, considerable insufficiency of polarization curve measurements still exists. One of the insufficiencies is showing the performance of the materials in a long run. Therefore, in the engineering sence, it is advisable to obtain the current-time curves at a selected potential before making a final conclusion. Current decay curves can be obtained by selecting a potential in the passive region and recording the current time relationship. At applied anodic potentials, the current begins at a high value, but drops rapidly during the first few seconds and finally slows down. Two linear relationships between the current density and time in log-log scale can be observed! the first one is in the early stage, the second one is in the later stage, and in between these two stages is the transition stage. That is, the current decay curves can be roughly divided into these three stages. The current density i passing at contant potential in the passive region is the summation of two different electrochemic al reactions at the interface oxide/eleotrolyte, i.e. the 233 formation of oxide layer i^. and the dissolution i^, i = if + id ( 6 - D By integrating the area under the decay curves ( Figures I4.l-I4.ll.). We can obtain the total number of coulombs p a s s e d during film growth. Those coulombs can be converted into the amount of iron dissolved ( W in gms/cm ) by Faraday*s law, » = QrT:nr M where is the area under the current decay curves in coulombs/cm , M the molecular weight of iron ( 55*85 Qm s/mole ), n the number of electrons involved in the reaction, and F the Faraday*s constant ( 96,500 coulombs/equivalent ). Then, supposing that iron formed an oxide instead of going into electrolyte, the weight from Equation (6,2 ) can be converted into film thickness. This can be done by considering only one oxide, Fe20^, and calculating how many grams of the oxide the above weight would yield. This value is divided by the density of the oxide, gns/cm^ to give a thickness. These results are given in Table 13. They show that the calculated film thickness is greater than the actual film thickness of the alloys. The actual film thickness, from the o Auger composition profiles, is in the range of 20 to 80 A. The difference between the calculated value and the actual value represent the amount of iron which was lost into the electrolyte instead of forming an oxide. Generally, this difference decreases for alloys with increasing chromium contents. These results indicate that when there is more Table 13. Film thickness (2) calculated from the total charge passed during current decay. Spec imen Fe-OCr Fe-5Cr Fe-l5Cr Fe-25.5Cr Electrolyte*''-*-^ IN H2S0^ 1000 163.5 162.1 75.2 IN Na2S0^, pH 3 690 107.0 106.8 67.3 IN Na2S0^, pH 6.5 810 66.2 53.0 43.8 IN Na2S0^, pH 10 670 125.2 116.8 93.5 chromium in the alloy to dominate the passivation reaction, then the iron is protected better. Hence* the lower the current density, the thinner the film, and the better the properties of the passive film. Table 11; gives the current density of each alloy after passivating in the electrolyte at 600 mV for 12 hours, and SvC it is designated as i , Though those current densities are 12 not in the steady state yet, they are in the very low range already. It seems reasonable to assume that, under those low current densities, the thickness of the passive films remains almost constant. The newly forming layer is just for compensating the loss due to the dissolution of the passive film. Figure 87 shows that i decreases with increasing 12 chromium content. The change of i with chromium content p12 is not as rapid as i (Figure 85) from polarization curves. Figure 87 also indicates that chromium has a slightly greater influence on i of Fe-Cr alloys in acid solutions than in p 12 the neutral and the basic solutions. Figure 88 gives the effects of pH on i of P'e-Cr alloys. p12 i decreases first and, then, increases with raising pH, and p12 it tends to minimize in neutral pH ranges. Also, Figure 88 shows that i_ curves of two high-Cr alloys are close toge- p12 ther, and i curve of the Fe-5Cr alloy is in the higher 12 level. Based on these results, it seems that the composition of the passive films of two high-Cr alloys are very similar*. Table lij., The current densities of Fe-Cr alloys after twelve hours of passivation and of Fe after one hour of passivation at 600 mV . see i . 2 x 2 p^,amp/cm p12, amp/cm Electrolyte Fe-OCr Fe-5Cr Fe-l5Cr Fe-25.5 Cr IN H„S0, i|.8xlO-7 2,6xlO-7 1.8xlO-7 2 U IN Na2S0^, pH 3 l.OxlO-5 il.i4.xlO-7 8.2xlO-8 7 .0xl0-8 IN Na2S0^,pH 6.5 i+.5xlO'6 1.2xl0-7 6.I4.XIO 8 5.OxlO-8 IN Na2S0^, pH 10 7.0x10 8 3.5xl0-7 2.lxlO-7 1.7xlO-7 Current density, amp/cm' 10 10 -8 6 - o Figure Figure 7 8 . 10 The i of of i The wt wt p12 % 15 Cr N a0,, H 10 , pH Nap00, O IN N Nao30, IN * N a0^ p 6.5 pH Na200^, °IN Fu-Cr H 0. pH H 6,5. pH 20 alloys. 237 i v r iue 8 Efc o p o te o eC alloys. Fe-Cr of i the on pH of Effect 88. Figute Current density, airp/cm 10 10 10 -8 6 - 0 2 PH Fe-23'.5Cr O Fe-l^Cr O Fe-5Cr A 6 P12 8 10 239 6.2.2 The Kinetics of Film Formation on Fe - Cr* Alloys An seen from the current decay curves, two sections of linear dependence of time are observed, and the section in the early stage is much steeper than that in the later stage. From the slopes of the log current density versus log time plots of current decay curves, one can sometimes deduce the corresponding film growth laws. A number of relationships have been suggested as given in Section 2.6.2. Those relation ships are the linear, the parabolic, the cubic, the logarithm ic, and the inverse logarithmic relationships. If we assume a uniform film growth model and a unity efficiency of the anodic current for film formation, then W « Qf (6.3) where is the amount of charge for film formation, and W the weight gain. Parabolic, cubic, and logarithmic film growth laws are respectively related to time as follows: j. At2 (6.ZO Qf ■ Atly/3 *f - (6.5) Ain t + B (6.6) Qf = where A and B are constants. Since the anodic current, If, is the derivative of Q^. Equations (6.^) to (6.6) lead to Equations (6,?) to (6.9)» respectively. where is the anodic current density, and S the area. Hence, in anodic current decay curves, slopes m= -1/2, -2/3, and -1 on a log-log scale indicate parabolic, cubic, and logarithmic rate laws for film growth, respectively. The slopes (m) and the intercepts (a) from the data obtained and given in Figures l+l-l|l|. are calculated by least square method. They are listed as following in the forms of equationsi 1. In IN H230^, at 600 mV i Fe-5Cr Early stage * log i=-2.Ul-0.12 log t . (6.10) Later stagei log i^-4. 99-0.30 log t. (6.11) Fe-15Cr Early stagei log i=-2.88-1.40 log t. (6.12) Later stage t log i=_3.!3_0.80 log t. (6.13) Fe-25Cr Early stagei log i=-3.20-1.53 log t. (6.1h) Later stagei log i=-3.52-0.85 log t. (6.15) In IN Na2S0^, at 600 mv i pH 3 . ; see Fe-5Cr Early stagei not determined * (6.16) Later stage t log i=_h.06-0.6l log t. (6 .1 7 ) Fe-15Cr Early stage i log i=-3.10-l.61 log t • (6.18) Later stage i log i=-3«64-0.72 log t. (6.19) Fe-25Cr Early stage t log i=-3 .36-2.98 log t * (6.20) Later stage i log i=-3•77-0.76 log t. (6.21) 2 1 + 1 3. In IN Na230^, , at 600 mV i pH 6.5 see Fe-5Cr Early stage: log i=-3.j+9-2.3i+ log t . (6.22) Later stage i log i=-3.79-0.80 log t. (6.23) Fe-15Cr Early s tage i log i=-3.21-1.5^ log t. (6.24) Later stage t log i=-3*80-0.86 log t. (6.25) Fe-25Cr Early stage * log i=-3.41-1.68 log t. (6.26) Later stage( log i=-3.62-0.96 log t. (6.27) 4. In IN Na2S0^, pH 10, at 600 mv i see Fe-5Cr Early stagei log i=-3« 29-1.10 log t. (6.28) Later stage i log i=-3.73-0.?2 log t. (6.29) Fe-15Cr Early stage i log i=-3.20-1.71 log t. (6 .30) Later stage t log i=-3*75-0.75 log t . (6.31) Fe-25Cr Early stage i log i=-2.94-1.94 log t . (6.32) Later stage t log i=-3*93-0.?4 log t. (6.33) The slopes and the intercepts of the above equations can also be plotted as the‘function of the chromium content and the pH values as shown in Figures 89-92. From these results we can make the following statements* 1. In (almost) all cases the slopes of the early stage are more negative than those of the later stage. A more negative slope of the current decay equations implies a great er change in reactions rate. Two different slopes of the early stage and the later stage might suggest two different mechanisms governing the growth of the passive film in two stages. In the long run, the current density decreases to a very low value. Eventually the growth rate of the passive 2 i;2 Early Stage * IN Na2S0^f pH 3 ° IN Na2S0^f pH 6.3 * IN NaPSO, , pH 10a - 0.2 - 1.0 - 1.2 -1.3- - 1.8 - 2.0 - 2.2 -2.il - 2.8 -3.0 0 10 20 30 wt % Cr Figure 89. The intercepts, a, and the slopes, m, of the early stage of the current decay curves of Fe-Cr alloys. 21+3 Early Stage 0 a Fe-5>Cr - 1 0 Pe-15'Cr o Fe-25.5Cr -2 < _ o -3 11 O o § ■ -1+ 0 - 0.2 -0-4 -OS - 0.0 - 1.0 - 1.2 -1.4 o - 1.0 o a - 1 . 0 o - 2.0 - 2.2 -2.4 -2.S - 2.1 -3.0 J fi I______I______L 0 2 6 8 10 12 PH Figure 90. Effect of pH on the intercepts, a, and the slopes, m , of the early stage of the current decay curves of Fe-Cr alloys. Later Stage -2 1 -21 -30 o 5 10 15 20 25 30 wt % Cr Figure 91. The intercepts, a, and the slopes, of the later stage of the current decay curves of Fe-Cr alloys. zk5 Later Stage a - 4 0 .4 - 0 .4 - 0- m - 1.4 - 2.2 - 2.6 2.8 - 3 . 0 0 10 12 p H Figure 92. Effect of pH on the intercepts, a, and the slopes, m, of the later stage of the current decay curves of Fe—Cr alloys* 21+6 film is so small that for many practical purposes the thick ness may be said to have reached a limiting value. 2. By increasing the chromium content, the current decay curve in the early stage becomes steeper ( more negative in the slope ), except the Fe-5Cr alloy in the neutral sodium sulfate solution. This trend can be attributed to the more active nature of alloying chromium. Chromium is a more active element than iron and chromium atoms have a stronger oxygen affinity than that of iron atoms. Alloying chromium atoms react faster with oxygen atoms than the iron atoms and lead to steeper slopes in the early stage as chromium content increas es 3. In neutral sodium sulfate solutions, particularly for the Fe-25»5Cr alloy, the slopes of the current decay curves are closer to -1 as compared to those in acid and basic solu tions. A -1 slope of the current decay curve shows the log arithmic process. All three alloys have the similar film growth mechanism in basic sodium sulfate solution which probably means that the same kind of oxide is formed on the surface of these alloys. 6.2.3 Summary From current-time measurements of Fe-Cr alloys, it follows 1 1. More reliable values of the passive current can be obtained after prolonged time of measurement. 2U'7 2. The passive current is extremely time dependent. After twelve hours of measurement, the passive current density (i in Table lij.) is about 2-3 orders smaller in magnitude p12 than the passive current density obtained from the polariza tion curve measurements for the same alloy in the same solution (Table 12). 3. For the same solution, the difference between i and i of the Fe-5 Cr alloy is greater than that of the two p12 high-Cr alloys. That is, after a long period of passivation, all alloys have close passive current densities. 4. Though, both i and i decrease with increasing P P 12 chromium content (Figures 85 and 86), there are several differences between them. First, the i -chromium content p12 curves are smoother than the i -chromium content curves. P Also, the i 's of all alloys are almost in the same p12 order of magnitude, but there is about one order difference in magnitude between the Fe-5Cr alloy and the Fe-25«5Cr alloy for i * s. Second, there is a general tendency that the i_ - p p12 chromium content curve is at the highest level for pKO and followed by that of pH 10, pH 3 and pH 6.5 in sequence. However, there is no such tendency for i -chromium content curves. 5 . It was difficult to see the true effect of pH on i (Figure 86). However, the effect of pH on i is clear: ^12 i tends to minimize at neutral pH (Figure 88). p12 6 . Roughly, a current decay curve can be divided into 21*8 three stages for film growthi the early fast growth stage, the transition stage, and the later slow growth stage. It is only, sometimes, possible to relate one of the established film growth laws for the later slow growth stage. 7. By comparing the calculated film thickness and the actual film thickness, it can be seen that most of the passive current density does not come from the current density for film growth but from the current density for film dissolution. 6.2.4 Conclusion In conclusion, long term current-time measurements are very useful for practical resasons. It will be more honest to reveal the real effects of chromium content and pH on the passive current densities of the alloys. However, it is not easy to find out the film growth kinetics from current decay curves, because, in addition to the film growth reactions, other reactions can also occur. 6.3 Auger Analysis The in-depth composition profiles of passive films formed on Fe-Cr alloy in sulfuric acid and sodium sulfate solutions were obtained by using AES technique. From the Auger analysis one can obtain the data concerning the distribution of the elements in the passive film and especially (1 ) the maximum of chromium enrichment zone in the passive film, (2) the average chromium centent in the passive film, (3 ) the composi 249 tion of the passive film, and (4) the sulfur content in the passive film. The thickness of the passive film is equal to ( the sputtering rate ) x ( the sputtering time needed to remove all the oxygen ). The sputtering rate estimated in this study o is about 1 to 2 A/min. Therefore, the film thickness of iron- o o chromium alloys in this study is from 20 A to 30 A depending on the environments. There is no great difference in the thickness of the passive films of all t!±ree alloys under the same conditions. Still, the films on the Pe-SC*1 v-lloy seem slightly thicker than those on the two high chromium alloys. Thickness alone is not very meaningful on the passivity of metals. There can be a protective and very thick film (e.g. anodized aluminum), a non-protective and very thick film (e.g. rusting iron), and a protective and very thin film (e.g. passivated chromium). On the outermost layer of the passive film, the AFS results (Figures 45-56 ) show the oxygen enrichment, the metal ions deficit, and more iron ions than chromium ions. The oxygen content diminishes continuously from the electroly te/oxide to the oxide/metal interface. The composition ratio 0/(Pe+Cr) curves (Figures 4 5 - 5 6 ) do not show a plateau region, which indicate that there is no region of finite thickness in which an oxide of constant composition exists. That is , the composition profiles indicate that the composi tion of the oxide film varies continuously with depth, v/hich 250 implies that the oxide may be non-stoichiometrie. The composition-depth profiles (Figures 1+5-56 ) indicate that the Cr content is low at the electrolyte/oxide interface but then rises and peaks inside the oxide; it then decreases continuously to a minimum value and rises to its value in the alloy at the oxide/alloy interface. All chromium-sputtering time curves are collected together as shown in Figure 93 for better viewing and comparator*. t U c position of the peaks changes more sensitively to pH than to The chromium content of the alloy (Figures 93-95). The most enriched chromium is observed in the layer close to the surface for acid and neutral solutions, uut it is deep inside the film Tor basic solution. Though iron oxide is the major constituent of the passive films of all three alloys. The collection of sulfur-sputtering time curves from Figures 1+5 - 56 is given in Figure 96. The details and the significance of this figure will be given in section 6.4. 6 .3 .I The Maximum Chromium Zone in the Passive Films Figures 1+5 --- 56 and 93 show that all three alloys reveal chromium enrichment within the passive film after expos ing to electrolyte solutions. However, two high chromium alloys have much higher degree enrichment in chromium than that of the low chromium alloy. The quantitative chromium atomic percentage of the most enriched chromium layer (or zone) in the film and the sputtering time required to roach that pH 10 o vn o vn ro n o 4 vn vn vn vn o Vn o r\> o vn vn Ratio ( Ratio Cr/Fe+Cr ( ) 0 * 0 o * o vn o * * o * o vn o * o vn r+ M o H- ro Figure 93* The chroniur. profiles of the passive filr.s cf re-Cr a llo y s Figure Figure Sputtering time required to reach tpe+gp ) x {in minute) 20 9^.. Required Required F+r max vFe+Cr ( . v crmu content. chromium vs ) — . 20 at sputtering tim e to to e tim sputtering N Na * IN N aS, pHIO NaoS0,O iN a IN NapS0j| pH65 IN % Cr Pli3 60 reach ru vn V■ o PO Cr Fe+Cr'max vn (in minute) vn Sputtering time required to reach ( CO o o a XI 3 CD rt- O 4 ffi 0 ct ct- CD 4 H* a o DJ CD 0 H- 4 CD a W CD H* vO Vn & xi cn 1 o 4 < to tJ 3 X Pe+Cr so pH 10 0.5 - 25,5cr 0 pH 6.5 .5Cr -l5Cr -V25.5Cr .2 o.5 pH 0 25-5Cr I5 0 r Cr 0 .0 20 Sputtering tir.e, m Figure 96. The sulfur profiles of the passive filns of Fe-Cr alloys, ^ vn layer are tabulated in Table 15. The amount of enriched chromium increases with increasing chromium content of the alloys as shown in Figure 97. The Fe-25-5Cr alloy may have more than 5 0 at Cr ( as referring to Figure 75 ) at the most enriched chromium layer. The degree of Cr enrichment of Fe- 15Cr is somewhat lower but may reach ^0 at % Cr. however, the enrichment of chromium of Fe-5Cr alloy is not very impre:j- sive in most cases, except in the IN KgSO^ solution, the enrichment of chromium is less than 10 at ',j Cr. That is, both the Fe-15Cr alloy and the Fe-25»5Cr alloy can reach high level of chromium enrichment. Furthermore, the Fe-25*5Cr alloy have a little bit broader chromium enriched zone than that of the Fe-15Cr alloy. Figure 98 demonstrates the effects of pH on the amount of chromium at the most enriched chromium layer, which is designated as (at °/» Cr) • Generally speaking, with the IH a X exception of the Fe-25»5Cr alloy in IN iigSO^ and 11/ (pH3) solutions, (at Cr) m&x decreases with raising pH. ( at % Cr)mc,„ is at higher lever in acid solutions than in TTiELX the neutral and the basic solutions. Higher (at % Cr) _ max does not necessary result in better corrosion resistance, since, as discussed in the previous section, the best corro sion resistance of Fe-Cr alloys is in the neutral pH ranges. However, under the same pH conditions higher (at Cr)max result in better corrosion resistance. Therefore, the corre- sion resistance of a metal depends on the composition of the Table 15. The chromium concentration and the required sputtering time for (-Cr- —) i*efur max Fe- 5Cr Fe-l5Cr Fe-■25.5Cr sputtering sputtering sputtering VTAPKts— at fcr» Cr 1 “fe+ur time in imi “1* 6 + 01* j-** % Cr time in minutes minutes Fe+Cr minutes 1NH SO 0.34 1 2 4 24.5 0.53 40 1 0.65 53 2 lNNa„S0, 2 4 0.12 9 2 0.49 37 2.5 0.69 58 1.5 PH 3 IN Ha£S0^ 0.12 pH 6.5 9 3 0.41 30 2.5 0.62 50 1.5 INK a-SO, 2 4 0.10 7 t pH 10 12 0.36 27.5 16 0.54 4- 12 ru vn O' iue 7 Te (at The 97- Figure at % Cr of the most enriched Cr layer 100 o k 20 60 80 0 0 V . mV s te lom^romiun u i m o r ^ alloym the vs. wle or o psiain t 600 at passivation of hours twelve see at at 20 % r n h alloy the in Cr % r i te asv film passive the in Cr) wt wt aJ. pHlO Na^JO. N I © * Na IN % Cr 1 3 otn after content pH6 pH3 60 2 ^ 8 100 rH O H r, 60 (D of-i 20 0 2 k 6 8 10 pH Figure 98, Effect of pH on the (at % Cr) in the passive films of Fe-Cr alloys after twelve hours of passivation at 600 mV . see 259 metal ( more closely depends on surface composition ) and the environment. Figures k 5 --- 56 and 93 also show a chromium depletion zone just adjacent to the metal/oxide interface for two high chromium alloys, which is not clearly shown for the he-5Cr alloy. The chromium in the chromium enrichment zone could partially come from this chromium depletion zone. 6 .3.2 The Average Chromium Content in tne Passive Films The average Cr/fFe+CrJ composition ratio in the oxide film is given by in which t is the sputtering time and t sputter to the oxide/alloy interface. The average fraction of chromium ions in the film, (at 'i Cr^ave calculated and listed in Table 16. (at Cr)ave is plotted as a function of the chromium content in alloys as given in Figure 99. The dashed line in this figure means the result that v/ould be obtained if (at Cr*)ave the passive film is equal to that in the alloy. It is seen that there is an excess of Cr in the oxide over that in the alloy for the case of acid and neutral solutions, but a deficency of Cr for the case of basic solution. The excess or the deficiency increases with increas ing alloy Cr content. The excess chromium in the film implies a loss of iron, and vice versa. This may be due to one of the Table 16, The average Cr content in the passive film of Fe-Cr alloys 5Cr l5Cr 25.5Cr Cr NAPHPe+Cr *tJ6Cp NAPHFe+Cr at%Cr NAPH|r^+cr at%Cr IN H 2 S0^ 0*12 9 0 . 2 6 1 9 0.43 31 .5 IN Na 2 S0^,PH3 0.09 7 0 . 2 8 2 0 .1+. 0 . 4 2 3 0 . 6 IN Na2 S0^,pH6.5 0 . 0 8 6 0 . 2 4 17.5 0. 4 1 30 IN Na2 S0^fpH 10 0 . 0 6 lj.,8 0.19 14 0.33 24 ro O' c O -M OJ F h > iue 9 Te vrg C cnet n h passive the in content Cr average The 99. Figure in the film 20 1+0 30 0 0 N aS, H 10 pH Na„S0, IN N Na230+(_ IN 6. pH 3 pH Na230^ IN N H2S0^ IN im s te lo C content. Cr alloy the vs. film at r C % wt wt n h alloy the in 20 r C % 26 26 1 262 following: (i) chromium passivates more easily than iron under the acid and the neutral environments,and iron passivates more rapidly in the basic solution, or (ii) the iron oxides or hydroxides are more soluble than the chromium ones in the acid and the neutral solutions. The reverse is true in the basic solution, figure 99 also shows that, in the film, (at Cr) is higher in acid solutions than in the neutral and the ave basic solution. That is, aeiJ conditions promote the chromium enrichment in the erosive film. figure 100 indicates that (at > Cr; ave decx cores v/ith raising pH, though the point for the Fe-15Cr alloy at pH 3 is higher than that at pH 0. Still, it can be seen that (at S Cr) is higher in acid solutions than that in the BJV G neutral and the basic solutions. Both (at c/j Cr) and max (at Cr)ave are higher for the cases of acid solutions than those for the cases of the neutral and the basic solutions. 6 .3.3 The Composition of the Passive Films In addition to the determination of composition- depth profiles, the AES technique can also be used to measure the changes in che binding energies of electrons. Experiment al results ( 25-?) indicate that energy shifts of Auger peaks involving valence electrons reflect more the redistribution of electrons within the valence band than core level energy shifts. The low energy Auger peaks show energy shifts Figure {at % Cr) in the film ave 1 0 20 30 0 0 100. 100. film vs pH values. vspH film h aeae r otns n the in pas contents Cr average The 2 pH 6 8 •* 10 t 26 and characterize surface chemical states, since the corres ponding transitions frequently involve valence electrcns.The representative low energy Auger spectra of the passive films of Fe-Cr alloys are given in Figures 57 - 68 . The magni tudes of the low energy peaks are tabulated in Tables 17-20. The magnitude of the Auger peak-to-peak height of those low energy peaks is in arbitrary unit for the purpose or observing the variation of each peak, the major low energy spectra (corresponding to ll0 _7V tranuLLions ) for Fe, Cr, and ^ t J and their oxides are listed in Table 21. Comparing Figures 5? - 63 to the Auger peak positions of the standards (Table 21), the compositions of the passive films of Fe-Cr alloys are not straightforward, though in some cases matches among them are observed. For the comparison, an experimental error of t 1 ev, though the reproducibility is +Q.i ev, should be considered. Before getting into details,several general thing should be noted: (1) For the low energy spectra of spinel iron-chrom ium oxide, the lowest one (at 33*5 eV) represents the chromium ions, the intermediate one (at A5-5 ) indic ates the mix of iron ions and chromium ions, and the highest one (at 52.5^7) is for iron ions. (2) During progressive sputtering, the ^5*5 eV peak is gradually moving toward the A8 eV peak ( for pure iron). In between this two extremes, the variation of the mamituue Table 17 . The peak-to-peak heights' of several low energy peaks observed in the passive films of Pe-Cr alloys after exposure to IN HoS0i for twelve hours at 600 mV — ______2 4______see guttering 13.5 ev 36 eV ^5*5->48 eV 52.5 eV 28 eV 41 eV CM (J ^ 10 Ji«« 5 C r - In in) 15Cr25C, 5C| 15Ci|25Cr 5C, 15C, 5Cr 1 5Cf 25 Cr 5Cr I S C ^ S C r 1 50, l1 1(55 TVT 0 3.8 N N 4-4 '* 1 90CJJ il N N T -Q*5 1Q4 9.5 1 # p N 17 119 1 2 2 4.4 3,7 26 N N N N T T 1 1 1 2 157 N N 20 1 2.1 12.7 2.7 1 2 2.9 N u N T T T 16lB 1.5 1 ^ 2 i5.3 N N 26 25,3 11-7 08 n4 0.9 N u ' N T - IT T 2 14,2 — 17.4 N — N 1 3 2 0.4 — 0.4 N - N T — N 2.5 i3 92 N — — 4 5 5 46.3 0.8 as N N N N 2 — 17 ^ — T — 20.0 — a? — N N 3 . 5 5> 2j0 — 65.8 — 0.9 H — N 4 i4.o 35 2.7 T 7Z1 25.7 as 0.7 N N N N 4 - 5 —* N - 5 11^ 4 1 . 1 98.4 322 0.4 05 N N N N 6 — 7 N 92 5.4 4,4 78.7 44 0.8 0.9 N N N N 8 N 2 95 0.8 t N N 9 1 10 i 11 12 N 2 Q?4 0.2 N N . 1 3 1 1 14 1 1 *5 N N 6,3 I6.7 i 78.5 63.6 0.8 0 .9 N N N N N: Negative; ND: Not Determined; T: Trace ru o w. Table 18. The peak-to-peak heights of several low energy peaks observed in the passive films of Fe-Cr alloys after exposure to IN Na^SO) , pH 3 for twelve hours at 600 mV ______see * 36 eV SpaltBrinj 13.5 eV 4-5 • 5—> i+8 eV 52.5 eV 28 eV 41 eV I"?* Ui it) 5 C f15C,J25C, 5C, 15Ci 25Cr 5Cr 1 5 c, 250, 5Cr 15Cr ?5Cr 5C, 15CJ 25C, 5C, 1 5 C r2 s q 0 05 as N N "IT m uif "TF ~ 5 F J. J. W N N N N N T 0.5 06 i\t N TT N 1 07 15 5 141 6,8 4,5 32 0.4 N N N N T 1 N — yi 1 0 ‘ 22,3 N N 21,6 19.4 15.7 7.1 4.6 2.6 0.4 NN NN T ND 10.5 2y.y ND N N 1 63 .... l.J m !7.? ND 37 2? ND TT N ND T N 2 A7 ND w t N NDN 32.5 ND 19.8 k5 ND N NDN NND N ,2.5 12$ ND NN ND 2.2 ND TT lkt ND T ND 3 20.4 N 15,0 126 % 06 N N T N NDND ND ND 3 0 ND ND 4 4.8 14,k N N 56,1 152 05 0.3 N N TN 4.5 -U I T 5 T I N 6 1 # 1.3 a? 721 26," 08 0,7 N N TT W 8 10; 1<* T 7.4 4?.2 34.6 0.7 17 N N T N N 1 0 5.6 7M U 60.5 63; 06 09 NN N N 12 2.5 2,8 71.7 14 N N ' 14 08 3# VOt 0.8 N N 16 18 20 N 13 94 05 _I N 22 L -U [ f‘ I 24 1 N N 1 83 14. € 1 79f 645 I 0.8 0.7 NN - t - N N N: Negative; ND: Not Determined; T: Trace Table 19, The peak-to-peak heights of several low energy peaks observed in the passive films of Pe-Cr alloys after exposure to IN NaoS0. , pH 6 . 5 Tor twelve hours at 600 mV • ______<— j see fyuttiriitg 33.5 eV 36 eV 4-5.5**U8 eV 52.5 eV 28 eV 4-1 eV lam) 5 C rJ15Cr2 5 0 5Cr15Ci ]25Ci 5 Q 15C, 25Cr 5Cr 15Cr25 C, 5C, 15C, 25<^ 5C, 15C, 25CV 0 07 01 cl£ N N N 110 Z Q lio T ?1TT ■■ iL ±-Li hr N n N N N N 0.5 04 - 3 2 11.3 N N N 524 24.1 20.1 227 35 01 N N N N N N 1 — tXi 18 — N N — 24.8 2318 2.7 __ NN N N 1 . 5 04 66 20.3 NH N 5&2 18,5 19,5 20,7 70 0,3 07 NN N N N 2 2.4 ND ND NND ND 60.6 ND ND 19J6 ND ND 05 ND NDN ND ND _ 2.5 31 1 U 21,3 -X H H 63,7 18^7 19 17,5 16 03 0.5 NN N T N 3 « NN 76^ 20,1 76 U •N N I N T N Ni) 3 - 3 9,2 ND 9\5 ND 17 ND N ND 4 — N ND 4 f m 20 N N T 116? 27,4 30.8 0.7 05 09 N N N N T N 4 . 5 Nl) 1 5k ND 05 NT?" \ kyi^J** 1ND op wn ND N ND N 5 NDN ND 40.0 NET 07 ND N ND N ND 6 3.? T 4 $k$ 5? U 1 r '"IT N N 8 T 22 63a °t? N N 10 i 12 14 16 18 20 N 1*3 i; N N 1 30 I N N 7.1 1 14.5 708 63.8 1 15 05 N N N N N: Negative; ND: Not Determined; T: Trace Table 20, The peak-to-peak heights of several low energy peaks observed in the passive films of Fe-Cr alloys after exposure to IN Na^SO, , pH 10 for twelve hours at 600 mV * _ * - 4 see pottering 13.5 eV 36 eV +5.5-*48 ev 5 2 .5 eV 28 eV 41 eV Tinu Sin i n 1 5 Ci- 15Cr25CJ 5Cr 15Ci 25Cr 5Q 15c, |25Cr 5Cr 15Cr 25 Cr 5C| 15CJ25Q 5C, 1 s c b s q 0 IT" N N N "IP N QT 4? '/TT 4o -Jfr. ‘v- H'.y - 3 ? 22 N N r n N N N 0 .5 ND N qo4 ND N NIX N ND 7.8 10.5 4.8 59 ND N T ND N N l T a3 N | N 1^,8 131 &3 t I 6,6 N T N ,-i*5 T — H — 11-4 — 56 N N 2 T N u N N 17,4 12.4 145 7r9 Lp T N TJ N N N U05 — N 2 .5 17.7 — 7,9 T N ND 3 q- ND ...W NR 1 4fi NT) hs> NT) N ND N 3 .5 U05 — N 17.1 72 T N 4 N 1 frl 3.7 N N 4 .5 1 5 0.0 ^ 182 N 25 T N ""h" 7 U V ff N N 1<1? l f a 14.3 1.7 TP T" N IP PT T ' $ 19 OS N N $L N 2 £$16.9 136 ^4 I™! li) T N N N N T 11 N N to 16,6 2,4 y? m 27.3 132 1.8 1.0 0.4 T N N N N T 13 4,0 12.0 N N 16,4 21,3 0.9 9.3 N N m N - —1 15 n ? N t QO N: Negative; ND: Not Determined; T: Trace 269 Table 21. The major low energy spectra of Fe, Cr, and their oxides. Peak positions Specimen Reference in eV Fe 1+8 233 y -FeOOH 1*4.7 52.5 253 a -Fe^jO^ 1+1+.5 52.5 253 y -Fe2 0^ 43.0 51.0 253 51.0 253 Pe3 \ 46.5 5i.o 253 F e x ° 46.5 Cr 36 233 Cr oxide 32.8 45.4 251 ; Cr2 °3 31.5 1+7.0 Present CrCOH)^ 32.0 1+6.5 Present Present Fei . 5 Cro.5°3 33.5 1+55 52.5 270 due to the increase of chromium ions and iron ions, or the increase of chromium ions and the decrease of iron ions, or the decrease of chromium ions and iron ions and the increase of pure iron atoms. Whichever is the case can be distinguish ed by observing the variations of 33«5 and 52.5 G'^ peaks. (3) The electron energy at 41 eV does not correspond to any well known Auger transition of iron oxide or chromium oxide. However, Seo,.Lumsden and Staehle's results (253) have shown a shoulder at 4 4 . 5 for Fe^O^ and Fex0 but not for other iron oxides. Here, the presence of 41 eV peak ( or just as a shoulder ) probably means the dispersion of Fe 0 in that A layer, since chromium has stronger oxygen affinity than iron. Then, the formation of Fe 0 is more likely than the formation of Fe3<^. (4) A small peak at 28 eV appeared in several cases: the Fe-5Cr alloy in IN Na230^ ( pH 3, 6.5# 10 ) solutions and the Fe-15Cr and the Fe-25Cr alloys in IN Na^SO^ ( pH 10 ) solution. It seems that this Auger peak is associated with an iron oxide since this peak can be present without the presence of chromium oxide (Figure 54)* Which iron oxide can produce this 28 eV Auger peak is not known. This 28 eV Auger peak has been skipped from the Auger Spectra of iron oxides in the literature. (5) In the final steps of sputtering, only pure iron, pure chromium and small amount of FeOOh (141,253) (at 53-5 e^) have been left. The FeOOH can penetrate quite deep into all 271 the three tested alloys. Wo attempt is made to remove all of the FeOOH. The results of composition analysis of the passive film on Fe-Cr alloys show that the passive film is not uniform but is composed of several different kinds of compounds. Generally» the composition of the passive film of Fe-Cr alloys can be classified into two types according; to the pH of the solution rather than the chromium content of the alloyi (1) In the acid and the neutral solutions ( Figures 57- 65 and Tables 17-19)» the outermost layers are made up of the spinel iron-chromium oxides. The composition may not be exactly the same as Fe-^ ^CrQ c®j* rather something like one of i;'ei+xCri_x0-5* The layers of spinel iron-chromium oxides are very thin (about 4-8 R in thickness). The inner layers are chromium oxide. The chromium oxide may be non- stoichiometric. The thickness of chromium oxide is about 20 S. A scheme for this case is given in Figure 101a. (2) In the basic solution (Figures 66 - 68 and Table 20), the outermost layers are iron oxide. The thickness of the iron oxide is about 2-10 R and it decreases with the increas ing chromium content of the alloy. The intermediate layers are one of the spinel iron-chromium oxides, and its thickness is about 10-15 The inner layers are chromium oxide and its thickness is about 25-40 fi. The arrangement of those compounds is shown in Figure 101b. 272 (Fe-Cr Joxide Cr-oxide Fe-oxide Matrix (a) For Fe-Cr alloys in the neutral and the acid solutions. Fe-oxide (Fe-CrJoxide Cr* oxide F e •oxide | Matrix I II * (b) For Fe-Cr alloys in the basic solutions. Figure 101, The composition of the passive films on Fe-Cr alloys. 273 In the following, the composition of the passive films of the alloys for each cases will be examined. Figures 57 - 59 and Table 17 indicated that, in IN solution increased or grew the chromium ion peak (at 33*5 e V ) of all three alloys during progressive sputtering from the outermost surface. The first couple layers, this 33*5 e'/ peak together with the 45*5 eV peaks characterized the spinel iron-chromium oxides. After the 52*5 peak became very small and the 45,5 e/ peak moves to about 48 eV, those inner layers characterized the chromium oxides. Further sputtering exposed that there was an inner zone in the passive film of the Fe-15Cr and the Fe-25Cr alloys containing chromium oxide, pure chromium and pure iron. This kind of zone was not observed in the passive film of the Fe-5Cr alloy. In the final steps of sputtering,only pure iron, pure chromium and small amounts of FeOOH had been left. The general pattern of the composition for the alloys in IN HgSOj^ as discussed in the previous paragraph also holds true for the alloys in IN Na^SO^ (pH 3 and 6 .5 ) solutions (Figures 60 - 65 and Tables 18-19)* However, two more things can be addedt (1) The chromium enrichment in the film of the Fe-5Cr alloy in IN NagSO^ (pH 3 and 6 .5 ) solutions is not as prono unced as that in IN HgSO^ sdution. (2) Figures 60 - 65 and Table 19 indicate that the 52.5 e'/ peak of the Fe-5Cr alloy increases quite a bit before 27i^ diminishing. On the other hand the 52*5 eV peak of the Fe- 15Cr and the Fe-25Cr alloys decreases quickly without any subsequent rise. This signifies the competition between the iron oxide and the chromium oxide for the alloy serving in the neutral solution! for the low-Cr alloys, the iron oxide is the major constitutent in the film of the alloy; for the high-Cr alloys, the chromium oxide is the major constitutent in the film of the alloy. Figures 66 - 68 and Table 20 show that the outermost layer of the passive film for the Fe-Cr alloys in the basic solution is principally iron oxide with an unnoticeable amount of chromium oxide. The iron oxide is most likely the (37, 253). As sputtering advanced to expose the inner layers of the passive film, it shows an increasingly strong iron oxide signal (at 52.5 eV) and a gradually increased chromium oxide peak (at 33*5 eV), When both iron oxide and chromium oxide signals are strong enough, we can say that the spinel iron-chromium oxide is the compound in those layers of the passive film. Further sputtering results in the diminishment of the iron oxide peak, and chromium oxide remains in the passive film. In addition, for the Fe-15Cr and the Fe-25Cr alloys, there is a zone in the passive film showing the mix of chromium oxide, pure chromium and pure iron. It takes longer sputtering time to remove all the oxide formed in the basic solution as compared to that required to remove the passive films formed in the acid and the neutial solu 275 tions. Generally, the passive film is thicker if the major constituent is iron oxides in stead of chromium oxides. 6.3*^ Summary 1. All alloys have an enriched chromium zone within the passive film after exposure to electrolytes. The chromium content of the most enriched chromium layer increases with increasing chromium content in the alloys. 2. For the same alloy, the alloy has the highest chromium content in the most enriched chromium layer in the acid slou- tions and the lowest in the basic solution. 3. The most enriched chromium layer is very close to the surface of the alloy in the acid and the neutral solutions, but this layer is deep inside the passive film of the alloy for the basic solution. Both the Fe-15Cr and the Fe-2jCr have high level of chromium enrichment, however, in addition the Fe-25Cr alloy has a broader chromium enriched zone than the Fe-15Cr alloy. 5- For the acid and the neutral solutions, the average chromium content in the film is higher than the chromium content on the alloy, but for the basic solution, the reverse is true. 6, The compositions of the passive films on the alloys, after exposure to the acid and the neutral solutions, arc quite similar* the outermost layers are composed of iron- chromium oxide and the underlying layers are the chromium 276 oxide. The compositions of the passive film on the alloys, after exposure to the basic solution, are: the outermost lay ers are relatively thick iron oxide and the middle layers ara iron-chromium oxide, and the inner layers are chromium oxide. 6 .h The Sulfate Ion Effect The electrochemical behavior of a metal is influenced by the electrolyte in which it is exposed. However, the anion effect is not wholly clear. AES results (Figures 1+5- 56 ) indicate that sulfur is engaged into the passive films of Fe-Cr alloys. Pure iron and the Fe-5Cr alloy have very high current densities at the critical potential of passivatio.i In all the sulfate solutions (Figures 37-1+0), including the neutral and the basic solutions. Florianovich, et al. (255) postulate that, in sulfate solutions, the sulfate ions directly take part in the process es of iron dissolution. Gibbs and Cohen (256) and Smialowska, et al. (25?) have found that the anodic films formed on iron contain iron-oxy-sulfate, an<* F-FeOOH. Present AES results are in accordance with those results. Smialowska, et al. (2 5 7 ) have suggested that the presence of sulfate ions in the passive film reduces its protective ability. Several points for this behavior can be drawn from the AES results (Figures 1+5-56.) as the follows: 1. In most cases, the sulfur content decreases from the surface of the passive film with increasing sputtering time. 2?7 In the neutral and the basic sulfate solutions, the sulfur content of the Fe-5Cr alloy increases from the surface, peaks inside the passive film, and then decreases. The sulfur content is roughly proportional to the oxygen content. Therefore, the sulfur is rather in the form of sulfate ion than in the form of sulfide ion. 2. In acid solutions, the sulfur content increases with the increasing chromium content of the alloys; in the neutral and the basic solutions, the reversed tendency is true. The relatively high and increasing sulfur content with increasing chromium content of the alloys in acid solutions suggest that chromium sulfate may be one of the components of the passive film. It seems that the increasing chromium enrichment in the passive film can overcome the degradation of the film due to the increasing sulfur content with increasing chromium content of the alloys. 3. In acid solutions, the high sulfur content in the passive film of the alloys will lead to poor protection and stability 4. In the neutral solution, the films on two high-Cr alloys exhibit little evidence of sulfur; they are the alloys with the most protective films. FINAL CONCLOSION3 1. Some correlations can be found between the electro chemical and the Auger analysis results* (i) Taking into consideration the thermodynamic data and polarization curve measurements for Fe-Cr alloys, it can be predicted that metastable chromium oxide in acid solutions and stable chromium oxide in neutral solution should form first at more active anodic potentials than iron oxides. Therefore, there should be chromium enrichment within the passive film of the alloys. The AE3 results support this point of view. (iij The polarization curve and the current decay curve measurements imply that, in the acid and the neutral solutions, the compositions of the passive films of the studied alloys have similar properties after a long term of passivation. The properties of the Fe-15Cr and the Fe-25Cr alloys are especial ly close. Hence, it should be predicted that the compositions of the passive films of those alloys are similar. The AEo results are in agreement with these results. The properties of the alloys in basic solutions are also very similar (based on the measurements of current decay curve in the passive region of potentials) in spite of the fact that the composition of the passive films found by the Auger analysis is different.276 279 2. In the basic solutions but with pH less than 12.6, electrochemical results cannot really predict chromium enrichment or iron enrichment in the passive film of the alloys, since both chromium oxide and iron oxide are stable. The A2S results show that the average chromium content in the passive film is lower than that in the alloy. But there is an enriched chromium zone deep inside the film on the surface of the alloy. The average iron content in the film is higher than that in the alloy. Therefore, it seems that iron oxide is more stable than chromium oxide in basic solutions. 3 . The comparison of i vs. chromium content in alloys (i is the minimum current density taken from the measure- P ments of polarization curves) with i vs. chromium content p12 (i is the current density obtained from the current p12 decay measurements after twelve hours of passivation at constant potential, 600 mV ) show that the rapid decrease SCc of the current at 15 wt Cr was observed in the first case but only small current decrease was observed in the second case. At the same pH, all i 1s of the alloys are almost p12 of the same order of magnitude, but there is about one to two order of magnitude difference between i of the r'e-5Cr P alloy and the Fe-25Cr alloy. These results show that increasing the chromium content in the alloy increases the passivation rate of the alloy. Therefore, there is a large difference in the current densities in the early stage of 280 the passivation for alloys with low and high chromium contents. However, after the stable passive film is formed, the difference of the passive current densities for alloys of various concentrations of chromium decreases. This characteristic is true in all tested solutions. It seems that, in the pH 10 solution, the dissolution rate of the chromium oxide is greater than that of the iron oxide, and this results in a loss of chromium in the passive film of the alloy. In the acid and the neutral solutions, the dissolution rate of the chromium oxide is less than that of the Iron oxide. 4. The AES results show that the composition of the passive film of the Fe-Cr alloys can be classified into two types* (i) in the acid and the neutral solutions, the very thin outermost layers are the spinel iron-chromium oxides and the inner layers are the chromium oxides; (ii) in the basic solution, the relatively thick outermost layers are the iron oxides, the middle layers are the spinel iron- chromium oxides, and the inner layers are the chromium oxides. The AES results also indicate that those compounds nay be non-stoichiometric• 5. Both the Fe-15Cr and the Fe-25Cr alloys have a zone within the passive film with high level of chromium content, but the Fe-25Cr alloy also has a little bit broader chromium enrichment zone in the passive film than that of the Fe-15Cr alloy. 281 6. Higher (at Cr)__„, which is the maximum chromium lu a X content in the passive film, of an alloy does not necessary result in better corrosion resistance of the alloy than the alloy with lower (at '/j Cr) a , The corrosion properties of ITlaX an alloy depend very much upon the pH of the solution, for example, the alloy has higher (at ^ Cr) after exposure max to the acid solutions than to the neutral solution, but the alloy has a lower corrosion rate in the neutral solution (Figure 102). That is, even if an alloy has a passive film with higher chromium content, its corrosion resistance could be lower in acid solutions than the same alloy with lower chromium content in the passive film in the neutral solution. 7. As follows from the Auger analysis, the average chromium content in the passive film is higher in acid solutions than in the neutral and the basic solutions. However, the corrosion resistance of the Fe-Cr alloys is best in neutral pH ranges (Figure 103). Therefore, it seems that the dissolution rate of the passive film is one of the primary factors in the corrosion processes of the alloys. 8. The AES results show that sulfur is incorporated into the passive film of the alloys. For the same alloy the sulfur content in the passive film decreases with raising pH. 9 . In the acid solutions, the sulfur content in the passive film increases with increasing chromium content in the alloy. It is generally agreed that the incorporation of anions into the passive film loweres the quality of the Current density, ar.p/cm' 10 10 Figure Figure -8 -6 0 102 Te v (t C) . Cr) ^ (at vs i The . 15 - max P-~ at at pH10 pH3 pH 6.5 % Cr H 0 pH aS, pH6.5 NapS0, N I ■ N a3, pH3 NaP30, * IN • IN IN NapS0i pH] pH] 0 NapS0i IN 60 Current density, amp/cm 10 10 Figure Figure -6 8 0 103 Te v. (at vs. i The . 10 D 4 *v © v e* o pH 3 pH10 at % 20 Cr pHO IN Na IN 6.5 pH Na^SO, IN % r . Cr) 30 H 10 pH 3 pH ko 283 2814- film. The presence of sulfate ions in the film may be an additional reason for less protective film formation in acid solutions. In the neutral solution, the sulfur content in the passive film decreases with increasing chromium content in the alloy. It seems that the chromium enrichment in the film and small incorporation of sulfate ions in the film give the best quality of the passive film. In the basic solution, the sulfur content in the passive film decreases with increasing chromium content in the alloy, however, the composition of the passive film on the alloy is different! iron oxides are the major constituent of the passive film. Alloying chromium with iron in this case improves only slightly the corrosion resistance of the Fe-Cr alloys• REFERENCE 1. J. Keir, Phil. Trans., 80, 3T2+ (1790). 2. C. T. Schoenbein, Pogg. Ann., 590 (1836); 4k, 70 (1838). 3. M. Faraday, "Experimental Researches in Electricity, Vol. II", University of London, (1844.). I4.. W. H. J. Vernon, "Metallic Corrosion and Conservation" The Conservation of Natural Resources, In3t. of Civil Engineers, London, p. 105 (1957). 5. L. Colombier and J. Hochann, "Stainless and Heat Resis ting Steels", p. 4-* English edn., St. Martin's Press, New York (1968). 6. C. A. Zapffe, "Stainless Steels", Am. Soc. Metals (1949). 7. L. Guillet, Rev. Met., 1, 155 (1904-); 2, 350 (1905), 2, 372 (1906). 8 . P. Monnartz, Metallurgie, 8 , 161 (1911). 9. M. G. Fontana and N. D. Greene, "Corrosion Engineering", p. 16, McGraw-Hill Co., New York, (1967). 10. "Corrosion Handbook", edited by H. H. Uhlig, p. 20, Wiley, New York, (194-8). 11. C. Wagner, Corros. Sci., 751 (1965). 12. C. Wagner, "Discussions at 1st Int. Symp. Passivity", Heiligenberg, West Germany, (1957). 286 13. M. J, Pryor and R. W. Staehle, in "Treatise on Solid State Chemistry", Vol. i|., Edited by N. S. Hannay, p. i|_8l, Plenum Press, New York, (1976). lif. N, D. Tomashov and G. P. Chernova, "Passivity and Protection of Metals Against Corrosion", p. 10, Plemum Press, New York, English ed., (1967). 15. E. 3. Hedges, J. Chem. Soc., p. 975 (1928). 16. N. D. Tomashov, "Theory of Corrosion and Protection of Metals", p. 325, MacMillan, New York, English ed., (1966). 17 . K. G. Adamson in "Corrosion", Vol. 1, Edited, by L. L. Shreir, p. Newnes-Butterworths, London (1976). 18. "Corrosion Tests on Metallic Molybdenum", Engineering Experimental Station, The Ohio State University, Project lij.2 (1957-1959). 19. C. R. Bishop, Corrosion, 1^, 308t (1963). 20. J. O'M. Bockris and A. K. N. Reddy, "Modern Electro chemistry", Plenum Press, New York, (1970). 21. J. P. Hoare, "The Electrochemistry of Osygen", Inter science, New York, (1962). 22. U. R. Evans, "Collection of Reports of the International Coloquium on Passivity of Metals, I", Darmstadt, (1957). 2 3 . H. H. Uhlig, "Corrosion and Corrosion Control", p. 60, Wiley, New York, 2nd ed,, (1971). 2Lp. P. A. Cotton and. G. Wilkinson, "Advanced Inorganic Chemistry", p. 831# Wiley, New York, 3rd ed., (1972). 2 8 7 2^. T. P. Hoar, in "Corrosion", Vol. 1, Edited by L. L, Shreir, p. 1:114, Newnes-Butterworths, London (1978). 26. U. R. Evans, "Metallic Corrosion, Passivity and Protection", p. 4* Edward Arnold, London (1937). 27. L. Tronstad, Trsns, Farad, Soc., 29, 502 (1933). 28. E. S. Hedges, "Protective Films or. Metals", p. 112, D. Van Nostrand,New York (1937). 29. S. Glasstone, "Textbook of Phyical Chemistry", p. 1011, D. Van Nostrand, New York (1940). 30. R. B. Mears, Trans. Electrochem. Soc., (1940). 31. K. F. Bonhoeffer, Z, Elecktrochem., 55, 151 (1951); Z. Metallk., 77 (1953). 32. U.F. Franck, Z. Electrochem., 154- (1951); 62* 614-9 (1958). 33* K. J. Vetter, Z. Electrochem., ^8, 230 (1951+)* 64* 642 (1958); Z. Phys. Chem. N. F. I±, 165 (1955). 31+. U. R. Evans, J. Chem. Soc., p. 1021; (1927). 35. N. Cabrera and N. F. Mott, Rep. Prog. Phys., 163 (1948). 3 6 . R. Ghez, J. Chem. Phys., £8, 1838 (1973). 37. N. Sato and M. Cohen, J. Electrochem. Soc., Ill, 512 (1964). 30. F. P. Fehlnes and N. F. Mott, Oxidation of Metals, 2, 59 (1970). 39. C. Lukac, J. B. Lumsden, S. Smialowska, and R. W. Staehle, J. Electrochem, Soc., 122, 1571 (1972). 2 8 8 40. K. G. Weil, Z. Electrochem., £9, 711 (1955). 14-1 . K. J. Vetter, "Electrochemical Kinetics Theoretical and Experimental Aspects", English Edition, Academic Press, New York (1967). 42. K. J. Vetter, J. Electrochem. Soc., 110, 597 (1963). 43. N. Nagayama and K. Cohen, J. Electrochem. Soc., 109, 781 (1962). 44. N. Nagayama and M. Cohen, ibid., 110, 670 (1963). 4 5 . N. Sato and M. Cohen, ibid., Ill, 512 (1964); 111, 519 (1964). 46. H. Gohr and E. Lange, Naturwissenschaften, 12 (1956). Z. Elektrochem. Ber. Bunsenges. Physik Chem. 6l_t 1242 (1957). 47. T. II. Geballe, B. T. Matthias, V. B. Compton, E. Corenzwit, G. W. Hull, Jr. and L. D. Longinotti, Phys. Rev., l^Z* 119 (1965). 48. Schrader and Euttner, Z. anorg. allgenu Chemie. 320, 205 (1963). 4 9 . J. s. L. Leach, Surf. Sci., 25? (1975). 50. H. H. Uhlig, Ann. N.Y. Acad. Sci., 5 8 , 843 (1954). 51. H. H. Uhlig, Z. Elektrochem., 62, 626 (1958). 52. H. H. Uhlig and P. P. King, J. Electrochem. Soc., 106. 1 (1959). 53. T. P. Hoar, Corros. Sci., 1, 341 (1967). 54. M, Pleischmann and H. R. Thirsk, J. Electrochem, Soc., 110, 688 (1963). 289 55* U. R. Evans, Z. Elektrochem., 62, 619 (1958). 56. M. J. Pryor, J. Electrochem. Soc., 106, 557 (1959). 57* K. P. Bonhoeffer and U. P. Franck, Z. Elektrochem., 55. 180 (1951). 58. H. Remy, "Treatise on Inorganic Chemistry, Vol. II," P. 750, (Translated by J. S. Anderson), Elsevier, New York (1956). 59. I. Langmuir, J. Am. Chem. Soc., 2H» 2273 (1916). 60. I. Langmuir, Trans, Electrochem. Soc., 29, 260 (1916). 61. I. Langmuir, J. Chem. Soc., 62, 517 (191+0). 62. J. N. Stranski, Z. Elektrochem., 1^, 393 (1929); i^6 , 25 (1930). 63. B. W. Erschler, Doklady Acad. Nauk, 258 (191+2); 2 1 262 (19i+2). 61+. Ya. M. Kolotyrkin and V. M. Kayasheva, J. Phys. Chem. U. S. S. R., JU, 1990 (1956). 65* Ya. M. Kolotyrkin, Z. Elektrochem., 62, 66!+ (1958). 66 . V. V. Gerasimov, Zh. P. Khimii, 22* 1 0 9 (196!+). 67 . G. Tammann, Z. anorg. Chem., 107. 10!+ (1919); 107. 236 (1919). 6 8 . M. G. Fontana, Corrosion 2* 5 ^ 7 (19!+7)* 69. H. H. Uhlig and. S. Lord, J. Electrochem. Soc., 100, 216 (1953). 70. H. H. Uhlig, in "Proc. Third Internet. Cong, on Metal. Corro.", Vol. 1, p. 25, MIR, Moscow (1969); Corros. sci., x, 325 (1967). 290 71. B. V. Ershler, Proceedings of the 2nd Conference on Corros. Of Metals at U. S. S. R. Academy of Sciences, II (19J+3). 72. N. D. Tomashov and R. M. Alftovskii, "Collection: Corrosion of Metals and Alloys", p. li+l, Metallurgizdc, Moscow (1963). 73. T. P. Hoar, Corros. Sci., ]_, 3J4.I (1967). 7b • H. H. Uhlig, "Corrosion and Corrosion Control", p. l8Jj., Wiley, New York, 2nd ed., (1971). 75. L. V. Wanyukowa and B. N. Kabanov, Zhur. Fiz. Khim., 28, 1025 (19514-J - 76. N. Hackerman, Z. Elektrochem., 62, 632 (1958). 77. W. A. Mueller, J. Electrochem. Soc., 107, 157 (I960). 7 8 . V. V. Andreeva, Corrosion, 20, 35t (1961).). 79. R. P. Frankenthal, J. Electrochem. Soc., 116, 580 (1969). 80. F. Finklestein, Z, Phys. Chem., ^1, 91 (1902). 81. A. Smits and A. Aten, Z. Phys. Chem., £0, lj.6 (1915). 82. A. Smits, "Theory of Allotrop", Longmans, Green and Co., New York (1922). 8 3 . A. Smits, Trans. Faraday Soc., 1£, 772 (192k). 81).. E. Grave, Z. Phys. Chem., JJ, 513 (1911). 85. G, Schmidt and W. Rathert, Trans. Faraday Soc., 257 (1934). 86. E. Muller and V. Cupr, Z. Elektrochem., l^, 1+2 (1937). 87. M. LeBlan, "Lehre der Elektrochemie", 6th ed., p. 299> Leipzig (19114-). 291 8 8 . 0. Sachur, Z. Elektrochem., _10, 8J4.I (19^0); 12, 637 (1906). 89. P. Foerster, "Elektrochemio w&ssrige Losungen", Leipzig (1922). 90. A. Russell, Nature, ll£, 455 (1925); 111, 47 (1926). 91. A. Russell, D. Evans, and S. Rowell, J. Chem. Soc., p. 1877, (1926). 92. R. Swinne, Z. Elektrochem., _^1, 4^2 (1925). 93. U. Sborgi, Atti Lincei (6 ), 1, 315 (1925); 1, 388 (1925). 94. U. Scorgi, Gazz., £ 6 , 532 (1926). 95. H. H. Uhlig and J. Wulff, Trans. AIME, 135. 494 (1939). 96. H. H. Uhlig, Trans. Electrochem. Soc., 85, 307 (1944). 97. H. H. Uhlig, in "U.S.-Japan Seminar on Passivity and its Breakdown on Iron and Iron Base Alloys", RACE, Houston, Tx. , p. 21 (1976). 98. R. F. North and M. J. Pryor, Corrosion, 10, 297 (1970). 99. U. R. Evans, Trans. Faraday Soc., 42* 1^9 (1944)* 100. U. R. Evans, Electrochim. Acta, 16, 1825 (1971). 101. P. Landau, Trans. Flectrochem. Soc., Gl, 559 (1942). 10,7. N. D. Tomashov, in "Proceedings of Third Interna. Cong, on Metallic Corrosion, Moscow, 1966", Vol. 1, p. 37 Moscow (1969). 103. G. Tammann, Z. anorg. Chem., 169. 151 (1928). 104. G. Tammann, Stahl u. Eisen, 577 (1922). 105. H. Remy, "Treatise on Inorganic Chemistry", Vol. II, Chapters 5 and 7* (Translated by J. S. Anderson), 292 Alsevier, New York (1958). 106. D. Nichools, in "Comprehensive Inorganic Chemistry", Vol. 2l(., Chapter 6.0, Pergamon Press, New ¥ork (1973). 107. "Metals Handbook", Vol. 1, 8th ed., p. 1200 and p. 1206, ASM, Metals Park, Ohio (1961), 108. U. K. Evans, Quart. Revs. (London), 21, 29 (1967). 109. C. L. Rollinson, in "Comprehensive Inorganic Chemistry", 110. "Metals Handbook", Vol. 8, 8th edn., p. 291, ASH, Metals Park, Ohio (1973). 111. H. W. Paxton and T. Kunitake, Trans. AtI!E, 218, 1003 (I960). 112. A. J. deBethune and N. A. S. Loud, "Standard Aqueous Electrode Pltentials and Temperature Coefficients at 25°C", Clifford A. Hampel, Skokie, 111. (196i±). 113. M. G. Fontana and N. D. Greene, "Corrosion Engineering", p. 32, McGraw-Hill Co., New York (1967). 116-. Ya. M. Kolotyrkin, in "Proc. 1st Int. Cong. Met. Corros., London, 1961", p. 10, Butterworths, London (1963). 115- G. Aronowitz and N. Hackerman, J. Elektrochem. Soc., 110. 633 (1963). 116. J. Kruger and J. P. Calvert, J. Electrochem. Soc., 11U. 6.3 (1967). 117. W. Whitman and E. Chappell, Ind. Eng. Chem., Ij3, 533 (1926). 118. H. H. Uhlig, N. Carr, and P. Schneider, Trans. Electro chem. Soc., 21* 111 (196-1). 293 119. U. F. Franck. Z. Naturforschung., )±A» 378 (19if9). 120. H. Rocha and G. Lennartz, Arch. Eisenhuhuttenw., 26, 117 (1955). 121. P. F. King and H. H. Uhlig, J. Phys. Chem., 6^, 2026 (1959). 122. R. Olivier, in "Int. Committee Electrochem. Therm. Kinet. 6th Meeting, Poitier, 195V> Rutterworths, London (1955). 123. R. Olivier, Dissertation, Leiden (1955). i2i|. A. M. Sukhotin, E. I. Antonovskaya, and A. A. Pozdeeva, Russ. J. Phys. Chem., ^6, 1284 (1962). 125. R. L. Beaschamp, Dissertation, The Ohio State University (1966). 126. B. Lovrecek and J. Sefaja, Electrochim. Acta, 17., 1151 (1972). 127. M. Romand, et al., J. Elec. Spec, and Related Pheno., 11, 185 (1977). 128. L. Colombier and J. Hochmann, "Stainless and Heat Resisting Steels", p. 254, English edn., St. Martin's Press, New York (1968). 129. U. R. Evans, J. Chem. Soc., p. 1022 (1927). 130. U. R. Evans and J. Stockdale, J. Chem. Soc., p. 2658 (1929). 131. W. H. J. Vernon, F. Wormwell, and T. J. llurse, J. Iron Steel Inst., 150. 81 (1944)* 291+ 132. N. A. Nielsen, E. M. Mahla, and T. N. Rhodin, Trans. Electrochem. Soc., 89, 167 (191+6). 133. Y. N. Rhodin, Corrosion, 12, l+65t (1956). 131+. Th. Heumann and W. Rosener, Z. Elektrochem., 59, 722 (1955). 135. Th. Heumann and H. S. Panesar, J. Electrochem. Soc., 110, 628 (1963). 136. H. Weidinger and E. Lange, Z. Elektrochem., 61+, 1+68 (1960). 137. M. Prazak and V. Prazak, Z. Elektrochem., 62, 739 (1958). 138. H. H. Uhlig, Z. Elektrochem., 62, ?00 (1958). 139. P. E. King and H. H. Uhlig, J. Phys. Chem., 6jl, 2026 (1958). ll+O. M. Okuyama, et al. , in "5th Int. Cong, on Metall. Corrosion, Tokyo, 1972', p. 197t NACE, Houston, Tx. (1971+) - 11+1. I. Olefjord and H. Fischmeister, Corros. Sci., 1^, 697 (1975). 11+2. J. B. Lumsden and R. W. Staehle, Scripts Met., 6, 1205 (1972). llj.3. A. E. Yaniv, J. B. Lumsden and R. W. Staehle, in "U.S.-Japan Seminar on Passivity and its Breakdown on Iron and Iron Base Alloys", p. 72, HACE, Houston (1976). 1U+. J. B. Lumsden and R. W. Staehle, ASTM STP 596, p. 39, Amer. Soc. For Testing and Materials (1976). 295 145. G. Okamoto, Corros, Sci,, 6, 471 (1973). 146. K. Asami, et al., Corros. Sci., Ij3, 909 (1970); 18. 151 (1978). 147. K. Hashimoto, et al., in "U.S.-Japan Seminar on Passivity and its Breakdown on Iron and Iron Base Alloys", NACE, Houston, Tx,, p. 34 (1970). 148. M. Pourbaix, "Atlas of Electrochemical Equlibria in Aqueous Solutions", English Edition, NACE, Houston, Tx. (1966). 149. R. P. Prankenthal, in "U.S.-Japan Seminar on Passivity and its Breakdown on Iron and Iron Base Alloys", p. 10 NACE, Houston, Tx. (1976). 150. H. J. Yearian, J. M. Kortright, and R. H. Langenheim, J. Chem. Phys., 22, 1196 (1954)- 151. R. H. Newnham and Y. M. deHaan, Z. Krist., 117. 235 (1962). 152. deWolff, Techn. Phys. Dienst, Delft, Holland. 153. Christensen and Jensen, Acta Choir,. Scand., 21_, 1 (1967); 21, 121 (1967). 154. G. L. McBee and J. Kruger, Electrochim. Acta, 1 J ,, 1337 (1972). 155. T. P. Hoar, J. Electrochem. Soc., 117. 17c (1970). 156. T. N. Rhodin, Corrosion, 12, 123t (1956). 157. A. B. Winterhottom, J. Scient. Instrum., lJt» 203 (1937). 158. Th. Heumann and F. W. Dickolter, Z. Elektrochem., 62. 745 (1958). 296 159. V.V. Andreeva, Corrosion, 20, 35t (1964). 160. M. A. Genshaw and R. S. Sirohi, J, Electrochem. Soc., 118. 1558 (1971). 161. G. M. Bulmaw and A. C. C. Tseunir, Corros. Sci., 12, 415 (1972); lj, 531 (1973). 162. K. N. Goswami and R. W. Staehle, Electrochim. Acta, 16, 1895 (1974). 163. V/. J. Mueller, 2. Elektrochem, ££, 4°1 (1927). I6J4.. A. K. N. Reddy, M. G. B. Rao and J. O'M. Bockris, J. Chem. Phys., 42, 2246 (1965). 165- J. O'M. Bockris, A. K. N. Reddy and M. G. B. Rao, J. Electrochem. Soc., 113. 1133 (1966). 166. M. Fleischmann and H. R. Thirsk, Frans. Faraday Soc., £1, 71 (1955). 167. A. Berwick, M. Fleischmann and H. R. Thirsk, Trans. Faraday Soc., £8, 2200 (1962). 168. M. Fleischmann and J. A. Harrison, J. Electroanalyt. Chem., 12, 183 (1966). 169. M. Fleischmann and J. A. Harrison, Electrochim. Acta, 11, 749 (1966). 170. R. D. Armstrong, J. A. Harrison and H. R. Thirsk, Corros. Sci., 10, 679 (1970). 171. N. F. Mott, Trans. Faraday Soc., ££, 1175 (1939); 36. 472 (1940). 172. T, P. Hoar and N. F. Mott, J. Physics Chem. Solids, £, 97 (1959). 29? 173. K. Hauffe and B. Ilschner, Z. Elektrochem., ^8, 382 (1954). 1 - J. P. Dewald, J. Electrochem. Soc., 102. 1 (1955). 175. A. T. Fromhold., J. Electrochem. Soc., 115. 882 (1968). 176. C. Lukac, J. B. Lumsden, S. Smialowska, and R. W. Staehle, J, Electrochem. Soc,, 122, 1571 (1975). 177. A. H. Lanyon and B. M. W. Trapnell, Proc. Roy. Soc., 227A. 387 (1955). 178. D. D. Eley and P. R. Wilkinson, Proc. Roy. Soc., 254A. 327 (1960). 179. M. G . Fontana and N. D, Greene, "Corrosion Engineering", p. 355, McGraw-Hill, New York (1967). 180. "CRC Handbook of Chemistry and Physics", Edd. by R. C, Weast, CRC Press, Cleveland, Ohio. 181. R. K. Freier, "Aqueous Solutions: Data for Inorganic and Organic Compounds", Vol. 1 (1976); Vol. 2 (1978), Walter de Gruyter, Berlin. 182. J. B. Lumdsen and R. W. Staehle, Progress Report for the Second Contract Year, May 1, 1971-April 30, 1972, AISI Project No. 61-244, The Ohio State University. 183. M. J. Pryor and. U. R. Evans, J. Chem. Soc. (London), 3330 (1949). 184. M. J. Pryor and U. R. Evans, ibid, 1259 (1950). 185. M. J. Pryor and U. R. Evans, ibid, 1266 (1949). ]86. I. I. Baram, J. Appl. Chem. USSR, 42, 1439 (1974). 2 9 8 1 8 7. K. Azuma and. H. Kametani, Trans. TMS-AIME, 230. 853 (1964). 188. H. Gerischer and F. Beck, Z. Physik. Chem. (Frankfurt), 2^, 113 (I960). 189. H. Gerischer and F. Beck, ibid., 2}±, 378 (I960). 190. H. Gerischer and W. Mirdt, Electrochim. Acta, l^, 1329 (1968). 191. II. Gerischer, "Physical Chemistry an Advanced Treatise" Academic Press, New York (1970). 192. D. A. Vermilyea and W. Vedder, General Electric Co., Report 69-C-152, New York (1969). 193. D, A. Vermilyea, J. Electrochem. Soc., 116. 1179 (1969). 194. D. A, Vermilyea and C. F. Kirk, ibid., p. 1487. 195. K. Nii, Corros. Sci., 10, 571 (1970). 196. J. J. Gilman, W. G. Johnston and G. W. Sears, J. Appl. Phys. 29, 747 (1958). 197. A. R. C. Westwood and H. Rubin, J. Appl. Phys., 33. 2001 (1962). 198. M. A. Heine and M. J. Pryor, J. Electrochem. Soc., 110. 1205 (1963). 199. S. Abe and R. W. Staehle, Second Annual Report, Office of Naval Research, Contract No. H00014-67-A-0232-0006. 2 0 0 . P. W. Palmberg, in "The Structure and Chemistry of Solid Surfaces", Edd. by G, A. Somorjai, p. 29-1, Wiley, New York (1969). 299 201. C. C. Chang, Surface Sci.. 2£# 53 (1971) - 202. P. W. Palmberg, in "Electron Spectroscopy", Edd. by D. A. Shirley, p. 835* North Holland, Amsterdam, (1972). 203. R. E. Weber, Research/Development, Oct. 1972. 204 . T. E. Gallon and J. A. D. Matthew, Rev. Phys. Technol., Ill, 1, 31 (1972). 205. J. C, Riviere, Contemp. Phys., 1^, 513 (1973). 206. C. C. Chang, in "Characterization of Solid Surfaces", Edd. by P. F. Kane and C. B. Larrabee, p. 509, Plenum Press, New York (197il)* 207. A. Joshi, L. E. Davis and P. W. Palmberg, in "Methods of Surface Analysis", Edd. by A. V/. Czanderna, p. 218 Elsevier, New York (1975)* 208. P. Auger, J. Phys. Radium, 6, 205 (1925). 209. P. Auger, Surface Sci., !^8, 1 (1975)* 210. J. J. Lander, Phys. Rev., 21, 1382 (1953). 211. L. A. Harris, J. Appl. Phys., _^2, lij.19 (1968). 212. R. E. Weber and W. T. Peria, J. Appl. Phys., ^8, 1^-355 (1967). 213. P. W. Palmberg, G. K. Bohn and J. C. Tracy, Appl. Phys. Lett., 12, 251i (1969). 2II4.. P. W. Palmberg, J. Vac. Sci. Technol., 2* 160 (1972). 215* N. C. MacDonald, Appl. Phys. Lett., 16, 76 (1970). 216, H. E. Bishop and J. C. Riviere, J. Appl. Phys., UP. 171*0 (1969). 300 217. R. G. Musket and W. Bauer. Thin Solid Films. 1^, 69 (1973). 218. J. F. Hannequin and P. Viaris de Lesegno, Surface Sci., 42, 50 (1974). 219. PHI DATA SHEET 1 0 4 4 3"76 10M; 1051 6-77 15M. 220. (a) M. L. Tarng and G. K. Wehner, Phys. Elect. Conf., 1972; (b) D. E. Eastman, P. E. C. Alburquerque, 1972; (c) J. W. T. Ridgeway and D. Haneman, Surface Sci., 24, 451 (1971); 26, 683 (1971); (d.) P. W. Palmberg and T. N. Rhodin, J. Appl. Phys., 2 1 , 2425 (1966). (e) K. Jacobi and J. Holz, Surf. Sci., 26, 54 (1971); (f) Y. Baer, et al., Solid State Commun., 8, 1479 (1970). (g) R. G. Steinhardt, et al., Proc. Int. Conf. Electron Spectrosc., Asilomar, 1971, p. 557; (h) M. P. Seah, Surf. Sci., J2, 703 (1972). 221. P. W. Palmberg, Anal. Chem., 549A (1973). 222. W. Reuter, Surf. Sci., 2£, 80 (1971) 223. R. E. Weber and W. T. Peria, J. Appl. Phys., ^8, 4355 (1967). 22i|. R. E. Weber and A. L. Johnson, J. Appl. Phys., 40. 414 (1969). 225. S. Thomas and T. W. Hass, J. Vac. Sci. Technol., 10. 218 (1973). 301 226. M. P. Seah, Surf. Sci., 12, 703 (1972). 227. M. P. Seah, Surf. Sci., 4 0 , 595 (1973). 228. C. C. Chang, Surf. Sci., 4 8 , 9 (1975). 229. E. N. Sickfaus, J. Vac. Sci. Technol., .11, 299 (1974). 230. E. Meyer and J. J. Vrakking, Surf. Sci., H , 271 (1972). 231. J. J. Vrakking and P. Meyer, Surf. Sci., \±J_, 50 (1975). 232. M. L. Tarng and G. K. Wehner, J. Appl. Phys., 44* 1534 (1973). 233. "Handbook of Auger Electron Spectroscopy1'. 2nd Edu., Edina, Minn. (1976). 234- M. A. Lis tengarten, Bull. Acad. Sci. USSR PH ,/s. Ser. , 24, 1050 (I960). 235. M. P. Chung and L. H. Jenkins, Surface Sci., 22, 479 (1970). 236. T. W. Hass, J. T. Crant, and G. Phys., 41, 1853 (1972). 237. F. J. Szalkowski and G, A. Aomorjai, J. Chem. Phys., 16, 6079 (1972). 238. C. C. Allen and R. K. Wild, J. of Elect. Spect. and Relat. Phenom., 1, 409 (1974). 239. M. Seo, J. B. Lumsden, and R, W. Staehle, Surf. Sci., kg, 337 (1974)- 2 4 0 . W. M. Mularie and W. T. Peria, Surf. Sci., 26, 125 (1971). 241. P. W. Palmberg, J. Vac. Sci. and Technol., 9, 160 (1972). 214-2 . D. F. Stein, W. C. Johnson, and C. L. White, Cana. Met. Quar., 13(1). 79 (19714-). 2lj.3. "VacSorb Pumps", Varian Vacuum Instructions, Varian Vacuum Division, May 1970. 2144., "Titanium Sublimation Pump Control Unit", Varian Vacuum Division, July 1969. 2l|5. "Noble Vaclon Pump", Varian Vacuum Division, August 1967. 2I4.6 . H. E. Chaung, Dissertation, The Ohio State University (1976). 214.7* R. E, Weber and A. L. Johnson, J. Appl. Phys., UP . 3H 4- (1969). 214.8 . W. J. Muller, "Die Bedeckungstheorie der Passivitat der Matalle and ihre experimentelle Begrundung", Verlag Chemie, Berlin (1933). 214.9 * T. P. Hoar, in "Proc. Third Internat. Cong, on Metal. Corro.", Vol. 1, p. 88, MIR, Moscow (1969). 250. M. Prazak, ibid., p. 336. 251. M. Prazak, V. Prazak and V. Cihal, Z. Elektrochem. 62, 739 (195>8). 252. F. J. Szalkowski and G. A. Somorjai, J. Chem. Phys., £6, 6097 (1972). 253* M. Seo, J. B. Lumsden and R. W. Staehle, Surf, Sci., £0, 5U1 (1975). 25I4.* G. C. Allen and R. K. Wild, J. Electron Spectroscopy and Related Phenomena, 14.09 (197lt). 255. G. M. Florianovich, L, A, Sokolova and Y. M. Kolotyrin, Electroehim. Acta, Y2, 878 (1967). 256. D. B. Gibbs and M. Cohen, J. Electrochem. Soc., 119. J+16> (1972). 257. S. Szklarska-Smialowska and G. Mrowczynski, Brit. Corro. J., 10, 187 (1975).