Kinetics and Mechanism of Copper Dissolution In Aqueous Ammonia

Berichte der Bunsengesellschaft fur physikalische Chemie (friiher Zeitschrift fiir Elektrochemie) Band 67, Heft 4, 1963 (Seite 402-406) Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia By Fat hi Hahashi Verlag Chemie, GmbH., Weinheim/Bergstr. 402 F. Habashi: Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia Berichte der Bunsengesellschaf Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia By FATHI HABASHI NRC Research Fellow, Department of Mines and Technical Surveys, Extraction Metallurgy Division, Ottawa, Canada (Eingegangen am 26. September 1962) The dissolution of copper in aqueous solutions of ammonia is a corrosion process in which the cathodic reaction is the reduction of oxygen at the surface of the metal according to: V8Oa + H20 + 2e~ -* 20FT and the anodic reaction is the oxidation of copper according to: Cu + 4NH3 Cu(NHs)f + 2 e~. Superimposed with these reactions is the cuprous-cupric equilibrium: Cu(NH.,)f+ + Cu ^ 2 Cu(NH,)+ Cu(NH3)+ + 2NH3 -> Cu(NH3)f+ + e". A theoretical derivation of the velocity equation of the dissolution process has been obtained, which describes quantitatively the experimental facts. The equation is as follows: where klt k2, ks and/ are constants, A is the surface area of the metal in contact with the aqueous phase and [Os], [NH3] and [Cu2+] are oxygen, ammonia and cupric ion concentrations respectively. Die Auflosung von Kupfer in waBriger Ammoniaklosung ist ein KorrosionsprozeB, bei dem als kathodische Reaktion die Reduktion von Sauerstoff an der Metalloberflache gemaB: V202 + H.,0 + 2e" -+ 20H" auftritt, wahrend die anodische Reaktion in der Oxydation des Kupfers besteht: Cu + 4NH3 Cu(NH3)f + 2e". Diesen Reaktionen ist das Gleichgewicht zwischen Cu(II)/Cu(I)-Ionen iiberlagert: Cu(NHx)- + + Cu ^ 2 Cu(NH3)+ Cu(NH3)+ + 2NH3 Cu(NH3)f + e". Es wird eine theoretische Ableitung einer Geschwindigkeitsgleichung fiir die Kupferauf losung gegeben, die quantitativ die experimentellen Erfahrungen wiedergibt. Die Gleichung lautet: Geschwindigkeit = + M[CW' — ./([NH,]). Es bedeuten: kl3 fe2, fe3 und/Konstanten, A die Metalloberflache in Kontakt mit der waBrigen Phase sowie [Oa], [NH3] und [Cu2+] die Konzentrationen der betreffenden Stoife. Introduction to purify the synthesis gas from carbon monoxide, The solubility of metallic copper in aqueous ammonia which otherwise would poison the catalyst, solutions was known since 1858 by Peligot1). The The actlon of aqueous ammonia on metallic copper blue solution obtained from this reaction was used also received attention in extractive metallurgy, and it extensively to dissolve cellulose to manufacture the was suggested as a possible process for recovering copper cuprammonium rayon, Schweizer"). In the ammonia from ores> Benedict*). In the USA this process was synthesis industry, this blue solution is used universally aPPlled successfully to leach native copper from tailings _ too low in copper (0.4% Cu as metal) to be effectively !) E. Peligot, C.R. hebd. Seances Acad. Sci.47,1034 (1858). 2) E. Schweizer, J. prakt. Chem. 76, 344 (1859). 3) C. H. Benedict, U. S. Patent 1,131,986 (1915) Bd. 67, Nr. 4 1963 F. Habashi: Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia 403 recovered by any other process, Benedict and Kenny4). Similar processes were also applied to recover nickel INCO staff5) and cobalt Car on6) from ores after a simple reduction treatment to the metallic state. A knowledge of the mechanism of these leaching processes is therefore of interest to the chemical as well as the metallurgical industries. In this paper, the mecha- nisms previously suggested are criticized and a new mechanism is discussed which overcomes their draw- backs. The similarity between this reaction and the cyanidation of gold and silver is noted. Chemistry of Dissolution It was realised since the discovery of this reaction that no dissolution takes place in the absence of air, Schwei- zer2). The reaction was formulated as follows: Cu + 4NH3 + VsOj + H20 Cu(NH3)f + 2 OFT. (1) There is evidence that some nitrite is formed during the dissolution; copper acting as a catalyst for the oxidation of ammonia, Schonbein7): 2NH3(aq) + 3°2 + 20H" - 2N°7 + 4H20. (2) No other oxidation-reduction products such as H2Oa or cuprous ions were detected in solution, Halpern8). Ammonium ion alone does not dissolve copper; but when added to NH3 solutions it increases the rate, Schweizer2). This was confirmed by Fisher and Halpern9), who found further that beyond a certain NH4/nh3 ratio further addition of NHJ has no effect; apparently due to a buffering action. Halpern et al.10) found further that when NaOH is added to the solution, the rate decreases. This leads us to attribute the accel- erating effect of NHJ to be two-fold; first increasing the concentration of free ammonia, and secondly decreasing the OH" concentration as given by the following two reactions: NHa + HzO <=> NH+ + OH', (3) NHJ + H20 <=>NHS + HaO+. (4) A thermodynamical study of the reaction was pres- ented by Halpern11). Mechanisms suggested 1. The auto-catalytic mechanism The dissolution was assumed to take place in three steps a) Formation of cuprous-ammine complex 2 Cu + 4 NH, + VtO, + H20 ->• 2 Cu(NH3)+ + 2 OFT . (5) 4) C. H. Benedict and C. H. Kenny, Trans. Amer. Inst. Mining Metallurg. Engr. 70, 595 (1924). 5) The International Nickel Company Staff. Canad. Mining Metallurg. Bull. 59, 201 (1956). *) M. H. Caron, Trans. Amer. Inst. Mining Metallurg. Engr. 188, 67 (1950). ') C. F. Schonbein, Ber. Akad. 580 (1856) and later papers. 8) J. Halpern, J. electrochem. Soc. 100, 421 (1953). 9) J. I. Fischer and J. Halpern, J. electrochem. Soc. 103, 282 (1956). 10) J. Halpern, H. Milants and D. R. Wiles, J. electro- chem. Soc. 106, 647 (1959). ") J. Halpern, J. Metals 9, 280 (1957). b) Oxidation of the cuprous-ammine to cupric- ammine by oxygen: Cu(NH3)+ + 2NH3 + VjOjj + H20 -» Cu(NH8)f + 2 OH". (6) c) The reduction of the cupric-ammine to cuprous by copper: Cu(NHs)f + Cu *=fc 2 Cu(NH8)+. (7) It is apparent from this mechanism that the cuprous +± cupric equilibrium plays the role of a catalyst. Supporters of this mechanism were: Yamasaki12), Zeretskii and Akimov13), and Lu and Graydon14). 2. The copper-oxide-film mechanism In this mechanism the dissolution of copper was believed to follow the following steps: a) Adsorption of dissolved oxygen onto the copper surface fast Cu + V.O, —*■ Cu-O. (8) b) Reaction of an NH3 molecule with the copper- oxygen complex Cu ... O + NH slow fCu< XT fast Cu(NH,)2+ + 2 OH" Cu(NH3)2+ + 3NH3 -> Cu(NH3)f (9) This mechanism was originally suggested by Lane and McDonald15)16) who thought that CuO was formed, and later modified by Halpern8) who stressed that oxygen was only adsorbed on the copper surface. This view was maintained by Fisher and Halpern9), Halpern et al.10), and Sircar and Wiles17). The last two studies were concerned with complexing agents other than ammonia. The first mechanism was criticized by Halpern8) on the following grounds: 1) No cuprous ions were detected in solution. 2) The addition of cupric ion did not affect the rate. The first point is not strong enough, because if the rate of oxidation of Cu+ to Cu2+ is faster than its rate of formation, as would be expected since it is a homo- geneous reaction, then no Cu+ would be detected. The answer to the second point was clarified by Lu and Graydon14), who found that in order to observe catalytic action, the cupric ion must be added in large amounts; (they did not however mention how large), and the oxygen concentration should be low as compared with the ammonia concentration. 12) E. Yamasaki, Sci. Rep. Tohoku Imp. Univ. Ser. 1,9, 169 (1920). 13) E. Zeretskii and G. Akimov, Zhur. prikl. Khimii //, 1161 (1938). 14) B. C. Y. Lu and W. F. Graydon, J. Amer. chem. Soc. 77, 6136 (1955). 15) R.W.Lane and H. J. Mc Donald, Corrosion & Material Protection 2 (5), 17 (1945); 2 (6), 15 (1945). 16) R.W.Lane and H. J. Mc Donald, J. Amer. chem. Soc. 68, 1699 (1946). 17) S. C. Sircar and D. R. Wiles, J. electrochem. Soc, 107, 164 (1960). 404 F. Habashi: Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia Berichte der Bunsengesellschaft The second mechanism in our opinion, is not likely to take place because one would expect that cupric oxide would dissolve more rapidly than copper, which is not the case. If copper is covered by any form of oxide or adsorbed oxygen, dissolution will be inhibited, Schwei- zer2), Yamasaki12). In this paper a new mechanism is discussed which overcomes all these difficulties and at the same time is in complete agreement with the experimental data. The Electrochemical Nature of Dissolution The generally accepted theory of corrosion of a metal, Evans18), is that when a metal comes into contact with an aqueous salt solution to which oxygen is accessible, oxygen takes up electrons at one part of the surface (the cathodic zone) while the metal gives them up at another (the anodic zone). In this way the attack of metal proceeds at an appreciable rate at room temperature. These principles are well established and they were successfully demonstrated in many cases, e.g. the dissolution of zinc in sodium chloride solution in contact with air, or gold in cyanide solution saturated with air, Thompson19). The same mechanism is supposed to take place during the dissolution of copper in ammonia solutions saturated with air. Two simultaneous reactions are believed to take place, (Fig. 1): Fig.l Schematic representation of the electrochemical dissolution process of copper in ammonia solution 1) The cathodic reduction of oxygen at the surface of copper: VaOj + H20 + le — 2 OH". (10) 2) The anodic dissolution of copper in the presence of the complexing agent: Cu + 4NH3 -> Cu(NH?)|+ -f 2e . (11) The overall reaction will be equation (1). 1S) U. R. Evans, The Corrosion and Oxidation of Metals. Arnold, London (1960). 19) P. F. Thompson, Trans, electrochem. Soc. 91, 41 (1947). There is no doubt that superimposed, with these reactions is the cupric-cuprous equilibrium which causes more copper to dissolve: Cu + Cu(NH„)f <=± 2 Cu(NrL,)+, (7) Cu(NH3)+ + 2NH3 -> Cu(NH3)f + e". (12) It is apparent that these last two reactions are of importance only at high cupric ion concentration. Let us consider the simple case when the concen- tration of Cu2+ is negligible. We make two assumptions, whose validity will be proved later. These assumptions are: 1) The cathodic reaction is a diffusion-controlled process. 2) The anodic reaction is a chemically-controlled process. From the first assumption it follows that the rate of oxygen diffusion through the Nernst boundary layer will be given by: d (02) DQ -If = -f ^iltOJ - [OJJ (13)*) where Doi = the diffusion coefficient of dissolved oxygen d = the thickness of the boundary layer A1 = the surface area at which the cathodic reaction takes place [02] = oxygen concentration in the bulk of the solution [02]s = oxygen concentration at the surface of the metal Since diffusion is the controlling factor, then the rate of oxygen reduction at the surface is very rapid as compared with the rate at which it diffuses, i.e. oxygen is consumed as soon as it reaches the surface, or [OJs = 0. Therefore, d (02) D0 -j^-- f^tP,]. (14) From equation (1), two equivalents of copper are dissolved when one mole of oxygen is consumed, therefore, Rate of copper dissolution = 2 = 2fe1^[02] (15) where Do fex=-p. (16) From the second assumption it follows that the rate of chemical reaction between ammonia and copper is much slower than the rate at which NH3 diffuses to the surface of the metal through the boundary layer, since the chemical reaction in this case is rate-controlling. This can be expressed approximately**) as follows: Rate of copper dissolution = fe2-42[NH3] (17) *) The curved brackets ( ) represent moles, while [ ] re- present moles/liter. **) The dependence of the rate of dissolution on ammonia concentration is not exactly linear, due to a side reaction, namely, the oxidation of ammonia by oxygen leading to nitrite formation. Bd. 67, Nr. 4 1963 F. Habashi: Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia 405 where fe2 = velocity constant of the chemical reaction (expected to be much lower than the diffusional velocity constant) A2 = the surface area at which the anodic reaction takes place [NHS] = the concentration of free ammonia At the steady state, the rate of cathodic reaction (diffusion of oxygen) = the rate of anodic reaction (chemical attack by ammonia) i. e. 2fe1i41[02] = fej-duPSTHa]. (18) But since A = A1 + A1 (19) where A is the total surface area of the metal in contact with the solution, therefore Rate of copper dissolution = 2m2^[02HNH,] (20) + fe2[NH3] From equation (20) it is seen that at low oxygen concentration and high ammonia concentration, the term 2 kx [Oz] can be neglected in the denominator, and the velocity equation becomes: Rate = 2fe1^4[Os] (21) (i.e.) the rate of copper dissolution under these con- ditions depends only on the oxygen partial pressure. 2 4 6 8 02 PRESSURE- ATM. Fig. 2 * Effect of oxygen pressure and ammonia concentration on the rate of dissolution of copper in ammonia solution, Halpern8) Table 1 Effect of Speed of Stirring and Temperature (Activation Energy) on the Rate of Dissolution of Copper in 0.5 mol/1 NHa, Halpern8) Speed of stirring rpm PQi = 1.4 atm*) P0a= 7.8 atm**) Rate mg • cm-2]!!-1 AE Kcal Rate mg • cm-'hr1 AE Kcal 470 15.0 30.0 545 17.6 29.0 660 19.3 1.33 30.0 5.54 820 21.6 31.9 *) Region of low oxygen pressure and high ammonia con- centration; rate dependent only on oxygen partial pressure. **) Region of high oxygen pressure and low ammonia con- centration; rate dependent only on ammonia concentration. This is in agreement with experiments as shown in Fig. 2. Further, the term fex is a diffusion velocity coefficient which therefore should depend markedly on the rate of stirring of the solution and only slightly on the temperature, (i.e. the activation energy is low). This is in agreement with the experimental results shown in Table 1, thus verifying the first assumption. Similarly, at high oxygen pressure and low ammonia concentration the term k2 [NH3] can be neglected in the denominator and the velocity equation becomes: Rate = fe2^4[NH3] (22) which is also in agreement with Fig. 2. Here, the con- stant k2 is a true chemical reaction velocity constant, therefore it should be expected that the rate is in- dependent of the speed of stirring and the activation energy should be higher than 4 Kcal. This is in agree- ment with experimental results shown in Table 1, thus verifying the second assumption. On the other hand, when 2fe1[02] = fe2[NH3] the velocity equation becomes: Rate,, V, (23) (24) =lim = l/2feife2^[C2]/UNH3 This represents the critical points shown in Fig. 2 when the curve changes its slope. Under these condi- tions it is seen that [NH8] 2fet constant. (25) [o2- Table 2 gives the values of ammonia and oxygen concentration at the critical points, from which it is seen that the ratio [NH3]/[Oa] is actually constant and equals Table 2 [NHjJ/fOJ Ratio at Limiting Rate of Dissolution of Copper at Different Ammonia and Oxygen Concentrations. Calculated from data by Halpern8) [NH3] Moles/1 PQ2 atm Decreaset) in 02-Solubility % [OJ Moles/1 [NHS] [OJlim 0.26 0.8 9.0 0.9 • 10-3 290 0.52 1.7 19.6 1.7 • 10-3 306 0.74 3.3 25.0 3.1 • 10-3 240 1.00 4.0 33.6 3.5 • 10-3 285 Average ratio 280 t) Estimated from data given in A. Seidell, Solubilities of Inorganic Compounds, Volume I, p. 1355, Van Nostrand, New York 1953, for the Solubility of Oxygen in Aqueous NaOH Solutions. 280*). Thus supporting the mechanism suggested. Since be the value of DQ is known (= 2.79 • 10~6 cm2 sec then the thickness of the boundary layer 6 can *) Halpern et al. obtained curves similar to those shown in Fig. 2 for the dissolution of copper in aqueous solutions of ethylenediamine solutions. From these curves we found also that there is a constant ratio between the amine and oxygen concentration at which the limiting velocity occurs: [Ethylenediamine] [OJ 15.6. 406 F. Habashi: Kinetics and Mechanism of Copper Dissolution in Aqueous Ammonia Berichte der Bunsengesellschaft calculated from equation (21) (fej = DQJd). Using Hal- pem's data8), we calculated*) 8 = 1.1 • 10'3 cm, from which kx = 2.54 ■ 10~2 cm sec~x. Therefore 2fe, fe. 280 = 1.81 ■ lO-'cmsec"1. The value of k2 so obtained is much smaller than a diffusion velocity constant, as already expected. From equation (10) it follows that the rate of dissolu- tion should decrease with increasing OH- ions (reaction product). This was proved by Halpern et al.10), who found a decrease in rate by adding NaOH. Since NH3 concentration is a function of [OH~], therefore equa- tion (20) should be empirically corrected as follows: 2k1kiA[Q!S] [NH, Rate -/([NH,]). (26) 2fe1[Oa] + fe2[NH3 When considering now the dissolution of copper in cupric ammine solution, we make the following assump- tions : 1) The rate of oxidation of Cu+ to Cu2f is very fast, ([Cu+] in the bulk of solution = 0). 2) The rate of dissolution is chemically-controlled and first order with respect to Cu2+ ion concentration, and the velocity constant = k'3. When equilibrium is established, then the rate of chemical attack of copper by Cu(NH3)|+ equals the rate of diffusion of Cu(NH.,)£ away from the surface, i. e. Rate of dissolution = fe3 A [Cu2+], Dr 'Cu+ A [Cu+j D, 2+1 '/j = M[Cu2+] K'>A[CvP] 2+1 '/s (27) Where k3 is a constant and K is the equilibrium constant of the reaction Cu + Cu(NH3)f 2 Cu(NH,)+ K - [Cut? [Cu2+] (7) Since the two dissolution processes are superimposed, then the apparent rate of dissolution will be the sum of these two velocity terms: r- - +fe^iCu2+]1/2 It is seen from this overall equation that at high ammonia and oxygen concentration, the catalytic term can be neglected. While at high ammonia and low oxygen concentration the velocity equation becomes: Rate = 2fej A [O,] + * ksA[Cu'+]lk k3A[Cu*+] '« - /([NH,]) (27)**) which was found experimentally by L u and G r a y d o n14) under similar conditions. *) Prof. C. V. King, New York University, New York; (personal communication). **) This equation is valid only when the first and third terms are of equal magnitudes, which implies a certain [02]/ [NH3] ratio. Conclusions The dissolution of copper in aqueous ammonia solutions can be represented by the following elementary reactions: VjOjj + H20 + 2e" -> 20H", Cu —>■ Cu+ + e~ Cu+ -» Cu2+ + e" Cu+ + 2NH3 — Cu(NH,)+ Cu5* + 4NH, -» Cu(NH,)f Cu(NH,)+ + 2NH3 Cu(NH,)f + e" Cu(NH,)|+ + Cu«=±2Cu(NH8)+. At high oxygen and ammonia concentration the overall reaction can be expressed as follows: Cu + 4NH3 + VjOj + H20 = Cu(NH3)f + 2 OH". The dissolution process is similar to a corrosion pro- cess in which minute parts of the metal acts as a cathodic region where oxygen is reduced to OH" ions, and anodic regions where the liberated Cu+ or Cu2+ ions react with the complexing agent. The velocity equation of dissolution was derived on the basis of this model, as follows: »- - + ^l'" -*■"*»• The second term (the autocatalytic term) can be neglected at high oxygen and ammonia concentrations, while at high ammonia and very low oxygen concentra- tion, it becomes predominant. Maximum rate of dissolution occurs when the ratio [NH, = 280. If [OJ [NH,] >280[O2] the dissolution will be controlled only by the rate at which dissolved oxygen will diffuse to the surface of the boundary layer. If [NH3]« 280[O2] the dissolution will be controlled by the rate of chemical interaction of free NH3 molecules with the metal. This is consistent with kinetic measurements. There is a close similarity between this velocity equation and that for the dissolution of gold and silver in cyanide solution, Habashi20): 2ADCN- D0 [CN-] [OJ Rate = -/(log [CM"]) 6{DCN- [ON"] +4D0i [O,]} - k log [OH-] with the exception that in the case of gold, both the anodic and cathodic reactions are diffusion-controlled. It is also possible that the dissolution of nickel and cobalt in aqueous solution of ammonia follow similar mechanism. Acknowlegments This study was carried out during a Fellowship awarded to the author by the National Research Council of Canada to which the author wishes to extend his thanks. 20) F. Habashi, Kinetics and Mechanism of Gold and Silver Dissolution in Cyanide Solution. Paper presented at the A.I.M.E. Symposium on Unit Processes in Hydrometallurgy. Dallas, Texas, USA. February 1963.