Thermochemistry and oxidation potentials of manganese and its

Note: In lieu of an abstract, this is the article's first page. Click to increase image size Free first page. View: PDF | PDF w/ Links. Related Conten...
0 downloads 12 Views 987KB Size
Download PDF
THERMOCHEMISTRY AND OXIDATION POTENTIALS OF MANGANESE AND ITS COMPOUNDS T. A. ZORDAN

AND

L. G. HEPLER

Department of C h i s t r y , University of Louim’lle, Louisville,Kentucky Received March 1, 1968 CONTENTS

737 738 738

....................................... I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Descriptive Chemistry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. “Simple” Compounds and Aqueous Ions.. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Complex Compounds and Ions... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I. INTRODUCTION I n this review we are primarily concerned with the chemical thermodynamics of manganese, its compounds, and its aqueous ions. The resulting thermodynamic data are of a kind that has long been known to be useful in electrochemistry, analytical chemistry, and many applications of chemical principles to other disciplines. We have been critical in the tabulation of data and have recalculated many of the published results cited. When data from different sources are not in good agreement, we have attempted to justify our choices. We have been explicit about the sources of data and also our treatment of data from the literature so that interested readers can check the steps leading to the tabulated values of thermodynamic properties and form their own opinions about reliability and accuracy. When we have not read the original papers cited, we have listed the Chemical Abstracts reference or the name of the journal in which the article has been translated. For several compounds and aqueous ions we have combined experimental data from a cited source with our estimate of some thermodynamic property. Numerical values of our estimates are given in parentheses. Whenever possible, we have used auxiliary thermodynamic data from the National Bureau of Standards Technical Note 270-3.’ Other auxiliary data are taken from sources such as NBS Circular 5002 or Latimer’s “Oxidation potential^"^ that are explicitly cited at appropriate places in the discussion. The matter of uncertainties in thermodynamic properties is difficult to handle. For example, the uncer(1) D. D. Wagman, W. H. Evans, V. B. Parker, I. Halow, S M. Bailey, and R. H. Schumm. “Selected Values of Chemical Thermodynamic Properties,” National Bureau of Standards Technical Note 270-3, U. S.Government Printing Office, Washington, D. C. 1968. (2) F. D. Rossini, D. D. Wagman, W. H. Evans, S. Levine, and I. Jaffe, “Selected Values of Chemical Thermodynamic Properties,” National Bureau of Standards Circular 500, U. S. Government Printing Office, Washington, D. C., 1952. (3) W. M. Latimer, “Oxidation Potentials,” 2nd ed, PrenticeHall, Inc., Englewood Cliffs, N. J., 1952.

743

tainty in a heat of formation calculated from a paxticular heat of reaction depends on the cumulative uncertainties of all of the heats of formation used in the calculation. Many errors in thermochemical investigations are due to difficulties in the compounds or processes being studied (Le., unidentified hydrolysis phenomena, nonhomogeneous phases, impurities, etc.), which are more difficult to assess than are measurement uncertainties. Our treatment of uncertainties is simply to list results to a “reasonable” number of figures. In some cases we have indicated that an uncertainty is larger than usual by placing an approximate sign (-) in front of the cited quantity. Although this review is not primarily concerned with high-temperature chemistry for its own sake, we have made considerable use of data obtained at high temperatures in calculating desired thermodynamic properties. Methods of making such calculations have been de~cribed.~ In spite of considerable discu~sion~-~ in recent years, general agreement is still lacking on “sign conventions” for potentials. Much of the confusion arises because “sign” can be either electrical or algebraic. A l l potentials tabulated in this review are oxidation halfreaction potentials with algebraic signs. This choice, which has been discussed previo~sly,~ permits straightforward use of the familiar AGO = -RT In K = -nFE”. The half-reaction potentials are based on the usual references Hz(g) = 2H+(aq)

Hz(g)

+ 20H-(aq)

+ 2e-

= 2H20(liq)

Eo E 0V

+ 2e-

E o = 0.828 V

(4) G. N. Lewis and M. Randall (revised by K. 9. Pitzer and L . Brewer), “Thermodynamics,” McGraw-Hill Book Co., Inc., New York, N. Y., 1961. ( 5 ) F. C. Anson, J . Chem. Educ., 36,394 (1959). (6) A. J. deBethune, J. Electrochem. SOC.,102, 28812(1955). (7) T. S. Licht and A. J. deBethune, J . Chem. Educ., 34,433 (1957). (8) J. B. Ramsay, J . Electrochem. Soc., 104, 255, 691 (1957); J. Chem. Educ., 38,353 (1961). (9) R. N. Goldberg and L.G. Hepler, Chem. Rev., 68,229 (1968).

737

738

T. A. ZORDANAND L. G. HEPLER

We have reserved the symbols E" and K for potentials and equilibrium constants that have been determined in very dilute solutions or in such ways that activity coefficients could be considered in treating the experimental data. Similarly, we use AHo and AS" to indicate data that refer to the usual standard ~ t a t e s ' J *and ~ use AH and A S to indicate data that refer to other states such as a solution maintained at constant ionic strength by means of some "inert" supporting electrolyte. The results of such investigations lead to equilibrium quotients represented by Q rather than equilibrium constants represented by K . We have used such symbols as Qo.5 and AH0.5 to indicate that the equilibrium quotient or heat of reaction was determined in solutions having ionic strength 0.5.

11. DESCRIPTIVE CHEMISTRY Species containing manganese in formal oxidation states ranging from -3 to +7 have been reported. We have no relevant data for the negative oxidation states and therefore limit our discussion to the zero and positive oxidation states. For the zero oxidation state we have the well-characterized carbonyl compound of formula Mnz(C0)lo. Related compounds such as (C0)5MnCo(C0)4 and Mn(NO)&O are known, but we have no thermodynamic data for them. Reduction of K4Mn(CN)6 in liquid ammonia apparently yields K6n/In(CN)6, but we have no data for this substance. Electrolytic or chemical reduction (with sodium amalgam) of Mn(CN)e4- in aqueous solution yields Mn(CN)e-6 and thence such solids as KJIn(CN)G containing Mn(1). Carbonyl halides of formula Mn(C0)gX are also represented as containing Mn(1). We have no data for these compounds. There is some evidence that the hydrocarbonyl HMn(C0)b is most realistically described as a hydrido complex containing Mn(1). I n this connection, it is interesting that HR/~I(CO)~ ionizes to H+(aq) and Aln(CO)5-(aq) with pK Ei 7. Most of the compounds and ions for which we have data are those that contain manganese in the +2, +4, or +7 oxidation states. These compounds and ions account for most of the aqueous chemistry of manganese, and discussion of them is the principal subject of this review. We do, however, have some data for species in other less familiar positive oxidation states. I n most acidic solutions and under some other conditions Mn(I1) is the most stable oxidation state of manganese. Most salts of Mn(I1) are quite soluble in water, but a poorly defined sulfide and a well-defined hydroxide can be precipitated. Complexes of Mn(I1) are generally less stable than corresponding complexes of Fe(II), Co(II), etc. Solids such as MnF3 and various complexes of R4n(111) are known. I n general, the reasonably stable

compounds of Mn(II1) are those that are only slightly soluble or slightly dissociated in solution. Although several compounds of Mn(1V) are known, &In02is certainly the most important, both as a source of manganese in nature and as a useful oxidizing agent. Heating RhOz causes it to decompose to hfn3O2 or Mn304, depending upon conditions. The latter compound is apparently most realistically described as Mn'11CIn'11204 (a distorted spinel). The dark green ion ktn043- contains Mn(V). This ion, which is reasonably stable only in strongly alkaline solutions, has been postulated as an intermediate in various oxidation reactions. For the +6 oxidation state, we have data for the dark green manganate ion, Mn042-which is stable only in alkaline solution. The violet permanganate ion, Rho4-,and its salts contain Mn(VI1). Permanganate apparently gives ILln03+in cold concentrated sulfuric acid, but it should be noted that solutions of permanganate in concentrated acids have exploded.

111. "SIMPLE"COMPOUNDS AND AQUEOUSIONS For the heat of sublimation of metallic manganese we follow h/lahlo and adopt A H o 2 9 8 = 67.2 kcal/mole from the work of RfcCabe and Hudson," whose results are in substantial agreement with those of a more recent investigation.12 This value is open to some question because higher values have been reported on the basis of other vapor pressure data and we can also calculate a higher (but less certain) value from the AHfO, heat of sublimation, and heat of dissociation of MnO. The So2g8 values for Rln(c) and ?t/ln(g) are taken from Kelley and King,13 and the AGfO for Mn(g) is calculated by combination of its AHf" with the entropies. The A H ~ Ovalues for the gaseous ions of manganese are taken from NBS Circular 5002 after adjustment for the new AHfO for JSn(g) we have adopted. We now consider the oxides RInO, MnzO3, lInaO4, and R1nOz. On the basis of earlier calorimetric investigations, Mahlo has adopted AHfo values for MnO(c), hln304(c),and NnOz(c), which we also adopt. We choose AH*' of ?tlnzOs(c) from the high-temperature equilibrium measurements of Otto. l4 Entropies of MnO(c), lClnzOs(c),and MnOz(c) are taken from Kelley and King13while that for Rfn304(c)is taken from Otto.14 Free energies are calculated by cambination of heats (10) A. D. Mah, U. S. Bureau of Mines Report of Investigations No. 5600, U. S. Government Printing Office, Washington, D. C., 1960. (11) C. L. McCabe and R. G. Hudson, J. Metals (Japan), 9, 17 (1957). (12) P. L. Woolf, G. R. Zellars, E. Foerster, and J. P. Morris, U. S. Bureau of Mines Report of Investigations No. 5634, U. S. Government Printing Office, Washington, D. C.,1960. (13) K.K. Kelley and E. G. King, U. S. Bureau of Mines Bulletin No. 592.U. 5. Government Printing Office, Washington, D. C., 1961. (14) E.M.Otto, J . Electrochem. Soc., 111, 88 (1964).

CHEMICAL THERMODYNAMICS OF MANGANESE AND ITSCOMPOUNDS and entropies. High-temperature emf datal6 lead to values for MnO in agreement with those adopted here. Several high-temperature investigations of the manganese oxides have led to other thermodynamic data that we do not accept. For example, Matsushima and ThoburnI6 have employed DTA to obtain 0 2 pressures at various high temperatures and have calculated AH" values by means of the van't Hoff equation. We have combined their reported AH" values (at 835 and data from Kelley" to obtain 1200°1InOa2-

-0. SB8

Xt1O43---0.3MnO:2-

I

XnOa'

I

--0.5jsd04-'

-0.603

L 3Mn02-(aq)

+ 2H20 = 2MnOd-(aq) + MnOz(c) + 40H-(aq)

is only slightly greater than unity. Thus Mn042is reasonably stable in solutions containing moderate concentrations of OH- and becomes more stable as the pH increases. Although the equilibrium constant for disproportionation of MnOd3- is large, its instability is diminished by increasing concentration of OH-. Although there is no necessary simple relation between potentials and reaction mechanism, the potential diagram for acidic solution accounts for the ordinary reduction of Mn04- to Mn2+,since every intermediate species that might be formed is also a strong oxidizing agent. On the other hand, reduction of i\ln04- in alkaline solution commonly yields MnOz, which is consistent with the observation that neither N n 0 2 nor ?vfn(OH)3 is a strong oxidizing agent in alkaline solution. Mechanisms of oxidations by permanganate (and other high oxidation-state species of manganese) have been reviewed by Ladbury and C u I l i ~ . ~ ~ We now turn to consideration of AHf" and .!Lo values for the aqueous ions of manganese along with the related thermodynamic properties of compounds of manganese. We combine the solubility23of KMn04 with T~ = (0.6) estimated by comparison with activity coefficients of other 1:1 electrolytes a t the same concentration to obtain its K,, and thence its AGO of solution. Then combination of this value with the AGf' value for Mn04-(aq) quoted earlier and AGr" for K+(aq)l gives AGr" = -176.8 kcal/mole for KMn04(c). We also calculate Sz" = 46.5 gibbs/mole for ?\ln04-(aq) by combining the AGO of solution with the heat of solutionz4 and entropies of K+(aq) and KMn04(c).13 (22) J. W.Ladbury and C. F. Cullis, Chem. Rev., 58,403 (1958). (23) A. Seidell, "Solubilities of Inorganic and Metal Organic Compounds," 3rd ed, D. Van Nostrand Co., Inc., New York, N. Y., 1040. (24) T. Nelson, C.Moss, and L. G. Hepler, J . Phys. Chem., 64,376 (1960).

-0.34

1

The AGf" and SZofor hln04-(aq) lead to AH!" = - 129.9 kcal/mole for this ion. We have several paths to the AHf" of MnZ+(aq). From the calorimetric determination of the heat of reduction of Mn04- reported by Hugus and Latimer?j and the AHf" for bln04-(aq), we calculate AHf" = -53.4 kcal/mole for Mn2+(aq). Another approach to this quantity is by way of AHf" and AH" of solution of IllnC12(c). Two i n v e s t i g a t i ~ n s are ~ ~ *in~ ~good agreement for the heat of solution, which we combine with a small estimated heat of dilution to obtain the desired AH" = -17.5 kcal/mole for solution of R/lnClz(c). Koehler and Coughlin28have reported AHfO = - 115.2 lical/mole for MnClz(c), while high-temperature data29*30 lead to AHf" = -116.9 kcal/moIe. Combination of these latter values with the heat of solution leads to AH!" = -52.8 and -54.5 kcal/mole for ?\lnz+(aq). Walliley3' has reported AH!" = -53.2 (his best) and -52.8 (his second best, our recalculation) kcal/mole for Rfn2+(aq). We also use the data of Southard and S h ~ m a t for e ~ ~MnSOl(c) with the most recent auxiliary datal to calculate A H ~ O = -254.9 kcal/mole for MnS04(c). Next we combine the heat of solution reported by T h ~ m s e with n ~ ~ corrections to 25" and infinite dilution estimated by comparison with

(25) G. Z.Hugus and W. M. Latimer, J . Electrochem.SOC.,98,296 (1951). (26) P.Ehrlich, F.W. Koknat, and H. J. Seifert, 2. Anorg. AZZgem. Chem., 341,281 (1965). (27) P.Paoletti and A. Vacca, Trans. Faraday SOC.,60,50 (1964). (28) M.F. Koehler and J. P. Coughlin, J . Phys. Chem., 63, 605 (1959). (29) L. Rossemyr and T. Rosenqvist, Trans. A I M E , 224, 140 (1962). (30) H.Schafer, Z . Anorg. Allgem. Chem., 278, 300 (1955). (31) A. Walkley, J . Electrochem. SOC.,93, 316 (1948); 94, 41 (1948). (32) J. C. Southard and C. H. Shomate, J.Am, Chem. SOC.,64,1770 (1942). (33) F. R. Bichowsky and F. D. Rossini, "The Thermochemistry of the Chemical Substances," Reinhold Publishing Corp., New York, N. Y.,1936.

CHEMICAL THERMODYNAMICS OF MANGANESE AND ITSCOMPOUNDS

741

dataa4 for ZnSO,, CdSO1, and CuSO4 to obtain the (OH)z(c). Berg and K o v y r ~ i n a *have ~ investigated dissociation of solid Mn(OH)z into solid MnO and HzOstandard heat of solution and thence AHf" = -53.0 kcal/mole for Mn2+(aq). We adopt AHt" = -53.2 (g). Their reported thermodynamic properties are not internally consistent, and their results are in poor agreekcal/mole for Mn2+(aq) and use this value in subsement with the values already cited for thermodynamic quent calculations. properties of MnO(c) and Mn(OH)z(c). We combine the AGt" and AHt" for Mn2+(aq) to For MnFz(c) we adopt the values cited b y Mah'O and obtain its 3," = -17 gibbs/mole. All other paths to Kelley and King.la Both the AGf" and AHt" values this entropy involve uncertain estimates of entropies of appear to be uncertain by several kilocalories/mole. solid compounds. Combination of the previously cited heat of soluCalorimetric data ( f1.5 kcal) from Hill and Williamti0n2~-~7 with the AHf" of Mn2+(aq) leads to AHt" = sona6lead to AHf" = - 157 kcal/mole for Mn04z-(aq). By combination of the AGt" and AHt" values, we obtain - 115.6 kcal/mole for MnClz(c), which is the value we adopt. This value falls between the values cited Sz" = 14 gibbsjmole for this ion, which is about the earlier.28-30 Combination of this AHt" with S O 2 9 8 same as the more certain values for Cr02-(aq) and from Chisholm and Stout41gives AGr" = - 105.9 kcal/ Mo02-(aq). Using our values for the thermodynamic properties mole for MnClz(c). Ifahlo has cited earlier calculations of AHt" of Mnof the aqueous ions, we now turn to systematic discusBrz(c) from gaseous bromine. The more recently detersion of compounds, beginning with the zero state and mined heat of solution42 of this compound combined proceeding to the +7 state. with an estimated heat of dilution and our AHt" of Although the solution chemistry of manganese carI\ln2+(aq) permits us to calculate AHr" = -92.6 bonyl, Mnz(CO)lo, has not been much investigated, kcal/mole for formation of lllnBrz(c) from liquid broits thermodynamic properties are of special interest nine. We combine this AHf" with our estimated S O 2 9 8 because of the Mn-Mn bond. The heat of formation to obtain the tabulated AGf". has been determined calorimetrically by Good, FairFor MnIz(c) we adopt AHf" = -64.2 kcal/mole from brother, and WaddingtonlX6who report AH? = the heat of solution43 combined with our estimated -400.9 kcal/mole for Mnz(CO)lo(c). Cotton and Monheat of dilution and AHt" values for Mn2+(aq) and champ37have made vapor pressure measurements and with AHt" leads to the I-(aq).l Our estimated SoZ98 calculated AH" = 15.0 kcal/mole for vaporization. tabulated AGt ". We combine these data to obtain AHr" = -385.9 hydrates ~J~ of kcal/mole for Mnz(CO)lo(g). Cotton and RiI~nchamp~~ Results of several i n v e s t i g a t i o n ~ ~of~ ~ manganous halides are in poor agreement and we do not have used these data in calculating that the Rln-Mn bond energy is 34 f 13 kcal/mole. A mass spectrotabulate thermodynamic properties for these commetric investigation by Bidinosti and M ~ I n t y r ehas ~ ~ pounds. Hayes and Martin44have determined the solubility led these workers to report 18.9 f 1.4 kcal/mole for of R/In(IO&(c) a t several temperatures and report the Rln-Mn bond energy. For manganous hydroxide we adopt K,, = 1.6 X K,, = 4.79 X lo-', AGO = 8.61 kcal/mole, AH" = 10-13 from references cited by and calculate the 2.28 kcal/mole, and AS" = -21.1 gibbs/mole for the standard free energy of solution. Combination of this solution reaction a t 298°K. We combine these values withdata for 1 0 3 - (as)' and llln2+(aq)to obtain AGt" = value with the free energies of the ions gives AGf" = - 145 kcal/mole for solid manganous hydroxide rep-124.9 kcal/mole, AHf" = -161.3 kcal/mole, and resented by Mn(OH)z(c). Two different calorimetric S O 2 9 8 = 61 gibbs/mole for ;l/ln(103)z(c). Heat capacity data leading to a third-law entropy are needed and measurements by Thomsen give data33that we have combined with known and estimated heats of dilution would thence lead t o a new Sz"for Mn2+(aq). in obtaining AHf" = - 167 kcal/mole for Rln(OH)z(c). Manganous sulfide, MnS, has been investigated by a These values lead to S O 2 9 8 = 23 gibbsjmole for A h number of workers. I < e l l e has ~ ~ ~calculated AGozgl = 15.4 kcal/mole, AH029s = 28.9 kcal/mole, and ASozgs = 35.3 gibbs/mole for (34) J. W. Larson, P. Cerutti, H. K. Gsrber, and L. G. Hepler, J . P h y s . Chem., 72,2902 (1968). (35) R. A. W. Hill and J. F. Williamson, J . Chem. SOC.,2417 (1957). (38) W. D. Good, D. M. Fairbrother, and G. Waddington, J . P h y s . Chem., 62,853 (1958). (37) F. A. Cotton and R. R. Monchamp, J . Chem. Soc., 533 (1960). (38) D. R. Bidinosti and N. S. McIntyre, Chem. Commun., 555 (1966). (39) L. G.Sillen and A. E. Martell, "Stability Constants of MetalIon Complexes," Special Publication No. 17,The Chemical Society, London, 1964.

(40) L. G.Berg and V. P. Kovyrsina, Zh. Neorg. Khim., 8 , 2041 (1983);Russ. J.Inorg. Chem., 8, 1065 (1963). (41) R. C. Chisholm and J. W. Stout, J. Chem. Phys., 36, 972 (1962). (42) P. Paoletti, Trans. Faraday SOC.,61, 219 (1965). (43) P. Paoletti, A. Sabatini, and A. Vacca, ibid.,61, 2417 (1965). (44) A. 11. Hayes and D. S. Martin, J . Am. Chem. SOC.,73,4853 (1951). (45) K.IT. Kelley, U.S. Bureau of Mines Bulletin No. 406, U. S. Government Printing Office, Washington, D. C., 1937; reprinted as Bureau of Mines Bulletin No. 601,1962.

T. A. ZORDANAND L. G. HEPLER

742 MnClz(c)

+ HzS(g) = MnS(c) + SHCI(g)

This value gives AHt" = -329.2 kcal/mole for MnSO4.H20(c). K e l l e ~quotes ~ ~ an equilibrium dissociafrom the high-temperature data of Jellinek and Podatm at 25". This pressure tion pressure of 2.64 X jaski. We use our data for h/InCl2with the above AGO leads to AGO = 2.15 kcal/mole for the dissociation reacand AH" values in calculating AGr" = -50.0 kcal/mole tion and AHf" = -47.5 kcal/mole for RfnS. The third~ this reaction is 30.5 gibbs/mole, which law A S o Z Sfor MnSO4.H*O(c) = MnS04(c) + HzO(g) differs significantly from that cited above. We have Combination of this AGO with AH" from AHr" ~ leads ~ ~ to ASozg8 = 48 gibbs/mole for the distherefore combined this latter AS"298 with the A G ~values calculated by K e l l e ~ ,to ~ obtain AH"298 = 27.5 kcal/ sociation. We have reliable S " ~ 9 8values for both Mnmole for the reaction and thence AHr" = -48.9 kcal/ S04(c) and H2O(g) and are confident that an estimole for MnS(c). Since this last value depends on mated Sozg8 = (37) gibbs/mole is satisfactory for Mnthe free energy of reaction but not its temperature S04-H20(c),leading to AS"298 = 35 gibbs/mole for the derivative, it may be more reliable than the first value dissociation. We combine this AS" with AHr" values given here. to obtain AGO298 = 6.1 kcal/mole for the dissociation We have recalculated the results of calorimetric reaction and thence a dissociation pressure of 3 X 10-5 measurements by Thomsen and BertheloP to obtain atm. This pressure seems more consistent with the AHr" = -48 and -49 kcal/mole for MnS. Combinastability of the hydrate than does the larger pressure tion of these values with the third-law So298= 18.7 cited above. Our own interpretation of the equilibrium gibbs/mole leads to AGt" = -49 and -50 kcal/mole ,~ data of Carpenter and Jette49cited by K e l l e ~ suggests for MnS. A/Iah10 has adopted AHr" = -49.5 kcal/ that the quoted dissociation pressure actually refers to mole from the calorimetric work of Jeff es, Richardson, a higher hydrate. We adopt a AGr" for MnS04. and Pearson. Combination of this value, which may HzO(c) based on the AHr" and our estimated So. be the most reliable of those we have cited, with the Although there are data for higher hydrates of Mnentropy leads to AGf" = -50.5 kcal/mole for h h S . SO,, there are so many uncertainties that we are not We adopt AGr" = -50.5 kcal/mole, AHf" = -49.5 listing thermodynamic properties for these compounds. kcal/mole, and S O 2 9 8 = 18.7 gibbs/mole for MnS(c). For h4n(N03)2*6H20(c,liq), we combine the results Combination of this AGr" with that for S2-(aq)1*46 of Shomate and YoungSowith the most recent auxiliary for MnS. Solubility measureleads to K,, = 2 X data' to obtain AHr" = -566.9 and -557.3 kcal/mole, ments have led various investigators to reporV9 "exrespectively. Although there have been difficulties perimental" K,, values covering a wide range, but it associated with determining the melting point of this appears to us that the "best" values are in reasonable compound, it is certainly close to the 25.8" cited in NBS agreement with this calculated value. Circular 500.2 For I~Lfn(N03)~(c)and Mn(N03)2 A recent investigation of NnS is that of Adami and 3H20(c) we adopt AHr" = - 167 and -355 kcal/mole, King4' who have reported AHr" = -51.0 kcal/mole respectively, from data of uncertain accuracy cited in for hlnS (alabandite). Taking SZ98" = 18.7 gibbs/mole NBS Circular 50Ou2 as before, we obtain AGf" = -52.0 kcal/mole and calFor R/lnCO3(c) we adopt AHr" = -214 kcal/mole culate K,, = 2 X 10-13 for the alabandite modification on the basis of investigations cited in NBS Circular 500.2 of MnS. It might also be noted that sat^^^ has taken Our recalculations with the data of Thomsen and BerAGr" = -50.0 kcal/mole for alabandite from Robie. theloP lead to AHr" = -211 kcal/mole for MnC03This AG," leads to K,, = 5 X (ppt), in agreement with the value listed in NBS CirAlthough it is known that MnS can exist in several cular 500.2 Anderson's low-temperature data have led forms and solubilities have been reported for both NBS Circular 500,2 Kelley and King,13 and Mahlo to pink and green forms, we have been unable to disadopt 8 ' 2 9 8 = 20.5 gibbs/mole for MnC03(c). Kostinguish thermodynamically between these forms. tryukov and Kalinkinasl have reported 5'298 = 27.0 For RlnS04(c), we adopt AHfO = -254.9 kcal/ gibbs/mole for a "synthetic" sample (rather than the molei132 as mentioned earlier. Combination of this mineral used by Anderson) of MnC03. This entropy AHf" with the entropyl0,l3leads to AGf". is larger than entropies of similar carbonates, but we Heats of solution measured by T h ~ m s e nlead ~ ~ to know of no other reason to question it. AH" = -6.0 kcal/mole for Although it is not entirely clear how the AHr" and e

MnSOd(c)

+ HzO(liq) = MnSOd.HtO(c)

(46) J. W.Kury, A. J. Zielen, and W. M. Latimer, J . Electrochem. Soc., 100,468 (1953). (47) L.W.Adami and E. G. King, U. 9.Bureau of Mines Report of Investigations No. 6495,U. S. Government Printing Office, Washington, D. C., 1964. (48) M. Sato, EEectrochim. Acta, 11, 361 (1966).

values above should be combined, we have chosen t o associate the lower e n t r ~ p y ~ ~with ' ~ * MnC03(c) '~ SO298

(49) C.D.Carpenter and E. R. Jette, J. A m . Chem. Soc., 45, 578 (1923). (50) C.H.Shomate and F. E. Young, ibid., 66,771 (1944). (51) V. N.Kostryukov and I. N. Kalinkina, Zh. Fiz. Khim., 38,780 (1964).

CHEMICAL THERMODYNAMICS OF MANGANESE AND ITSCOMPOUNDS and the larger entropy51 with MnCO~(ppt). Thus we obtain AGf" = -196 kcal/mole for MnC03(c) and AG," = - 194 kcal/mole for MnCOa(ppt.) The corresponding solubility products are K., = lo-" for MnCO,(c) and K,, = 10-lo for MnC03(ppt). Previously r e p ~ r t e d ssolubility ~ ' ~ ~ products fall between 3 X and 2 X lo-". It should also be noted that Garrels and Christ62quote AGt" = - 195.7 kcal/mole for natural rhodochrosite, which corresponds to K,, = 3 X 10-11. Calorimetric measurements by bar an^^^ give AHf" = -284.8 kcal/mole for MnMo04(c). Combination of this value with our estimated entropy leads to the tabulated AGf". We use this AGf" with that for Mo042-(aq)54 to calculate K,, = on the assumption that the solid phase in equilibrium with saturated solution is indeed MnMo04(c) We have several highly uncertain paths to AGr" for Mn(OH)3. First, we may combine the AHr" = -212 kcal/mole for i\h(OH)3 listed in NBS Circular 5002 with our estimated So2%= (24) gibbs/mole to obtain AGt" S - 181 kcal/mole. We may also adjust the heat of formation listed by Bichowsky and Rossini33 for different AH*' values for reference compounds to obtain AHfO g -224 kcal/mole and thence AGf" E -193 kcal/mole for Mn(OH)3. All of these values are from the calorimetric data of Petersen, which are d a c u l t to interpret. Our view inclines toward the more exothermic value. Still another approach is to take the AGO for hydration of Mn201 to Mn(OH)3 by liquid water to be small, which leads to AGrO E -190 kcal/mole for Mn(OH)3. Adopting an uncertainly weighted average AGf" E -190 kcal/mole, we obtain potentials cited earlier and also calculate K,, S for Mn(OH)3. Although this K,, is smaller than several published estimates, it is not unreasonable, and we know of no experimental data that contradict (or confirm) this value. For the solubility of barium manganate (BaMnOJ, Latimer3 has given K,, = 1.5 X 10-lo. Combination of this K,, with our AGf" for Mn04+(aq) and AGt" for Ba2+(aq)2gives AGt" = -268 kcal/mole for BaMn04(c). We estimate SoZs8= (37) gibbs/mole for this compound and combine with the AGf" to obtain its AHt" = -293 kcal/mole. The AHf" = -282 kcal/mole listed in NBS Circular 5002 appears to be in error. We have no further data for solid manganates, but a comment on estimation of AHr" of KzMn04(c)may be worthwhile. Hill and W i l l i a m ~ o n ~have ~ combined (62) R. M. Garrels and C. L. Christ, "Solutions, Minerals and Equilibria," Harper and Row, Publishers, New York, N. Y., 1965. (53) R. Barany, U. 8. Bureau of Mines Report of Investigations No .6618, U. 8. Government Printing Office, Washington, D. C., 1965. (54) R . L. Graham and L. G. Hepler, J. Am. Chem. Soc., 7 8 , 4846 (1966).

743

experimental data on Mn0d2-(aq) with their estimate that the heat of solution of K2Mn04(c) is about 10 ~I AHf" kcal/mole [as it is for K M ~ O ~ ( Cto) ~calculate of K2Mn04(c). It seems more reasonable to estimate AH" = (4) kcal/mole [by analogy to K ~ C ~ O ~ ( C ) ~ ~ ~ ~ for the heat solution of KzMn04(c),which leads to its AHf" = -282 kcal/mole. The heat of formation of KMn04(c) is calculated from its heat of solution2* and combined with its entropylOJato yield its AGf".

IV. COMPLEX COMPOUNDS AND IONS The best data on hydrolysis of Mn2+(aq) appear to be those of P e r ~ k , who ~ ' has reported the following. Mnz+(aq)

+ H,O(liq) = MnOH+(aq) + H+(aq) K

=

2 . 6 X lo-"

We use this equilibrium constant in calculating AGof = -97.3 kcal/mole for MnOH+(aq). PerrinS7 also reported K values at other temperatures, but we do not believe the accuracy justifies calculation of AHfO or S2"for MnOH+(aq). I n addition to the stability data and references cited by Sillen and Martell,39we call attention to some other investigations of complex ions of Mn2+. Cell measurements by Nair and N a n ~ o l l a shave ~ ~ led to AGO = -3.07, AHo = 3.37 kcal/mole, and AS" = 21.6 gibbs/ mole for the association of Mn2+with sod2-as in K = 181 Mn*+(aq)+ SO,*-(aq) = MnSOd(aq ion pair) Combination of these data with our data for Mn2+(aq) and those for SOd2-(aq)' lead to the tabulated values for MnS04(aq ion pair). Atkinson and PetrucciSe have measured conductances of manganous sulfate and manganous m-benzenedisulfonate in water and mixed solvents. Their data indicate that the latter may be very close to a true strong electrolyte in dilute aqueous solution, while the former was found to be associated, in satisfactory agreement with the results cited above.68 We also call attention to recent ultrasonic investigationsmV6'of solutions of these same salts. Although the experimental results and their interpretation are still controversial, the method is ultimately capable of providing detailed information about association of ions in solution. Nair andNancollad2 have investigated association of Mn2+(aq) with malonate at several temperatures and have calculated A H o = 3.53 i 0.2 kcal/mole for the association reaction. Later, McAuley and N a n c o l l a ~ ~ ~ (55) (56) (57) (58) (59) (60) (61) (62) (63)

L. G. Hepler, ibid., 80, 6181 (1968). C. N. Muldrow and L. G. Hepler, ibid., 79, 4046 (1957). D. D. Perrin, J. Chem. Soc., 2197 (1962). V. S. K. Nair and G. H. Nancollas, ibid., 3934 (1959). G. Atkinson and S.Petrucci, J . Am. Chem. SOC.,86,7 (1964). G. Atkinson and 8. K. Kor, J. Phya. Chem., 70,314 (1966). L. G. Jackopin and E. Yeager, ibid., 70,313 (1966). V. 8. K. Nair and G. H. Nancollas, J . Chem. SOC.,4367 (1961). A. McAuley and G. H. Nancollas, ibid., 989 (1963).

T. A. ZORDANAND L. G. HEPLER

744

TABLEI investigated the same system calorimetrically and THERMODYNAMIC DATAFOR MANQANESE AT 298"Ka reported AH" = 3.68 f 0.12 kcal/mole for the same AHfO AUfo, SO, kcsl/mble kcal/mole gibbs/mole association reaction. McAuley and N a n c o l l a ~ ~ ~ ~Substance ~~ Mn(c, a) 0 0 7 , 6413 have also investigated association of Mn2+(as) with 7.7213 Mn(c, Y ) succinate in these same ways and reported AH" = Mn(g) 67.210 57.1 41 .4913 Mn+(d 239.72 2.95 i 0.1 kcnl/mole and AH" = 3.02 f 0.15 kcal/ 600.42 Mn2+(g) mole. This agreement of resdts obtained by dif-53.295,27-80 -55.110.1.3 -17 Mn2+(aq) ferent methods is characteristic of the best investigaMnOH +(as) -97.357 9% MnS03(aqion pair) -267.1" -236. l m tions, but is all too uncommon. Other data for asMna+(aq) --213 sociation reactions are presented in a book by NanMn0d3-(aq) ~-127m collas.66 - 15785 - 120.520 MnO,*-(aq) - 129.92' - 107,619 MnO r-(aq) Brannan, Dunsmore, and NancollasB7have investiMnO(c) -92.0'0 -86.7 gated association of Mn2+(aq)with glycine, and Desai -228.71' -210.1 MnzOdc) and NairQ have done an unusually thorough investi-331,310 -306.2 Mn30r(c) MnOz(c) - 124,410 - 111 , 3 gation of complexes with phthalate. Atkinson and - 16723 - 14g39 Mn(OH)t(c) Bauman69 and Davies and Dunning" have employed >In (013)3 ( e ) - 221 N-190 emf and calorimetric measurements in investigating MnFz (c ) N-19010 N-17910 -115 636t27 -103.9 MnCldc) association of &In2+with 2,2'-bipyridine, Data for -228.8'5 CsMnCl2(c) oxalate complexes have been reported." - 333. sze CszMnC14( c) Ciampolini, Paoletti, and S a c c ~ nJ i3 ~have ~ carried -43.5 ,934 Cs3RfnCls(c) - 224,725 RbMnC&(c) out detailed calorimetric investigations of the associaKMIICI~ (C ) - 223,425 tion of Mn2+ (and ions of other transition elements) NaMnCldc) - 214.596 with ethylenediamine and various other amines. The MnBrz(c) -92.6Q -89 MnIl(c) -64.2'3 - 6.5 association of manganous thiocyanate has been studied -161.3'' -124.9" M n (I0 3 )z(c) by Nancollas and Torrance.74 Thermodynamic paramhInS(c) -49.510 -50.5'O eters for the association of Mn2+(aq) with adenosine MnS(alabandite) -51.047 -254.91.8' -229.1 MnSOa(c) tripho~phate'~and h i ~ t i d i n e ~have ~ , ~ also been reMnSOc .&O(c) -329.2a3 -289.9 ported. - 1673 Mn(NO3)dc) Various studies of the association of MnZ+(aq) with - 35*53 i\/ln(NOa)z 3&o(c) - 566.95'3 Mn(NO& .6H20(c) aqueous halide ions have been made. Ciavatta and Mn(?J0~)z~611zO(liq)-557.350 GrirnaIdP have investigated the competition of &In2+RfnCOs(c) -2149 - 1966' 20~52,10,13 (as) and H+(aq) for F-(aq). They report the followMnCOdppt) -21123 - 194 27.061 hlnMo04(c) -284, S58 - 261 (32) ing. ( -282) XnMnOifc) Mn*+(aq) F-(aq) MnF+(aq) Qt = 6 . 2 KMnOd(c) -200.624 -176.8 41.013 N

+

i=

Morris and Short7ghave reported hln*+(aq)

+ Cl-(aq)

= MnCl+(aq)

Q0.7

= 3.85

BaMnO4(c) -293 -268' Mnz(C0)dc) -400.9*6 MnACO)io(g) 385.936,37 6 Superscript numbers are references.

(37 )

-

Approximate equilibrium quotients for complexes con(64) A. MoAuley and G. H.Nancollas, J . C h n . SOC.,4458 (1961). (65) A. McAuley, G.H. Nancollas, and K. Torrance, Inmg. Chem., 6,136 (1987). (66) G. H. Nancollaa. "Interactions in Electrolyte Solutions." Elsevier Publishing Co.,Amsterdam, 1966. (67) J. R. Brannan, H. 5. Dunsmore, and G. H. Nancollas, J . Chem. Soc., 304 (1964). (68) I. R. Desai and T'. S. K. Nab, ibid., 2360 (1962). (69) G.Atkinson and J. E. Bauman, Inorg. C h m . , 1, 900 (1962); 2,64 (1963). (70) R.L.Davies and D. W. Dunning, J. C h m . SOC.,4168 (1965). (71) A. ~VcAuleyand G. 13. Nancollas, ibid., 2215 (1961). (72) M.Ciampolini, P. Paoletti, and L. Sacconi, Nature, 186, 880 (1960); J . Chem. SOC.,4553 (1960);2994, 5115 (1961);3589 (1963). (73) P. Paoletti and M. Ciampolini, Inorg. Chem.,6, 64 (1967). (74) G.H.Nancollas and K. Torrance, ibid., 6, 1567 (1967). (75) M. M.T. Kahn and A. E. Martell, J . Am. Chem. Soc., 88, 668 (1966). (76) H.Kroll, ibid., 74,2032 (1952). (77) D. Perrin and V. 9. Sharma, J . Chem. SOC.,A , 724 (1967). (78) L. Ciavatta and M. Grimaldi, J. Inorg. ,?'ztcZ. Chem., 2 7 , 2019 (1965). (79) D. F.C.Morris and E. L. Short, J . Chem. SOC.,5149 (1961).

taining more C1- indicate that these species are significant only in concentrated solutions. Heat of solution measurements by Ehrlich, Koknat, = -9.73, and SeifertZ6permit calculation of -6.21, -3.63, and -0.71 kcal/mole for reactions of type MCl(c)

+ MnCb(c) = MMnCls(c)

in which %1 represents Cs, Rb, K, and Na. We use these results with our AHtO for hiInClz(c) and data for J.ICl(c) compounds from NBS Circular 5002 to calculate our tabulated AHrO values. Their heats of solution permit calculation of the following. 2CsCl(c)

+ MnCl,(c) = CstMnCL(c) A H " B ~= -11.21 kcal/mole

3CsCl(c)

+ MnClz(c) = Cs&fnCla(c) AH0z9* = - 9 . 8 1 kcal/mole

CHEMICAL THERMODYNAMICS OF MANGANESE AND ITSCOMPOUNDS We also calculate AHtO values for CszMnClr(c) and CsaMnC16(c). The AH data are relevant to the interesting phase diagram22for the CsC1-MnC12 system. Heats of solution of compounds of type (NR& M n X 4 have been determined27e42g43 and discussed, but auxiliary data required for calculation of heats of formation are not available. Wells and Daviesso have investigated the hydrolysis of Mna+(aq) and report Mna+(aq)

+ HpO(aq) = MnOH*+(aq) + H+(aq) Q4.o

=

0.88

It thus appears than MnJ+ is more extensively hydrolyzed than any other +3 ion. Wells and DaviessO also determined the equilibrium quotient (Q4.0) at several other temperatures and calculated AH and A S values for the hydrolysis reaction in this solvent system. Data required to convert their values to our ~~

~

745

standard state are not available, so we tabulate no values for MnOHz+(aq). Suwyn and H a m s 1 report results of interesting kinetic and equilibrium investigations of the reaction MnIIIEDTA(OHz)-(aq)

+ Na-(aq) = MnlllEDTA(Nt)z-(aq) Qo.25

= 32.1

AH and A S values for this reaction are also given. Theys1 also review and discuss evidence for sevencoordination of Mn(II1). Table I presents a summary of thermodynamic data for manganese at 298°K ACKNOWLEDGMENT.-We thank the National Science Foundation for support of this work and Professor C. A. Wulff of the University of Vermont for making his literature search available to us.

~

(80) C . F. Wells and G. Davies, Nature, 205, 692 (1965).

(81) M. A. Suwyn and R. E. Hamm, Inorg. Chem., 6,2150 (1967).