Oct. 20, 1963
DONORPROPERTIES OF SULFOXIDES, ALKYLSULFITES, A N D SULFONES [CONTRIBUTIOX FROM
THE
3125
WM. A . NOYESLABORATORY, UNIVERSITY OF ILLINOIS,URBANA, ILL.]
Donor Properties of Sulfoxides, Alkyl Sulfites, and Sulfones BY RUSSELLS. DRAGO,BRADFORD WAYLAND, AND ROBERT L. CARLSON' RECEIVED M A Y2 , 1963 The equilibrium constants and enthalpies for formation of adducts of dimethyl sulfoxide, tetramethylene sulfoxide, tetramethylene sulfone, and diethyl sulfite with the acids iodine and phenol have been measured. The donor properties of these materials have been interpreted in terms of inductive and r-bonding effects in the donor molecules. These data are compared with results from similar measurements on other donors commonly employed a s nonaqueous solvents. In an earlier article a linear relationship was reported t o exist between - A H of formation and the phenol frequency shift AVOHfor some donor-phenol adducts The tetramethylene sulfone adduct deviates from this relationship. I t is proposed t h a t both oxygens of the sulfone are coordinated t o the phenol protoii and a proposal for a low frequency shift is based on this inodel. a n "enthalpy procedure" previously r e p ~ r t e d after , ~ t h e change Introduction in molar absorptivity with temperature is evaluated. Recent interest in the unusual properties of sulfoxides Results and sulfones as aprotic solvents has prompted our inIodine Equilibrium Constants and Enthalpies,vestigation of their donor properties toward the acThe complex maxima for the blue shift iodine peaks ceptors iodine and phenol. We report in this article occur a t 450 mp for TMSO and DMSO. The shifted thermodynamic data for adduct formation of dimethyl I1 peak occurred between 473 and 480 mp in the diethyl sulfoxide (DMSO), tetramethylene sulfoxide (TMSO), sulfite and tetramethylene sulfone complexes. The and tetramethylene sulfone (TMS02) with the Lewis criteria for selecting the best uave length a t which t o acids iodine and phenol. For the sake of comparison, determine the enthalpies of 1 2 complexes have been similar data were obtained for the donors diethyl sulfite reported.56 The wave length 550 mp was selected and acetonitrile. These results are compared with refor the sulfoxides, which show large blue shifts. Enthported data for other nonaqueous solvents. alpies for the sulfite and sulfone were determined In order to interpret trends in the donor properties a t the complex maxima. Measurements a t 450, 4G0, of the sulfones and sulfoxides, the problem of the extent and 470 mp gave constant enthalpy values. Ethalpies of sulfur-oxygen T-bonding must be considered. This of formation for the dimethyl sulfoxide and tetrageneral problem has been reviewed recently,? b u t no methylene sulfone adducts with phenol were measured definite conclusions can be drawn because of the ama t 284 m p . The TMSO phenol enthalpy was estimated biguity in interpreting existing data. The main cause from the reported relationship between the phenol of the ambiguity arises from complications introduced AVOHand the AHof adduct formatiom3 by solvent effects. Most of the existing data t h a t bear on this problem have been obtained in water. The donor properties measured in carbon tetrachloride for TABLE I acetone, sulfoxides, sulfones, and sulfites can be raDATAA N D EQUILIBRIUM COXSTANTS FOR THE INTERACTION OF tionalized in terms of changes in T - and n-bond interS O M E DONORSC O N T A I N I K G AX s-0 GROUP\\'ITH IODINE actions. ( T = 25 OD, X = 520 mp) I t has been reported that a linear relationship exists between the enthalpy of adduct formation of phenol CAI CD, -(A with several donors and the shift in the 0-H stretching Donor mole/l mole/I A @ ) rc - S A K , I /mole vibration of phenol that results from c ~ m p l e x a t i o n . ~ Dimethyl sulioxide 0 01498 0 16,5 The d a t i obtained for the adduct of tetramethylene 0 00171 01873 200 -766 161-1 0 sulfone with phenol do not fit this relationship. This 03746 ,329 is the first exception reported t h a t we are confident 07492 ,518 Tetramethylene 00968 193 exists and an explanation for the deviation is offered. sulfoxide 0 00237 01212 237 -651 1 7 1 0 8 I t is believed t h a t the proton of phenol is simultaneously 405 0242.1 hydrogen bonded to both oxygens of T M S 0 2 . This 64 1 04847 type of interaction is found in some transition metal ion Diethyl sulfite 4983 113 0 00093 6229 complexes4where ThISO2 behaves as a bidentate ligand. ,137 -816 0 3 4 i o o 1 1 2458
Experimental Purification of Reagents.-Dimethyl sulfoxide and tetramethylene sulfoxide were purified by shaking over barium oxide for 4 days and distilling from calcium hydride under reduced pressure. Diethyl sulfite, obtained from t h e Columbia Chemical Co., was purified by distillation under reduced pressure just prior t o using. The center fraction consisting of about 5O5!;; of the material was collected for use. Fisher Spectroanalyzed carbon tetrachloride was used without further purilication. Apparatus and Procedures.-The apparatus and procedure employed in making these measurements have been described in previous articles3>jalong with t h e methods employed for calculation. The enthalpies are calculated by employing a least squares procedure and errors are reported at the 9O',C confidence level. Enthalpy results (accurate t o 1 0 . 2 kcal. mole-') are obtained h y (1) Abstracted in p a r t from t h e P h D. thesis of Robert I,. Carlson, University of Illinois, 1962 (2) C C. Prirc and S Oae, "Sulfur Bonding," T h e Ronald Press C o , S e w Y o r k , S Y , 1UG2 ( 3 ) h I . D. Jt,esten and R . S . Drago, J . A m C h r m SOL.,84, 3817 (1Yfi2); R . hl Badger and J . H B a u r r , J Chi,ii P k y s , 5, 838 (1937). (4) C H. Langford a n d P 0 . Langiord, l i i ~ i g C k e m . , 1, 184 (1962). ( 5 ) R I< Carlson and R . S. Drago, J . A m . Chern. Soc., 84, 2320 (1962).
Tetramethylene sulfone
2 4916 0 5447 0 00133
0 6809 1 3618 2 7236
,228 3.51 ,211 244 366 .io9
-553
0 73 1 0 05
Discussion Table V contains a summary of all the data pertinent to this discussion. The donor sites for many of the adducts in Table V are obvious. I t has been established that oxygen is the donor atom in amide^.^ -4 decrease in the S-0 stretching frequency in the iodine and phenol adducts with sulfoxides, compared to free sulfoxide, also indicates oxygen c o o r d i n a t i o ~ i . ~A~ ~ very slight decrease in the S-0 stretching frequency of ( A ) K S Drago, R I, Carlson. N J Rose, and D A . K'enz, ibid , 8 3 , ;3572 (1961). 17) R S Drago and D 1%' XIeek, J P h y r Chem . 65, I446 11961). and references contained therein ( 8 ) W D . Horrocks, Jr , and F. A Cotton, Spectrochim. Acta, 17, 134 (1961).
RUSSELLS.DRAGO,BRADFORD WAYLAND, A K D ROBERT L. CARLSON
3126
VOl. 85
TABLE I[
CHASGES FOR DATAASD ENTHALPY
THE
INTERACTION O F SULFOXIDES, DIETHYLSULFITE, A N D TETRAMETHYLENE SULFOSE WITH IODINE - AIi,
Donor
Dimethyl sulfoxide
Tetrarnethylene sulfoxide
C A mole/l.
Cr). mole, 1
0.00176 001 74 00 173 00172 00171 00237 00236 00235
0 0313 0310 0308 0306 0304
Tetramethylene sulfoiie
0,308 273 ,241 2 19 195 334
0179
,0177
PHENOL IYITH A SULFONE ANI) A S ~ L F O X I D E ( T = 25.11°, X = 284 m p ) c.4. CU, e(' -mole11
Dimethyl sulfoxide 0 000727
Tetramethylene sulfone 0 002996
A -
moWI
A0
0 00:3094 0 357 006188 536 76R Ol.i-170 0 00ss7s 280h 022196 517 044392 ,860
K , 1 /mole
E.%
182
1-133
706
+1
17 1
I
0 7
C.4 is the concentration of acid; CD is the concentration of donor; rl - A D is the absorbance of the complex ininus the absorbance of the acid; cc - e.4 is the difference in the molar absorptivity between the complex and acid. XI1 measurements were made in carbon tetrachloride. A - A o corrected for small absorbance due t o tetrainethylene sulfone (e11 = 0.44 a t 281 Ill,).
K , I.,'mole
Temp., ' C .
12,3/ 10.66
23 4 28 9 35.4 40.6 46.1 24.8 29 2 3j.0 39 5 45.1 23.5 27.3 32.6 36 2 38.3 44.1 25.0 32 1 39.3 45.5
9.20 8.27 7.25 14.90
13.57
- 686
550
812
470
i35
470
220 212 233 ,221 ,211 ,208 ,194 ,279 ,259 .239 223
E Q U I L I B R I L X COSSTANTS FOR THE IiYTEKACTIOS OF
Donor
550
247
.I1176 0175 1 218 1,245 1.234 1.229 1 226 1217 0.545 540 ,i:3 1 .i31
TABLE 111" DATAAND
X , mp
- e.4
- 638
308 271
0178
0023' 000986 000983 000975 000971 000969 GO0971 00133 ,00132 00131 00130
Diethyl sulfite
cc
- Ai')
-(A
11.75 10.63 9.39 0 347 ,331 . 3 14 ,297 ,293 ,2/2 ,731 ,673 62 1 573
kcal /mole
4 4 f0 . 3
4 . 4 =t0 . 3
2 . 2 f0 . 3
2.2 f0 . 3
density into the ring than a methyl group. This occurs by a conjugative mechanism in which lone pair electrons on the oxygen form a a-bond with carbon, placing electron density on the o- and $-positions of the ring. The substituent constant for a m-substituted ethoxyl group indicates it is electron withdrawing relative to a methyl group. The conjugative mechanism is not as important in this position and the withdrawal of electron density by the electronegative oxygen is more important than con jugation effects. Since diethyl sulfite is a poorer donor than dimethyl sulfoxide, the ethoxyl group must be withdrawing electron density from the sulfur. I t can be concluded that sulfuroxygen a-bonding involving the d-orbitals of sulfur and the ethoxyl group is much less effective than carbonoxygen x-bonding in the p-position of a phenyl ring. Because of poorer a-bonding the electronegativity of oxygen results in the withdrawal of electron density
TABLE I\-* INTERACTION OF PHENOL WITH DIMETHYL SULFOXIDE AND TETRAMETHYLESE S U L F O S E ( A = 284 Inp)
DATAA N D ENTHALPY CHANGES FOR Donor
Dimethyl sulfoxide
C.A m o l e 3 I1
Tetramethylene sulfone
(10727 00725 00722 007 15 001913 00 1901 001895 001889 0 0 1884
THE
CD, mole,'l.
1 006188 006169 006145 006083 02327 02516 02503 02496 02488
(
'4
-
Ao
0 536 514 480 422 40Sh 381
350 33 1 3 13
cc
- c.4
1435
706
K , l./mole
183 0 169 3 149 5 121 5 17 53 15 91 14 44 13 48 12 59
Temp., O C .
25 27 30 36 25 28 32 34 37
0 3 7 8 0 4 5 8 3
-AH, kcal./mole
6 5 f O 2
4 9 f 0 3
Footnotes the same as in Table 111
the sulfite adduct with iodine is consistent with coordination on the sulfoxyl oxygen and a weak interaction. This conclusion is uncertain, but if the acid coordinates on the "ethoxyl" oxygen the sulfoxyl oxygen would be an even poorer donor and the following discussion is still pertinent. X comparison of the donors dimethyl sulfoxide and diethyl sulfite shall be considered first. The Hammett substituent constants indicate that an ethoxyl group in the p-position of a benzene ring releases more electron
from sulfur, which decreases the donor properties of the sulfoxyl oxygen. If a-bonding were the dominant term the sulfoxyl oxygen would be a strong donor. The stronger donor ability of the sulfoxides than acetone is also consistent with less effective sulfur-oxygen a-bonding. The valence state electronegativity for tetrahedral sulfurg (3.21) is greater than that for a (9) This value represents a lower limit for it pertains t o sulfur with t w o lone pairs which is less electronegative than a similarly hybridized sulfur in the sulfoxide with only one lone pair,
DONORPROPERTIES OF SULFOXIDES, ALKYLSULFITES, AND SULFONES
Oct. 20, 1963
TABLE V
SUMMARY OF PERTINENT THERMODYNAMIC DATAFOR FORMATION OF PHENOL Donor (CHajLCO" (CHdrSO (CHa)%SO (CHd4SOz (CH30)zSO HCON(CHa)a CHCOS(CHaj9 (CiHs)2 0 (CzHdyS CHaCS C6He ( C H d rCO (CHI)~SO (CHdzSO (CHdcSO? HCON(CHd2 CH3COS(CHdz (CiHdzO (CzHdzS CHaCS C6He
Acid I2 12
12 I2 12
12 I2 1% I2
I2 I2
C6HiOH CeHaOH CeHaOH C8HsOH CeHaOH CeH6OH CeHsOH CoHaOH CeHsOH CeHsOH
THE
AND IODINE ADDUCTS
K , I./mole 0.85 (25') 14 7 i 0 . 8 (25') 11 6 i 1 0 ( 2 5 ' ) 0.73 zt 0 05 (2.5") 0 34 i 0 . 0 1 ( 2 5 0 ) 2 . 9 i 0 . 2 (250) 6 9 0 2 (25') 1 16 (20') 210 (200) 0 40 =?= 0 03 (28") 0 . 1 5 (2.30) 13.5 1 . 0 (25')
*
+
182 i 1 ( 2 5 ' ) 1 7 . 1 i 0 . 7 (25') 64 i 1 (25') 134 * ( 2 5 ' )
5 0 i 0 . 2 (25")
-AH, kcal./mole
4.4 4.4 2.2 2 2 3 7 4.0 4.2
+ 0.3 i ,3 i .3 + .3 i 3 i ,1
7.8 2.3 i
3 1.5 3.3 i . 5 7.0 6.5 i .2 4.9 i 3 6.1 i . 2 6 4 1 2 5 0 4.6 3.3 i .5 1.5
Rei. 100 10 10 10 10 11 12 13 14 10,15 16, 17 2 18 10 10 2 19
20 18 2 18
O- Due t o the formation of triiodide, indicative of a more complex reaction, the acetone-12 enthalpy could not be determined accurately.
trigonal carbon ( 2 . i 5 ) . If electronegativity were the sole effect operative, acetone would be a stronger donor than sulfoxide. The stronger donor properties of sulfoxide can be attributed to less effective sulfuroxygen s-bonding than the carbon-oxygen s-bonding in acetone. Less effective x-bonding in the sulfoxide results in more electron density residing on the sulfoxyl oxygen, and as a result it is a stronger donor than acetone. The different properties discussed above are thus attributed to less effective s-bonding between second and third row elements than between two second row elements, presumably because of less effective x overlap in the former case. I t is informative to compare the donor properties of diethyl sulfide, diethyl ether, tetramethylene sulfoxide, and tetramethylene sulfone. As can be seen from the data in Table V, diethyl sulfide is a better donor toward iodine than diethyl ether, but the reverse order of donor strengths applies toward the acid phenol. In the interaction with iodine (a very polarizable acid with no dipole moment) the ease with which the lone pair on the donor can be distorted is an important property of the donor affecting the magnitude of the interaction. Toward phenol, a polar acid, the lone pair dipole moment of the donor, in a configuration similar to that in the adduct, becomes an important parameter. Keeping this in mind it is interesting to note that both iodine and phenol coordinate on the oxygen atom of sulfoxide instead of the sulfur. Compared to dimethyl sulfide, considerable electron density has been drained off the sulfur of sulfoxide by the electronegative oxygen atom (10) This work. (11) H. S. Drago, D. A. Wenz, and R . L Carlson, J . A m Chem. Soc., 84, 1106 (1962). (12) R . S. Drago, R . L. Carlson, N . J . Rose, a n d D . A. Wenz, ibid., 89, 3572 (1961) (13) M. T a m r e s and M. Brandon, ibid., 82, 2134 (1960). (14) H. Tsubomura and R P. Lang, ibid , 89, 2085 (1961). (15) Subsequent t o our measurements values of K = 0.42 zt 0.005 and -AH = 1.9 i. 0.6 were reported b y W. B. Person, W. C . Golton, and A. I. Poyov, J A m Chem. Soc., 86, 891 (1963). (16) L J Andrews and R M . Keefer, ibid., 74, 4500 (1952). (17) R M Keefer and L. J . Andrews, i b i d , 77,2164 (195.5). (18) Enthalpies determined by t h e 0 - H stretching frequency shift method reported in ref. 3 ; A H = O.O16A,yo~ 0.64. (19) M D . Joesten a n d R. S. Drago. J A m . Chem Soc., 84, 2037 (1962). (20) R. West, reported a t t h e 140th National Meeting of t h e American Chemical Society, S t . Louis, Mo., M a r c h , 1961. (21) J. Hinze a n d H . H . Jaffe, J . A m . Chem. Soc., 84, 540 (1962).
+
3127
and by rehybridization of the sulfur atom. No detectable sulfur-coordinated complex is observed with iodine in excess sulfoxide, so the equilibrium constant for such a complex must be orders of magnitude less than for the oxygen complex. In the sulfone, the weakly basic pair of sulfur electrons present in the sulfoxide is bonded to an oxygen. If we view this as coordination of the sulfur lone pair by oxygen, the net effect would be to drain electron density off the sulfur and from the sulfur oxygen bond. An increase in the S-0 stretching frequency results. This causes a decrease in electron density on the sulfone oxygens relative to the sulfoxide oxygen and a n appreciable decrease in electron density on the sulfur in the sulfone. 4 s a result the sulfone is a poorer donor than sulfoxide. In addition to the decrease in donor properties by this redistribution of electron density in the a-system, a decrease is also expected from more effective oxygen to sulfur a-bonding in the sulfone because of the d-orbital contraction22 that occurs as electron density is removed from sulfur. The above effects are also manifested in the sulfur-oxygen stretching force constants in sulfoxides and sulfones. A value of 6.54 X lo5 dynes em.-' has been reported for the S-0 force constant in dimethyl sulfoxide* and 9.86 X lo5dynes cm.-' for sulfones.23 The donor strength of the sulfone toward the polarizable acid iodine is of the same order of magnitude as that of the weak donor benzene. Since the sulfone has a high dipole moment, the low heat indicates that the electron density in the molecule is not easily distorted in keeping with a high formal charge on sulfur. This is equivalent to saying t h a t the sulfone forms only weak covalent bonds with iodine. The S-0 stretching frequency of the coordinated sulfone in both the iodine and phenol adducts occurs a t about the same frequency as in the uncoordinated sulfone. This also indicates very little polarization of the electron density on the oxygen upon coordination. The low equilibrium constant and heat of formation of the acetonitrile-iodine adduct can be similarly rationalized. As indicated by the data in Table V, the donor properties of dimethyl sulfoxide and N,N-dimethylacetamide are very much alike. The equilibrium constants for the sulfoxide are slightly greater. There is no detectable difference in the enthalpy of formation of adducts of dimethyl sulfoxide and tetramethylene sulfoxide. The equilibrium constant for formation of the dimethyl sulfoxide adduct is slightly lower. The similarity in the coordinating abilities of these donors toward nickel(11) is also reflected in the similarity in the reported Dq-values of these materials as ligand^:^^,^^ Values of 769, 773, and 775 cm.-' are reported for DMA, DMSO, and TXISO, respectively. Since the dielectric constants of DMA and DMSO are also similar (38.9, 48.9), one would expect them to have many similarities as nonaqueous solvents. Ionization will be facilitated to a slightly greater extent in DMSO. A similar conclusion could be made regarding expected similarity in solvent behavior of acetonitrile and tetramethylene sulfone. However the possibility that the sulfone can behave as a bidentate ligand will lead to appreciable differences. The shift of the phenol 0-H stretching frequency upon adduct formation has been shown to be linearly ( 2 2 ) D . P. Craig, A. Maccoll, R . S. Nyholm, L. E . Orgel, and L E . Sutton, J . Chem. Soc., 332 (1954); D . P. Craig,and C Zauli, J . Chem P h y s . , 37, 601, 609 (1962). (23) K . Fujimori, Bull Chem. SOC.J a p a n , 91, 1374 (1959). (24) D. W. Meek, R S . Drago, and T. S. Piper, I n o r g . Chem , 1, 285 (1962). (25) R. S. Drago, D. W. Meek, M . D. Joesten, a n d L . LaRoche, ibid., 4 , 124 (1963).
3128
RUSSELLS . DRAGO.BRADFORD WAYLAND, A N D ROBERT L. CARLSON
related to the enthalpy of adduct formation for a wide variety of donors with phen01.~ The DMSO-phenol adduct obeys this relationship very well. All of the bases studied in determining the AVOHv s . A H relationship form linear or near linear hydrogen bonds, ;.e., the angle 0 in Fig. 1 is 180'. The anomalously low 0-H stretching frequency shift for the phenol-TMS02 adduct suggests t h a t the bonding or geometry of the sulfone adduct differs from the adducts used in determining this relationship. Tetramethylene sulfone has recently been reported to act as a bidentate ligand in metal ion c ~ m p l e x e s . ~Nre tentatively propose a similar model for the phenol-TAMS02 adduct which places the acidic phenol hydrogen between and equidistant from the two sulfonic oxygens (Fig. I ) . Using
Fig. 1.-Proposed
structure of the adduct between TMS02 and phenol.
this model the phenolic oxygen-sulfonic oxygen lone pair electron-electron repulsions are enhanced, but this repulsion term must be more than compensated for by the phenol dipole interacting which each of two oxygen lone pair dipoles maximizing electrostatic interaction. The hydrogen bond in the structure proposed in Fig. 1 is not linear as indicated by the angle 0, which is 180° in a linear hydrogen bond. The nonlinear hydrogen bonds suggested for the sulfonephenol adduct must be relatively ineffective in polarizing the 0-H bond when compared to a linear hydrogen bond. This may be the result of reduced lone pair-
Vol. 85
hydrogen overlap. The repulsions of nonbonding lone pair electrons on the sulfone and phenol prevent the overlap from being as great as in a linear hydrogen bond. There are unusually lzrge contributions to the heat of interaction in this adduct from the electrostatic term and this accounts for the deviation of this adduct from the AH-AVO-H correlation. The deviation of the tetramethylene sulfone adduct from the -AH us. AVO-H correlation may also be in part due to a low frequency shift resulting from the nonlinearity of the hydrogen bond. As the phenolic 0-H distance increases during the stretching vibration in a linear hydrogen-bonded adduct, overlap of the donor lone pair electrons and the proton increases, partially compensating for the decreased overlap in the phenolic 0-H bond. This results in a lower energy and lower frequency for the 0-H stretching vibration than would be obtained in the absence of this effect. In a bent hydrogen bond, the proton moves a t an angle to the donor lone pair during the 0-H stretching vibration (see Fig. 1) and lone pair proton overlap is not increased as much when the bond is stretched as in a linear hydrogen bond. As a result the energy of the vibration is decreased less and the frequency shift is smaller for a bent bond than for a linear bond. These two reasons for the deviation require that there be significant contributions to the frequency shift from covalency. Acknowledgment.-The authors wish to thank the Chemistry Branch of the Atomic Energy Commission for their generous support of this research through Contract No. AT(11-1)75S. They also wish to thank Richard L. Mddaugh for reading the manuscript and offering several helpful comments.
