J. CHEM. SOC. FARADAY TRANS., 1994, 90(9), 1227-1231 Complexation and Precipitation Equilibria in the System Ni"-Cr"'-H,O R. Castaiio," M. A. Olazabal, G. Borge and J. M. Madariaga Kimika Analitikoaren Departamentua, Euskal Herriko Unibertsitatea, P.K. 644,E-48080 Bilbao, Spain A potentiometric and spectrophotometric study of the complexation equilibria between Cr"" and Nil has been performed. Formation of the soluble NiCrO, complex has been found and the thermodynamic formation constant has been calculated (log b';, = 2.40 f0.03), as well as its molar absorptivity values at the different wavelengths studied. The precipitation equilibria in the system Ni"-CrV'-H20 have also been investigated. A mixed precipi- tate, NiCrO, * 3Ni(OH), , has been found and its thermodynamic solubility constant has been determined (log K io= -51.1 & 0.2).Hydrometallurgical processes produce industrial wastes con- taining moderate or high metal concentrations, some of which have a high toxicity, making necessary their elimi- nation before waste disposal. In some cases, the elimination or recovery of the metals is hindered because of the lack of information about the equilibria taking place between the dif- ferent components present. One of the processes that produces waste waters with high toxicity is that corresponding to chromium plating, owing to the high toxicity of Cr". In these kinds of waste waters some transition metals like Fe'", Ni" and Cu", and Ba" are present. There is enough information about the CrV'-Ba" interaction in the literature'.' to conclude the formation of the BaCrO,(s) precipitate.In the case of Feu' and Cu" there is very little information in the literature, except for the prob- able presence of the soluble complex FeCrO,' in the Fe"' system3v4 and the CuCrO,(s) precipitate in the Cu" system.' However, no information has been found on the equilibria in the Nin-Crv' system. This lack of knowledge makes necessary an experimental study of this system. Taking into consideration the similar behaviour of the Cr0,' -anion and other inorganic anions like SO,' -, HPO,'-and C0,2- and the available information about the formation of soluble complexes and precipitates between these anions and Ni2+ (SO,'-gives a soluble complex with Ni", NiSO,;' HPO,'-a neutral complex' and CO,'- forms basic non-soluble carbonates, which are mixed precipi- tates with carbonate and metal hydroxideg), the formation of soluble and mixed solid species in the Ni1'-CrV1-H2O could also be expected.The present work was designed in order to elucidate the complexation and precipitation equilibria in the Nin-CrV'-H 20system. Knowledge of the thermodynamic equilibrium constants allows further information on the system to be obtained for different ionic media provided that an adequate method for activity coefficient estimation is available. Instead of this, information on the stoichiometric stability constants is valid only for the specific medium studied. The problem with the activity coefficients can be solved if the ionic strength used for the study of the equilibria is low, because estimation of the activity coefficients can then be performed in a very simple way.In this work the experimental conditions selected are such that data treatment can be carried out directly on the molar activity scale in order to obtain the thermodynamic stability constants of the system. Experimental Reagents The chemicals used were all of analytical grade. Stock solu- tions were prepared: K2Cr04 (Merck, pa) solutions at 5.0 x and 1.0 rnol dm-, concentrations. The Ni(NO,), solutions at 9.51 x and 0.1 mol dm-, were stan-dardized through a complexometric titration against EDTA as titrant and murexide as indicator." Solutions of HNO, (Fluka, pa) at 1.0mol dm-, and KOH (Merck, pa) at 0.1 and 1.0 mol dm -were standardized volumetrically against tris(hydroxyrnethy1)aminomethane" and potassium phthalate" using methyl red and phenolphthalein as indica-tors, respectively.Experimental Technique Two kinds of experiments were performed in this work corre- sponding to the complexation study and to the precipitation equilibria study. Complexation Equilibrium A potentiometric-spectrophotometric study of different solu- tions containing Ni" and Ni"-Crv' mixtures was carried out at different pH and metal concentration levels. A conventional pH cell: Ag I AgCl(s)IKC1 (3 mol dm- ') !test solution I glass electrode (0 was used to measure the activity of H+ in solution. The test solution had a variable composition of A mol dm-3 Ni(NO,), + B mol dm-, K,CrO, + C, mol dm-, HNO, or A mol dm-, Ni(N0,)' + B mol dm-, K,CrO, + C2 mol dm-3 KOH without constant ionic strength, but this was always known and <0.1mol drn-,.The Ni" concentrations studied were 1.0 x lop3, 5.0 x lop3 and 1.0 x lod2 mol drn-,, and for each Ni" concentration another three levels of CrV' concentrations (2.40 x lo-,, 3.68 x lo-, and 5.06 x lo-, mol dm-,) were studied. According to Bates,I2 and considering that the electrodic system was calibrated using buffer solutions of known activ- ity, the activity of H+ was calculated as {H+} = lopPH. The spectrophotometric measurements were carried out using a Hewlett-Packard 8452A Diode Array spectropho- tometer.In all cases the whole spectrum in the range 240-700 nm was recorded at 2 nm intervals. Prior to the study of the Ni"-CrV'-H20 system, the Ni"-H,O system was studied under the same experimental conditions. It was found that the spectrum did not vary with pH in this concentration range, which is in agreement with the absence of significant hydroxy complex formation in the pH range studied. Precipitation Equilibrium In order to minimize the formation of nickel hydroxide pre- cipitates, the test solutions were prepared in the following order: K,CrO, ,Ni(NO,), ,KOH. Table 1 Proposed results of the thermodynamic formation con-stant and molar absorptivities for the NiCrO, complex ~ results log fi;lo &274nm '350 nm '400 nm -eqn.(I) 2.409 k 0.035 3367 f 14 2558 9 1796 k 5 eqn. (11) 2.386 L-0.031 3460 k 19 2601 & 13 1818 11 proposed 2.40 i-0.04 3399 f 48 2573 & 33 1796 13 Reagent addition was performed dropwise under contin- uous stirring to avoid, as much as possible, local precipitate formation. The possibility of contamination and evaporation was minimized. Three sets of experiments were carried out covering the Nil'. CrV1 ratios: 2 : 1, 1 : 1 and 1 : 2. The test solutions in contact with the solid phase were kept at 25 1"C during the time needed to reach equilibrium, i.e. ca. one week. After that, the pH values of the saturated solu- tions, in contact with the precipitates, were measured with the galvanic cell described above [cell (i)].Analysis of nickel in solution was performed by atomic absorption spectros- copy (AAS) measurements using a Perkin-Elmer 560 spectro- photometer. On the other hand, analysis of chromium was carried out by redox titrations using Fe(NH,),(SO,) 6H,O as titrant and ferroin as indicator because there were inter- ferences of Ni" in the analysis by AAS. Practically all of the experiments revealed the presence of two kinds of precipitate, one blue (of nickel hydroxide appearance) and a second brown. Therefore, the solubility data might obey the saturation conditions for these two pos- sible precipitates. Results Complexation Equilibria The complexation equilibria in the Ni"-CrV1-H,O system can be explained by the following general equilibrium: pNi2+ + qOH-+ rCr042-Bwr (Ni),(OH)q(Cr04)~P-4-(1)Ir Owing to the variation of ionic strength in the pH range studied, graphical treatment of the data cannot be performed using the concentration scale.J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 The (-log{H+), AA1,A,, , . . . , data at different wavelengths and total concentrations of Ni" and CrV' were treated with the SPECA program,13 which takes into account the variation of the activity coefficient at each data point. Since the ionic strength is always <0.1 mol dm-3, calcu- lation of activity coefficients can be performed using the Debye-Huckel extended law. l4 A further analysis of system-atic errors was performed. SPECA minimizes the difference between the experimental absorbance and the absorbance calculated at each wave-length, assuming a set of species and their thermodynamic equilibrium constants. An iterative procedure must be solved for each data point because the absorbance equation uses equilibrium molar concentrations, while species formation is defined in terms of thermodynamic equilibrium constants.The program calculates a set of stoichiometric stability constants and a value of the ionic strength from the initial values of the thermodynamic constants and the total concen- tration of all of the species. With this set of constants and solving the mass balance equation it is possible to obtain the detailed composition of the system (i.e. the free concentration of each species). Once the system composition is obtained, it is possible to recalculate the ionic strength, I, of each solu- tion.With the value of the ionic strength and making use of the Debye-Huckel extended law the activity coefficients can now be calculated. Then, with this set of activity coefficients, it is possible to recalculate a new set of stoichiometric con- stants and restart the cycle of I and y calculation. This process is continued until the value of I and y converge. Then it is possible to obtain the calculated value of the absorbance for each point because the free concentration, of the species are known and the absorptivities, E, are given in the first instance. The program then varies the values of the ther- modynamic constants and E, until a minimum of the error- square sum is found.Two minimization strategies were employed: in the first, the sum of the squared deviations, NP and in the second, the sum of the relative squared deviations, were used for all the experimental points, N, . For each of the three selected wavelengths, 274, 350 and 400 nm, the (log{H+), A), data as well as the total concentra- tions of all of the non-reacting ionic species (K' and NO,-) are required by the program for ,each experimental point in order to calculate the ionic strength and perform the correc- tions for activity variations. In all cases, the absorbance due to the Ni contribution was subtracted from the values corre- sponding to the Ni"-Crv1 system in order to simplify the data treatment. The different statistical parameters, U, 0 (standard deviation) and R (Hamilton factor) obtained for several models with different (p, q, r) stoichiometric values, employed for the two minimization strategies, (J= -3 4 5 6 7 (N, N,)'120.3i -log{ H+} Fig.1 Experimental data, A =f( -log(H+}), for the Ni"-Cr"'-H,O system at I = 350 nm, for CCrvI= 2.40 x mol i= 1 dm-3 and different CNil1:(0)1.0 x (A) 5.0 x and (0)1.0 x mol dm-3 where N, is the number of computed constants. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Table 2 Thermodynamic formation constants of the hydrolysis, complexation and precipitation equilibria used to ascertain the stoi-chiometry of the mixed precipitate between Nil' and CrV1 species Ni2+ H+ CrO,'-log /?" ref.HCr0,-0 1 1 6.58 16 Cr,0,2 -0 2 2 14.55 16 NiOH+ 1 -1 0 -9.86 15 Ni( OH), 1 -2 0 -19.0 15 Ni,(OH)3+ 2 -1 0 -10.7 15 Ni,(OH),*+ 4 -4 0 -27.74 15 NiCrO, 1 0 1 2.40 a OH-0 -1 0 -14 15 Ni(OH), 6) 1 -2 0 -12.4 19 -10t -,-12 hNI Q -14-!$ h ;--16-z z 0'-18 -%I \{1-201 1 I t7.0 7.5 8.0 8.5 -log{H +} Fig. 2 L0g((Ni~+)~{Cr0,'-))as a function of -log{H+} 7 -4 . I1 I I 6.0 6.5 7.0 7.5 8.0 8.5 -log{H +I Fig. 3 Solubility of Ni" as a function of pH for three different Nil' :Cr"' ratios: (0)2 : 1, (0)1 : 1, (H)1 :2 1229 I I I I I , -1.5 " e >,:-2.0- -0 0 -2.5 I . I - Precipitatioa Equilibrium The solids obtained for each data point were checked by X-ray diffraction and the absence of NO3- and K+ in the precipitates was confirmed.Therefore, the presence of only Ni", CrV1 and OH- (or H+)can be assumed. Even if the nickel hydroxide is present, the formation of new insoluble species from the Ni2+, H+and Cr042- com- ponents can be expressed by the following general equi- librium : pNi2+ -4H+ + e(Ni),(H)-,(CrO,),(s) (2) with the corresponding thermodynamic solubility product: Kb* = {Ni'-}J'(H'}-~{CrO,"}' and the electroneutrality condition expressed by : 2p -q -2r = 0 (Iv) Taking into account the variability of the ionic strength, it was decided to work with activities to define the stoichio- metry of the precipitates as well as the thermodynamic con- stants. The MASBAC program" was used to compute the activity of each species in solution making use of the experi- mental {H'} data, the total concentration of all components (Ni", Cr", K+, NO3-) in the saturated solutions and taking into consideration the chemical model shown in Table 2.With knowledge of the {Ni2+}, {H+} and {Cr042-} values for each data point, a graphical treatment can be performed to ascertain the stoichiometries of the precipitates. Several log({Ni2+}p{Cr0,2-}~ us. log{H+} models were tested in order to obtain a suitable straight line with a slope adequate to fulfil the requirement of the electroneutrality condition. Fig. 2 shows the simplest combination with integer numbers which gave a straight line, with a slope equal to -6 for = 1 and P = 4.Thus, the Proposed stoichiometric indexes are: p = 4, q = 6, r = 1. The log Kh* value was obtained from the origin intercept of the plot in Fig. 2: log Ki: = 32.9 f 0.2. If the formation of the solid is defined with OH- as a com- ponent, its stoichiometry could be written as NiCrO, * 3Ni(OH), and its thermodynamic solubility product would be expressed as: KL = {Ni2c}4{OH-}6{Cr042-) with a value of pK,", = 51.1 f0.2. Discussion The thermodynamic solubility product of the mixed precipi- tate together with the thermodynamic formation constants of the soluble species, collected in Table 2 have been used in MASBAC to plot the theoretical solubility curves of Nil' and CrV1 shown in Fig. 3 and 4. Fig. 5 shows the distribution of the Ni" species with pH.This plot has been constructed with the ISP software package,18 keeping the ionic strength fixed at 0.2 mol dm-3 and considering all the species of the system collected in Table 2. As can be seen, the two precipitates are nearly always present, which agrees with the experimental observa- tions. As is shown in Fig. 3, the solubility of Ni" decreases con- tinuously with increasing pH but with a varying slope, owing to the different mean composition of the solid phase. At pHs between 7.2 and 7.7, the mixed precipitate micro, 3Ni(OH),] is the major species in the solid phase, while NiCrO, is the major complex in solution (see Fig. 5). The variation of the solubility of Ni" is controlled by the ratio of Ni" and OH- in both major species.When the pH is higher than 7.7, Ni(OH),(s) becomes the major species in the solid phase, so the variation in the solubility changes accord- ing to the new ratio. In the case of Cr"' (Fig. 4), the solubility decreases only at the beginning of the precipitation, owing to the formation of the mixed precipitate. The solubility curve for Cr"' will then display a zero slope and subsequently a positive one, until the conditions of saturation of this mixed precipitate disappear. There is a very narrow pH range (<0.5 pH units) where the NiCrO * 3Ni(OH)2 precipitate is the most important Ni" species, as can be seen in the predominance area diagram for the Ni" species shown in Fig. 6, i.e. although this precipitate is present as a minor, but important, species at a higher pH range, as can be seen in the predominance area diagram for the CrV' species shown in Fig.7. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Ni2+ -log{H +} Fig. 6 Predominance diagram of the Ni" species as a function of pH and CCrvlfor constant CNiu= 2.0 x lop2mol dm-' The soluble NiCrO, complex is the most important for the Ni" species not only near saturation conditions but also at more acidic pHs, especially when the CrV' concentration increases. Although this species (NiCrO,) does not have a very high formation constant value, it is very important because of its predominance in the Ni" system. This is valid for the mixed precipitate as well, and both species are very important for modelling the Ni"-CrV'-H,O system and should be taken into account when new treatments for waste waters with these two components are designed.Finally, one of the most important contributions of this work is the proposed experimental methodology and data treatment for the calculations of thermodynamic equilibrium constants. These constants are usually calculated by means of different correlations from the experimental stoichiometric constants, obtained previously with a more classical experi- mental methodology. Knowledge of the thermodynamic equi- librium constants has a larger applicability, but they are difficult to determine experimentally because of the variation of the activity coefficients with ionic strength. As a result, there are no computer programs available for calculation of I-log{ H +} -3.04 6 a 10 12 14 Fig.5 Distribution diagram of Nil' species as a function of pH; -log{H +} CNiu= 4.0 x loA2mol dm-' and C,,,, = 5.6 x mol dm-'. (a) Fig. 7 Predominance diagram of the CrV1 species as a function of Ni2+,(b) NiCrO,, (c) Ni,CrO,(OH),(s), (d)Ni(CH),(s). pH and CNiufor constant CCrVl= 4.0 x mol dm-' J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 equilibrium constants with data at variable ionic strength. Therefore the graphical method presented has been devel- oped in order to treat the data for the precipitation equi- libria. In order to simplify the calculations of the activity coefficients, the ionic strength of the solutions was kept at values <0.1 mol dm-3.Financial support of the University of the Basque Country, through Project UPV/EHU 1713 1O-E142/90, is gratefully acknowledged. References G. L. Beyer, and W. Rieman, J. Am. Chem. SOC., 1943,65,971. 0.Lukkari, and H. Lukkari, Suomen Kem., 1972,45B, 6. M.J. Burkhart and R. C. Thompson, J. Am. Chem. SOC., 1972, 94,2999. J. H.Espenson and S. R. Helzer, Inorg. Chem., 1969,8, 1051. S. Peterson and 0. W. Cooper, Trans. Kentucky Acad. Sci., 1951,13, 146. N.N.Greenwood and A. Earnshaw, in Chemistry of the Ele- ments, Pergamon Press, Oxford, 1984. S. Katayama, Bull. Chem. SOC.Jpn., 1973,46, 106. 1231 8 C. M. Fry and J. E. Stuehr, J. Am. Chem. SOC., 1972,94,8898. 9 F. Burriel, F. Lucena, S. Arribas and J. Hernandez, in Quimica Analitica Cualitativa, Paraninfo, Madrid, 1983. 10 G. H. Jeffery, J. Bassett, J. Mendham and R. C. Denney, in Vogel’s Textbook of Quantitative Chemical Analysis, Wiley, New York, 1989. 11 R. G. Bates and H. B. Hetzer, Anal. Chem., 1961,33, 1285. 12 R.G.Bates, CRC Crit. Rev. Anal. Chem., 1981, 247. 13 R. Cazallas, M. J. Citores, N. Etxebarria, L. A. Fernandez and J. M. Madariaga, Talanta, 1993,submitted. 14 R. A. Robinson and R. H. Stokes, in Electrolyte Solutions, Butterworths, London, 1959. 15 C. F. Baes and R. E. Mesmer, in The Hydrolysis of Cations, Wiley, New York, 1976. 16 M. A. Olazabal, G. Borge, R.Castaiio, N. Etxebarria and J. M. Madariaga, J. Solution Chem., 1993,22,825. 17 J. M. Madariaga, in preparation. 18 I. Puigdomenech, TRITA-OOK-.JOZO,Dept. Inorg. Chem., The Royal Inst. Technol. (KTH), Stockholm, September, 1983. 19 A. Ringborm, in Complexation in Analytical Chemistry, Wiley-Interscience, New York, 1963. Paper 3/06309G; Received 22nd October, 1993