Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The reaction of copper-gold alloys in aqueous ammonia under oxygen pressure Fisher, James Irwin 1953

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1953_A7 F5 R3.pdf [ 4.93MB ]
Metadata
JSON: 831-1.0106438.json
JSON-LD: 831-1.0106438-ld.json
RDF/XML (Pretty): 831-1.0106438-rdf.xml
RDF/JSON: 831-1.0106438-rdf.json
Turtle: 831-1.0106438-turtle.txt
N-Triples: 831-1.0106438-rdf-ntriples.txt
Original Record: 831-1.0106438-source.json
Full Text
831-1.0106438-fulltext.txt
Citation
831-1.0106438.ris

Full Text

THE REACTION OF COPPER-GOLD ALLOYS IN AQUEOUS AMMONIA UNDER OXYGEN PRESSURE by JAMES IRWIN FISHER A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in the Department of MINING AND METALLURGY We accept this thesis as conforming to the standard required from candidates for the degree of MASTER OF APPLIED SCIENCE Members of the Department of Mining and Metallurgy THE UNIVERSITY OF BRITISH COLUMBIA September, 1953 V ABSTRACT An investigation was conducted of the corrosion of copper-gold alloys in ammonia solutions under oxygen pressure. The reaction, which took place in an autoclave, was followed by sampling and analy-sis of the solutions. It was found that only copper was dissolved from the alloys, the gold being left behind in a film on the surface of the corroding specimen. The reaction of pure copper as well as of four alloys rang-ing in gold content from 2 to 15 atomic percent were studied. Other variables examined include the concentrations of NH3 and NHj in the solution, the oxygen pressure and the temperature. It was found that while the rate curves for the dissolution of pure copper were linear, those for the alloys were generally para-bolic in shape. The rate of dissolution of copper from the alloys ap-pears to be determined by the transport of reactants and products through the gold rich films. Some copper oxide may also be precipit-ated in the pores of the film or in the region between the film and the underlying metal, further impeding the transport processes and contributing to the lowering of the rate. ACKNOWLEDGEMENT The author is indebted to the National Research Council of Canada for the financial aid which enabled this project to be carried out. The author is grateful to the members of the Department of Metallurgy for their assistance throughput this work, and is especially grateful to Dr. J. Halpern, vbo ably directed this invest-igation. TABLE, OF CONTENTS Page: Introduction 1 Previous Work on the Corrosion of Similar A l l o y Systems; 2 Object and Scope of the Present Investigation 4 Experimental 5 Ao Preparation of the Alloys 5 Bo Apparatus 7 Co Chemical Reagents and Solutions 7 D« Measurement of the Rates of Dissolution 9 E 0 A n a l y t i c a l Procedures 9 (1) Copper 9 (2) Gold 10 (3) Ammonia 12 Results and Discussion 13 Nature of the Reactions and Rate Curves 13 Effect of NH3 Concentration 16 Effect of (NH4)aS0j„ 24 Effect of Varying NH3 i n the Presence of Constant NH+. 34 Effect of Oxygen P a r t i a l Pressure 41 Effect of Temperature 46 Structure and Composition of the Films 50 Conclusions 59 Summary of Experimental Results 59 Interpretation of the Results 61 TABLE OF CONTENTS (continued) References LIST OF FIGURES Figure No. Page 1. Diagram of Pressure Vessel 8 2. Calibration Curve for Copper Carbamate 11 3. Typical Rate Curves for Copper and Alloy 5 i n NH3 14 4o Rate Curves for Copper i n NH3 17 5. Rate Curves for Alloy 5 i n NH3 18 6. Parabolic Plots for Alloy 5 i n NK3 19 7. Comparative Rate Curves for the Alloys in NH3 20 8. Effect of NH3 on the Hate of Copper Dissolution 21 9. Effect of Gold on the Rate of Copper Dissolution i n Solutions containing NH3 Alone 22 10. Rate Curves for Pure Copper with NH3 Constant and NH£ Varying 25 11. Rate Curves for Alloy 10 with NH3 Constant and NH£ Varying 26 12. Parabolic Plots for Alloy 10 with NH3 Constant and Ml Varying 27 13. Comparative Rate Curves for Alloys with Constant NH3, Constant NH£ 28 LIST OF FIGURES (continued) Figure No. 14. Effect of NH£ on the Rate of Copper Dissolution at Constant NH3 15. Effect of Gold on the Rates of Copper Dissolution with NH3 Constant and NHJ Varying 16. Rate Curves for Copper with NHlt Constant and NH3 Varying 17. Rate Curves for Alloy 5 with NHj Constant and NH3 Varying 18. Rate Curves for Alloy 10 with NHj Constant and NH3 Varying 19. Parabolic Plots for Alloy 10 with NH£ Constant and NH3 Varying 20. Effect of NH3 in Solutions Containing Constant NH£ 21. Effect of Gold on the Rates of Copper Dissolution with NH£ Constant and NH3 Varying 22. Rate Curves for Copper at Various Oxygen Pressures 23. Rate Curves for Alloy 10 at Various Oxygen Pressures 24. Rate Curves for Copper at Various Solution Temperatures 25. Rate Curves for Alloy 2 at Various Solution Temperatures LIST OF FIGURES: (continued) Figure No. Page 26. Surface Structure of Alloy 2 51 27. Surface Structure of Alloy 2 51 28. Surface Structure of Alloy 5 52 29. Surface Structure of Alloy 10 52 30. Surface Structure of Alloy 10 53 31. Surface Structure of Alloy 10 53 32. Surface Structure of Alloy 15 54 33. Surface Structure of Alloy 15 54 LIST OF TABLES Table No. Page ,1 Data on Experiments with Solutions Containing NH3 Alone 23 II Data on Experiments with Solutions in which NH£ was Varied and NH3 Held Constant 29 III Data on Experiments with Solutions in which NH3 was Varied and NHj. Held Constant 40 IV Data on Experiments in which. Oxygen Pressure was Varied 45 V Data on Experiments in which Temperature was Varied 49 VI Analyses of Films 55 THE REACTION OF GOLD-COPPER ALLOYS WITH AQUEOUS AMMONIA UNDER OXYGEN PRESSURE INTRODUCTION This study of the kinetics of the dissolution of copper-gold alloys in aqueous ammonia and ammonium salt solutions under oxygen pressure had i t s inception i n a similar study on the kinetics of the dissolution of copper recently conducted i n this laboratory ( l ) . In the course of the latter investigation, i t was shown that the t o t a l rate of dissolution of copper in ammonia solutions was the sum of the rates of two independent reactions, f i r s t order i n relation to the concentrations of ammonia and ammonium ion respectively. These rates were found to be independent of the oxygen concentration, provided the oxygen was present i n excess so that i t s transport to the copper surface did not limit the rate of the reaction. It was also found that the rates were independent of the concentrations of hydrogen ion and dissolved copper. On the basis of these results, the follow-ing mechanism for the dissolution of copper was proposed; (i) Adsorption of dissolved oxygen onto the copper surfaces fast Cu + 1 / 2 0 2 — Cu....O . , , , . . . ( 1 ) ( i i ) Reaction of an ammonia molecule or ammonium ion with the copper-oxygen complex on the surface? 2 . slow NH3 fast Cu » p • © 0 + NH3 + HOH Cu(NH3) % 20H" (2) Cu O • e e 0 + NH* + OH" — Cu(NH3) + 20H « « O s o • • .(3) Following this work on copper, i t seemed of interest to examine the effect on the dissolution reaction of varying the composit-ion of the metal phase by suitable alloying, copper remaining the principal or only dissolving metal. Gold was chosen as the alloying metal because: (a) i t is inert to attack by a l l the reagents contemplated. (b) i t forms a continuous solid solution with copper, and the structural properties of this system have been extensively investigated. (c) there appeared to be some possibility of studying the^  effect of ordering on the corrosion properties of the alloy. Previous Work on the Corrosion of Similar Alloy Systems.. studied by Tammann (2) and Muller, Freissler and Plettinger (3). How-ever, their work was mainly restricted to establishing parting limits of the alloys in various corroding media. Landau and Oldach (4) have compiled a comprehensive summary of qualitative and quantitative data on the corrosion properties of various solid solution alloys, and have noted several interesting-correlations. In particular, they The parting limit of an alloy may be defined as the maximum conc-entration of the less noble element of an alloy at which selective oxidation can occur. Beyond this limit the alloy behaves like the pure noble metal (5). The corrosion properties of copper-gold alloys have been 3. emphasized the importance of passivity limits , and noted the exist-ence of minima in the corrosion rate - composition curves of certain alloy systems0 Several theories relating to the corrosion of alloys have been proposed and, although the present work was not undertaken with the idea of applying or confirming any particular one of these theories, i t may be useful to refer briefly to some of them0 In 1921 Tammann (2) proposed a correlation between parting limits and the atomic ratio of the metals in the alloy, noting certain exceptions and explaining these i n terms of space-lattice isomerides. Russell (6) noted a somewhat similar correlation between alloy structure and passivity. Uhlig and Wulff ( 8 ) , extending Russell's ideas, proposed the Electron Configur-ation Theory in which passive metals are characterized by unfilled d energy bands. These passive metals are able to transmit their charact-e r i s t i c s to other less passive metals in alloys by electron sharing. Landau (9), treating theories based on alloy structure with consider-able reserve, extended the local c e l l theory for the corrosion of pure metals to alloys. It should be noted here that most alloy corrosion theories treat passivity rather than parting. Since i n the present system the alloying metal (gold) i s thermodynamically immune to corrosion in the media used, the concept of parting i s of greater significance than Uhlig and Mears (7) give the following definition of passivity; "a metal or alloy i s passive i f i t substantially resists corrosion i n an environment where thermo-dynamically there is a large free energy decrease associated with i t s passage from the metallic state to ap-propriate corrosion products , , 0 that of passivity, Ob.ject and Scope of the Present Investigation This study was undertaken with the object of establishing the effect of alloying copper with an inert metal on the dissolution kinetics of the former. Specimens of copper-gold alloys were exposed to ammonia solutions i n an autoclave maintained under controlled temp-erature and pressure conditions, and the rate of dissolution of copper determined by periodic sampling and analysis of the solution. The variables examined included alloy composition, NH3 and NH£ concentrat-ions, oxygen pressure and temperature. The experimental procedures used and the results obtained are described and discussed below. 5. EXPERIMENTAL A« Preparation of the Alloys: The copper used in this work was supplied as half-inch rods by Ellett Copper and Brass Company. The gold was supplied as fine gold shot by Birks Limited. Spectrochemical analyses of these metals and of the alloys are given in Appendix I. At the outset, i t was thought that three alloys at 5 atomic percent intervals (i.e. corresponding to 5, 10 and 15 atomic percent gold) would be sufficient to establish a pattern of results which would suggest what other alloys should be prepared and studied. The metal constituents for these alloys were carefully weighed, and then degassed and melted by induction heating in a carbon crucible under 150 microns of argon pressure. On cooling, a cylindrical ingot about 1.25 inches in diameter and weighing about 25 grams was obtained. It was subsequently decided to prepare and study another alloy contain-ing 2 atomic percent gold. Since a carbon crucible was not easily obtainable at that time, an alundum crucible was used in preparing this alloy. It was also desired to conduct some comparative dissolution experiments on pure copper, and a sample of pure copper was melted and heat treated under the same conditions as the alloys. " 1 ' - — — _ Kindly carried out by the Department of Mines, Victoria, B.C. Henceforth the alloys will be referred to as alloy 2, alloy 5» etc., where the number corresponds to the gold content in atomic percent. 6. A l l the alloys were homogenized in an argon atmosphere at 850°C. for periods ranging between 60 and 75 hours. They were then furnace cooled to 500°C. and water quenched. This treatment was designed to produce disordered structures for, i f studies were to be made on the effects of order on the dissolution rates of the alloys, i t was decided to treat disordered alloys f i r s t . Although ordering is most significant in a 25 atomic percent gold alloy, a consistent treatment for a l l the alloys was desirable. The c r i t i c a l temperature for the order-disorder transformation i n a 25 atomic percent alloy i s 391°C The ingots were turned on a lathe to a uniform diameter which was measured with a micrometer. They were then sawed in half through a plane perpendicular to the cylindrical axis, and mounted in bakelite with the circular face exposed. Samples from the saw cut-tings and from d r i l l i n g s , which were taken at the completion of the work, were assayed for gold. The assays were found to correspond very closely to the intended compositions (see Appendix II). Metallographic tests failed to show any evidence of coring or inhomogeneity in the ingots, but in some cases interdendritic shrinkage cavities were obs-erved toward the upper end. It was assumed that this shrinkage did not affect the surface area appreciably. Prior to each experiment the mounted alloy samples were refaced in a lathe to remove any f i l m or surface inconsistencies in composition which might have resulted from the previous experiment. About 3/1000 inches were machined from the exposed surfaces of the alloys, which were then polished using standard techniques and stored 7. i n a dessicator for subsequent experiments. B„ Apparatus; The experiments were conducted in a stainless steel auto-clave (see Figure 1) designed for working pressures up to 150 psig. The bakelite mount (which also served as el e c t r i c a l insulation) cont-aining the metal specimen was held by a stainless steel rod, tapped into the l i d of the autoclave, so that the surface of the alloy was exposed to the solution about 0.25 inches above the blades of the impeller. Fresh solution was forced against the face of the alloy by the impeller, which was 10.9 cm. i n diameter and rotated at 720 RPM. The temperature of the solutions was regulated by a Wheelco controller coupled to a platinum resistance thermometer, which Was placed i n a stainless steel well extending from the l i d of the auto-clave into the solution. The thermometer leads were connected i n the controller across a double pole, single throw microswitch which activ-ated solenoid valves on gas and water lines. The gas was fed to a ring burner mounted under the autoclave, and the water line was con-nected to cooling coils running through the corroding solution. A cyclic temperature variation resulted which could be held to within ±1°F. at 77°F. (25°G:.) and ±3°F. at 132°F. (50*©.). Pressure was controlled by a standard pressure regulator on the oxygen cylinder. G. Chemical Reagent s and.. Solutions... Chemically pure grade ammonium hydroxide and ammonium sulphate, supplied by the Nichols Chemical Company, were used. Oxygen was supplied i n cylinders by the Canadian Liquid Air Company, 8 . Figure L, Diagram of Stainless Steel Pressure Vessel and Internal Parts. A - Cooling Coil D - Sampling Tube B3 -• Impeller £ - Thermometer Well C - Gas Inlet Tube F - Metal Specimen in Bakelite Mount and was used without further purification. Solutions were prepared by diluting a measured volume of ammonium hydroxide to 3»0 l i t r e s with d i s t i l l e d water. Ammonium sulphate was weighed and added to the solution in the autoclave. D. Measurement of the Rates, of. Dissolutions. After the desired specimen and solution were placed i n the autoclave, the latter was sealed and the solution brought to temp-erature. When the correct temperature was attained, the oxygen was introduced and maintained at the desired value. The f i r s t sample of the solution was taken within one minute of the pressure increase. To follow the rate of the reaction 25 ml. samples were withdrawn through the sample tube shown in Figure 1 at 30 minute intervals and analysed for copper. ' Since the surface area of the alloy and the vol-ume of solution were known, the amount of copper dissolved per unit area could be calculated. E. Analytical.Procedures.:. (1) Copper; Copper concentrations were determined spectrophotometric-a l l y by the Carbamate method (11). Sodium diethyldithiocarbamate forms a stable, colored chelate compound with cupric ions, showing a maximum light absorption at 437 millimicrons. A Beckman model DU spectrophoto-meter was used to determine the optical density of the solutions at this wavelength. The optical density was found to be proportional to the copper concentration. (A calibration curve for this analytical 10. procedure i s given in Figure 2). To prepare the samples for analysis, 7 ml. of a carbamate solution were added to an aliquot of the sample taken from the autoclave, and the mixture diluted with d i s t i l l e d water to 50 ml. in a volumetric flask. The aliquot used varied from 1 to 5 ml. depending on the copper concentration. Since the st a b i l i t y of the copper carbamate complex i s affected by daylight ( l l ) , analytical solutions were prepared in an a r t i f i c i a l l y l i t room. With careful technique, results obtained by this method could be duplicated within 1 percent. (2) Gold: It was also thought advisable to develop a method for determining gold concentrations in the corroding solutions. A color-imetric procedure was again.adopted using p-dimethylaminobenzalrhodan-s ine which forms a colored chelate complex with gold ions. A 15 ml. aliquot of the solution was made 0,1 N i n HCl and 1 ml. of a 20 per-cent saturated solution of rhodanine in ethyl alcohol was added. The resulting solution was diluted to 50 ml. and i t s absorption measured i n s i l i c a cells at 300 millimicrons. While i t was found that this method did not give quantitatively reproducible results, i t was very sensitive and made i t possible to detect visually less than ly of gold per ml. by the purple color developed. Copper had no effect on the visual test, although i t could be detected by the spectrophoto-: — _ _ _ _ _ Preparation of the carbamate solutions Solution As 0.4 grams carbamate in 400 ml. d i s t i l l e d water. Solution Bs 10 grams gum arabic and 1 ml. toluene in 1000 ml. d i s t i l -led water. Solutions A and B were mixed and fi l t e r e d , and stored in the dark. 11 . 20 h 0 6 0 8 0 100 Cu f y / m l . Figure 2 . Calibration curve for copper carbamate. 12. meter. It was found that the optical densities of copper and gold rhodanine were additive. (3) Ammonia: Ammonia concentrations were determined by potentiometric t i t r a t i o n with standard HCI. 13o RESULTS AND DISCUSSION Nature of the Reactions and Rate Curves. In a l l the experiments, specimens of copper or of a copper-gold alloy were exposed to solutions containing NH3 (and, i n some instances, (NH*)aSO*,.) under oxygen pressure. In general, i t was found that the specimens were corroded under these conditions and that copper was dissolved. At no time could any gold be detected i n the solutions. The reaction involved thus appears to be represented by the equation: C u(ioo-n) A u(n) + 7 / 2 ° 2 + W 3 + y H * ° ~* Cu(ioo-.r)M(r5 + yCu(NH 3)r • 2y0lT ..(4) where r> n, n and r being the atomic percent gold i n the alloys before and after corrosion. A l l the results described below thus relate to measurements of the rate of dissolution of copper. The shapes of the two rate curves, shown i n Figure 3, are typical of those obtained for pure copper and for the copper-gold alloys, respectively, under most conditions. The rat<e plots for the dissolution of pure copper were always found to be linear, the amount of copper dissolved being proportional to the time, i.e. Cu - fcjt (5) On the other hand, the rate of dissolution of copper from copper-gold alloys generally f e l l off with time, corresponding i n most cases 14. ALLOY O — TIME - HOURS Figure 3. Typical rate curves for the dissolution of copper for pure copper and alloy 5, showing relative rates. T <= 259C; 0 2 - 6.8 Atm.j NH3 - 0.5 <a/l. 15. to a parabolic relation, i.e. £cuj = a + k2T 00.0.0 . ( 6 ) where Cu is the amount of copper dissolved per square centimeter of surface i n time T, and k p kg and a are constants 0 Such a relation usually denotes the buildup of a surface f i l m on the corroding specimen which slows the reaction as i t thick-ens by impeding the transfer of reactants and products. It was concluded that this film was due to the gold which was l e f t behind at the surface of the alloy when the copper dissolved. This was confirmed by subsequent analyses of some of the films which showed them to be much richer i n gold than the original alloys. Although i t was found that the rate curves obtained i n most of the experiments on alloys conformed closely to a parabolic relation, there were not infrequent exceptions to this rule. This, together with the fact that the rate curves for pure copper were linear, made i t impossible to compare the rates for the different specimens i n terms of a uniform rate constant. For these reasons i t was decided to adopt the amount of copper dissolved during the f i r s t hour of any experiment as a measure of the rate. A l l subsequent correlations of, the effects of variables are expressed i n terms of this quantity. Where possible, such correlations were compared with those based on parabolic rate constants, and in general the trends were found to be similar. In the following sections are described the results of studies on the different variables. Unless otherwise stated, the 16. pressure and temperature i n a l l the experiments were 6.8 atmospheres of oxygen and 2 5°C, Effect of NHS Concentration. The corrosion of copper and each of the four copper-gold alloys was investigated using a series of solutions containing d i f -ferent concentrations of NH3 generally ranging from 0 . 1 5 to 2 . 5 m/1. No NHi£ salts were added to these solutions. Some typical rate curves obtained during this series of experiments are shown in Figures 4 to 7, and the results are summarized i n Figures 8 and 9 and in Table I. In general i t was found that increasing the concentration of NH3 resulted in a higher rate of dissolution of copper for the alloys as well as for pure copper. This i s shown in Figures 8 and 9 . Figures 4 and 5 demonstrate the rate curves obtained for copper and for an alloy with increasing NH3 concentrations. The parabolic character of the alloy rate curves i s shown i n Figure 6, vAiere the square of the amount of copper dissolved is plotted against time. It should be noted that the extrapolations of the lines in Figure 6 do not go through the origin, but intersect the ordinate, indicating a high i n i t i a l rate of copper dissolution. This effect was particular-l y pronounced with alloy 1 0 ( s e e Figure 7)} and increased with i n -creasing NH3 concentrations. However, a f a l l i n g off of the rate was always observed following this i n i t i a l period, indicating that ultim-ately the film became rate controlling in every case. The effect on the rate of increasing the gold content of the alloy i s shown in Figures 8 and 9 o Particular reference may be 17. 160 \ -120 o I S o CO CO 0 o TIME - HOURS Figure 4. Rate curves for pure copper in solutions containing NH3 alone. T - 25°C; 0 2 = 6.8 Atm. 18 s o 1 o CO CO M Q o TIME - HOURS Figure 5. Rate curves for the dissolution of copper from alloy 5 in NH3 solutions, T * 25°C; 0 2 =6.8 Atm. Figure 6„ Parabolic plots for the dissolution of copper from alloy 5 in NH3 solutions. T = 25°Cj 0 2 <= 6„8 Atm. 20. TIME - HOURS Figure 7. Comparative rate curves for the various alloys under similar corroding conditions in NH3 solutions. NH3 = 0.5 m/1; T = 25°C; 0 2 - 6.8 Atm. ALLOY • - Cu NH3 - M/L Figure 8. Effect of NH3 on the rate of copper dissolution. T = 25°C; 0 2 = 6.8 Atm. 22. Figure 9. Effect of gold on the rate of copper dissolution i n solutions containing NH3 alone. T=25°C; 0 2 = 6.8 Atm. 2 3 . TABLE I. Data on Experiments wi th So lu t ions Conta in ing NH^ Alone. ± ±kk Run NH 3 A l l o y NH 3 Rate L inear Parab. • Cu No. T i t r a t e d Type Rate Rate Dissolved i n 1 hour m/1 m/1 mg C-6 0.5 Cu 0.47 L 21.0 20.5 C-34 0.5 2 0.48 I (0.019) 2.89 C-3 0.5 5 0.46 P - 0.02 4.41 G-4 0.5 10 0.48 P - 0.013 5 .25 C-5 0.5 15 0.47 P - 0.013 4.0 C-35 0 .75 5 0 .75 P - 0.60 10.4 C -24 1.0 Cu 0.99 L 66.6 62.0 C-36 1.0 2 0.96 P - 0.19 12.6 C-21 1.0 5 1.00 P 0.15 18.3 G-22 1.0 10 1 .05 P — 0.08 16.6 G-23 1.0 15 1 .04 I - (0.11) 10.0 C-99 1.5 5 1 .57 P _ 0.39 27 .1 C-39 1.5 15 1.60 P = 0.60 18.0 C-103 2 ,0 2 2.00 36.0 C-60 2 .0 Cu 1.89 L 82.2 82.2 2 .0 5 2 . 0 0 P 0.88 42 .5 c-27 2.0 10 1.99 I (0 .34) 39.8 c-25 2,0 15 2.15 P - 0.81 22.1 0=42 2.3 5 2.28 P 1 .03 45.6 C-41 2.3 10 2.30 I = — a . 5 ± Rate typess L = L i n e a r | P = P a r a b o l i c ; I = Intermediate; ±k Rates i n parentheses are d o u b t f u l . Uni ts of parabo l ic ra tes are [mgCu/ cm2j 2 /hour. ±M Uni ts of l i n e a r ra tes are mgCu/cm 2 /hour. 24. made to the pronounced decrease i n ra te which occurs when as l i t t l e as 2 atomic percent gold i s added t o the copper. Under the condi t ions of these experiments ( i . e . when s o l u t i o n s contained NH3 but no N H 4 , s a l t s ) f u r t h e r a d d i t i o n o f gold had r e l a t i v e l y l i t t l e e f f e c t on the r a t e . Fo l lowing a minimum at about 2 atomic percent g o l d , the ra tes appeared to increase s l i g h t l y with f u r t h e r gold a d d i t i o n , reach a maximum i n the reg ion o f 5 t o 1 0 atomic percent g o l d , and then de -crease again as the gold was increased to 1 5 atomic percent . Since the shapes of the rate curves f o r the var ious a l l o y s i n NH3 are complex and d i f f e r s i g n i f i c a n t l y f o r the d i f f e r e n t a l l o y s , the use of the amount of copper d i s s o l v e d i n the f i r s t hour as a measure of the rate i s p o s s i b l y not j u s t i f i e d i n t h i s case . The r e s u l t s should there fore be in te rp re ted with c a u t i o n , s ince the d i f f e r e n c e s invo lved are r e l a t i v e l y s m a l l . E f f e c t of (NH/.) 9S0A, I t was found that s o l u t i o n s conta in ing NH* s a l t s o n l y , wi th no f r e e N H 3 , d i d not react with copper. Apparently some f ree NH3 i s necessary to form the so lub le cuprammine i o n . The e f f e c t o f the NH4 i o n was there fore inves t iga ted us ing a s e r i e s of s o l u t i o n s i n which the f ree NH3 concent ra t ion was kept constant at 0 . 5 m/1 and the concent ra t ion of (NH*) 2S04 v a r i e d between 0 . 0 1 and 0.06 m/1 ( co r re -sponding to NH£ concentrat ions ranging from 0,02 to 0 . 1 2 m / l ) . T h i s s e r i e s of s o l u t i o n s was used to study the cor ros ion of copper and o f each of the a l l o y s . Some of the t y p i c a l ra te curves obtained i n these experiments are shown i n F igures 10 to 1 3 , and a l l the r e s u l t s sum-+ marized i n Table II , The ra tes are p lo t ted as funct ions of the KH/,. TIME - HOURS Figure 10. Rate curves for pure copper in solutions with constant NH3 and varying SJH+. NH3 = 0.5 m/1; T = 25°Cj 0 2 » 6.8 Atm. Figure 11. Rate curves for alloy 10 in solutions with constant NH3 and varying NH*.. NH3 - 0.5 m/lj * = 25°Cj 0 2 - 6.8 Atm. 27 Figure 12. Parabolic plots for alloy 10 in solutions with constant NH3 and varying NH?. NH3 - 0,5J T = 25PC; 0 2 - 6.8 Atm. T J M E - HOURS Figure 13. Comparative rate curves for the alloys in solutions with constant NH3, constant NHJ. Nh*£ • 0.04 m/1; NH3 - 0.5 m/1: T m 25°Cj 0 2 -» 6.8 Atm. 29. TABLE II. Data on Experiments with Solutions in which + NHfe was Varied and Nfo Held Constant. ft NH3 Rate Linear Parab. Cu Run No, NH* NH3 Alloy Titrated Type Rate Rate Dissolved in 1 hour m/1 m/1 m/1 mg C-73 0.006 0.5 Cu 0.49 L 33.3 33.3 C-80 0.012 0.5 Cu 0,51 L 52.5 - 52.5 C-18 0.02 0.5 Cu 0.49 L 62.2 — 62.2 C-37 0.02 0.5 2 0.50 I - (1.51) 31.5 C-19 0.02 0.5 5 0.50 I - (1.41) 28.8 C-20 0.02 0.5 10 0.48 P - 0.53 17.2 C-17 0.02 0.5 15 0.49 - 0 0 0 C-10 0.04 0.5 Cu 0.49 L 76.8 __, 76.8 C-43 0.04 0.5 2 0.53 L 60.0 _• 60.0 C-8 0.04 0.5 5 0.53 P - 3.60 45.0 C-9 0.04 0.5 10 0,50 P - 1.14 25.5 C - l l 0.04 0.5 15 0.51 - 0 0 0 C-31 0.06 0.5 Cu 0.53 L 80.8 „ 80.8 C-45 0,06 0.5 2 0.50 I 68.3 - 68.3 C-14 0.06 0.5 5 0,50 P 5.47 55.0 C-16 0.06 0.5 10 0.51 P - 1.72 32.8 C-13 0.08 0.5 Cu 0.52 L 83.3 83.3 C-46 0.08 0.5 2 0.52 L 67,8 - 67.8 C-26 0.08 0.5 5 0.49 P mm 5.64 54.8 C-15 0.08 0.5 10 0.51 P - 2.36 37.8 C-12 0.08 0.5 15 0.51 - 0 0 0 C-56 0,10 0.5 5 0.51 P mm 6,86 61.0 C-38 0.10 0.5 10 . 0,52 P mm 2.74 44.0 C-44 0.12 0.5 Cu 0.51 L 89.6 mm 85.5 G-59 0.12 0.5 2 0.51 I (59.0) mm 68.0 C-29 0.12 0.5 5 0.45 P - 7.04 59.0 C-30 0.12 0.5 10 0,49 P — 2.67 44.9 ± L = Linear; P = Parabolic? I 3 Intermediate! M Rates in parentheses are doubtful. Units of parabolic rates are {mgCu/cm2/2/hour. Units of linear rates are mgCu/cm2/hour. 30. ion concentration and of the a l l o y composition i n Figures 14 and 15, respectively. An examination of the results reveals the following effects? (1) A l l the rate curves f o r pure copper were again found to be l i n e a r . Most of those for the alloys (but not a l l ) were parabolic (compare Figures 10, 11 and 12). (2) Increasing the gold content of the a l l o y resulted i n a lowering of the rate of disso l u t i o n (see Figures 13 and 15). The rate decreased continuously with increasing gold concentration u n t i l the reaction ceased at about 15 atomic percent gold. In no case, when the solutions contained NH]£ ions as w e l l as free NH3, could any reaction be detected with a l l o y 15. These results are i n marked contrast to those observed with solutions containing NH3 only and discussed e a r l i e r , where i t was found that the i n i t i a l addition of a small amount of gold to the copper resulted i n a large decrease i n the rate, while further gold addition had r e l a t i v e l y l i t t l e effect (compare Figures 9 and 15). (3) Increasing the gold content of the a l l o y also tended to make the rate curves more parabolic as i s shown i n Figure 13. In a s o l -ution containing 0.08 m/1 of NH^, i t was found that the rate curve f o r a l l o y 2 remained l i n e a r u n t i l a considerable amount of copper had dissolved. The rate curves f o r alloys 5 and 10, on the other hand, were parabolic from the s t a r t . (4) The rates of disso l u t i o n f o r pure copper and f o r alloys 2, 5 32. - M/L O- 0.02 • - 0.04 O - o.o6 V- 0.08 \ \ \ \ \ \ \ • \ \ \ \ \ \ v \ \ \ N \ \ \ 10 15 ATOMIC % GOLD Figure 15. Effect of gold on the rates of copper dissolution i n solutions with constant NH3 and varying NH£. NH3 = 0.5 m/l> T = 25°C; 0 2 =6.8 Atm. 3 3 . and 10 increased with increasing NH4 concentration u n t i l the latter reached about 0.06 m/1. Further addition of NH£ did not increase the dissolution rates (see Figures 14 and 15)« The limiting value of the rate decreased with increasing gold con-centration i n the alloys, and became zero for the 15 atomic percent alloy. For pure copper, there i s evidence (l) that the attainment of a limiting value of the rate is due to control by oxygen transport through the solution to the surface of the metal. With the alloys, i t i s apparent that other considerations are involved. (5) With alloy 2 i t was noted that the rate curves tended toward linea r i t y when the NH^ ion concentration was increased. Thus the rate curve was parabolic below 0.02 m/1 of NH^ ion and linear at higher concentration, but with a tendency to become parabolic i n later stages of the reaction. Without going into detailed considerations of the mechanism at this point, some general conclusions relating to the effect of NH^ may be ,noted. The presence of NH£ appears to have a large effect on alloys of low gold content. This effect i s to increase the rates very markedly and to make the rate curves tend toward linearity, or i n other words, to offset the effects which were observed when only NH3 was present in the solutions and which were attributed to the presence of a film. Since i t i s not l i k e l y that NHj£ could have any effect on a f i l m consisting only of gold or a gold-rich metallic phase, there i s some suggestion here that an oxide, such as copper oxide whose s t a b i l i t y i s very sensitive to NH^, i s constituting part of the 34. fi l m structure. It should be emphasized that this applies only to + the low gold alloys. For the alloys higher in gold, NH* had a much smaller effect and i n the case of alloy 15 actually inhibited the reaction. The Effect of Varying NH^  i n the Presence of Constant NH* In view of the differences noted above between solutions containing NH3 only and those containing both NH3 and NH*, i t was decided to undertake a series of experiments on a l l the alloys i n which the NH* concentration was held constant at 0 .02 m/1 and the NH3 concentration varied between 0,25 and 1 ,0 m/1. Some typic a l results of these experiments are represented by Figures 16 to 20 and summarized in Table III. These results present a very striking feature, namely that increasing the concentration of NH3 in solutions containing both NH3 and NH* has a very similar effect to that obtained by increasing the concentrations of NH* i n such solutions and discussed earlier. In particular, the following points of similarity may be noted5 (1) Increasing the NH3 concentration caused the rate curves to ap-proach linearity. This was noted with both alloy 2 and alloy 5 . It was shown earlier that increasing the NH* concentration had a similar effect. (2) The rate increased with increasing NH3 concentration, and reach-ed a limiting value at 0.75 m/1 of NH3. At higher NH3 concen-trations a tendency was observed for the rate to f a l l off again, but the effect was small and i t s v a l i d i t y i s open to question in 35. TIME - HOURS Figure. 16 Rate curves for pure copper i n solutions with constant NH+. and varying NH3. NH* = 0.04 m/l> T = 25°Cj 0 2 «• 6.8 Atm. : 36. TIME - HOURS Figure 17. Rate curves for alloy 5 in solutions with constant NH£ and varying NH3. KH* = 0.04 m/lj T = 25°Cj 0 2 - 6.S Atm. 37 Figure IB. Rate curves for alloy 10 in solutions with constant NH* and varying NH3. NH* - 0.04. m/1; T = 25°C; 0 2 - 6.8 Atm. Figure 19. Parabolic plots for alloy 10 i n solutions with constant m% and varying NH3. NH* = 0,04 m/1; T - 25°C; 0 2 = 6.8 Atm. 39 ALLOY Figure 20. Effect of NH3 i n solutions containing constant NH* NH* = 0.04 m/1; T = 25°C; 0 2 = 6,8 Atm. f 3 40. TABLE I I I . Data on Experiments with Solutions i n which  NH^ was Varied and NHt Held Constant. •k ±±k Run NH3 Rate Linear Parab. Cu No. NH£ NH3 Alloy Titrated Type Rate Rate Dissolved i n 1 hour m/1 m/l m/1 G-63 0.04 0.15 Cu 0.14 L 48.2 48.2 G-51 0.04 0.25 Cu 0.26 L 68.4 68.4 C-62 0.04 0.25 2 0.24 I f~ - 22.8 C-47 0.04 0.25 5 0.26 P _. 1.21 26.1 C-50 0.04 0.25 10 0.25 P ea 0.42 14.3 C-97 0.04 0.35 Cu 0.35 L 75 75 C-10 0.04 . 0.5 Cu 0.49 L 76.8 „ 76.8 C-43 0.04 0.5 2 0.53 L 60.0 =. 60.0 C-8 0.04 0.5 5 0.53 P 3.60 45.0 C~9 0.04 0.5 10 0.50 P 1.14 25.5 GJ-ll 0.04 0.5 15 0.51 - 0 0 0 .C-52 0.04 0.75 Cu 0.74 L 86.6 . =. 82.8 C-68 0.04 0.75 2 0.71 I 67.2 =. 67.2 C-49 0.04 0.75 5 0.72 P - 8.29 60.5 C-53 0.04 0.75 10 0.80 P - 2.36 38.5 C-58 0.04 1.0 Cu 1.02, L 80.9 , „ 78.3 C-70 0.04 1.0 2 0.92 I ~ - 61.0 C-54 0.04 1.0 5 1.02 I (54oO) - 54.0 C-55 0.04 1.0 10 1.00 P 3.72 45.0 C-48 0.04 1.0 15 0.99 " 0 0 0 C=56 0.12 0.12 10 0.12 P 0.50 17.5 C-61 0.12 0.20 10 0.21 P — 0.91 24.5 C-62 0.12 0.25 10 0.24 I - 26 .0 C-30 0.12 0.5 10 0.49 P — 44.5 C-64 0.12 1.0 10 0.93 I — — 47.0 L - Linear; P = Parabolic! I = Intermediate; Rates i n Parentheses are doubtful. Units of parabolic rates are [mgCu/ cm^ 2/hour. Units of l i n e a r rates are mgCu/cm2/hour. view of the errors involved i n both the measurements and the method used to estimate the rate. More s i g n i f i c a n t i s the fact that the l i m i t i n g rates obtained on increasing the NH3 concen-t r a t i o n were very close to those which resulted o n increasing the NH]£ concentration. The value of t h i s l i m i t i n g rate thus appears to be c h a r a c t e r i s t i c only of the composition of the a l l o y . I t might also be noted that the effect of NH3, while q u a l i t a t i v e l y s i m i l a r to that of NH%. was smaller, considerably higher concen-t r a t i o n s of NH3 being required to achieve the l i m i t i n g rate (see Figures 14 and 20). (3) At very low NH3 concentrations the pronounced i n i t i a l decrease i n rate on addition of a small amount of gold to the copper was again observed, with an apparent minimum i n the rate occurring at about 2 atomic percent gold (see Figure 21). A s i m i l a r effect has already been noted f o r solutions very low i n NH^. Effect of Oxygen P a r t i a l Pressure. The effects of varying the p a r t i a l pressure of oxygen from 1,7 to 6.8 atmospheres were investigated f o r copper and f o r each of the a l l o y s . In t h i s series of experiments the compositions of the solutions were held constant at 0.5 m/1 NH3 and 0.02 m/1 NH£. The rates were thus w e l l below the l i m i t i n g values attained at higher NH3 or NH4 concentrations. Some t y p i c a l rate curves depicting the effects of varying the oxygen pressure on pure copper and on one of the alloys (the effects f o r the other alloys were s i m i l a r ) , are shown i n Figures 22 and 23. The results f o r t h i s series of exper-iments are summarized i n Table 17. NH^  - M/L 2 5 1 0 1 5 ATOMIC PERCENT GOLD Figure 2 1 , Effect of gold on the rates of copper dissolution in solutions with constant NH* and varying NH3. NH* = 0 , 0 4 m/1; T = 25°C; 0 2 - 6 . 8 Atm. PRESSURE O- 1.7 ATM. • - 3.4 1 2 TIME - HOURS Figure 22, Rate curves for pure copper at various oxygen pressures. NH3 = 0.5 m/1; NH* - 0.02 m/1; T = 25°C 44. TIME - HOURS Figure 2 3 . Rate curves for alloy 10 at various oxygen pressures. NH3 = 0 . 5 m/1; NH* = 0 . 0 2 m/1; T = 25°C. 45. TABLE IV o Data on Experiments in which Oxygen Pressure was Varied. ft M Run mt NH3 Rate Linear Parab. Cu Diss. No. Alloy Pressure Temp. NH3 Titrated Type Rate Rate in 1 hr. Atm. "C^ m/l m/i, m/1 me. C-88 Cu 1.7 25 0.5 0.02 0.49 L 20.0 20.0 C-83 2 1.7 25 0.5 0.02 0.51 I - 25.5 C-69 10 1.7 25 0.5 0.02 0.50 I - - 12.2 C-92 Cu 3.4 25 0.5 0.02 0.49 L 42.0 42.0 .J3=89 2 3.4 25 0.5 0.02 0.50 I - ..... ..... 31.4 C-71 10 3.4 25 0.5 0.02 0.49 P •- 0.69 18.7 C-96 Cu 5.1 25 0.5 0.02 0.47 L 53.9 mm 53.9 C-94 2 5.1 25 0.5 0.02 0.49 P - 2.0 39.0 C-74 10 5.1 25 0.5 0.02 0.50 P - 0.78 19.6 C-18 Cu 6.8 25 0.5 0.02 0.49 L 62.2 _ 62.2 C-37 2 6.8 25 0.5 0.02 0.50 I (1.51) 31.5 C-19 5 6.8 25 0.5 0.02 0.50 I - (1.41) 28.8 C-20 10 6.8 25 0.5 0.02 0.48 P - 0.53 17.2 C-101 Cu 8.5 25 0.5 0.02 0.50 L 61.1 - 61.1 C-79 10 1.7 15 0.5 0.02 0.52 I C O 11.1 C-85 10 3.4 15 0.5 0.02 0.51 P 0.64 16.8 C=76 10 6.8 15 0.5 0.02 0.50 P - 0.54 17.7 C-87 10 1.7 35 0.5 0.02 0.48 I 15.4 C-91 10 3.4 35 0.5 0.02 0.48 P 0.98 24.7 C=93 10 5.1 35 0.5 0.02 0.49 I - - 21.7 C-78 10 6.8 35 0.5 0.02 0.49 I — 22.4 ± Rate typess L = Linear; P = Parabolic; I = Intermediate; Rates in gparentheses are doubtful. Units of parabolic rates are [mgCu/cm J 2/hour. t±k Units of linear rates are mgCu/cm2/hour. 4 6 . For pure copper It was found that the rate increased with the oxygen pressure and levelled off at a constant value when the pressure was increased beyond 6 .8 atmospheres. The significance of this has been discussed elsewhere ( l ) . At low oxygen pressures the rate i s determined by the transport of oxygen to the metal surface and i s therefore proportional to the oxygen pressure. At higher oxygen pressures, the chemical reaction at the metal surface becomes rate controlling and the rate becomes independent of the oxygen pres-sure. With the alloys,however9 the effect of varying the oxygen pressure was found to be somewhat different. The rate increased with the oxygen pressure up to a maximum value, generally at about 5.1 atmospheres, and then f e l l again as the oxygen pressure was further increased (see Figure 2 3 ) . This was the case for a l l the alloys. Effect of Temperature. The effect of temperature on copper and on alloys 2 and 10 was investigated using solutions of the same composition as those used i n the investigation on pressure (i.e. 0 .5 m/1 NH3 and 0 .02 m/1 NH*). The effect of temperature on alloy 5-was investigated using solutions containing 0 o 5 m/1 NH3, but without NH*. The temperature was varied between 15 and 50°C with the pressure held constant at 6 . 8 atmospheres. Typical rate curves for copper and for an alloy are shown i n Figures 24 and 25 and the results are summarized i n Table V. While increasing temperature was found to increase the rate 48. TEMPERATURE: I O - 15°C • - 25°C T I M E - HOURS j Figure 25. Rate curves for alloy 2 at various solution temperatures. NH3 - 0.5 m/1; NH* = -.-2 m/1; 0 2 - 6.8 Atm. 49. TABLE V» Data on Experiments i n which Temperature was Varied t ±± Run NH3 Rate Linear Parab. Cu Diss. No. Alloy Temp. Pressure NH3 ml Titrated Type Rate Rate i n 1 hr. °C. Atm. m/1 m/1 m/1 C-77 Cu 15 6.8 0.5 0.02 0.50 L 56.2 56.2 C-75 2 15 6.8 0.5 0.02 0.52 I - = 29,8 C-76 10 15 6.8 0.5 0.02 0.50 P = 0.54 17.7 C=18 Cu 25 6.8 0.5 0.02 0.49 L 62.2 O B 62.2 C=37 2 25 6.8 0.5 0.02 0.50 I = (1.51) 31,5 C-20 10 25 6.8 0.5 0.02 0.48 P 0,43 17.2 6=84 Cu 35 6.8 0.5 0.02 0.51 L 61,6 61,6 C-81 2 35 6.8 0.5 0.02 0.51 P C O 0,75 32.5 C-78 10 35 6.8 0.5 0.02 0.49 I = - 22.4 C-98 Cu 50 6.8 0.5 0,02 0.49 L 82.2 C O 82.2 G=95 2 50 6.8 0.5 0,02 0.48 P - 0.33 22.7 C-82 10 50 6.8 0.5 0.02 0.51 P - 0.19 15.4 C-79 10 15 1.7 0.5 0.02 0.52 I 11.1 C-85 10 15 3.4 0.5 0.02 0.51 P ... 0.48 16.8 C-86 10 15 5.1 0.5 0.02 0.49 I - - 14.2 C-99 5 25 6.8 1.5 1.57 I 27,1 0=100 5 35 6.8 1.5 - 1.51 I — 20.5 0=102 5 50 6.8 1.5 — * 1.51 I — — 13.3 t Rate typess L = Linear! P = Parabolic; I = Intermediate; ±& Rates in parentheses are doubtful. Units of parabolic rates are [mgCu/ cm2? 2/hour. Units of linear rates are mgCu/cm2/hour. 50. for pure copper, the effect was slight (corresponding to an activation energy of the order of 320 cal/mole). This indicates some measure of rate control by transport of oxygen under these conditions. For the alloys, the effect of temperature was strikingly different. Here i t was found that, with the possible exception of a short i n i t i a l period, the rate generally decreased with increasing temperature. These results suggest that as the temperature i s increased the influence of the films in retarding the reaction becomes more pronounced. This i s supported by the results of physical examinations of the films, which indicate that the films formed at higher temper-atures are denser and would therefore be expected to be less permeable to reactants or products. Structure and Composition of the Films. Following most of the experiments, the surfaces of the a l -loy specimens were examined microscopically and i n some cases the films which had formed during the reaction were removed for analysis. A soft lead spatula was used to remove the films, which were very fragile and, when scraped from the underlying metal, broke up into a fine apparently metallic powder. Photomicrographs of typical films are shown i n Figures 26 to 33, and the results of chemical analyses of some of the films are lis t e d in Table VI. Several attempts were made to determine the crystal structure of the f i l m particles using X-ray diffraction meth-ods, but due either to very fine particle size or imperfect crystal Figure 26. Surface structure of alloy 2. Experimental conditions: NH3 =0.5 m/1; Nh*J = 0.06 m/1; T = 25°C; 0 2 = 6.8 Atm. Magnification X75. Figure 27. Surface structure of alloy 2. Experimental conditions: NH3 - 0,5 m/1; NH£ = 0.02 m/1; T = 50eC; 0 2 = 6.8 Atm. Magnification X600. Figure 28. Surface structure of alloy 5. Experimental conditions: NH3 - 0.5 m/1; T - 25°C; 0 2 - 6.8 Atm. Magnification X1000. Figure 29. Surface Structure of alloy 10. Experimental conditions: NH3 - 0.5 m/1; T - 25°C; 0 2 - 6.8 Atm. Magnification X1000. Figure 30. Surface structure of alloy 10. Experimental conditions: NH3 - 0,25 m/1; NHj - 0.04 m/1; T « 25°Cj 0 2 - 6.8 Atm. Magnification X?5. Figure 31. Surface structure of alloy 10. Experimental conditions: NH3 - 0.5 m/1; NH& - 0.02 m/1; T - 3^ 'C; 0 2 - 6.8 Atm. Magnification X500. 54. Figure 32. Surface structure of alloy 15. Experimental conditions: NH3 = 1.5 m/1; T = 25°C; 0 2 = 6.8 Atm. Magnification X85o TABLE VI. Analyses of Films. Run No. Alloy Sample wt.-mg wt. % Copper C-8 5 23.88 11.05 C-9 10 35.02 10.45 C-3 5 23.78 10.95 C-14 5 28 .74 9.50 C-16 10 40.90 9.98 C-15 10 50.56 8.82 C-20 10 19.78 10.72 C-56 5 3 0 . 4 0 8.32 C-49 5 24.62 11.50 C-54 5 43.40 12.91 C-55 10 60.44 9.12 C-48 C-46 2 21.52 8 . 5 0 C-57 C-61 10 25.86 9 . 6 0 * C-81 C-83 1 * 18.46 3 7 . 0 This was a sample of the unusual f i l m mentioned i n connection with Figure 30. 56. structure, no interpretable results were obtained, < It i s not possible at present to explain i n any detail the structures that are represented in Figures 26 to 33$ and the reasons for the variations observed, but a few features of interest may be noted. In Figure 26 i s shown the surface of alloy 2 after i t had been exposed to a corroding solution at 25°C„ V For this alloy the sur-face structure was the same whether the solutions contained NH^ or not. This i s perhaps surprising i n view of the very large effect of NH]£ on the rate for this alloy. The surface appears to be porous and coarse. Figure 27 shows the surface of the same alloy following reaction at 50°C. This f i l m was observed to be much thinner and denser than those formed at lower temperatures and can be seen to have a very d i f -ferent surface structure. This could account for the inverse temper-ature effect on the rate of dissolution of copper from the alloys which was noted earlier. The surface structures of alloys 5 and 10 after reaction i n solutions containing only NH3 at 25°C,, are shown in Figures 28 and 29o These are seen to be very similar to each other, but marked-l y different from the typical structures for either alloy 2 (Figure 26) which appears to be much more porousr or alloy 15 (Figure 32) which seems much more dense. In contrast to the observations for alloy 2 , i t was found that the presence of NH^ i n the solution had a considerable effect on the appearance of the surfaces of the higher gold alloys. This can be seen i n Figures 30 and 31 , which show surface structures of alloy 10 after exposure to solutions containing both NH3 and NH£. 57. Figure 30 illustrates an interesting effect which was often observed with the higher gold alloys, particularly after they had been exposed to reaction conditions which gave a low rate ( i.e. low NH3 concentrations, oxygen pressures or temperatures). In these instances, after the specimens were removed from the reaction vessel, their sur-faces appeared to be uncorroded except for a very thin tarnished film. However, after a short exposure to the atmosphere, cracks developed i n this f i l m and fragments of a surface layer, about 1/1000 inches thick, and metallic i n appearance, separated from the underlying material. An example of such a specimen i s shown i n Figure 30. The removal of this surface film, which analysis showed to be very high in gold, exposed another very thin layer of a black material which could readily be removed as a very fine powder to disclose the unat-tacked metal. It was not found possible to collect a sufficient amount of this powder for analysis, but i t s appearance suggested that i t could be an oxide rather than a metal. Figure 32 shows a photomicrograph of the surface of an alloy 15 specimen which had been corroded i n a solution containing NH3 alone. After the reaction, the surface appeared to the eye to be only s l i g h t l y tarnished. Under the microscope the surface f i l m appeared to be very compact, showing fewer cracks than the films of alloys lower in gold content. When the tarnished layer was removed by polishing, the underlying surface, shown i n Figure 33, was observ-ed to be metallic and obviously very rich i n gold. Further, when this gold-rich layer was removed by machining, the shavings were found to contain a fine black powder. 58, No evidence of any surface changes, such as tarnishing or f i l m formation, could be observed either visually or under the microscope, on specimens of alloy 15 which had been exposed t o sol-utions containing both NH3 and NH*. Table VI l i s t s the results of the copper analyses which were carried out on some of the surface films formed during corrosion experiments. Most of the films are seen to contain about 10 percent by weight of copper, corresponding to about 25 atomic percent- copper i f the f i l m material i s assumed to consist of a copper-goId alloy. Qualitative tests confirmed that a l l the films were very rich i n gold. The apparent constancy of the compositions of the films i s perhaps surprising i n view of the wide differences in their physical appear-ance and i n the compositions of the alloys from which they w e r e formed It i s clear from the above discussion that while the form-ation of surface films i s an essential feature of the corrosion of the copper-gold alloys, the films which are formed are very complex in nature, often consisting of more than one layer or component. Despite an apparent constancy of the composition (as reflected in the copper content) of these films, their detailed chemical and physical struct-ures have not been determined. Nor has i t been possible to relate their observable properties i n any simple way to the compositions of the alloys, the conditions of the reactions or the reaction rates. 59. CONCLUSIONS Summary of Experimental Results The general pattern of the experimental results which were obtained i n the course of this investigation may be summarized as follows;. (1) A l l the rate curves for the dissolution of pure copper were linear. In general the results obtained on pure copper were i n good agree-ment with those found i n earlier studies ( l ) . (2) Most of the rate curves for the alloys showed a decrease i n the rate with time. In many cases the shapes of the rate curves con-formed to a parabolic relation, but sometimes they were more complex, showing a greater or smaller tendency for the rate to f a l l off with time, (3) In general, increasing the gold content of the alloys resulted in lower rates of dissolution of copper and i n a greater tendency for the rates to decrease with time. This was particularly ap-parent i n solutions high in NH3 and NH£. In solutions low i n NH3 and NR\, the rates for a l l the alloys were s t i l l much smaller than for pure copper, but the rate appeared to be less sensitive to the gold content of the alloy, (k) Increasing the concentration of NH3 i n the solution (in the presence of constant NH£) resulted i n higher dissolution rates for a l l the alloys. The rates increased up to a maximum value beyond which further NH3 addition had l i t t l e effect. The 60. 1 limiting rates were inversely dependent on the gold content of the alloys. Increasing the NH3 concentration also tended to make the rate curves, particularly for the lower gold alloys, m o r e nearly linear i n shape. Increasing the concentration of NH* in the solution (in the presence of constant NH3) had a very similar effect. The rate increased with the concentration of NH*, approaching a limiting value which was the same i n the case of each alloy, as that reach-ed by increasing the NH3 concentration. Much smaller amounts of NH* than of NH3 were required to produce a comparable effect. The rates for the low gold alloys (i.e. alloy 2) were particular-l y sensitive to the NH* concentration. Those for the higher gold alloys showed a smaller dependence, and for alloy 15 the depend-ence was actually reversed, no reaction taking place i n the pres-ence of NH£ salts. The rates for a l l the alloys were found to increase with the oxygen partial pressure up to a maximum at about 5 atmospheres. At higher oxygen pressures the rates fell off again. While the rates for pure copper increased with temperature, those for a l l the alloys showed an inverse dependence, falling off as the temperature was raised. It was found for a l l the alloys that the reaction was accompanied by the formation of films, generally very rich in gold, on the surface of the corroded specimens. It was not possible to estab-l i s h the detailed structure of these films. 61 Interpretation of the Results From the nature of these results i t i s apparent that the system involved i s of such complexity and the number of variables so great, that a far more extensive study than the present one would be required to elucidate f u l l y the reactions which take place and their mechanisms. The investigation which has been carried out can there-fore be considered to be only of an exploratory nature, intended to reveal the general features of the system. Many aspects remain to be studied i n greater detail. Nevertheless, the results which have been obtained form a sufficiently complete and consistent pattern that i t i s possible to account for some of the observed effects and to draw some conclusions, of a general nature at least, concerning the kinetics and mechanisms of the reactions involved in the attack of ammonia solutions on copper-goId alloys. (1) With pure copper the reaction has been shown to proceed by chemical attack of NH3 and NH^ on the metal surface, following the rapid adsorption of oxygen, and resulting i n the dissolution of copper as the cuprammine ion. The surface of the corroding copper remains essentially unchanged, with no indication of any fi l m formation. Consequently, the rate curves for copper are linear. (2) With the alloys there i s some indication that the i n i t i a l attack and dissolution of copper proceeds in a somewhat similar fashion, except that some electrochemical reaction due to the presence of regions of different composition (i.e. gold content) on the metal surface may also be involved. This might account for the fact 6 2 , that under certain conditions, slightly higher i n i t i a l rates were observed with increasing gold content of the alloy. As the corrosion of the alloy proceeds, a film is soon formed on its surface. This film appears to be due to the accumulation of gold, as the copper in the surface region dissolves, resulting in the formation of a gold rich deposit which is itself relative-ly inert to attack and which retards the reaction by presenting a barrier to the transfer of reactants and products between the -underlying metal and the solution. This film thickens as the reaction proceeds, accounting for the parabolic shape of the rate curves (i.e. for the fact that the rate for the alloys falls off with time). There is some indication, both from examination of the specimen surfaces after corrosion and from the nature of the kinetic results, that a deposit of oxide (either Cu20 or CuO) may also form on the surface of the corroding alloys and that this deposit contributes in the same way as the gold film itself to the slowing down of the reaction. Such a deposit might be expected to accompany or to result from the formation of the i n i t i a l gold rich film, since such a film would give rise to stagnant regions (i.e. in the pores of the film or between the film and the under-lying metal) in which the solution would be deficient in NH3 and NH* and in which the reaction products (Cu + + and OH") would accumulate. These conditions would favour the precipitation of CuO and Cu20. This picture is supported by the results of exper-iments made at low NH3 and NH* concentrations, in which i t was 63. found that the rates for the low gold alloys were very low (comp-arable to those for the higher gold alloys) despite the fact that the gold film was apparently very porous. The fact that the rate passes through a maximum with increased oxygen pressure and then falls off again might also find an ex-planation in the above picture. At low oxygen pressures, the transport of oxygen through the film may control in part the rate of the reaction. However, as the oxygen pressure is in-creased, the rate is increased to a point where there is insuf-ficient NH3 and NH£ to prevent the accumulation of Cu + + and OH™. The resulting precipitation of copper oxides thus causes the rate to drop again. This is comparable to the well known system involving the corrosion of iron in contact with an aqueous solution, where increasing the concentration of oxygen may first accelerate corrosion, but at high enough oxygen concentrations a passivating film of iron oxides is formed and corrosion inhibit-ed o The inverse effect of temperature on the rates of reaction of the alloys may be similarly explained. At higher temperatures the formation of denser film structures would be favoured by mechanical failure, as would certainly the precipitation of cop-per oxides. The increase of the rate with NH3 and N H 4 concentrations may be due to a higher rate of transport of these reactants through the films to the underlying metal which is dissolving. At the same time higher NH3 and NH£ concentrations would reduce the tendency for copper oxides to precipitate, and this would also cause the rates to increase. Ultimately the rate must become controlled by the transport of oxygen through the gold film which is always present, and whose thickness and permeability depend primarily on the composition of the alloy. This would correspond to the limiting rate region, in which further increases in the N H 3 or NH* concentrations are without effect„ It might be ex-pected that in this region the rate would increase with the oxygen pressure. No explanation can be offered at this point for the fact that the alloy containing 15 atomic percent gold did not corrode i n solutions containing both NH3 and NH*, while i t did undergo re-action in solution containing NH3 only,, 65. APPENDIX I. S.pectro chemical Analyses of Metals and Alloys Used. Pure Copper Pure Gold Alloy 5 Alloy 10 Alloy 15 Cu Major component 0.001-0.005 Major component Major component Major component Au - Major component Major component Major component , Major component S i 0.001-0.01 0.001-0.01 0.001-0.01 0.001-0.01 0.001-0.01 Mg 0.001-0.01 0.001-0.01 0.001-0.01 0.001-0.01 0.001-0.01 Ca 0.001-0.01 0.001-0.01 0.001-0.01 - 0.001-0.01 Ag 0.CO1-0.01 0.001-0.01 0.005-0.01 0.005-0.01 0.001-0.01 Fe 0.0001-0.001 - 0.001-0.01 0.001-0.01 0.005-0.01 Cr 0.0001-0.001 - 0.0001-0.001 0.001-0.01 0.0001-0.01 V - 0.001-0.005 T i - - 0.001-0.005 Mo - 0.0001-0.001 -The above figures are percent. Due to the wide variations i n the compositions of the samples, even the percentage ranges quoted for the impurities can only be regarded as giving the approximate order of magnitude of the concentrations present. APPENDIX II Assays for Gold on Copper-Gold Alloys. Alloy No, Cuttings Drillings Average Atomic $ Au Wt. #Au Wt. % Au Wt. % Au 2 6.00 6.00 2.02 5 13.99 14.12 14.05 5.01 10 25.73 25.64 25.69 10.00 15 35.56 35.56 35.56 15.11 67 REFERENCES 1, Jo Halpern , J . E lectrochem. S o c , i n p r e s s . 2 0 Go Tammann, Z . anorg. a l lgem, Chem., 142, 61=72 (1925)o 3. W.JoMul ler , H. F r e i s s l e r , and E:. P l e t t i n g e r , Z*. E lek t rochem. , 42 . 366-71 (1936). 4P RO Landau and G.SoOldach, Trans . Electrochem. S o c , 81. 521-558 (1942). 5o U .R.Evans, ' M e t a l l i c C o r r o s i o n , P a s s i v i t y and P r o t e c t i o n " , A r n o l d , p. 229. 6. A . S . R u s s e l l , Nature 115, 455-6 (1925); 117, 47-8 (1928), J . Chem. S o c , 1872-81 (1926). 7. H . H . U h l i g , " C o r r o s i o n Handbook", John Wi ley and Sons, p. 21. 8. H .H .Uh l ig and J . W u l f f , T r a n s . , A.I.M .E... 135. 494-521 (1939). 9. R. Landau, T rans . E lectrochem. S o c . , 81, 559-571 (1942). 10. C.Sykes and F.W.Jones, P r o c , Roy. S o c L o n . , A 157. 213-233 (1936). 11. E . B . S a n d e l l , " G o l o r i m e t r i c Determination of Traces of Metals", I n t e r s c i e n c e , (1950). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0106438/manifest

Comment

Related Items