[Avenican JOURNAL oF SciENcE, Vor. 281, SuMtER, 1981, P. 935-962] THE THERMODYNAMICS OF CATION HYDROLYSIS* BAES, JR. and R. E, MESMER Chemistry Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 87830 fc! ABSTRACT. Data for the hydrolysis of aqueous cations are reviewed, enthalpy and entropy values for the variows kinds of reactions ste compiled, and correlations ae rawa wth the object of predicting behavior at elevated tenipefatures, Such Informa tion is eowential for understanding the chemisuy and the transport of metallic clements in hydothermal environments. The enthalpy of the fist hydrolysis sep, M'"-+H,O Sa MOH Hts is always positive and taually neat the enthalpy of diswciaion of Sater (183 Keal/mioie): hence, hydvolysi in solution increases with temperature. ‘The ethapy for the precipitation reaction, that is, Me" + 31,0 <= M(OH),(@) 28, Is tisually positive at well, but as the temperature rises hydrolysis fn solution shoud pro ead further before precipitation occu. Polynucear species ate less important i sol tion as the temperature fncreates because of decreasing stability and because the ma nium concentration of the hydrolyzing fon in soliton is lotered by the decreasing Solubility of the oxide or hydroxide. ‘The limited data available for the formation of Mccesive mtovionuclear hydroxide complexes by the reaction, M(OH),t"-2" + HO <= M(OM),..0-"—" +H siggest the accompanying entropy change will decrease with the chaige on the reacting species and that the enthalpy of such successive hydrolysis eps will decrease for most cations. Consequently the pit range of dominance of exch species, especially the neutral one, should increase with temperature. The equilibrium iiween the oxide or hydroxide solid phase andthe neutral. species SMOW),(aq)- Which usually determined the minimum solubility of the solid, is Expected to. havea Small temperature dependence. In more alkaline solutions, where anionic complexes fre formed, the solubility should show a positive temperature coefficient at-a given DH. Projections based on these corvelaions of hydrolysis behavior at elevated. temperature depend on a knowledge of the stepwise hydrolyse constants for monomictear species, Sind move sc information is badly needed. INTRODUCTION When a metallic element enters an aqueous solution, its chemical behavior is of course determined by the nature of the ionic or molecular species that it forms. In the absence of strongly complexing ligands and depending on the acidity of the solution, the simple cation of the element often reacts with water itself (hydrolyzes) to form complexes with the hydroxide ion. These complexes may be cations, neutral molecules, or anions, They may be mononuclear or polynuclear; that is, they may con- tain one metal atom or several. These hydrolysis reactions play a dom nant role in the aqueous chemistry of most metallic elements. Moreover, their importance grows as the temperature is increased. Knowledge of this class of reactions therefore is essential in dealing with a wide variety of aqueous systems encountered in the environment, in geochemistry, and in numerous technological applications, The hydrolysis of metal ions in aqueous solution has been studied extensively at or near room temperature for the last 30 yrs (see reviews by Sillen and Martell, 1971; Baes and Mesmer, 1976; Smith and Martell, 1976). The potentiometric determination of pH has been the principal method of measurement, but its sensitivity has generally limited measure- * Research sponsored by the Division of Chemical Sciences, Office of Basic Energy wes, US. Department of Energy under contract W-7405-eng-26 with the Union je Corporation. Sci Cant 935936 C.F. Baes, Jr. and R. E. Mesmer ments to metal concentrations above 0.001 molal (m). Polynuclear hy- drolysis species often form in such solutions and consequently have been the principal subject of study. In the more dilute solutions produced by the hydrolytic precipitation that usually occurs with increasing pH or temperature, the dominant species are mononuclear ones and are often less well characterized. Increasingly, measurement of the solubility of the hydroxide or oxide phase has been the method of choice in the study of such solutions, Measurements are often tedious and time-consuming, however, and progress has been slow. As a consequence the stability of important hydrolytic species of many cations in aqueous solutions, es- pecially at elevated temperatures, remains unknown, In the absence of such information, the only alternative is to project behavior in dilute solutions or at elevated temperatures from the data available for the element of interest and to make use of whatever general- izations and correlations about the thermodynamics of the hydrolysis of cations that may be expected to apply. It is the primary purpose of this review to consider the extent to which estimates of behavior at elevated temperatures can be drawn from what is presently known of the thermo- dynamics of the various hydrolysis processes that can ocur. First of all, the important kinds of hydrolysis reactions will be review- ed briefly. Then from the available data some correlations will be made of the equilibrium constants, the enthalpies, and the entropies of these reactions. Finally an attempt will be made to project typical behavior for the hydrolysis of metal ions as the temperature is increased. KINDS OF HYDROLYSIS REACTIONS ‘The initial step in the hydrolysis of a cation is usually the formation of the species MOH@—)+, and is generally represented by the reaction Mt + H,0 = MOH@-»+ + H+, q) ic representation of this process is the reaction M(OH,),#+ == M(OH)(OH,),_,"-+ + H+, @ in which a proton is lost by a solvating water molecule, It is perhaps not surprising, therefore, to find that the enthalpy of this reaction, AHL,, is usually close to the enthalpy of dissociation of water (13.3 kcal/mole at 25°C). If the metal ion concentration is low enough to avoid precipitation as the pH is increased, additional hydroxide ligands are formed (that protons are lost by additional water ligands) in a stepwise fashion up to a total that often reaches four or more. M+ > MOH@-0+ > MCOH),@-2+., A more reali M(OH), 0+, ‘The equilibrium quotients for the formation of such mononuclear species from the unhydrolyzed cation will be designated as Qyy, Quae-Quyr = [MOH e+) [F+)/[Me+] = [M(OH),-) +) [H+}/[Me+] 8)J kmetes/as) The thermodynamics of cation hydrolysis 987 (wherein the bracketed quantities denote the molality of each species). More often we will refer to the limiting value of Q at zero ionic strength, which will be denoted by K. At concentrations >10-* m, polynuclear species often become dom- inant soon after the appearance of the initial mononuclear species. Their formation may be represented by the general reaction xM*+ + yH,O = M,(OH),(-"+ + yH+, (4) The equilibrium quotient for such a reaction will be designated Q,y, Quy = [Mx(OH),7—-7 + ][F+ 2 /[Me¥]s, ® A variety of such species have been identified, ranging from dimers to size- able polymers. For metals of valence IV or less, they are usually cations. ‘The extent of hydrolysis of a cation in solution at ordinary concen- trations is usually limited by the appearance of a hydroxide or oxide pre- cipitate, It will be convenient to consider the equilibrium of such a phase with the unhydrolyzed cation. M(OR),(¢) + 2H+ = M+ + 2H,0. © ‘The equilibrium quotient will be designated Quo, Quo = [MeV] (7) ‘The majority of such precipitates are gelatinous, amorphous, active solids of high surface energy, and this has greatly limited the usefulness of solu- bility measurements involving them. ‘The relationship that typically exists among the various hydrolysis reactions is illustrated in figures 1 and 2. In figure 1 the concentrations of species at equilibrium with an oxide or hydroxide phase are plotted on TT TTY 25°C | 2 mim; MOM moms / | on Li Neer i . 5 7 ; . 3 Fig. 1. Distribution of hydrolysis products in solution at equilibrium with a hydroxide solid. phase for a hypothetical trivalent cation. The heavy curve is the total concentration of M(III) in solution.938 C. F, Baes, Jr. and R. E. Mesmer a logarithmic scale versus the pH. ‘These concentrations are controlled by reactions of the type xM(OH),(c) + (xz-y)H+ = M,(OH),'#—-9)+ + (x2-y)H.O, (8) and, except for the effects of variation in water activity or ionic activity coefficients the resulting curves on such a plot are linear. The slope of each line is equal to —(x7—y), which is the negative of the charge on the corresponding species if no charged ligands other than OH— or O% are bound, The heavy curve shown in figure I represents the total concentra- tion of the metal (of valence 7) in a saturated solution. Over most of the pH range this total concentration is so low that only mononuclear species are formed in appreciable amounts, and, as can be seen, it is the tability of these species that largely determine the shape of this “solubility” curve, The solubility of the oxide or hydroxide phase initially decreases rapidly with increasing pH but levels off as hydrolysis in solution pro- ceeds, It reaches a minimum, often with the neutral species M(OH), dominant in solution, and finally increases again as anionic species ap- pear. Because the polynuclear species formed by metals of valence IV or less usually have a relatively high positive value for the quantity (x2—y), they generally contribute to the solubility curve in the region of lowest pH and highest solubility. Fig. 2. Regions of dominance of the hydrolysis products in solution for a hy shel vent cation, The boundaries ae dened by he equa Hed in able The curve (from fig. I) represents the solubility limit of the hydroxide solid. phase ‘Tats igure ait hgo'l, 9, ahd 8 were consrucicaastushing only values for Ke and the typteal spacing of 1.8'Tog units per each mononuclear hydrolysis step. ‘The other Boundlries were inferred trom correlations described in the text.The thermodynamics of cation hydrolysis 939 ‘The regions of stability of the various species in solution can be seen more clearly in figure 2. In this diagram the boundaries denote conditions of pH and total metal concentration under which the species in adjacent regions will be present in equal concentrations. Movement away from a boundary will, of course, cause one species to become dominant and the other relatively less abundant. The heavy curve in the figure again indi- cates the solubility of the oxide or hydroxide phase. At low metal con- centrations successive mononuclear species become dominant with in- creasing pH. ‘The polynuclear species, here represented by the typical dimer M,(OH).*, are generally confined to a relatively small region at the lower end of the pH range and at high metal concentration. More- ‘over, as can be seen, this region is usually supersaturated with respect to the oxide or hydroxide phase. In the more basic supersaturated solutions the mononuclear species of metals of valence IV or less usually become dominant, because they reach higher y/x (that is, OH/M) ratios than do the polynuclear species. ‘The hydrolysis behavior depicted in figures 1 and 2, with appropriate shifts in the pH scale, approximates the behavior of many M(II), M(III), and M(IV) metals. The regularity of stability suggested for the mononu- clear species has not been confirmed in many cases, but such regularity is ‘often exhibited by complexes of metal ions with fluoride ion, a ligand with the same size and charge as the hydroxide ion, Several post-transition metal cations, however, form neutral hydroxide complexes of relatively high stability. Metals of valence higher than IV are extensively hydrolyzed above pH = 0, and many form anions and polyanions at higher pH. ‘THE THERMODYNAMICS OF HYDROLYSIS REACTIONS, In the following sections we will consider what is known of the equilibrium constants, enthalpy changes, and entropy changes of the reactions (summarized in table 1) that determine the boundaries and intersections in figures 1 and 2 and hence their temperature dependence. The system of subscripts used with Q, K, AG, AH, and AS to define the reaction involved may be summarized as follows: Two integers (xy) indi- cate the general reaction (4), in which x is the number of cations M=* and y is the number of OH™ ligands in the species formed. If an s is present in the subscript (sxy), then the solubility reaction (8) is referred to. The M:+/MOH@-")+ boundary.—Generally, it is to be expected from consideration of the electrostatic energy changes involved in the first hydrolysis step (reaction 1) that the logarithm of the equilibrium constant K,, (that is, the value of Q,, as the ionic strength approaches zero) should increase linearly with the ratio of the cation charge z to the M-O distance d (Baes and Mesmer, 1976, p. 398). This indeed proves to be a useful correlation provided the various cations in the periodic table are grouped according to type (table 2). In the table the equilibrium con- stant for reaction (1), Kys, has been represented by the expression. log Ky, = A+ 11.0 (2/d). (O)C. F. Baes, Jr. and R. E, Mesmer 940 HA + (¢ay\(HOPW = O"HA + WE OTH —2) + 4¢¢—2)(HOW = ,H(A — 2) + (0)*(HOOW OTH? + NW =,H2+(0)(HO)W +H + (6 SHON 07H + 44 A(HOWN +H+4q-2nHOW =? 0"H+ .2W 4(4—2x)(HO) "W/W +6-2)(HOJN/()*(HO)W W/0)"(HOW 414 HOW 4 25(HOW 40-2 HOW/ 2 uonsey, suonavas siskoxpAy Jo Aaeurumng Tay, (20 1 By) ‘AuepuinogCation Mg set ‘sm°* Poe 3 BP Hg Sn* Par * Shannon and Prewitt (1969); The thermodynamics of cation hydrolysis Tams 2 The first hydrolysis step, 25°C Log Ky = A-+ 110 (@/d) MLO Distance aa) Log Kut 2.120 11.44 218 2.40 12.85 22.0 2.56 =13.29 219 2.76 13.47 214 1.930 4.97 -22.1 2.292 -17 -22.1 2.461 85 219 2.364 -19 219 2.248 16 223 24 13.64 188 2.42 14.18 -18.7 2.78 -14.46 18.4 165 5.40 -18.7 2.220 10.59 20.5 2.170 9.5 19.6 2.145 9.65 -19.9 2.100 9.86 =203 243 -193488 183 2.145 8.96 19.2 2.35 10.08 19.4 2.130 43 198 2.07 22 18.1 2.040 ~2.26 18.4 2015 40 20.4 2.045 219 183 1.923 -12see# 18.4 2.065 3.4 19.4 2.020 26 18.9 2.190 4.00 19.1 2.48 3.20 -20.9 243 0.84 -173 242 0.65 18.8 2.40 149 19.8 2.38 0.5 19.0 2.55 -120 163 2.90 =13.21 170 271 -70 158 2.42 0.62 -14.3 2.53 =1.09 -14.1 2.42 3.40 12.5 2.64 3.40 -17 2.02 21 13.0 * Baes and Mesmer (1976); * (1980); #### Sutclifle and Weber (1956). 941 Avg A -219 19.1 15.5 -124 Paulson and Kester942 C. F. Baes, Jr. and R. E, Mesmer ‘The numerical constant (11.0) has been adjusted such that cations of similar type will, insofar as possible, produce similar values for the con- stant A. It is seen that after allowance is thus made for size and charge, the “hard” presransition metal cations are generally the most resistant to hydrolysis. Thus the alkaline earth ions, and AP+, Y*+, and the tri- valent lanthanides give the most negative values of A in the above ex- pression, about 28. A second group of “hard” cations, slightly less resist- ant to hydrolysis, includes principally the alkali-metal ions, the transi- jon-metal ions, and the tetravalent actinide ions. Here the value of A is less negative by about two and a half log units. A third much smaller group, containing most of the soft transition cations, shows K,, values about four log units less negative for ions of the same size and charge — A Tanue 3 The enthalpy of the first hydrolysis step any Cation M Method mole) Sources* ut 0 pot 15-35 13422 from? Mr 30 pot 0-30 1443 I from3 ct 40 pot 0-40 186*0.7 1 from4 Mn 90 pot 1s-42 lasts Feit sol 25-300 2415 6 Ft dil sol 150-330 12720707 co I MKCI pot 25-200 146205 from8 Co 40) pot 25-200 146205 — from8 N™ ail sol, pot 15-300 11922 from9,10 zn? 90 pot 15-42 Batt ca* = 3MLICIO, cal. i420 2 Ca — 3MNiCIO, pot 60-100 nse? Hg" 3MLICIO, cal 7232018 12 am ail pot 10-40 119207 from 14 Sc TMNaCl0, pot 10-40 144214 fromls cet 30) pot 0-100 9721 fom 16 Fe Git pot? 18-32 10: from 17 Co* I MNaClO, spt. 12-28 10. is Bi 3MNACIO, eal 4lz14 19 Th — IMNaClO, pot 0-98 11242 from 20,21 UN OSMNaClO, spt 10-43 112205 from22 ee) spt 25-150 ioor04 3 Pu = 2MLICIO, pot? 15-34 72212 from24 +The abbreviations pot, wl, cal, and spt. denote, respectively, potentiometric, solubility ealorimetrc, and specrophotomeie melhods. "Sos (Tas and Meer (179) Taed aid Cops (198: (8) Harned and Mannwciler (1985); (Bates ave others. (1930); (@) Pern. (1962a)s (0) Swecton and Baes (1970); (2) Zicmniak and. Jones (1078): (8) Glassen and. Tevcand (1978) (0) Tremaine nd. Leblanc (1970) (10) Bern (1968); (11) Pervin (19625); (12) Arnek and Kakolowier (961); (18) Murkor, Garmashy-and Lilich (979); (V4) Volokhow and thers (1971), (19) Rilpatrck nd Pokeas (2853): (16) Dijereum (1907); (17) Nlilber 957); (is) "Suetine and. Weber (1956); (19) Olin, (19%5):_ (20) Bes, Meyer, and Roberts (1965): (21) Kraus and Holmberg (1954); (22) Kraus and Nelson (1958); 3) Nikotaeva (1978): (4) Rabideau (1957).The thermodynamics of cation hydrolysis 943. being about —15.5. The remaining ions, Hg*+, Sn*+, and Pd?+, are the most readily hydrolyzed cations with A values three log units higher still and over nine log units more positive than “hard” cations of the same size and charge in the first group. Values for the enthalpy of the first hydrolysis step, AH, are sum- marized in table 8. Since all but three (for Gd?+, Hg?+, and Bi*+) are derived from the observed temperature dependance of the equilibrium quotient Q,, (or, by extrapolation to zero ionic strength, of the equilib- rium constant K,,), they have fairly large uncertainties, usually in the range 0.7 to 2 kcal/mole. These AH,, values have been combined with the corresponding values of K,, to obtain the entropy change AS,, at zero ionic strength. AS, =AH,/T+RInK,. (10) In so doing it was assumed in a number of cases that the effect on AH, of changing the ionic strength from the values at which the equilibrium quotient was measured (up to 5 m) to zero is less than the uncertainty in AIT,,. In at least two cases (Fe*+ and Zn*+), where AH, or the tempera- ture coefficient of Q,, has been determined as a function of ionic strength, this seems to be true. 40 L soe" 4 Ty 20 x 4 ‘e ‘es é cS x Pn 4 & zat a ° ao 7 jew 4 -20 4 i i 10 15 20 2/8 Fig. 9. Correlation of the entropy of the first hydrolysis step (reaction 1) with the ratio of the cation charge (2) to the metaLoxygen distance (d). The line is the least squaves fit to the data weighted according to the uncertainties,944 C. F. Baes, Jr. and R. E, Mesmer “Tamed Calculated entropies and enthalpies of the first hydrolysis step, 25°C. as asi am, ai, Cation —Lobs)_ (eal), fobs) tale) (eal mole” (cal/mole) uw -17527 -152 3822 ut Nat Fi9s0 172,343 M2 we =187 ia) Ag’ 138 123 1 =168 0 ae =136 ua Me? 22 Mo ca 68 15.6 0.7 18.5) se -86 186 vat =100 isa Ma 0243 1s asst ao Fe oss os Basis 32 ~09 #24 127207 co® 7 07 18620538 Ne 5427 07982 137 pe 152 14 cat 39 120 Zn 3943 20 B4et 128 co 2287 slo ane? 133 =4227 1as#2 Hg" 8720.7 98 7.2202 76 sa 8s 12 Pb o4 10.7 an i2e24 09 9t07 130 se 28685 3 atid 116 yy" i 139 La 82 141 sm" 103 138 ua 120 140 te 23.8 10.1 ve 241 10.3 ce 42022 a4 one ns Fe 242021 102 cot 27827 m6 002 99 Rh 28 ad Aut 303 70 cr 2s 106 ne wt m2 1 226 16 Bit BBE 20.7 4th WT zt asa ss he 336 10.3 TH m947 200 at? no Pa 30 87 uo aria OS 2e0s 100 3ose14 1oos08 v2 107 219+ 313 72212 100The thermodynamics of cation hydrolysis 945 The resulting values of AS,, vary approximately linearly with z/d (fig. 8), though Hg’, the softest cation with the strongest tendency to hydrolyze, shows a positive deviation several times the small uncertainty (# 0.7 cal K-* mole-1) involved. This suggests that a dependence on Kyy should be included in the representation of AS,;. The following expres- sion was found to give a satisfactory fit to the observed values, AS,, = 1.772 log Ky, + 19.12 @/d), au) producing a standard error to fit 1.47 times the uncertainty in AS,,. The ing standard error of fit for the line in figure 3 is 2.68. Assum- ing this relationship, AH, can be calculated from K,, and 2/d for most of the simple cations (table 4). AH,, = —0.886 log K,, + 5.70 (2/d). (12) AIL 45 values of AH,, thus obtained are positive, 27 are within 2.8 keal/ mole of the enthalpy of dissociation of water (18.3 kcal/mole), and only 9 of the enthalpy values are less than 10 keal/mole, The lowest values are found for the “soft” ions Pd?+ (7.4), Hg?* (7.6), Sn?#(7.2), Aue (7.0), ‘Tis (7.6), and Bit+ (7.7). The M(OH),c)/M:+ boundary.—The enthalpy of reaction of the oxide or hydroxide phases to produce the simple cations in solution (AH, reaction 6) have been calculated (table 5) by combining values of the equilibrium constant, K.,9, with the corresponding entropy changes, AS.yo, derived mainly from the National Bureau of Standards compilation (Wagman and others, 1968). In a number of cases where values of the equilibrium constant and the entropy were not for the same solid phase, the required entropy value was estimated by assuming the entropy change for the reaction MO .-.)/-(OH) > MOg,-y/(OH), + (u-)/2 HO (18) to be 3.64 (u-v) cal K-1 mole~*, This value was derived by inspection from the NBS tables. The resulting 28 enthalpy values are all negative, signifying that the boundary in figures 1 and 2 between the solid phase and the simple cation will fall as the temperature is increased. For cations of the same charge, AH9 tends to be less negative for the softer ion: The M(OH).(c)/M?+/MOH@"+ intersection.—As we have already noted (Baes and Mesmer, 1976), there is a correlation between the equil- ibrium constant for the solubility of an oxide or hydroxide, K,,o, and the constant for the first hydrolysis step, K,,, such that Ky Kary = 10-99, (14) This corresponds to approximate constancy of the equilibrium constant for the reaction M(OH),(c) + (2-1)M*+ = 7MOHO—)+, (15)z ‘Tapte 5 Calculated enthalpies of reaction of oxides and hydroxides, 25°C Solid oo B80 2AH FAH Cation phase E°BK,9 2logK 108K, 10 (cal et souteest Ke! mote") (eacmote) ae YAO 6.29 £005 ~$7403 112202 52301 70210 1 Bet BeOH), 669 +002 4102 =10081 -1210203 107218 hilo Mg? MA(OH), 16.84 +003 “60202 =i44e1 26203 26218 ae c* ——CuOH, 22.8 +002 -2902 0994014 31.08 +005 1518 a1 og Mni* Mn(OH)s <15.2 <598 19222 >-231 49 nt & Fee Fe(OH): <1283 <62 -205 22 >28 moo Co CAO, 23 04 7.00 +0.14 =126 #3 62820 mF Ne N(OH), 10801 =892 012 =184222 T1e19 m8 Pa PUN, -08 £01 50402 & a ao 762 +005 824401 =1728403 ~15552010 assis re zm" 20 1.142003 67801 205405 21322016 43218 mf ca* Ca(OH), 13.65 £0.04 65202 65205 ~20.56 £0.16 60418 33> OB Hg" HO 2.560005 424015 8762021 6.102009 90418 33 oe sa S10 176201 50208 6821 036203 140421 330«S Pom P60 1272 #005, =27202 44902 1601 +009 53218 3300 an WOH), 85401 64201 20884 23521 15629 aa 8 se soo) 94801 =35203 40283 =254209 95229 nie ve YH), 17S #03 ~56207 =359 22 36.6207 71229 nie Le LyOH), 20303 -52207 Bris -3698 205 52229 ne Sm Sm(OHy «168 £03 =12207 =32722 3225 +07 93429 ne Lu Lyi) 185 #03, =83+07 40222 3175407 10.1#29 ne ce cxOH), — <120 <00Fee Feo(oH) aS #08 6.1208 58.15 +06 1802 1.1 126429 1at Rh RWOH, — <9.0 <-12 Au" uO), -5.51 £007 098 +06 Ge GOH) 39403 49209 60.223 291 loo+3 hae In InOH)s $070.1 69203 =125 23 =106+09 228429 me 1 4Tin0; 3.9401 3803 Bit MBO, 3.46 40.1 019 20.18 35423 153409 77228 st ze 2102 A198 -0721 ue HiOs -12e1 Le Th ThOs 6321 65213 83.224 ~334218 lies 66 Pt Pad 0623 4023 ue U0: -18#1 4421 84045 22642 13%4 ua Net pO: 4022 -100= W785 5 18023 25045 ms Pu Po: “6583 ~8523 =19325 1484 25246 78 *Sourees: Log Ky, values are from table 2 and Log Kay values are from Baes and Mesmer (1976), AS, values are the calculated values from table 4, the first number lisied indicates the source of the M& ion entropy used, the second number listed indicates. the c of the solid-phase entropy used: (1) Wagman and others (1968.); (2) Parker, Wagman, and Garvin (1976); (3) Cox (1977); (4) igway and Robie (1977); @) Vasilev and Glavina (1969); (6) Wagman, Schumm, and Parker (1977); (7) Fuger and Oetting (1976); (8) Wagman (1978), » The entropy of the oxide was used with eq 13 in calculating AS,y and AH, (see text). «The entropy of the hydroxide was used with eq 13 in calculating ASp and AHyy (see text). sistoxphy uoryn9 Jo soruvucpoussys 24.1 46948, C. F. Baes, Jr. and R. E. Mesmer and a corresponding constancy of the position of the M(OH),(c)/M*+/ MOH-»)+ intersection along the vertical axis in figure 2, independent of the cation involved. Of the 36 values for the equilibrium constant of this reaction listed in table 5, 22 fall within 1.6 log units of 10-** at 25°C, only 7 give values with deviations of more than 3 log units, While the deviations in most cases exceed the uncertainties in the values of the product K,,*K,,o, in view of the wide range of the K,, and Ky,» values in- volved, this appears to be a significant correlation, Since a similar correlation might well be expected for the correspond- ing enthalpy of the solution of the solid phase and the enthalpy of the first hydrolysis step, values of AH,9/z have been plotted versus the ob- served values of AH,, in figure 4. Clearly there is a decrease in AH.so/z as AH,, increases. The straight line shown in the figure corresponds to the expression AH: = z (9.30 — 1.588 AH,,), (16) obtained by a least-squares fit of the AH,19 values in table 5 using smooth- ed values of AH,, from eq (12). The correlation in this case approaches the uncertainties of the quantities involved. The resulting expression for the enthalpy of reaction (15) is AH(15) = 24H, + AHyo = Values for the enthalpy of reaction (15), calculated from values of AHy10 in table 5 and smoothed values of AH,, (from eq 12), are included in table 5. With the single exception of Ca*+ ion, enthalpies of reaction (15) are positive and, as eq (17) suggests, tend to increase with the charge of the cation. As will be seen, these correlations involving the stability of the initial hydrolysis product in solution, MOH@-"+, and the solubility of the hydroxide or oxide phase are fairly useful, The small value of the equilibrium constant for reaction (15) reflects the strong tendency for precipitation to occur shortly after hydrolysis begins. The positive heat means this tendency is weakened by increasing the temperature. Hence, the intersection of the M'+/MOH®-»+ boundary with the solubility curve in figure 2 will tend to rise with temperature, the more so the higher the charge on the cation. The M*+/MOH@~!)+/M{OH),“*~¥)+ intersection.—We have pre- viously noted (Baes and Mesmer, 1976) that the tendency of the species MOH») + to polymerize according to the reaction MOH6=1)+ = (1/y)M,(OH), 7-9) + + (1 ~ x/y)M*+ (18) (9.30 — 0.588 AH,,), a7) was about the same for all cations, At the high ionic strengths involved in most measurements (1-3 m), the equilibrium quotient for this reaction is usually given to within 0.3 log units by the expression 1og(Qur7/Qu) = 28 — (x ~ 1)/y log (55.5) (19) at 25°C, The last term on the right converts from the mole fraction con-The thermodynamics of cation hydrolysis 949 T T T sh 4 Or 4 , 3 Ha “gst gz st > L 2 uit 4 8 reg T S art 2 Se qd TAF. $ at oe Ni Co! sok 4 4 Sa ot cat Ma} Fe2* A5 4 cult ° 5 10 15 AH,, (kcal/mole) Fig. 4. Correlation of the enthalpy of solution of an oxide or hydroxide phase in acd (reaction 6) with the enthalpy of the frst hydrolyse step (reaction 1). ThE. presents the lea-squares fit (o the All, values in table 5 and the caleulated val Of Atl, rom table 4 The points are the observed values of AH, from cable 5.950 ©. F. Baes, Jr. and R. E, Mesmer centration scale, on which the equilibrium quotient for reaction (18) is a better constant, to the molal concentration scale. We also noted that the entropy change of the formation reaction for a polynuclear species (reac- tion 4) could be correlated with z/d, again on the mole fraction scale. On the molality scale the expression for this entropy change, AS,,, was AS,y/y = —17.5 + 16.2(¢/d) + (1 +y—)/y R In (65.5). (20) This expression and that for AS,, (eq 11) may be combined with the above approximation for the equilibrium quotient (eq 19) to estimate the en- thalpy of reaction (18). AH(18) = AH,,/y — AH), ~ 6.6 — 0.58 log Ky, — 0.9 (2/d). (21) The resulting enthalpy change for reaction (18) is negative and falls in the range —6 to —18 kcal/mole, Hence the effect of increasing temperature is to reverse the polymerization reaction (18) and to cause the intersec- tion of the M+/MOH@-+ and the M#+/M,(OH),(#-7)+ boundaries in figure 2 to rise, The M(OH),(2-+/M(OH),4,¢-¥-+ boundary—In order to pro- ject the effect of increasing temperature on the course of hydrolysis, it remains for us to consider the mononuclear species formed beyond the first one, MOH@-»+, Unfortunately there is relatively little information available concerning those species that lie between the first hydrolysis product and the anionic mononuclear species that often terminate the series. Such information almost invariably comes from measurements of the solubility of the oxide or hydroxide formed by the cation, and, as we have noted, there has been a limited supply of accurate data, Increas- ingly, hydroxide and oxide solubility measurements are being made at elevated temperatures (compare references cited in tables 6 and 7), where there is less difficulty from the traditional problems of long equilibration times and active, amorphous phases. Of particular interest to us is the enthalpy associated with each suc- cessive mononuclear hydrolysis step. M(OH), 9+ + HO. M(OH),4.¢-7-)+ + Ht, AH(22) = AH, y4.— AH. ‘The limited amount of information presently available (summarized in table 6) suggests that for the divalent cations the heat of the second hy- drolysis step is usually close to that of the first step, though perhaps a bit less positive. For these cations the enthalpy of hydrolysis of the neutral species M(OH),(aq) is usually lower still. The entropy changes accom- panying the hydrolysis of the various mononuclear species in table 6 show a fairly uniform decrease as the charge on the reacting species de- creases (fig. 5). By including the anions of C(IV), P(V), As(V), (VI), r(VI), and Se(VI), we see that the decrease in the entropy continues be- yond the neutral species. Until more data are available this correlation may be the best way to estimate the temperature dependence of the step- @)The thermodynamics of cation hydrolysis 951 wise hydrolysis of metal cations for which equilibrium constants are available. ‘The trend in figure 5 suggests the approximation AS(22) = -17.8 + 12.2 (2-y). @3) ‘This permits us to estimate the enthalpy of reaction (22) as a function of the stepwise hydrolysis constant K(22) (= Kyy44/K,s) and the charge on the reacting species (2-y), AH(@2) The results are summarized in figure 6, wherein the horizontal axis, —log K(22), corresponds to the pH at which M(OH),-*)+ and M(OH)ys; (=F are present in equal amounts. Since the uncertainty in the estimated entropy is about +10 cal K-* mole, the uncertainty in the resulting enthalpy is about * 3 kcal/mole. It can he seen from figure 6 that if the successive stepwise hydrolysis constants of a cation were to decrease by about 2.7 log units per step, then the enthalpy associated with each step would be about the same. As a consequence, except for heat capacity effects, the pH range over which each species is dominant would remain unchanged as the temperature in- =1.36 log K (22) — 5.8 + 3.64 (z-y). (24) so T T T T T T T ze vie. Act + H20 == ACHE 4 Ht rie aoe exc) | + eum ‘» e.enm é 7 OF \ 3 a -20b eae 1 L 4 3 2 1 ot “2 CHARGE ON REACTING SPECIES Fig. 5. Correlation of the entropy of stepwise hyrolyis with the charge on the952 C. F. Baes, Jr. and R. E, Mesmer Tans 6 “Lhe thermodynamics of stepwise hydrolysis of mononuclear species, 25°C M(OH),-?" + 1,0 = MOH), + HY Hydrolyzing an as ‘Sources Species Log k (kcal/mole) (cal K“'mole~) Log AH AS Fe 95201 1B2 08 1 2 FeoH" Us t 1221 3526 1 3 Fe(OH); (29) Basis 28217 2929 3 3 co 9654005 134 07 1 2 CoOH" =9.15#0.1 n2s.7 4346 1 4 Ne 9864003 137 o7 1 2 NioH™ =104407 782 -18427 5 s N(OH); (aq) 9.53 408 Dass 3562 11 5 5 an" “8962005128, 1 2 2001" =78 403 soe? 6 6 Zn(OH (aq) -11.5 ¥0.2 8542 6 6 20H)" =129201 ass 6 6 Hg 3402007 7.2340.18 ' 7 HOH 2.772008 2614026 1 7 Poe “771201 107 1 2 PeOH 9.41 202 1482 1 8 POOH): (aa) 094201 ott ' 8 So(OH)* =1.41 £0.08 5.1406 1 9 SHOH)s (ag) 118201 159205 1 ° POCOH)s (29) 148 =1.90 wo on PO;(0H),- 7.198 099) 2 ou POs(OH)* =1238 353 Boon AsO(OH)s (aq) -2.24 169 ou ‘As03(0H)- 694 on bon As0(OH)™ = 11.50 435 ou $05(0H)- -199 $24 nou e0,(0H)" 1.663 42 27 1 ou BOOH); (a) —--9.23620003 3.2220.) -31.46 404 1 S(OH), (a9) 9.825 0033 6.1203 2443-41 Iss Sources: (1) Baes and Mesmer (1976); (2) from eq 11; (8) Zieminak and Jones (1978); (4) from AHye (table 7) and AH, (table 5); (@) from values given by’ Tra- maine and LeBlane (1980); and Ky, and AH; (6) from results of Reichle and others (1973); (7) Atnek and Kakolowicz (1967); (8) from results of Tugarinov and others (1975); (Q) from results of Popova and others (1975); (10) Vanderzee and Quist (1961); (11) Wagman and others (1968.); (12) Bates and Acree (1943); (13) Bates (1951); (14) Mesmer and others (1972); (15) Busey and Mesmer (197).Tame 7 ‘The thermodynamics of reactions of oxides and hydroxides to form neutral aqueous species, 25°C M(OH),(c)* M(OH),(aq) MO, j9(c) + 2/2 HyO* M(OH),(aq) as 2 Solid AH Source Cation foe LogKs1z (ecalfmole) eel logk OH aS Agt 4Ag, 0 “5.71403 70742 2447 1 2 Fe Fe(OH): <-18 >58 ~-16 1 3 Co* Co(OH)2 6.5 +0.14 4141 16.0234 1 4 Ni** NiO 8.1405 =1.26#1 41344 5 3 Zn 2n0 562402 — -0.52 40.5 -27.542 6 6 Hg" Hg ~3.61 + 0.07 3.74 £0.18 4.0 40.7 1 7 Pb PbO (red) 4.40.1 5441 -2043.4 1 6 Bll) B(OH), 5.26 176 1 1 ast % AssOe ~0.68 + 0.01 3.584 (0.1) 8.9 +(0.3) 1 7 sb** % Sb2 05 ~4.24 40.1 7.7 £05 6421.7 1 9 Bi* % Bi, Os, 5.40 +015 Si(IV) SiOz (a-quartz) 4040.1 710 +05 5.2£1.7 10 10 *Sources: (1) Baes and Mesmer (1976); (2) from & Sy given by eq 11 and & Syip from Wagn jemniak and Jones, 1978); (4) from cempetature dependence of Ky: (1976); (6) Tien (978). aiid others (1975); (@) Popova and others (1973); (10) Walther and Helgeson (197). 1» and others (1968-); (8) from ies from Eguchi and Yazawa ¢ and Lebilane (1180); (6) Ky, and 4 Hy, values from table 6 were combined with Kqy and A Hy values from table 5; (7) Wagthan and others (1908-); (8) from temperature dependence of K, from Tugarinoy sistjoapdy wove Jo sormputpouuoy) 24.1, $96954 C. F. Baes, Jv. and R. E, Mesmer creases, For a few cations, ranges of stability of one mononuclear hydroly- sis product (usually the neutral one) wider than 2.7 pH units do occur, and in such cases this range should narrow with increasing temperature. For most cations, however, ranges of stability are narrower, typically 1.3 PH units per hydrolysis step. Stepwise enthalpies in such cases should de- crease as hydrolysis proceeds, and the range of stability for each inter- mediate species should widen with increasing temperature (fig. 7). The M(OH){c)/M(OH)aq) boundary.—The location of the boun- dary determined by the equilibrium between the hydroxide or oxide solid phase and the neutral hydroxide complex in solution M(OH),(c) = M(OH),(aq) Mo,,,(c) + 2/2 HO = M(OH),(aq) (25) of special interest since it determines the minimum in the solubility curve for many solids. While estimates of the corresponding equilibrium constant K,;, are generally of poor accuracy, for all but the soft cations they appear to decrease as the cation charge increases. Indeed it may be seen from the following argument that such should be the case: If a ca- tion produces a series of mononuclear hydrolysis products for which the oe « 6 6 0 © « wien K(22) Fig. 6, Estimated enthalpy of stepwise hydrolysis (reaction 22) as a function of the equilibrium constant and the charge on the reacting species. The abscissa cor- responds to the pH at which the reactant and the product species would be present at equal concentrations.The thermodynamics of cation hydrolysis 955 successive stepwise constants, K(22), decrease by a constant factor B (typ: ically about 1.3 log units) then it would follow that for the reaction Mt +2H,0 = M(OH),(aq) + 2H, (26) the equilibrium constant would be given by log Ky, = z log Ky, — 0.5 2(2— 1) B. 7) ‘The equilibrium constant Kj; (for reaction 25) is related to Ky, by log Karz = log Kary + log Ky. @8) Combining eqs (27) and (28) and introducing the approximation given by eq (14), we obtain Jog Kure 5.6 —0.5 2(z — 1) B. (29) For B= 1.8 log units this expression gives log K,,,~—5.6, —6.9, -9.5, and —15.4 respectively for z= 1, 2, 3, and 4, Since the variability of B is considerable, these are quite approximate estimates, as may be judged by comparison with the Kx, values listed in table 7. Those for Ag+, Fe**, Ni*+, Co'+, and Zn?+ are within one or two log units of the predicted values. The soft cations Hg+, Pb?+, As'+, Sbi+, Bit+, and Si(IV) give much higher values because of the extra stability of the neutral complexes they form in solution, Table 7 summarizes the limited information about the enthalpy and entropy changes associated with reaction (25). The enthalpy values, AH.» are in the range —1.3 to 7.7 kcal/mole, suggesting that the temperature 250 200 150 400 TEMPERATURE (°C) 4000/T(K"') 8 25 Fig. 7. Regions of dominance of the mononuclear species of the trivalent cation from figures I and 2 asa function of temperature. ‘The temperature scale is Tinear in 1/T K-, hence the curvature of the boundaries indicates the effect of estithated AC, Values (from table 8). The dished curve indicates neutral pH.956 C.F, Baes, Jr. and R. E. Mesmer dependence of such solubility is likely to be relatively small and, especially for a hydroxide solid, positive. The M(OH).(c)/M(OH).,,~ boundary.—The temperature depend- ence of the equilibrium M(OH),(¢) + H.O = M(H), ,.~ + H+ (80) may be estimated as follows: The enthalpy for this reaction is the sum of AH, and the enthalpy for the reaction M(OH),(aq) + H,O = M(COH),,..~ + H+. (31) ‘The approximate value of the latter enthalpy is given by eq (24) as AH(81) ~ — 1.36 log K(31) — 5.3, (82) and hence H(80) = AH,,, — 1.86 log K(81) — 5.8 (83) Since AH,,, is usually positive and of the order of the last term on the right in eq (82), then AH(80) ~ 1.86 log K(81). Log K for reaction (81) is generally less than —7, hence AH for reaction (80) should be positive and usually greater than 10 kcal/mole. Of perhaps greater interest is the equilibrium representing the tendency of the hydroxide (or oxide) phase to dissolve in basic solutions, M(OH),(c) + OH- = M(OH),..>. G4), Since the enthalpy of this reaction will be that just estimated for reaction (80) less the enthalpy of dissociation of water (13.3 kcal/mole), it is likely to be small. Estimating AC, for hydrolysis reactions.—It is of course desirable to estimate AC, for hydrolysis reactions in order to predict changes in the associated AH and AS values with temperature, In the virtual absence of information about the partial molal heat capacities of hydroxide com- plexes, the best that can be done is to make estimates from the partial molal heat capacities of other ionic species and from the values of AC, for other protolytic reactions. ‘The partial molal heat capacities for monatomic ions assigned by Griss and Cobble (1964) exhibit such a smooth variation with the ionic radius that it is possible to predict the heat capacity of a simple cation of specified size and charge with an accuracy comparable to the measured values. We will approximate the heat capacity of a mononuclear hydrox- ide complex M(OH),@-2)* as that of a monatomic cation of the same charge as the complex and the same size as the hydrolyzing cation M=. This seems reasonable since the hydroxide complex and the cation that produced it probably differ only by the protons that have been lost in hydrolysis (reaction 2). The resulting values of AC, for various hydrolysis reactions involving mononuclear species of cations of several sizes are summarized in table 8,Tanz 8 Estimated AC, values for hydrolysis reactions* Ave. AC, Pauling (cal K“! mole“) Reaction Radius of M8 | =“ (A) 25-28-60 25-100 25-150 25-200 CO) M** + H,O= MOH + Ht 0s =10 -8 -5 4 -1 08 ~9 -8 6 5 4 i 8 9 7 ~6 ~6 M(OH)}*, +H,0 + M@H)?*, +H* 0.5 =i -i1 -10 —10 -10 08 -10 0-11 38 38 38 Ld 38 9 6 -s -s M(OH)?* , + H20 = M(OH)?_, + H* 0.5 -1s -14 -10 -10 —10 08 =n -1 7 6 Ld -7 9 6 4 4 M(OH)S_, + H20*M(OH), (aq) + H* 0 0 0 0 0 ‘M(OH), (aq) + H20 * M(OH),, , +H 55 AT —44 —45 -51 H;0* H* + OH” “35 -AT 44 45 -sI “The AG, values for the first three yeactions were obtained by assuming that the partial molal heat capacity of each hyde complex is equal to that for a monatomic cation of the sime radius as M=" but with & charge equal to that of the complex. The fstimates of the average hese capacities of monatomic cations as'a-function of radius and tcipertture were made using’ the values ‘of Gris and Cobble (1964). AC, values for water dissociation are from Sweeton, Msmer, and Bucs (1974). sistjouphy uorjv2 Jo soruputpousoys oy. 196958 G. F. Baes, Jr. and R. E. Mesmer From a number of recent studies of protolytic reactions over the temperature range 0° to 300°C (Mesmer, and others, 1972; Mesmer and Baes, 1974; Hitch and Mesmer, 1976; Busey and Mesmer, 197). it appears that AC, is close to zero for reactions that produce no change in the number or in the magnitude of the ion charges. Accordingly AC, has been asumed zero (table 8) for the reaction M(OH),_.+ + H.0 = M(OR),(aq) + H+. (85) A similar assumption is made for the reaction M(OH),(aq) + OH- = M(OH),4..>, (36) which in combination with the dissociation reaction for water gives the final hydrolysis reaction listed in table 8. ‘The quantities listed in table 8 are average heat capacity values over the specified temperature interval. Enthalpies and entropies of reaction ‘ated with sufficient accuracy from the approximate relation- ASy = ASoog + AC, 1n(T/298.15) G7) AHy = AHkns + AG, (T-298.15). (38) ‘The accuracy of these estimated heat capacities depends of course primari- ly on the reliability of the assumptions employed, namely, (1) that the partial molal heat capacity of a mononuclear hydroxide complex is the same as that of a simple cation with the size of the unhydrolyzed cation and the charge of the complex, and (2) that AC, is zero for a reaction in which there is no change in the number or magnitude of the ion charges. ‘These assumptions are expected to produce AC, values reliable to 10 cal mole! K-* up to 200°C and hence should lead to errors of extrapola- tion in AH and AS no larger than the uncertainties in the observed values in most cases. SUMMARY AND CONCLUSIONS In summary we may make the following observations about the hy- drolysis of cations and its dependence on temperature, 1. The tendency for a cation to hydrolyze increases with its charge and decreases with its size. There is also a strong elfect of the position of the element in the periodic table, post-transition metals showing the greatest tendency to hydrolyze and the pre-transition metals the least. For cations of a similar type, the equilibrium constant K,, for the first hy- drolysis step (reaction 1) varies approximately linearly with the ratio of the cation charge z to the metal-oxygen distance d (table 2). 2. The enthalpy of the first hydrolysis step (AH,,, table 4) is positive and often close to the enthalpy of dissociation for water (18.3 kcal/mole). Hence the tendency of cations to hydrolyze increases with temperature. ‘The post-transition metal ions with the strongest tendency to hydrolyze exhibit the least positive values (as low as 7 kcal/mole). Expressions are suggested that give AH, (eq 12) and AS, (eq 11) as a function of K,, and 7/d.The thermodynamics of cation hydrolysis 959 3. The enthalpy (AH.y.) of the reaction of an oxide or hydroxide phase with H+ to produce the unhydrolyzed cation in solution (for ex- ample, reaction 6) is negative (table 5), anc hence such solids show a de- creasing solubility in acid as the temperature is increased. There appears to be a fairly good correlation between AH,,, and AH,, (eq 16 and fig. 4). 4, The tendency for the oxide or hydroxide phase to react with the cation to produce the first hydrolysis product (reaction 15) is comparable for all cations, the equilibrium constant being approx 10-*# at 25°C. Accordingly the intersection M(OH),(c)/M*+/MOH@—)+ in figure 2 typically should occur at a total metal concentration of about 10-*m, Since the enthalpy of reaction (15) is generally positive (table 5), this intersection should rise with temperature. 5. The entropy of reaction of the cation to form a polynuclear hy- drolysis product (AS,, of reaction 4) correlates well with the ratio 2/d (eq 20). The tendency of MOH'-+ to polymerize to such a species ac- cording to reaction (18) is about the same for all cations (eq 19). The en- thalpy change for this reaction is in the range —2.6 to —7.5 kcal/mole, and hence the intersection M'+/MOH@-+/M,(OH),@2—7)+ in figure 2 will rise with temperature, reflecting a decreasing stability of polynuclear species, 6. The limited data available for mononuclear hydrolysis products beyond the first one suggest the approximation that the entropy of step- wise hydrolysis (of reaction 22) decreases linearly with the charge on the reacting species (eq 23 and fig. 5). From this it can be predicted that for typical cations the enthalpy of each successive hydrolysis step should de- crease, and that the pH-range of dominance of each successive mononu- clear species (fig. 2) should increase with temperature. For unusually stable neutral species, the wide pH range of dominance should decrease with increasing temperature. 7. The solubility of the oxide or hydroxide phase to produce the neutral species, M(OH),, in aqueous solution usually determines the min- imum in the solubility curve (fig. 1) and generally decreases as the charge on the cation increases. Cations such as Pb?+, As+, Sb°+, and Bi*+ which exhibit extra stability for the neutral species show higher minimum solu- bilities for the oxide or hydroxide phase than would otherwise be expect- ed. The limited data available suggest that the temperature dependence of the solubility reaction that produces the neutral species is relatively small and usually positive (table 7). 8. Similarly, the tendency for the oxide or hydroxide to react with basic solutions to produce the anion M(OH),4,~ (reaction 34) is expected to have a small temperature dependence. 9, The heat capacity changes (table 8) are expected to be small for stepwise hydrolysis reactions involving mononuclear species, except for those producing anions. On the basis of these correlations it is possible to construct a diagram such as that in figure 7, which indicates as a function of temperature the regions of dominance of the various species produced in solution by hy-960 C. F. Baes, Jr. and R. E, Mesmer drolysis of the hypothetical trivalent cation in figures 1 and 2. The con- centration of the metal is assumed to be low enough that polynuclear species can be neglected. As the temperature increases the onset of hy- drolysis occurs at a rapidly decreasing pH because of the large positive value of AH,,, dropping about two log units by the time 100°C is reached. At the same time the region of dominance of each hydrolysis product — especially the neutral one —widens. The temperature scale on this dia- gram is plotted as the reciprocal of the absolute temperature and hence the boundaries of each region may be viewed as Arrhenius plots of the stepwise equilibrium constants. ‘The first three boundaries — M*+/ MOH?+, MOH?+/M(OH),+, and M(OH),+/M(OH),(aq) —are linear or nearly so, because AC, values for the reactions involved are small, The next boundary, M(OH),(aq)/M(OH),~, curves toward higher pH in the manner indicated because AC, for the reaction involved is appreciable, comparable to that for the dissociation of water. ‘The curves shown in figure 8 indicate the temperature dependence to be expected for the solubility of the hydroxide of our hypothetical cation (neglecting AC, for reaction 6). On the acid side the solubility falls rapidly with increasing temperature, At the minimum the temperature dependence is small. On the alkaline side the moderate positive tempera- ture coefficient of solubility is reduced and finally reversed by the heat capacity effect as the temperature rises. The contribution from the poly- nuclear species shown in figures 1 and 2 decreases as the temperature is increased. Projections to elevated temperatures of the sort indicated in figures 7 and 8 require primarily a knowledge of the formation constants of the successive mononuclear hydroxide complexes in solution at one tempera ture at least, The rest may be inferred, with useful accuracy we hope, from the correlations offered in this review. More accurate information pH Fig. 8. Solubility of the hydroxide solid phase from figure 1 as a function of tem- perature.The thermodynamics of cation hydrolysis 961 about the mononuclear species is badly needed, It will most probably come from measurements of the solubility of crystalline oxide and hy- droxide phases as a function of pH and temperature. Such measurements will not only yield needed information about specific cations, they will also further test and improve correlations needed to deal with the numer- ‘ous species of other metallic elements formed in dilute aqueous solutions at elevated temperatures. Rerrnescrs Amek, R., and Kakolowicr, W., 1967, Themmochemical studies of hydrolytic reactions ‘x ‘Thermochemical study of hydrolyzed Hg(ClO,), solutions: Aeta Chem, Scandina: vica, v.21, p. 49-2180, Bacs, ©. Fy Ji Meyer, J» and Roberts, C. E 1965, The hydrolysis of thorium(tv) at @ and 95°C: Inorgaite Chemistry, ve, p. 518-527 Bes, C.F, Jr and Mesmer, R. E> 1976, The hydolysis of cations: New York, John ‘Wiiey-intetscfence, 489 p. R. G, 1951, Fisst dlswociation constant of phosphoric acid from 0° to. 60°C: imitations of clectromotive force method for moderately stiong acids? Natl. Bur. Standards, Jour. Research, v.47. p. 127-194. Bates, R. G., aid Acree, §. F., 1043, pi values of certain phosphate-chloride mixtures and the second dissociation constant of phosphoric acid from 0* to 60°: Natl, Bur Standards, Jour. Research, v. 80, p. 129-55. Bates, R. Gy Bower, V-E,, Canant, R. G., and Prue, J. E., 1959, The dissociation Consiant of CaOH" from 0° to 40°C: Fatatay Soc. vans, v. 35, p Bjerrum, N., 1907, Studies on chromic chloride, Part 1: Zcitschr. Phys. Chemie, v. 59, 986-388, Barkow K. A, Garmash, L. A., and Lilch, LS. 1977, Hydrolysis of cadmium ions at clevated iemperatures! Vesnik Leningradskogo Univ, Fizika, Kimiya, v. 5, n. 16, p.3389, Busey, RUHL, and Mesmer, R. E., 1977, Ionization equilibria of silicic acid and poly- ‘tlicate formation in aqueous sodium chloride solutions to 800°C: Inorganic Chem fsury, v.16, p. 2441-2450, Cox, J. Dy, 1977, Codata task group on key values for thermodynamics, Tentative set ‘of key values for thermodynamics. Part VI: Internat, Council Sci, Unions, Commit: tee on Data for Science and Technology, CODATA Spec. Rept. 4 20 P. Cris, C2 M., and Cobble, J. W., 1964, The thermodynanie properties of high tempera- tare agus solutions. V. "The calnlton of Tone hea capaces ip Yo 200" Entroples and heat capacities above 200°C: Am, Chen, Soc. Jour v. 86, p. 3890: Enuop J P Eguchi, M., and Yazawa, A, 1975, Fundamental studies on high tempersture aqueous totic IT. The precipitation reaction of cobalt hydroxide at temperatures be tween 25 and 200°C: Nippon, Kogyo Kaishi, v. 91(1043)p. Fugen, J. and Octting, F. Le 1976, The chemical thetmod and ‘compounds, Part 3. The actinide aqueous fons! Vienna, Internatl. Atomic Eon Agency 67 Giasson, C., and Tewari, P. H., 1978, Hydrolysis of Co(ll) at elevated temperatures ‘Catadian. Jour. Cheinisry, v.50, p. 435-440, Harned, H. 8. and Copson, KU R., 1083, The dissociation of water in lithium chloride solitions: Am, Chem, Soe. Jour v.55, Ps 2206-221 Hamed, H. 8, and Mannweilei, G. £, 1933, The thermodynamics of ionized water in Socliuim chloride solutions: Am. Chem, Soc, Jour, v. 57, p. 1873-1876, Hemingway, B. 5. and Robie, Ro A,, 1977, ‘The entropy’ and Gibbs free energy of formation of aluminum font Geochim, et Cosmochim: Acta, v.41, p. 1402-1404, Hitch, B. Fevand Mesmer, Re E., 1976, The fonization of aqucous animonia to 300°C in KCI media, Jour. Solution Chemistry, v. 5, p. 667-680. Kilpatrick, M., and Pokras, L., 1058, Acid dissociation of the aquoscandium ions: Jour. ‘Electrochem, Soc. . 100, p 85.92. Kraus, KA and Holmberg, R. W, 1954, Hydrolytic behavior of metal ions. 1, Hy- “olysis of thotivm(1V): Jour, Phys. Chemistry, v. 58, p- 325-390. Kraus, KA, and Nelson, F, 1955, Hydrolytic behavior of metal ions. IV. The acid Constants of uranium IV'ag a function of temperature: Am. Chem. Soc Jours ¥. Tp. 321-3722, Bate:962 C. F. Baes, Jr. and R. E. Mesmer sociation equilibria in Mesmer, R, E., anc Bact, ©. F., Je, 1974, Phosphoric acid d aucots solutions to 300°C Jour. Solution Chemistry. ¥. 8, p. 807-322. Mesintn, Re Ey Bnes, ©, Fy. Jr and Sweeton, Fr Hs 1072, Acidity measurements at clevated temperatures. VI Borie acid equilibria: Inorganic Chem. . 11, p- 537-543. Milburn, R. My 1957, A spectrophotomettic study of the hydrolysis of fyon(iM}) ion. Tl, Heats and entropies of hyclralysie Am. Chem. Soc. Jury ¥. 9, p. 387-540. Nikolacvay N. Mo, 1978, The hydrolysis of US ions at elevated temperatures: Tr. Sibirskogo Oticteniya Akad. Nauk SSSR, Seriya Kimichesikh Nauk, v4 p. 9193. Otin, A. 1915, Thermochemical study of bytolyzed Bi(CIO,), solutions: Aeia Chem, Sean, ¥. 29, p. 07-910 Parker, V. i, Wagman, D:D, and Garvin, D., 1976, Selected thermodynamic data compatible with the CODATA recommendations!’ Washington, D.C, Natl, ur Standards Interim Rept, NBSIR. 75-068, 81p. Paulon, A.J and Keser DR 1080, Copper (tH fon hydrolysis fn aqueous solution Jour, Solttion Chemistry, v. 9, p.200-21- Perth, D. Dy 10624, The hyctroiysis of manganese(It) ion: J. Chemical Soc. p. 2197-2200. 1962, The hydrolysis of metal fons. Part HK, Zine: Chem, Soe. four. v. 87%, 4500-4502 1964, ‘The hydrolysis of metal ions, Part IV. Nickel: Chem. Soe. Jour, v.04, p. S044-3048, Popova, Ni Va, Khodakovsky, I. L., and Ozerova, N. Ax 1975, Experimental determin: ‘ation of the thermodynamic properties of hydvoxo- and hydrovolluoride complexes of antimony at temperatures up to 200°C: Geokhimiss, p. 885-845. Rabideau, 8, W., 1957, he hytreljss of plutonium): Am, Chem. Soc. Jour, v. 79, p. 3675-3677. Reichie, R.A, McCurdy, K. G., and Hepler, L. G., 1975, Zine hydroxide: Solubility Produce and hydroxy-compiex stability constants from 125-75°C: Canadian Jour Ghremistry, v.53, p. 3841-3845. shannon, Re Dy and Prewett, C. D., 1969, Elfetive tonic radit in oxides and fuori "heta Cryst, v. B25, p. 925-940, Sitlen, T. Gl, and Martell, A. E., 1971, Stability constants of meta-ion complexes: London, ‘The Chem. Soe, Spec. Pub. 17, Tip. 1864. Suppl. 1, Spec. Pub. 25, 865 Smith, A.M, and Martell, A. E., 1976, Creal stability consist? New York, Pl Press, v4, 257 Sutcliffe, Ly H., and Weber, J. R., 1956, The reaction between cerium ([11) and cobale ‘ataday Soe, Trai v.32, p, 1285-1281. HL, and aes, C.F Jr i970, The solubility of magnetite and hydrolysis ‘of ferrous on in aqucous soltion at elevated temperatures: Jour. Chem. ‘Therm dynam, v- 2, p. 479.500 Swecton, F. fly Mesiner, RE, and Bacs, C.F, Jo, 1974, Acidity measurements at levated temperatures, VIL" Dissociation oF waters Jour. Solution Chem ¥. 3, ps 191-214 ‘Tremaine, P. R., and LeBlanc, J. C, 1980, The solubility of nickel oxide and hydrolysis Of Nin water to 57BK: jour, Thermodynamics, v.12, p. 521-298. ‘Tugarinow, 1A, Gancyev, I Gs an Kodskovskiy, 11, 1975, Experimental determina- tion of iydvolysis constants of lead fons in ajicous solutions at elevated tempera tires up to. 300°C: Geokhimiya 0, p. 13431354, Vanderzee, ©. Ey and Quist, A. Sy i961, Lhe third dissociation constant of oxthophos photic acd: Jour. Phys. Chemistry, . 65, p. 1183128 vasifet, V-P., aid Giavina, &. Re, 05, ‘Thermodynantie characteristics of the Bi Elektrokbiimia, v5 pe A0417 Votokhoy, Vu. Ay Pavlov, L. N., nd Miranov, V. v.44, p. 216-249. Wagman, D. ns, W. I, Patker, V. B,, Halow, Ly Baile R. H., 1968, Selected values of chemical ‘thermodynamic D,, Natl. Bur. Standards Tech. Note 270-8 to 270-7, continuing. Wagman, D. D., Schumm, R. H., and Parker, V. B., 1977, A comptter-assisted evalua- tion of the thermochemical data of the compounds of thorium: Washington, D.C, Natl. Bur, Standards Rept. NBSIR 77-1300. Walther, J. V., and Helgeson, H.C, 1977, Calculations of the thermodynamic proper- ties ‘of aqueous silica and the solubility of quartz and its polymorphs at high pressures and temperatures: Am. Jour. Sct. v. 277, p. 1315-1351. Ziemniak, S. E,, and Jones, N. E., 1978, Solubility behavior of transition metal ferrites at high temperature in aqueous solutions: Miamisburg, Ohio, Knolls Atomic Power Lab, KAPL-I12, 40 p. | 1971, Hydrolysis of S.M, and Sch pertics: Washington,