erties and Electrochemical Stability of the Thio Solvents

John W. Biggle and D. Bogsanyi erties and Electrochemical Stability of the Thio Solvents. Dimethylthioformamide and Hexamethylphosphorothioic Triamide...
0 downloads 0 Views 434KB Size
John W . Biggle and D. Bogsanyi

1018

erties and Electrochemical Stability of the Thio Solvents Dimethylthioformamide and Hexamethylphosphorothioic Triamide John W. Diggle" and D. Bogsanyi The Research School of Chemistry Australian National University. Canberra. A C. T. Australfa

(Recefved October 29. 1973)

The thio solvents dimethylthioformamide (SDMF) and hexamethylphosphorothioic triamide (SHMPT) have recently been identified as two powerful solvators of soft cations (each being a stronger soft cation solvator than the corresponding oxygen analog). For this reason the physical properties useful in correlating ion-solvent behavior have been determined for each solvent. The electrochemical stability of SDMF and SHMPT has also been determined and found to be inadequate anodically presumably due to the oxidation of the sulfur moiety. The available voltage range for SDMF and SHMPT containing 0.1 M tetraethylammonium perchlorate is +0.15 to -2.5 V and f 0 . 4 to -1.3 V, respectively, us. the Pleskov electrode in acetonitrile as reference. The high dielectric constants of 47.5 and 39.5 for SDMF and SHMPT, respectively, suggest that they would make good solvents for strong electrolyte solutions, and the dipole moments, being 4.37 and 4.83 D for SDMF and SHMPT, respectively, indicate these solvents to be highly polarizable. From the dipole moment it is also concluded that both HMPT and SHMPT are slightly associated in the pure form, whereas DMF and SDMF are not associated. In contrast to HMPT, sodium or potassium metal when added to SHMPT does not liberate the blue coloration supposedly characteristic of the solvated electron.

Introduction Dimethylthioformamide (SDMF) and hexamethylphosphorothioic triamide (SHMPT) and their respective oxygen analogs DMF and HMPT permit a convenient comparison of ion-solvent interactions in terms of the Pearson1 hard-soft classification. For example,2,3 SDMF and SHMPT, both being classified as soft solvents, interact strongly with soft cations (Ag+, C u f ) resulting in these thio solvents being stronger solvators for soft cations than the oxygen analogs. Conversely, hard cations (Li+, Naf and K f ) interact much more weakly with the soft solvents SDMF and SFIMPT than they do with the harder oxygen analogs DMF and HMPT, thus the oxygen analogs are stronger solvators of the hard cations. Since data on the physical properties of these two thio solvents appeared sparse, these data were generated in the present work. It was also considered to be of interest to ascertain the electrochemical stability of SDMF and SHMPT, this was done using tetraethylammonium perchlorate and sodium tetrafluoroborate as solutes, Experimental Section Dimethylthioformamide was prepared from dimethylformamide and phosphorus pentasulfide, according to the method of Wilistatter and Wirth,4 and then twice fractionally distilled after drying with activated 4A molecular sieves. Hexamethylphosphorothioic triamide was prepared by the reaction between phosphorus trichloride and dimethylamine in an ether medium, according to Burg and Slota,s and then allowing the product (Me2N)sP to react with sulfur according to Vetter and Noeth.6 The product SHMPT was finally twice fractionally distilled after drying with activated 4A molecular sieves. SDMF and SHMPT prepared in the manner described above contained 45 and 19 ppm of HzO, respectively, as determined by Marl Fischer titration. Boiling point, melting point, and density (weighing botThe Journal ofPhysical Chemistry, Vol. 78, No. 10, 7974

tle technique) were determined in the conventional manner.7 Refractive indices (nD) were determined using an Abb6 60 refractometer; specific conductivity was determined using a Type B221 Wayne Kerr bridge and conductivity cell of cell constant 1.93 em-'; surface tension was determined using a Cambridge Du Nuoy tensiometer and viscosity was determined by employing a No. 6 Viscometer tube for SHMPT and a Kimax A53 Size 50 viscometer tube for SDMF. The dielectric constants ( e ) of SDMF and SHMPT were determined using a capacitance meter and probe,8 described in ref 9 in conjunction with a capacitance cell comprising of a 22 gauge platinum wire 25 mm in length axially disposed in a copper cylinder of 5 mm diameter and 25 mm length. The frequency of the capacitance measurement (2 MHz) is sufficiently high to permit accurate measurement, in the picofarad range, without significant error being introduced by dc conduction in the capacitance cell. This capacitance meter, cell, and probe was calibrated with nine solvents of known dielectric constant (ranging from 7.5 to 80) and a direct capacitance reading (in pF) us. dielectric constant calibration graph constructed. The dc resistance of all solvents studied in this work was >IO0 Kilohms when in the capacitance cell. The dipole moment ( p ) of SDMF and SHMPT was obtained by the following methods: (i) Onsager calculationlo using the physical properties as measured in this work; (ii) the Halvesstadt and Kumler methodll in which the values of t, nD, and density are determined for a series of solutions of the thio solvent in both benzene and carbon tetrachloride. The electrochemical stability of SDMF and SHMPT containing either 0.1 M tetraethylammonium perchlorate or 0.1 M NaBF4 (saturated solutions containing >0.1 M were used in some cases) was determined using a conventional three-compartment double H cell, each compartment being separated from the other by a porosity 4 sin-

Physical Properties OF Thio Solvents

1019

roperties of SDMF, SHMPT, DMF, and HMPTa Solvent

Prop er t Y

Bp, "(2 :cwHp, 042

Density, d, g/em3 1% D

SDMF

DMFb

67-70 (1mm) -8.5 1.0237 (27') 1.5741 (27')

153.0 (760 mm) -60.4 0.9440 1.4282

Specific co:nductivjty, M&os/cm

1.9

x

94 (1.5 mm)c 29.0 I . 043 1.5070 3.1

6 X

10-7

Hnwr

SHMPT

x

10-7

233 (760 mm) 7.2 1.0270 (200) 1.4570

1.9 X IO--' (23°)d

Surface tension, dyn/em ‘Viscosity, C P B

Onsager calcdf in stated solvent jBasicity, SbCljl :Basicity, CHC13i AGt,, Na-" Basicity,o A(%,, .f%gMolar polarizability,p em

Molecular polarizability,Q a%3

45.4 1.98 47.5 4.44 4.37 (CC14)Q 37.33

28.7 5.55 39.5 6.25 4.83 (CCla)"

35.2 0.802 36.7 4.07 3.86 (CsHe) 26. iP 0.78" -4.5 -5.2

111.7 -22.6 75.4

7.91

14.7

It " 83" -5.5

0.90" 45.9 - 18.8

71.3

33.8 (200) 3.47 ( Z O O ) 28 " 3'

1'77.7

158.a)

22.5

18.8

For SHMPT all measurements were a t 30"; all other measurements were a t 25O except where stated otherwise. Reference 25. This value agrees quite closely with that reported by Keat and ShawZg but does not agree with that value reported by Vetter and Noeth.8 Reference 26. e Reference 20. f Reference z.0. ,See text for discussion of these values. I? Reference 21. Basicity expressed as the negative of the enthalpy of interaction bet,ween SbCls and the stated solvent in high dilution in 1.,2-dichloroethane.'6 Reference 2. IC Reference 16. Basicity expressed as nmr chemical shifts of CHClr at infinite dilution in the stated solvent.2iiz8 Reference 27. Reference 28. O Basicity expressed as the free energy of transfer (kcal/mol) of either N a + or Ax+ from water to the stated solvent taken from ref 2 and 3 . P [ ( e - l ) / ( r 2 ) 1 ( M / d ) . [ ( n -~ l )~/ ( n d f 2)1[3/4rN1. Q

+

*

ter. The working electrode12 (a platinum disk of area 25 111111~)was situated in the center compartment, and the counter electrode ( a large surface area platinum spiral) in the right Compartment. Into the left compartment which contained the working electrolyte was dipped the reference electrode of A g ( 1 P 2M AgC104, 10-1 M TEAP, acetonitrile, contained within its own compartment and separated from the thio solvent in the left compartment by ;in additional porosity 4 sinter. The voltage range of each solvent containing Et4NC184 or NaBF4 was obtained using a Princeton Applied Research Model 170 electrochemistry system in the cyclic voltammetry mode at 100 mV/sec and employing iR compensation. The voltage range is recorded as those anodic and cathodic voltages at which the total current exceeds 50 PA. The measured physical properties of SDMF and SHMIPT along with solvent basicity values and molar and molecular polarizabilities are shown in Table I. The elecIrochemica.1 data from SDMF and SHMPT are shown in Table 11. In both tables the respective data for DMF and I-JMP'T are S ~ G V U Ifor the purposes of comparison.

Discussion The electrochemical stability of the thio solvents to oxitiation is expectedly inferior to that of the oxygen analogs, 1l-m being presumably due to the oxidation of the sulfur moiety and thus being independent of the anion present. The electrochemical stability toward reduction is dependent upon the cation, and essentially identical for each pair of thio and oxygen solvents. The liberation of the blue coloration, supposedly indicative of solvated e1ectrons,l3 was observed in HMPT + I\JaBFs at high cathodic potentials but was not observed for SHMPT + NaBF4. Both sodium and potassium metal were essentially insoluble in SHMPT with no blue coloration being observed (this is in contrast to the well-known moderately stable blue solutions of Na or K in I.I[R/IPT14J5). If the electron may be classified as a hard

TABLE 11: Electrochemical Voltage Range of DMF, HMPT, SDMF, and SHMPT Containing E i t h e r Et4NC104or NaBF4 of 0.1 M Concentration Except Where Otherwise Noteda Solute Solvent

DMF SDMF HMPT SHNIPT

0.1 M EtaNClOa,

0.1 M NaBFa,

V

V

+ 1 . 2 to -2.7 $0.15 to - 2 . 5 1-0.6to -1.3 4 0 . 4 to - ? ~ . 3 ~

4-1.2 to +0.2 to 1 0 . 5 to 1 0 . 5 to

-3.4 -3,5b, -2.5cld -3.1b

The reference electrode in each case being A g ' l 0 - 2 M AgC101, 0.1 M Et1NC101, acetonitrile. Saturated solution of the solute in this solvent was less than 0.1 M. Appropriate iR compensation in the experimentation was however still possible. cThis cathodic limit is equivalent to -2.2 V us. aqueous sce, the electrode Ag110-2 M AgClOa, C H G N being +0.300 V us. aqueous sce. This value of -2.2 V does however include junction potentials due to HMPTICHCN and CHICN /HzO junctions; however, these are considered to introduce only a 1 5 0 mV err0r.30 This cathodic limit is suspected to be low due to low level water contamination which, for HMPT, is known31 to produce a positive potential shift in the cathodic limit. Previously reported values*'-38 range from -3.0 to -3.2 V (aqueous sce)

.

acid then greater stabilization, i.e., solvation, will occur with the harder of the two solvents, i . e . , HMPT.2J As mentioned in the Introduction, SDMF and SHMPT are strong solvators, i . e . , strong bases, of soft cations, e.g., Ag+, while they are weak bases in their interactions with hard cations, e . g . , Na+ (see Table I). On these principles the greater donor number for SDMF on the Gutman scale of basicities16 shows SDMF to be a stronger base than DMF; therefore, SbC15:SDMF adduct formation must typify a soft-soft interaction. The basicities expressed as CHC13 chemical shifts would seem to represent, a t least in the case of SHMPT and HMPT, a hard-hard interaction since SHMPT has the lower shift of the two solvents. The thio solvent molecules are more covalent and more polarizable than their oxygen ana10gs.I~The dipole moThe Journal of Physical Chemistry, Vol. 78, No. 10, 1974

John W. Diggle and D. Bogsanyi

102

ments of both thio solvents are higher than their respective oxygen analogs (see Table I), the value for SDMF of 4.37 D being in good agreement with the previous literaturels value of 4.54 D obtained in CC14 solvent. The dipole moment results in Table I would seem to indicate that the greater polarizability of the thio solvent molecule more than compensates for the decreased ionicity of the P=S bond; thus, the dipole moments of thio solvents are greater than their respective oxygen analogs. Dipole moment results in benzene, at weight fractions of thio solvent up to 0.20, were substantially lower than those reported in Table I using cc14 as solvent, i.e., in benzene, dipole moments of 3.91 and 3.37 D were obtained for SHMPT and SDMF, respectively. This would seem to be in accord with the interactions previously observed by Keat and Shawl9 between HMPT and benzene; interactions between benzene and the solute being studied are known to produce a significant lowering of the dipole moment.:Ls,20 The choice of p = 4.48 D for HMPT requires some comment. The five literature values reported for the dipole moment of H:R/IP'I', being obtained by various methods in benzene solution, are 4.31,21 4.48,21 5.54,22 4.30,23 and 4.31.24It is to be noted t,hat these values, with the exception of 5.54 where no experimental procedures were stated, were obtained a t weight fractions of HMPT in benzene much less than 0.020. Since the majority of results seem to favor a dipole moment of HMPT around 4.3-4.4, the most recent value21 of 4.48 D is accepted and taken in Table 1. This value for HMPT is significantly lower than the calculated Onsager valuelo of 5.38 D, i.e., the Kirkwood correlation factorxo g > 1; this could perhaps indicate that HMPT is somewhat associated in the pure state, a conclusion in agreement reported by Gal and Moliton.. ouchetout.21 A similar conclusion may be made concerning SHMPT, whereas SDMF is apparently not associated since the Onsager value of 4.44 D is close to the measured vahie of4.37 D. The high dielect.ric constants of these two thio solvents, in conjunction with their high polarizabilities, should make SDMP ,and SHMPT excellent solvents for electrolyte studies and as a media for inorganic and organic reactions.

Acknouiledgments. The authors gratefully acknowledge the cooperation extended to them by Dr. J. S. Dryden of the National Standards Laboratory, Sydney, for the accurate measurements of c required in the Halverstadt and

The Journalof Physical Chemistry, Val. 78, No. 10, 1974

Kumler method for determining the dipole moments reported in this work. The authors are indebted to Dr. A. J. Parker for his encouragement and discussion during the course of this work. References and Nates R. G. Pearson, Surv. Progr. Chem., 5, 1 (1969) W. E. Waghorne, Ph.D. Dissertation, Australian National University, 1972. R. Alexander, D . A. Owensby, A. J. Parker, and W. E. Waghorne, Aust. J. Chem., in press. R. Willstatter and T. Wirth, Berichte, 42, 1908 (1909). A. B. Burg and P. J. Slota, Jr., J. Amer. Chem. SOC., 80, 1107 (1958). H. J. Vetter and H. Noeth, Berichte, 96, 1308 (1963). A . Finlay, "Practical Physical Chemistry," Longmans, Green and Co., London, 1955. The capacitance meter and probe used in this work differs from that described by Pikg in that the capacitance range was expanded from 50 to 150 pF. The circuit diagram as given by Pikg is incorrect in that diode OA 91 should be located at the IOOK resistor end to the 100-pF capacitor, and not after the 100-pF capacitor as shown. P. Pik, Electron. Aust.. 42 (March 1971). C. P. Smyth, "Dielectric Behaviour and Structure," McGraw-Hill, New York, N. Y . , 1955, p 226. I. F. Halverstadt and W. 'D. Kumler, J. Amef. Chem. Soc., 64, 2988 (1942). The platinum working electrode surface was pretreated by etching in hot concentrated nitric acid for 2 min followed by a 5-min immersion in a saturated aqueous solution of ferrous sulfate. The electrode was then washed in water and acetone, dried, and then finally rinsed in the electrolyte to be studied. (a) J. M. Brooks and R. R. Dewald, J. Phys. Chem.. 72, 2655 (1968); (b) R. Catterall, L. P. Stodulski, and M. C. R. Symons, J. Chem. SOC.A, 437 (1968). T. Cuvigny, J. Normant, and H. Normant, C. R. Acad. Sci.. Ser. C, 3503 (1964). H. Normant, T. Cuvigny, J. Normant, and B. Angelo, Bull. Soc. Chim. Fr., 1561 (1965). V. Gutmann, Chem. Brit.. 7, 102 (1971). W . Waiter and J. Voss, "The Chemistry of the Amides," J. Zabicky, Ed., Interscience, New York, N. Y.. 1970, Chapter 8, p 385. W. Walter and H. Huhnerfuss, J. Mol. Struct . 4,435 (1969). R. Keat and R. A. Shaw, J. Chem. SOC.A.. 703 (1968) I. Suzuki, Speclrochim. Acta. 16, 471 (1960). J.-Y. Gal and C. Moliton-Bouchetout, Euli. SOC. Chim. f r . . 464 (1973). J. E. Dubois and H. Viellard, J. Chim. Phys., 82, 699 (1965). J. P. Fayet, C. R. Acad. Sci.. Ser. C. 270, 9 (1970) Sr. M. Schafer and C. Curran, Inorg. Chem., 4, 623 (1965) J. A . Riddick and W. B. Burger, "Techniques of Chemistry." Vol. II, "Organic Solvents," 3rd ed, Wiley, New York, N. Y . . 1970. D, A. Owensby, A. J, Parker, and J. W. Diggie. J. Amer. Chem. SOC..in press. J , J. Delpuech, Tetrahedron Lett.. 2111 (1965). G. Martin and A. Besnard, C. R. Acad. Sci., Ser. C, 257, 2463 (1963). R. Keat and R A. Shaw, J. Chem. SOC.A . 4E02 i1965). J . W. Diggle and A . J. Parker, Aust. J. Chem . submitted for publication. J. E. Dubois, P. C. Lacaze, and A . M de Ficqueimont, C. R. Acad. Sci.. Ser. C, 262, 249 (1966). J. E. Dubois, P. C. Lacaze, and A. M. de Ficquelmont, 6. R. Acad. Sci.. Ser. C, 262, 181 (1966). Y. Kanzaki and S. Aoyague, J . Electroana/. Chem. 36,297 (1972).