June 20, 1959
THIOPHANE-BORANE AND RELATED ADDUCTS
Accurate comparison of the solvation entropies of the amide ion and symmetrical ions of like charge is difficult because of the ambiguity of the ami4e ionic radius which falls between 2.06 and 1.6 A. depending on the criterion employed. Assuming the upper limit and freely rotating ions, we note that the solvation entropies of the borohydride and amide ions are in good agreement as should be the case in view of the similarity of charge and radius. Since it is likely, however, that the actual radius displayed by the amide ion is somewhat less than 2.G6 A., which value is based on the NH2-H2N distance of neighboring ions with opposed hydrogens, comparison should probably be made with a smaller ion such as the chloride ion for which the solvation entropy is more negative by 5 e.u. In view of present experimental uncertainties, this discrepancy is probably without significance. I t is, of course, to be expected that the entropy of solvation of a monatomic ion should not be more negative relative to a polyatomic ion of similar charge and radius since loss of rotational motion of the latter would result in a more negative solvation entropy for the polyatomic ion. Ionization Constant of Liquid Ammonia.-\Vith the aid of the heat of ionization of liquid ammonia and the entropy of the amide ion determined in this research, the free energy of ionization of ammonia and the equilibrium constant of the reaction: NH3(1) = H f ( a m ) NHz-(am), can be calculated. At 21G°K., AHo = 2 @ . l Z 0 kcal.; ASo SO[H+] So[NH2-] - So[NH3(1)] = O.OC9 - 19.312 - 2O.Sso = - 40.1 cal. deg.-l
+
+
mole-'; AFO = 35.7 kcal. and K = 3.2 X Pleskov and M o n o ~ z o ncalculated ~~ K = 1.9 X from cell potential data a t -50". Correcting their value to 240°K. we obtain 1.3 X which differs substantially from the K obtained in this research. For the ionization of liquid ammonia a t 29S.15' K., ASo = So[NH2-] So[H+]- So[NH3(1)]= - 1 4 . 7 1 2 0 - 24.732= - 39.1 cal. deg.-l mole-'; AFO = 37.76 kcal. and K = 2.2 X For comparison with this value we correct K (Pleskov and Monoszon) to 298°K. giving K = 5.1 x 10-27. Although the difference between the constant obtained by Pleskov and Monoszon and the constant based on the thermodynamic properties of amides previously estimated by has been narrowed substantially with the use of experimental entropy data for sodium amide, a considerable difference still remains. This difference may be explained, however, by the presence of unknown liquid junction potentials in the cells employed by Pleskov and Nonoszon, by an incorrect degree of dissociation of potassium amide assumed for the calculation of their K or by the uncertainty in the free energy of solution of sodium amide employed in this research.
+
+
(30) R Overstreet and W. F. Giauque, THISJ O U R N A L , 69, 254 (1937). (31) V. A Pleskov a n d A. M . Monoszon, J . P h y s . Chem., (U.S.S.R.), 6 , 513 (1936). (82) W L Jolly, Chem. Revs., 60, 851 (1952); U S. Atomic Energy Commission, UCRI, Report h-0. 2201, M a y , 19.53, ref. 19h.
BOSTOS,MASS
(29) By convention Ta[H"(,m)] = 0 0
[CONTRIBUTION FROM
2989
THE
MALLINCKRODT CHEMICAL LABORATORY, HARVARD UNIVERSITY]
Molecular Addition Compounds of Boron. 11. Thiophane-Borane Adducts1s2 BY T. D. COYLE,3H. D.
KAESZ4 AND
and Related
F. G. A. STONE
RECEIVEDDECEMBER 20, 1958 T h e addition compounds (CHr)4S,BHs, ( C H J ) ~ S . B FEtS.BH3 ~, and EtsS.BF1 have been prepared, and their liquid saturation pressures and dissociations in the gas phase examined. Gas-phase dissociation of the adduct Me2S.BH3 has been reinvestigated. As might be expected, all the complexes are weak. Indeed, only the borane adducts are sufficiently associated in the gas phase to permit reliable determination of thermodynamic constants. Nevertheless, it was clearly demonstrated t h a t borane forms much more stable adducts than does boron trifluoride with the thioethers, whereas the reverse is true with ordinary ethers. Moreover, adducts like EtZS.BH3 are much more stable than their oxygen analogs. Furthermore, the order of coordination of the thioethers toward borane is Me2S"Et2S > (CH2)4S,in contrast to the order displayed by ethers, uiz., (CH2)40 > Me20 > EtaO. It is probable t h a t the thioethers display a similar sequence of coordination toward boron trifluoride but in this case, because of the weakness of the boron trifluoride adducts, experimental evidence is only qualitative,
Introduction During the period 1942-1951, the gas-phase dissociations of many molecular addition compounds were studied by Brown and his co-workers and by Coatese5 These addition compounds involved (1) Previous paper, W. A. G. Graham a n d F. G. A. Stone, J . Inorg. Nucl. Chem., 3, 164 (1956). (2) T h e work described in this paper was made possible by t h e award of a Grant (G5106) from the National Science Foundation. (3) N a t v a r Corporation Fellow a t Harvard University, 1958-1959. (4) Public Health Predoctoral Research Fellow of t h e National Heart Institute. ( 5 ) For a review of molecular addition compounds of t h e Group 111 elements see F. 0 . A . Stone, C h m t . Revs., 68, 101 (1957).
a variety of combinations of different Lewis acids with different Lewis bases, e.g., (CH?)40.BF3, Me2S.AlMe3,Me3N.BlIe3, Mepls.Gahle3 and MesP.BF3, etc. It was natural t h a t the relative stabilities of these compounds, usually compared in terms of relative heats of gas-phase dissociation, should excite interest. This was especially so on the realization t h a t if one compared all the adducts made by Brgwn, et ol., and by Coates, it was possible to make a generalization. This generalization has been expressed in several different ways by several authors, e.g., (i) "The order of stability of complexes formed by alkyl derivatives of the
elements of the fifth and sixth group actinl: II?S > R2Se and I i j N > I i , P > I RiSb (H. C Brown, personal cotiimunicatioii ; Coates, J . Chcm Soc., 2003 1 % j 1 ) ) ' 6 (ii) "Xccording to H. C. Brown, base strength, for UXI, .UXj and the like as reference acids, decreases in each of the series NR?> PI ASK > SbRI; OK? > SR, > SeRL > TeKL' So far as the dimethyl derivatives of the Group VI elements are concerned, later work by Graliariil showed that statements such as these do not apply when borane and trimethylborane are the acceptors Furthermore, in Group V Graham shou ed that the displaceiiicnt reaction h'Ie?Y BH,
+ Ale-P
--f
1lc.P B H 3
+ Ale \
took place, suggesting an order of affinity i\lc!P > Me toward borane L2nothcrresult of Graham's work was the demonstration that whereas boron trifluoride and borane were of comparable Len is acidity towards the first-row donor trimethylamine, toward second-row donors, borane coordinated much more strongly, w z . Me2S BHs
>M
e 6 BF3
hIerP BH? > I,le?P BI-9
The unusual orders of coordination toward borane suggest that the classical description of donoracceptor bonding in terms of a dative a-bond may be inadequate when applied to some borane adducts. At present, the exact nature of the boron-ligand bond is unknown ; however, the experimental results are consistent with an interpretation which attributes unusual bonding properties to the borane group I t has bcen suggested' that the three Is atomic orbitals of the three hydrogen atoms of the borane group coiiibiiie to form a pscudo p-orbital having a-symmetry. This p w d o pn-orbital could then overlap a vacant orbital of a ligand, if the ligand possessed such an orbital. Usually only donor atonis from the second or later rows of the Periodic Classification possess vacant orbitals. Thus whereas the donor-acceptor bonds in Me7Tu'.BH3and hIe20.BH7 are adequately described as being of the classical 0-type, the lipandboron bonds in ;21e2S.RH3 and lIe7P.BHI niay have u,n-character. This ~vouldaccount for their being unexpectedly strong. Graham had been concerned only with the methyl derivatives of the Group VI and Group V elements in establishing orders of Coordination t o u m d BH?. It was just possible that these orders of coordination were influenced by the nature o f the groups on the donor atoms. For this reason it was decided to compare stabilities of the adducts Et:O.BX?, Et?S BXI, (CH2)40.BX7and (CHL)4SBX?, where X = H or F, with each other and with those of the inethyl derivatives. The task was somewhat simplified because a few of the coninouiids had alreadv been adequately investigated by others, c.g., EtiO.BFI and (CH&O.RF+.' (0) D. P. Craig, A . nfnccnll. I 1Ie.O > E t 2 0 . It is of interest to establish an order of cobrdination for the thioethers toward boron trifluoride. Unfortunately, it is not possible to do this quantitatively by tensimetry because of the high degree of dissociation of the various adducts in the gas phase. However, it is possible to arrive at a probable order of donor power for the thioethers based on volatility considerations. For complexes of low stability it has long been customary to deduce relative stabilities from relative ~ o l a t i l i t i e s . ~If two addition compounds have similar structures and molecular weight, the less stable has the higher saturation pressure. If two addition compounds differ in molecular weight arid the heavier is the inore volatile, this is taken to indicate that it is also more highly dissociated. This rule is somewhat limited in its application since it is often desired to compare stabilities of two addition compounds where the heavier is the least volatile. For csatnplr, in the work described in
June 20, 1939
2993
THIOPHANE-BORANE AND RELATED ADDUCTS
TABLE IX this paper it is of interest to determine the relative stabilities of Me2S.BF3 and (CH2)4S*BF3. The SULFIDE ADDUCTS OF BORANE A P ~ DBORON TRIFLUORIDE former compound is the more volatile but since the Troulatter has a higher molecular weight, i t is not ton possible to infer that i t is the most stable. For a conB.p. stant situation of this kind we suggest t h a t relative of of cumAb log p (mm.) = B - AT-’ B.p.O base plexa (b.p.) stability may often be determined from a consideraCompound B A (“C.) (“C.) (e.u.) (“C.) tion of Trouton constant values, and the difference 9.220 2346 97 36 2 9 . 0 $61 in boiling point between that of the adduct and Me2S.BHS“ 92 2 8 . 0 4-27.3 Et?S.BHz 8.991 2398 1 1 9 . 3 that of the donor molecule.21 The concept may be illustrated by first consider- (CH2)dS.BH3 8.602 2321 132.5 121 2 6 . 2 $11.5 ing briefly trends in Trouton constant values for MeL3,BFt 9.806 2227 4 8 . 4 36 3 1 . 6 +12.4 molecular addition compounds of the Group I11 EtrS.BF3 8 . 5 3 5 1893 6 1 . 6 92 2 5 . 9 -30.4 elements. A summary of such Trouton constants ( C H & ~ . B F Z 8.045 li88 73.0 121 2 3 . 6 - 4 8 . 0 has been given el~ewhere.~For complexes which A(b. a By extrapolation of the vapor pressure equation. are highly stable (e.g., i\.leaN.BH3, Me3N.BF8, . ) = boiling point of complex (extrapolated) minus boiling Me:iP.AIMes, Me3N.Gallle~),Trouton constants ppoint of donor molecule. From Graham, Ph.D. Thesis, are greater than 21 but usually less than 30, and Harvard University, 1956. From reference 1. variations in the values undoubtedly depend on the extent of dipole associations in a condensed phase Application of these ideas concerning Trouton in equilibrium with a vapor. For complexes of constants and A(b.p.) values to the boron trimoderate or weak stability, those with high fluoride adducts of the thioethers (Table I X ) sugdegrees of dissociation in the gas phase and low or gests a probable order of coordination, MezS > immeasurable enthalpies of dissociation, Trouton Et2S,> (CH2)4S toward BF3. We hope to test constants vary from values as low as 17 to as high this idea by calorimetric studies. as 40. In such complexes other effects occur as The borane complexes of the thioethers are all well as dipole interactions. When high Trouton clearly more stable than the borane complexes of constants are observed, vaporization of the addition dimethyl ether, l8 diethyl etherz5 and tetrahydrocompound is being accompanied by an increase in furan.’e Indeed, the thioether complexes of borane, dissociation. When low Trouton constants are in contrast to the thioether adducts of boron triobserved, the addition compounds being highly fluoride, are sufficiently undissociated in the gas dissociated in the gas phase, the adducts are un- phase to permit evaluation of fairly reliable thermodoubtedly well dissociated even as liquids. There- dynamic constants for reactions of the type fore, in a series of weuk addition compounds,having the same reference base or same reference acid, >S BHdg.) >S(g) ‘/&Hdg) there appears to be a decrease in Trouton constant The results are summarized in Table X. Because with decreasing stability. There are many ex- of the high degree of dissociation of the adducts in amples of this behavior in the literature, e.g., the gas phase, the enthalpies of dissociation are not Me?HP.Bl\lea (Tc = 30 e.u., AH, 11.5 kcal.) so reliable as those for relatively strong adducts and hleH2P.BhIes (TC= 34 e.u., AH, too small to like hle3N.BMe314and hleJ’.BMe~.’~ It is possible be measured)*?; Me~S.AlMea (Tc = 27 e.u., that for (CH2)4S.BH3the uncertainty in AI€’ is as A H , S.5 kcal.), -l.lezSe~AIMe~ (Tc = 25.3, A H , G 1 kcal., but for the less easily dissociated kcal.) and Me2TeA1Me3 (To = 22 e.u., AH, too much as ?rle2S.BH3the uncertainty in AHo is less, probably small to be m e a ~ u r e d ) . ? When ~ vaporization of an not more than 0.5 kcal. Hence, within the addition compound is accompanied by dissociation, limits of error the adducts Ale2S.BH3 and Etzit is well known that extrapolation of the vapor pressure equation cannot give a true boiling point. S.BH3 are of equal stability measured in terms of enthalpies of dissociation. Thiophane-borane is The difference between the apparent boiling point of the complex and the boiling point of the clearly less stable than either Et2S BH3 or Me2S.BHr ligand is significant, however, and merits considera- therefore, toward borane, h l e 2 S e Et2S> (CH2)S tion. From a study of data for weak complexes in in base strength. This sequence of coordination the literature, it appears that for two adducts based on quantitative data is essentially in agreeinvolving the pairing of two bases B or B’ with a ment with the qualitative results summarized in reference Lewis acid A, the base strength of B is Table IX. For (CH2)4S.BH3, the Trouton constant and A(b.p.) are less than those for either greater than that of B’ if Me2S.BH3 or EtzS.BH3, suggesting that thiophane(“b.p.”B:A - b.p.B) > (“b.p.”,’:, - b.p.,’) borane is less stable than the other two borane adducts. Since the gas-phase heats of dissociation [A(b,p.) > A’(b.p.)I Many pairs of weak complexes illustrate this rule, of 1Lle2S.BHa and Et&BH3 within the limits of e . g . , Nle2HP.BF3 ( A H = 14.7 kcal., A (b.p.) = experimental accuracy are equal, i t might have been 86”) and MeH2P.BF3 (AH too small to be measured anticipated that their Trouton constants and A(b.p.) values (Table IX) would have been much A(b.p.) = 33°).24 closer than they are. The observed divergence in (21) For a fuller treatment of this suggestion see H. D. Kaesz, Ph.D. these values could be the result of different degrees Thesis, Harvard University, 1858. ( 2 2 ) See reference 15 and H. C. Brown, J . Chem. Soc., 1248 (1956). of dipole interactions for the two liquids, or per-
+
*
*
(23) G. E. Coates, ibid.,2003 (1951). (24) E. A. Fletcher, Ph.D. Thesis. Purdue University, 1952. also H. C. Brown, J . Chem. Soc., 1248 (1956).
See
( 2 5 ) H E Wirth, F E RIassuth and D X Gilbert J P h y s Chew , 62, 870 (1958).
2994
T. D. COYLE, H. D. KAESZAND F.G.*I.STUNE
haps more likely it is due to the heats of dissociatioii of the two addition compounds in solution being different from their heats of dissociation in the gas phase. TAULEX GAS-PHASE DISSOCIATION DATA"FOR Me2S,BHd, ( CH2)dS. BH3 A N D Et2S.BHd
Vol. h l
Therefore, for h1e-S BHj(g) --+ Me.S(g)
+ BHJ(g)
A I i ' = 20 1 h c d ASU = 35 2 c u
Illolc
'
For the dissociation (CHIIIS B W g ) --+ (CHi)AS(g) BHd(g) Ail" = 1 8 0 h c d A S " = 30 3 e u
+
IIIU~C
and for Me2S,BHa 1332 3.940 6 . 1 1 8 . 1 0.076 0.894 (CH2)rS.BHs 8 7 2 . 8 2 . 6 6 4 . 0 1 2 . 2 - ,061 1 . 1 0 Et?S,BH3 1312 3.890 6.O l i . 8 ,058 0 . 9 1 8
The values reported are averages, except for (CH?)dS, where constants have been calculated as described in the Experimental part. AHO, ASO, AFTO and K,, values are for the dissociation a
E t 3 BHd(g) + Et,S(g)
+ BH,(g)
A H u = 30 0 kcal niolc-' A S " = 34 9 e u
When compared to the order observed with the ethers (CH2)40> M e 2 0 > Etr0,8 27 the order of base strength Me2S"Et2S > (CH2)4S observed toward borane, and probably toward boron trifluoride, is in accord with the idea that steric effects are much diminished with second-row ligand dtoms. In this respect it is interesting that Graham' was able to isolate MezS BMeJ, but h l c ~ O . BMeddid not exist down to - i s " . It is very probable, however, that orbital hybridization of the donor atom in a sequence such as Me$-EtzS > (CHJaS plays an important part i n base strength.' 25 It is unfortunate that little information is available on the bond angles in EtJS and (CHJ4S. CAMBRIDGE, MASS
(20) (a) R. E. McCoy and S. H. Bauer, THISJ O U R N A L , 78, 2Otil (19563; (b) S. H . Bauer, ibid., 18, 5775 (19%).
( 2 7 ) B. Rice and €1. S. Uchida, J . Z'iiys C h c i i i . , 9 9 , 050 ( l K 5 ) (28) 3. H . Gibbs, J. Cht.rrt. Z'izys . 22, 1400 ( l g j l ) .
