UNFAMILIAR OXIDATION STATES, 1952-54' JACOB KLEINBERG University of Kansas, Lawrence, Kansas
Tm
large number of papers which now regularly appear on the subject of the less familiar oxidation states of well known elements attests to the great interest of the inorganic chemist in this subject. During the period covered by this report much has been added to our knowledge of this area of chemistry. From the preparative viewpoint, perhaps the most significant achievements in the past few years have been the following: the isolation by dry methods of 0x0 complexes containing 3d transition elements in unusual oxidation states, e. g., Cr(IV), Cr(V), Mn(V), Fe(IV), Co(IV), Cu(II1) ; the preparation of transition-element complexes formally related to the simple carbonyls and cont,aining the metal in the zero oxidation state, e. g., Cr(CNC&),, Xi[P(NCO)sb, Ni(CNCsHs)r, Mo(CNCs H&; the continued demonstration that solutions of alkali metals in liquid ammonia are extremely useful for effecting reductions leading to the formation of novel substances incapable of preparation and existence in Ir(NH&. One rewater medium, e . g., K~[CO(CN)~], port, namely, that of the isolation for the first time of a compound containing uninegative rhenium, is worthy of special mention, although it is concerned with one of the less common elements. ELECTROCHEMICAL EVIDENCE FOR UNIPOSITIVE MAGNESIUM
It has long been known that the anodic oxidation of magnesium in aqueous salt solutions is accompanied by t,he evolution of hydrogen a t or near the anode (1). Moreover, the metal dissolves anodically with an initial mean valence number appreciably less than two, as calculated from the loss in weight of the anode and the quantity of current passed through the solution (3). Although it was suggested many years ago (3) that the evolution of hydrogen arises from reaction of the solvent with an unstable unipositive magnesium ion formed in the anodic oxidation of the metal, it has been only recently that evidence, apparently conclusive in nature, has been obtained for this hypothesis. It has been demonstrated (4) that in nonoxidizing salt solutions the anodic evolution of hydrogen occurs in amount corresponding to oxidation from the meas.
For previous papers by the author on this subject see J. CHEM.EDUO.,27, 32 (1950); 29, 324 (1952). The material
included in the current paper is presented in the pattern employed in the above references. Progress for the period 1952.54 is reported, in the sense that, with a few exceptions, primary literature sources were obtained from the Chemical A b s t m t s of these years. Where primary references were unavailable, the Chemical Abst~aetscitation have been given. Major emphasis has again been placed an the less well known states of the more familiar elements.
ured initial mean valence state of the magnesium (ca. 1.3-1.6) to the familiar dipositive ion. I n the presence of any one of a variety of strong oxidizing agents, such as permanganate or chlorate ion, a reduction product other than hydrogen is observed in the neighborhood of the anode, and the quantity of anodic hydrogen is found to be decreased. No reduction product from the added oxidant is observed at the cathode when magnesium is used also as this electrode. Furthermore, nonelectrol.ytic interaction between the electrodes and electrolyte is negligible. When electrolyses are performed in a cell where electrolyte (e. g., sodium chloride or sulfate solut,ion) flows continually past the anode from a reservoir at a higher level into the oxidizing agent (silver ion or permanganate) which is not permitted to come in contact with the anode, a reduction product (metallic silver or manganese dioxide) is again found in the vicinity of the anode. It would appear extremely difficult to explain this last observation except in terms of the formation of unipositive magnesium on oxidation of the metal anode. The anodic behavior of magnesium in solutions of electrolytes in anhydrous pyridine, both in the absence and Dresence of the organic notential electron-acce~tor benzophenone, has been investigated (5). When electrolysis of pyridine solutions of sodium iodide is carried out betweell magnesium electrodes in a cell in which auolyte and catholyte are separated by a sintered glass disk, the metal is oxidized t o the normal dipositive state. In the presence of benzophenone, however, initial mean valence numbers in the neighborhood of 1.6-1.8 are found for the magnesium, but after electrolysis all the metal in the anolyte is in the +2 state. However, hydrolysis of the anolyte yields benzopinacol, a reduction product of benzophenone, in quantity corresponding to conversion of magnesium from its measured initial mean valence state t o the common dipositive condition. It should be emphasized that direct nonelectrolytic reaction between metal and the ele~trolyt~ic solutions does not occur during the length of time employed for electrolysis. The behavior of the magnesium anode in the presence of benzophenone can most reasonably be explained, just as in aqueous salt solutions, by the hypothesis that the metal-enters solution in both the unipositive and dipositive states; the former, being a potentreducing agent, is converted to the latter by reaction with the organic oxidant. THE
METALS
L ~ ~states ~ , of . ~ l ~ potentiomet~c ~ i ~titrs, ~ of liquid ammonia aluminum(lll) iodide with similar solutions of potassium show two
~
,
JOURNAL OF CHEMICAL EDUCATION
74
distinct end points, corresponding to the consumption of one and two equivalents of alkali metal per gram atom of aluminum (6). Although these results were tentatively interpreted to indicate reduction of AI(II1) t o Al(I1) and then to Al(I), the latest available information (7, 8) appears to eliminate the possibility of this stepwise reduction. TABLE .1 Thermodvnamic Data for Various Indium Species Ion
AF0w
cal./mole
Couple
Attempts to reduce aluminum(II1) polarographically in liquid ammonia a t potentials more positive than that required for electron dissolution have been unsuccessful (7). A single polarographic wave is observed with the half-wave potential approximately corresponding to that for reduction of ammonium ion. Titration of aluminum(II1) iodide in liquid ammonia by means of a variety of alkali and alkaline-earth metals to a visual end point gives isolable products which are best explained as arising from titration of ammonium ion formed by the ammonolysis of AI(II1) (8). Experiments designed to demonstrate the presence of lowervalent aluminum species in the reaction mixtures have yielded negative results (7,s). It has been concluded that AlzO is the principal formed in the gas phase when metallic alu. minum is heated with the sesquioxide (9). Volatilization of the sesquioxide alone gives A10, AH298for the re. action: AL08(s) = 2A10(g)
+ O(g)
No evidence was obtained for the formation of the previously reported In,S (13, 13). Lower States of Gallium and Indium. A new method for the preparation of gallium(1) and indium(1) oxides has been described in a brief report that lacks quantitative experimental details (14). Reaction between an excess of elementary gallium or indium with carbon dioxide a t 850' and 10 mm. of mercury pressure proceeds in the following fashion: 2M + C02 = MzO + CO Under the experimental conditions the metal oxide is volatile and distills from the site of reaction. The equilibrium constants for the various possible reactions between indium metal and In+3(aq.) have been determined experimentally by the equilibration of solutions of various concentrations of the tripositive ion (as perchlorate) with the metal, and the determination of the reducing power ava.ilable a t equilibrium in each of the solutions (15). On the assumption that the following were the only significant equilibria:
K1 was found to be 2.4 X 10-" and K 2 1.9 X From the values of the equilibrium constants the thermodynamic data shown in Table I were calculated. SUPEROXIDES
Both sodium and potassium superoxide have been show" to exhibit polymorphism (16, 17, 18). or the sodium compound, three crystalline forms exist (16). One, stable above -50°, has a disordered pyrite stmcture, with a lattice constant of 5.49 R . 2 a t 25'; another, existing from -50 to -77', possesses theopyrite structure, with a unit cell distance of 5.46 A.2 a t -70'; and the third. ~,stahle helow -77". has a marcasite structure, with the a axis equal to 4.26 R., b = 5.54 R., and c = 3.44 A. at -100". The tetragonal form of potassium superoxide, stable a t room temperature, undergoes a transition in the region 60-100' to a cubic form having a lattice constant of 6.09 A. (18). I n addition, there is a transition occurring at approximately -80' to a structure (not yet determined) less symmetrical than the tetragonal form. Little success has been achieved in the few attempts to prepare pure superoxides other than those of the alkali metals. The action of concentrated (from 30 to 90 per cent) aqueous hydrogen peroxide solution on the octahydrates of calcium, strontium, and barium peroxides has been reinvestigated (19). The maximum yield of superoxide obtained in each case was approximately 13 per cent. It should be pointed out that this. appears to be the first report of the preparation of strontium superoxide. The possibility of preparing new superoxides by metathetical reactions in liquid ammonia between a ~
~
~
~~~
heing calculated to be 456 + 10 kcal. The reactions: 4Al(l) AhOds) = 3A1,0(s) (1) Al(I) AbOs(s) = 3A10(s) (2) have been investigated in the temperature range 10002000" by means of a high temperature X-ray technique (10). Reduction of sesquioxide by metal does not occur until 1100". Between this temperature and 1500°, Also is formed in accordance with reaction (1). In the range 1500-1600° reactions (1) and (2) take place simultaneously, whereas above 1600' only the latter reaction occurs. On cooling or rapid quenching, the lower oxides disproportionate to the original reactants. Both AIZOand A10 have cubic crystal stmctures, with lattice constants of 4.98 A. for the former a t 1110" and of 5.67 A. for A10 a t 1700". The system In-In2& has been studied by thermal analysis, the data obtained heing supplemented by metallographic m d X-ray examination of the products (11). The sulfides In& and InS are definitely found to exist in this system, and the spinel-like compounds InS.InzSa and 3InS.In& probably also are formed.
+ +
~
~
~~
~
~
'The unit cell distance reported in (17) is not in good agreement with this value.
VOLUME 33, NO. 2, FEBRUARY, 1956
75
variety of anhydrous metal salts and potassium and sodium superoxides has been investigated, with negative results (80). The salts studied included the nitrates of lithium, magnesium, calcium, cadmium, and zinc, and the tetrammoniates of aluminum chloride and copper(I1) nitrate. I n every case in which the reactants are mixed in the stoichiometric proportions necessary for metathesis, and in which reaction occurs, the superoxide ion is unstable and decomposes with the liberation of oxygen. In most instances of reaction, solid products are obtained containing large proportions of metal peroxide. CATIONIC IODINE AND BROMINE
The electropositive character of iodine and bromine is sufficiently great that the unipositive ions of these elements are capable of existence. These ions may be stabilized by coordination with such nitrogen bases as pyridine ($1). Recent work apparently has demonstrated that iodine(1) can exist to a limited extent in solutions of molecular iodine in water and ethanol. The free-energy changes and equilibrium constants for the reactions Xz(aq.) H20(1) +H20.X+(aq.) X-(aq.) have been calculated from data available in the literature ($8). At 25' the equilibrium constants in the order X = C1, Br, and I are lo-", 10-to, agd 10-lo. These values illdicate that of the halogens only t,he hydrated I + ion should he sufficiently stable to exist in detectable quantity in aqueous solution. The equilibrium constant for the above reaction with iodine has been determilred experimentally by use of cells of the type: Pt, Iz,Agi, H + IKNO.1 I; L, H+, Pt and has been found to be 1.2 X lo-", a value in good agreement with that calculated. The hydrated iodine cation may react with the solvent in the following manner:
+
+
essentially ionic nature of iodine and bromine in INOI and BrNOa (24). Solutions of the compounds are prepared by reaction between the halogen and silver nitrate in absolute ethanol; some tripositive iodine (e. g., I(NO&) is formed in the process, but may be converted to the unipositive state by the presence of excess halogen. Halogen may he removed from the solutions of the nitrates by passage through the cationic resin noted above. With iodine nitrate, fixation of the halogen on the exchanger is practically quantitative, 0.0208 g. of I + being removed from a solution containing 0.0223 g. (per 100 ml. of solvent). The highly polar character of the iodine-pyridine addition compound Ip.Py is brought out by its behavior toward a cation resin (85). Iodine is removed from the complex when a methanolic solution is passed through the hydrogen form of the resin Permutite RS. The conclusion is reached that the complex is only slightly dissociated into I , P y + and I- ions in methanol, since a maximum of about 10 per cent of the total iodine is fixed on the resin. COPPER. SILVER, AND GOLD
Copper(III). The existence of copper (111) in the form of the cuprate(II1) ion, CuOz-, has been definitely established. An attempt ($6) to repeat the reported (87)preparation of copper(II1) oxide, Cu203, by the action of sodium hydroperoxide, NaOOH, on the cuprate(I1) ion, Cu(0H)r--, in concentrated sodium hydroxide solution, has been unsuccessfnl. However, when the product of reaction between cnprate(I1) and sodium hypobromite in concentrated aqueous sodium hydroxide is added to aqueous barium chloride, a fairly stable, red, diamagnetic compound of the composition B ~ ( C ~ O Z ) ~ .isH precipitated. ~O About 96 per cent of the original copper(I1) is converted to the tripositive state in the form of the barium compound. Anhydrous potassium cuprate, KCu02, has been obtained by the oxidation of copper(I1) oxide by means of potassium H?O.I+ + H 2 0 e HOI + HIOt superoxide at 400-500' ($8, $9). The potassium com3HOI + 3H20 1 0 3 + 2 1 + 3H80+ pound is a diamagnetic, steel-blue substance which These equilibria can be repressed by high acid concen- decomposes above 500' with the evolution of oxygen. trations. Oxygen is also liberated in aqueous solution, hut no The presence of cationic halogen in solutions of iodine hydrogen peroxide is formed. The diamagnetic charor bromine in absolute ethanol has been inferred on the acter is evidence that the cnprate(II1) ion is an inner basis of ion-exchange studies ($3). Passage of the orbital (or penetration) complex. Periodate and tellurate complexes of copper(II1) solutions through the hydrogen resin Amherlite IRlOOH results in the fixation of some of the halogen and the (see ( d l ) , pp. 60-1 for a discussion of these compounds) liberation of hydrogen ion. The following reaction has have recently been prepared by a new method, namely, the oxidation of copper(I1) by means of sodium hypobeen proposed for the process. chlorite in strongly alkaline solution, followed by the Res-H+ X2 Res-St + H + + X addition of an acidic solution of either sodium periodate, The positive halogen may be eluted by means of iodide Na2H110s, or sodium tellurate, NapHITeOs (SO). The ion. The relatively small extent of apparent ioniaa- compounds isolated from the resulting solutions possess tion of iodine in ethanol is illustrated by the fact that the compositions Na7Cu(106)2, 16H20 and NasCuonly 0.024.03 g. of the halogen is fixed on the resin from (TeOe)z.20HtO. Data are presented which indicate a solution containing 0.5 g. (per 100 ml. of solvent). that, in the presence of excess alkali, Cu(I1) is conNo quantitative data are given for bromine. verted by hypochlorite to the Cu(OH)&-ion, and that Ion-exchange experiments also appear to show the the latter forms 1:l and 1:2 complexes with both
=
+ =
76
JOURNAL OF CHEMICAL EDUCATION
periodate and tellurate. The 1: 2 complexes appear to pyridine solution has been reported (52). The yellowhave the formulas CU(IO~)~-'and Cu(HTeO&-'. brown compound is insoluble in water and common At 40' the equilibrium constants for the reversible dis- organic compounds, and unstable in air and in solutions sociation of these ions are 8.0 X lo-" and 1.1 X lo-", of acids. respectively. The 1:l complexes are less stable and A general method for the formation of fluoro comhave dissociation constants of 3.4 X lo-= (for periodate) plexes containing certain transition elements in high and 1.8 X lo-' (for tellurate). The heat of formation oxidation states involves the high-temperature fluorinaof the 1:2 periodate complex from the simple ions is tion of the appropriate stoichiometric mixtures of an calculated to be about 7.5 kcal., and for the tellurate alkali metal halide and transition metal halide (5.9). about 20 kcal. It is interesting that stannate(IV), This procedure has been utilized to prepare yellow stibnate(V), and selenate(V1) ions are incapable of fluoro complexes of Ag(II1) of the type MAgF, (M = K, Cs). The compounds are extremely sensitive to forming analogous Cu(II1) complexes. Silver(1II). Oxidation of silver(1) in a strongly moisture. There is an interesting difference in magalkaline medium by means of persulfate ion yields netic behavior between these substances and the copper silver(I1) oxide, Ago. The nature of the substances (111) fluoro complex K3CuFs. The latter is paramagformed when the oxidation is performed in a neutral netic, with a molar susceptibility of 2.8 Bohr magnetons, or acidic solution has been a matter of some question, a value in excellent agreement with that expected for an although there is no doubt that the product contains outer orbital (or normal) complex. On the other hand, tripositive silver. The most recent communication (Sf) KAgF4 is diamagnetic, a strong indication that the on the oxidation of silver(1) nitrate by persulfate in complex is of the inner orbital type. acid medium states that the black precipitate produced It is noteworthy that the nature of the alkali metal has a composition which may be expressed either as appears to play an important role in determining (Ag,0a)..AgS04 or (Ag304)2a.Ag2S208, where a is 2-2.5. whether fluoro complexes containing transition metals As expected, the material is a powerful oxidizing agent. in higher oxidation states will be formed by the method The preparation of his-8-hydroxyquinoline silver(II1) described above. With halides of the larger alkali hydroxide, A~(C~HGNO)~OH.'/?C~H~N.H~O by r e metals potassium, rubidium, and cesium, the procedure action of silver(I1, 111) oxide with the oxine in cold works satisfactorily; in the few cases I$-here lithium and sodium halides have been used, the preparation of the fluoro complexes appears t o be much more difficult. TABLE 2 Analogues of the Simple Volatile Metal Carbonyls Gold(l1). Although the dipositive st,at.e of silver is well established and a fair number of compounds conMetal Compound Physical properties taining this species have been isolated, no substance Chromium Cr(CNCrHd6 Red; m. p. 178.5" containing gold(I1) has been definitely prepared. It Cr(CNCeH,CI-p), Bright red; m. 21525" (decomp.? has recently been shown, however, that the kinetics of Orange-red; m. p. 156' Cr(CNCsHPl+x)s Cr(CNCeH40CHa-p). Intense red; m. p. exchange between CI- ion and AuC14- in the presence of Fe++ ion are consistent with a mecha.nism whereby 124.4" Cr(CNCsHLX-p), Red; m. p. 152.8" the latter reduces Au(II1) to Au(I1) which exchanges C I ~ C N C ~ H ~ C ~ ~ - ZRed; , ~ ) ~m. p. 166" associated chloride very rapidly with C1-, then underCr(CNCsAsClr4,2)~ Red; m. p. 172' Cr(CNCeH.CI.-3,2)a Scarlet; m. p. 1924' goes rate-determining exchange with Au(III), and is Nickel Xi(CNCeH& Canary-yellow solid; finally destroyed by disproportionation (34). decomp. 105' Ni(CNC1H&Hs-p), Ni(CNCinHd)r Ni(CNCH& NXPCl,), , .. Ni(PBrr)r Ni(PF&
Molybdenum
Tungsten
Fe++
Pale-yellow solid: deeomp. 120'' Orange-red solid; decamp. 80' Colorless liquid; b. p.
7o ."..7'
+
+ + +
A u C k = Fe(II1) Au(I1) A u ~ ~ c I -C1*- = A U ' ~ I * - C1Au'~cI*- AuCL- = Au%AuClI*2Au(II) = Au(1) Au(II1)
+
+
+
CHROMIUM AND MANGANESE
Colorless solid; deChromium (0). One of the most interesting recent camp. 200' developments in unfamiliar oxidation state chemistry Yellow; deeomp.~15'; has been the preparation and characterization of m. p. 130 + 2 Yellow; m. p. 86.5' transition-element compounds which may he regarded Colorless; m. p. 98" as analogues of the simple volatile metal carbonyls and Red solid Orange solid thus as substances containing the central element in the Brick-red solid zero oxidation state. Among the ligands bound to the Golden-yellow eolid metal in its zero oxidation state, one finds phosphorus Orange-yellow solid Red; m. p. 120-30' Red. m. p. 205-7' (d'ecamp.)
trihalides and pseudohalides and a large variety of aromatic isonitriles. It is worth noting that these are all ligands for which double-bonded structures with the metal atom, involving d electron pairs of the latter,
VOLUME 33, NO. 2, FEBRUARY, 1956
may be written. A summary of the compounds characterized is given in Table 2. The chromium(0) derivatives which have been prepared are all hexaaryl isonitriles (35). They are made by the addition, in an inert atmosphere, of the isonitrile to a suspension of cbromium(I1) acetate in ethanol, the hexaarylisonitrile chromium(0) compound formed being insoluble in the alcohol. The stoichiometry of the reaction has been shown to be the following:
chromate(V1) and barium carbonate in an oxygen-free nitrogen atmosphere a t 1000' (37). The stoichiometry of reactmionis the following: 2BaCrO