J. Chem. Sac., Faraday Trans, I , 1982, 78, 1199-1207 Infrared Spectroscopic Study of the Effects of Different Cations on NN-Dimethylacetamide and Fully Deuterated NN-Dime t hylformamide BY W. EARLE WAGHORNE* AND HECTOR RUBALCAVA Department of Chemistry, University College Dublin, Dublin 4, Ireland Received 27th May, 1981 The infrared spectra of NN-dimethylacetamide and fully deuterated NN-dimethylformamide complexed to a variety of metal cations in propan-1-01 have been measured. The OC-N stretching and 0-C-N bending frequencies were found to vary systematically with the electrostatic potential at the surface of the cation. The 0-CN stretching frequency was primarily determined by the electronic structure of the complexed cation. The data indicate that complexation takes place via the carbonyl oxygen for all of the complexes studied, and support the idea of electrostatic stabilization of negative charge at the carbonyl oxygen by cations.The data also indicate that back bonding may take place between transition metal cations and complexed amides. In a recent p.m.r. study1 it was found that the barrier to rotation about the OC-N bond of NN-dimethylacetamide (DMA) complexed to different cations increased systematically with the charge and radius of the complexed cation. The silver ion, which lowered this rotational barrier, was exceptional among the cations studied. It has also been found that the lH and 13C chemical shifts of complexed amides2v vary systematically with the electrostatic potential arising from the complexed cations.Surprisingly, while amide complexes have been the subject of several infrared spectroscopic studie~,~-lO no such correlation with the nature of the complexed cation has been reported for the observed changes in any of the amide bands. Thus it was decided to carry out an infrared spectroscopic study of simple amides complexed to a wide range of metal cations. To avoid complications arising from the participation of water in the cation-amide interactions1l7l2 it was decided to carry out the work in non-aqueous media, and propan-1-01 (PrOH) was chosen as solvent. DMA and fully deuterated NN- dimethylformamide ([2H,]DMF) were chosen as amides for the study. EXPERIMENTAL Propan- 1-01 was dried over anhydrous CaSO, and fractionally distilled. DMA was purified by fractional distillation under reduced pressure as described previously;13 CCl, (Merck Spectroscopic grade) and [,H,]DMF (Fluorochem 99%) were used without further purification. The spectra of CCl, and propan-1-01 showed no significant impurity peaks. The perchlorates of barium, calcium, cobalt, nickel and strontium were prepared by reacting the carbonates with perchloric acid.The other electrolytes were analytical grade or better, except Zn(C10,), (Alpha, 98.9%) and Pb(ClO,), (Alpha, 95 %) and were not further purified. All of the electrolytes were dried at 60 OC under reduced pressure for 24 h and stored over P,O,. Subsequent analysis' indicated that AgClO,, LiClO, and MgCl, were anhydrous, Mg(C10,), existed as the monohydrate and Cd(C10,),, Cu(ClO,),, Pb(C10,), existed as the 11991200 INFRARED S T U D Y OF AMIDE COMPLEXES hexahydrates.In making up solutions it was assumed that the other electrolytes existed as the stable hydrates :I4 Al(ClO,), 6H,O, Ba(C10,), 3H,O, Ca(ClO,), 2H,O, Co(ClO,), 6H,O, Ni(C10,), 6H,O and that Sr(ClO,), was anhydrous. The molecular sieves (B.D.H., 4A) used to dry the electrolyte solutions were exposed to 1 mol d n r 3 solutions of the appropriate electrolyte until no further cationic exchange was detected, washed with distilled water and activated at 250 "C under reduced pressure for 12 h. Infrared spectra were measured using a Perkin-Elmer 125 spectrometer and matched liquid sample cells having a nominal path length of 0.1 mm and AgCl windows. The spectra were recorded from 4000 to 400 cm-', and the range from 2000 to 800 cm-l was recorded in an expanded mode.The experimental precision, for a sharp absorption band, was 2 cm-l for the former and 1 cm-l in the expanded range. Peak positions were measured relative to those of a sample of polystyrene. Solution spectra were measured using either the solvent or the appropriate M(C10,), solution as reference. The latter was preferred for the electrolyte solutions as it reduced the interference caused by the ClO; absorption bands, particularly in the region 1000 to 1150 ern-'. RESULTS The possibility of effects arising from the formation of amide dimers in solution7* was investigated by recording the spectra of DMA solutions in PrOH for the concentration range 0.1-1 .O mol dm-3. No change was observed in the frequencies of any of the amide absorbance bands, indicating that dimerization of the amides should not affect the results of the present study.The spectra of DMA solutions (0.5 mol dm-3) containing varying amounts of water (to 2 mol dm-3) were measured. The only significant change in the DMA spectrum was a broadening of the absorption band at 1635 cm-l. This could be removed by using PrOH containing the same concentration of water in the reference cell, indicating that the observed broadening resulted from a superposition of the water and amide bands and that the presence of water did not affect the amide bands in PrOH. The spectra of DMA and [,H,]DMF solutions containing AgClO,, LiClO,, Cd(ClO,),, Cu(ClO,),, Mg(ClO,),, Pb(ClO,), and Zn(ClO,), were recorded after they had been dried with the appropriately treated molecular sieves and after the addition of known amounts of water (to 1 mol dm-3).The frequencies of all the amide bands remained unchanged. Therefore it was concluded that the presence of small amounts of water in the solutions did not affect the measurements. The only differences between the spectra of DMA and [,H,]DMF solutions exposed to the various molecular sieves and those of the undried solutions was the presence of the water bands in the latter. Thus it is clear that, over the required drying time, neither the amides nor the PrOH were significantly decomposed by the molecular sieves. The spectra of DMA solutions (0.5 mol dm-3 DMA) containing varying concentra- tions of AgClO, and LiClO, (to 2 mol dm-3), Mg(ClO,), (to 1 mol dm-3), Pb(ClO,), and Zn(C10,), (to 0.5 mol dm-3) were recorded.The bands of the uncomplexed amide became progressively weaker and those of the complexed amide more intense with increasing electrolyte concentration. The positions of the bands associated with the complexed amide, where they were sufficiently removed from those of the uncomplexed amide to be clearly resolved, were independent of the electrolyte concentration (see fig. 1). In all cases the positions of the bands of the complexed amides were determinable for solutions having a cation to amide ratio of 1 to 1 [i.e. 0.5 mol dm-3 M(ClO,),, see fig. 11. The spectra of 0.5 mol dm-3 [2H7]DMF solutions containing 0.5 mol dm-3 Mg(C10,), and MgCl, differed only by the presence of the Cloy bands in the former.This indicated that the anion was not influencing the positions of the absorption bands of the complexed amide. Since the frequencies of the amide bands were not affected by changes in theW. E. WAGHORNE AND H. RUBALCAVA 1201 1700 1600 1700 1600 v/cm-' FIG. 1.-Variation of vco of DMA (0.5 mol dm-3) in the presence of varying LiC10, (I) and Pb(ClO,), (11): DMA ratios. I, [LiClO,]:[DMA] = (a) 0; (b) 0.5; (c) 1.0; ( d ) 2.0: 11, [Pb(ClO,),]:[DMA] = (a) 0; (b) 0.125; (c) 0.25; ( d ) 0.50; (e) 1.00. TABLE IN INFRARED FREQUENCIES OF DMA AND [2H,]DMF DMA [2H,]DMF neatb in CCldC in PrOHC neatb in CClaC in PrOHC 30 1 5dm 2925ds 1746m 1650%~ 1545w 1 500dm 1410s 1390s 1350 1260s 1 180s 1055s 1030w 1008w 953vw 730dw 585ds 469dm - 2930d 1742 1 660d 1546 1 494d 1410 1392 1352 1264 1184 1055 1030 1008 - 584d 463d - 1635d 1555 1 500d 1414 1400 1260 1185 - - 735d 590d 470d 22 1 Odm 2 1 30dm 2 1 05dm 2065dm 1 700ds 1 650evs 1390vs 1260s 1123s 1070m 1050m 1035m 918m 890s 835w 765dw 615ds 22 1 Od 2130d 2 1 OSd 2065d 1 69lId 1665,d 1644sh 1380, 1388sh 1264 1123 1070d 1050d 1032d 916 890 835 612d - 22 1 Od 2150d 2 1 05d 2070d 1 700d 1644d 1400 1260 1123 - 892 - 620d a Units are cm-'; uncertainties f 1 cm-l unless otherwise stated; w, m, s, vs, sh indicate weak, medium, strong, very strong and shoulder, respectively.* As their films between KBr discs. As 0.5 mol dm-3 solutions, 0.1 mm path length. f 2 cm-l. 5 cm-l.1202 INFRARED STUDY OF AMIDE COMPLEXES concentrations of the amide or electrolyte, by the presence of water nor by the nature of the anion present, it was concluded that the observed changes resulted from direct cation to amide interactions. Table 1 lists spectral data for DMA and [2H,]DMF as thin films formed between KBr discs and as 0.5 mol dm-3 solutions in CCl, and PrOH.Several of the amide bands were not observable in the solution spectra because of dilution or interference by the solvent bands. Tables 2 and 3 list the frequencies of the observable bands, below 2000cm-1, in the spectra of DMA and C2H,]DMF (0.5 moldm-3) in solutions containing 0.5 mol dm-3 of the different electrolytes. Several of the amide bands were obscured by those of CIO,. TABLE 2.-INFRARED FREQUENCIES OF DMA IN Mn+-DMA COMPLEXES IN PrOHa complexed cation ~~ Li+ 1 640b 1 50Sb 1418 1400 1264 596b Ag+ Mg2+ 1630b 1520 1418 1404 1260 cu2+ 1 604b 1511 1420 1402 1259 592b Zn2+ 1610b 1514b 1420 1402 1260 595b Cd2+ 1610b 1512b 1419 1401 1260 595b Pb2+ 1 596b 1508 1418 1401 1259 590b 1 608b 1 50gb 1414 1498 1260 - - a Measured for solutions containing 0.5 mol dm-3 DMA and 0.5 mol dm-3 M(ClO,),; units are cm-l; uncertainties f 1 cm-l unless otherwise noted.Values are f 2 cm-l. TABLE ~.-INFRAFED FREQUENCIES OF [2H,]DMF IN Mn+-[2H,]DMF COMPLEXES IN PrOHa complexed cation Li+ Ag+ Mg2+ Ca2+ Sr2+ Ba2+ co2+ Ni2+ CU2+ Zn2+ Cd2+ Pb2+ ~ 1 3 + 1642,b 1648sh 1 632b 1 6Ub 1 640b 1 640b 1 639,b 1630sh 1633b 1 632b 1 630b 1 632b 1 627b 1618,b 1612sh 1644,b 1650sh 1412 1404 1421 1418 1412 1408 1414 1415 1412 1414 1412 1409 1420 1260 1260 1251 1249 1250 1260 1250 1250 1245 1250 1248 1243 1240 893 894 900 898 897 896 900 900 902 900 900 900 904 630b 622b 64Sb 635b 628b 620b 645b 650b 666b 645b 640b 632b 685b a Measured for solutions containing 0.5 mol dm-3 [2H,]DMF and 0.5 mol dm-3 M(ClO,),; Values are units are cm-l; uncertainties f 2 cm-l.1 cm-l unless otherwise noted; sh, shoulder. DISCUSSION The simplest 'zero-order' description of the vibrations of an amide molecule, using non-interacting valence force coordinates, predicts several localized vibrations, including a carbonyl stretching, an OC-N stretching and an 0-C-N bending among others. Clearly a more accurate description would involve mixing of theseW. E. WAGHORNE A N D H. RUBALCAVA 1203 deformations to give the more realistic normal vibrational modes.Nevertheless, the simple zero-order description is sufficient for our present purposes, since only the major contributions to three of the amide absorption bands are required. Thus a complete vibrational assignment of the spectra of DMA and [2H7]DMF was not required nor was one attempted. However, it is necessary that the major contributions to the bands of interest are known before the effects of complexation can be understood. The strong bands at 1654 and 1635 cm-l in the spectra of [2H7]DMF and DMA, respectively, have been assigned to the amide carbonyl stretching vibration ( ~ ~ ~ ) , ~ - ~ 9 15* l6 although some participation by other deformations, most importantly the OC-N stretching vibration (vCN), has been suggested in the case of [2H7]DMF.15 The strong band at 1400 cm-l in the [2H,]DMF spectrum and the weaker one at 1500 cm-l in the DMA spectrum have been assigned to the OC-N stretching vibration (vCN).,-'9 1 5 9 l6 The moderately strong band at 620cm-l in the [2H7]DMF spectrum has been assigned to a vibration which is predominantly the 0-C-N bending vibration (docN) with some contribution from the rocking and stretching vibrations of the -N(C2HJ2 group.No corresponding band in the spectrum of DMA was observed. It is clear from the data in tables 2 and 3 that complexation affects the positions of the 1645 cm-l [2H7]DMF and 1635 cm-l DMA bands similarly, as it does the 1400 cm-l [2H7]DMF and 1500 cm-l DMA bands. Thus the assignment of these pairs of bands to similar vibrational modes is reasonable. None of the observed DMA bands showed variations similar to that of the 620 band of [2H,]DMF.It is necessary to establish, as far as possible, the predominant site of the cation to amide interactions before the effects of this interaction are discussed. The oxygen and nitrogen atoms of the amide molecule are the most likely sites because of their relatively high electronegativities. However, considerations of the amide structure in terms of simple resonance theory indicate that the oxygen should be the preferred site. Thus the bonding in an amide molecule can be described as a hybrid of structures I and II,17a \ / R R C=N+ R C-N R I \ R -0 I I1 each making a significant contribution. Clearly the resulting negative charge at the oxygen and positive charge at the nitrogen will strongly favour interactions at the oxygen atom.The coulombic effect of complexation at the oxygen should be to increase the relative importance of structure I1 to the description of the amide bonding. That is, the 0-CN bond order should decrease and that of the OC-N bond increase. Conversely, complexation at nitrogen should greatly decrease the relative importance of 11, thus reducing the OC-N bond order and increasing that of the 0-CN bond, relative to the unperturbed molecule. It is clear that, in the absence of complexation, the bonding in the amide molecule will be influenced by the amide-solvent interactions. For example, interactions with the hydroxyl proton of PrOH should have effects similar to those of complexation to a cation. However, in all cases, the vc0 frequencies of the complexed amides are lower and the vCN frequencies higher than those of the neat amides or of the amides dissolved in CCl,.This allows one to tentatively conclude that complexation occurs via the amide oxygen. / \ 01204 INFRARED STUDY OF AMIDE COMPLEXES The question of the site of complexation can be approached in a second independent way. It has been shown' that complexation of DMA to Na+, Li+, Mg2+, Cd2+, Pb2+ and Zn2+ raises the activation energy for rotation around the OC-N amide bond, indicating an increase in the OC-N bond order. This unambiguously places the interaction at the carbonyl oxygen for these complexes. It is clear from the preceding discussion that the infrared spectra of oxygen-complexed amides should differ markedly from those of nitrogen-complexed amides.However, there is a general consistency among the spectra of the amides complexed to the closed-shell cations, including Li+ and Mg2+, and among those of amides complexed to the transition-metal cations, including Cd2+ and Zn2+ (cf. below). Thus we can conclude that each of the cations is complexed via the carbonyl oxygen of the amides. This agrees with the conclusion of other w o r k e r ~ ~ ' ~ ~ that the amide oxygen is the predominant site of complexation to cations. 0 5.0 10.0 9/10 J C-' FIG. 2.-Variation in the frequency vCN for complexed [2H,]DMF with the electrostatic potential y at the surface of the complexed cation ; solid line, transition-metal cations ; broken line, closed-shell cations. Fig. 2 shows the variation in the frequency of the vCN band of [2H7]DMF with changing electrostatic potential, ry, at the surface of the complexed cation.The electrostatic potential was calculated via 2 1 ry=-X-- rc ~xE,, where 2 is the ionic charge, rc is the ionic radius 17b and E, is the permittivity of free space (E, = 8.85 x 10-l2 F m-l). As noted above the variation of vCN for DMA is similar to that for [2H,]DMF (cf. tables 2 and 3). Since ry is a purely electrostatic parameter of the complexed cation the increase in vCN with ry is in agreement with the simple electrostatic argument. However, the data in fig. 2 suggest that the nature of the complexed cation is also important. There is a linear variation (solid line) of vcN for [2H7]DMF complexed to those cations which have d electrons in their outer shell and a separate variation (broken line) for those which do not.At low values of ry this latter variation is much more pronounced than the former. It is interesting to compare the variations in vCN of the complexed amides with thoseW. E. WAGHORNE AND H. RUBALCAVA 1205 in the activation energy, AE,, for rotation around the OC-N bond of DMA complexed to different cati0ns.l The AE, values for DMA complexed to Li+, Cd2+, Mg2+, Pb2+ and Zn2+ were also found to increase with increasing ly,l and there is a good linear correlation between the vCN and AE, values for these DMA complexes. Since the AE, values refer unambiguously to the OC-N amide bond this correlation strongly supports the assignment of the vCN band. 680 660 I 5 51 --- z b o 640 r - - - 1 ($g%:; , , ., , , , 62 0 P B O 0 5.0 10.0 $/lo J C-' FIG. 3.-Variation in the frequency of 6,,, for complexed [2H,]DMF with the electrostatic potential y at the surface of the complexed cation. The AE, value of DMA complexed to Ag+ is unusual, in that it is less than that of the uncomplexed amide.lV 1 9 9 2o However, the vCN values of the amides complexed to Ag+ are not anomalous, which indicates that the predominant Ag+ complex involves the expected increase in the OC-N bond order. Thus the lower AE, value observed for DMA complexed to Ag+ does not result from an anomaly in the bonding of the predominant species in solution. This strongly supports an earlier suggestion19* 2o that the observed lowering of AE, by Ag+ results from the presence of a small but kinetically significant proportion of the amide molecules being complexed to Ag+ via the nitrogen atom.This is similar to the accepted mechanism for the lowering of AE, by the protonation of amides21 Fig. 3 shows the variation in the frequency of the dOCN band of [2H,]DMF with ly. There is a relatively good linear correlation for these data. However, the values for the Ba2+, Li+ and Sr2+ complexes lie below the line, which suggests that there could be separate correlations for complexes of the cations having available d electrons and those without, as in the case of the vCN data. The value of 666 cm-l for the Cu2+ complex is a striking anomaly which we cannot presently explain. Fig. 4 contains the vco data for complexed [2H,]DMF as a function of ly.There is no apparent correlation between these spectral data and ly. Rather, the frequency of the vco band appears to be related to the electronic structure of the complexed cation. Thus vco is 1642 _+ 4 cm-l for complexes with closed-shell cations, 1630 _+ 4 cm for complexes with the transition-metal cations and 1618 cm-l for the Pb2+ complex. The fact that the vco frequencies of the amides complexed to the transition-metal cations and Pb2+ are lower than those of the amides complexed to the closed-shell1206 INFRARED STUDY OF AMIDE COMPLEXES cations is particularly interesting. In each case the vco frequency is relatively independent of ly, indicating that interactions other than simple coulombic ones are involved. The lowering of vco by Pb2+ and the transition-metal cations is consistent with a degree of covalent bonding involving the metal d electrons and the amide carbonyl 1640 - I 1630- .ou - T T I 0 Cd 1621 OPb I T 1 OAl 1 1 0 5.0 10.0 $/lo J C-' FIG. 4.-Vanation in the frequency of vco for complexed [2H,]DMF with the electrostatic potential y at the surface of the complexed cation. n antibonding orbitals (back bonding).22 Such an interaction would weaken the amide 0-CN bond and result in the observed lowering of the vco frequency. This interaction could also give rise to the observed difference between the dependences of vCN on ly for the transition-metal and the closed-shell cations. We thank Mr G. Flynn for making the infrared measurements. W. E. Waghorne, A. J. I. Ward, T. G. Clune and B. G.Cox, J. Chem. SOC., Faraday Trans. I , 1980, 76, 1131. A. Fratiello, D. P. Miller and R. Schuster, Mol. Phys., 1967, 12, 111. Ch. P. Rao, P. Balaram and C. N. R. Rao, J. Chem. SOC., Furaday Trans. I , 1980, 76, 1008. W. E. Bull, S. K. Madan and J. E. Willis, Znorg. Chem., 1963, 2, 303. A. J. Carty, Can. J. Chem., 1966, 44, 1881. E. W. Randall, C. M. Silcox Yoder and J. J. Zuckermann, Inorg. Chem., 1966, 5, 2240. M-H. Baron, C. de Loze and G. Sagon, J. Chim. Phys., 1973,70, 1509. P. Combelas, M. Costes and C. Garrigou-Lagrange, Can. J. Chem., 1975, 53, 442. lo C. N. R. Rao, H. S. Randhawa, N. V. R. Reddy and D. Chakravorty, Spectrochim. Acta, Part A , 1975,31, 1283. l1 M. J. Adams, C. B. Baddiel, R. G. Jones and A. J. Matheson, J. Chem. SOC., Furaday Trans. 2, 1974, 70, 1114. M. J. Adams, C. B. Baddiel, G. E. Ellis, R. G. Jones and A. J. Matheson, J. Chem. SOC., Furaday Trans. 2, 1975, 71, 1823. ' M-H. Baron, J. Corset, C. de Loze and M. L. Josien, C.R. Acad. Sci., Ser. C, 1972, 274, 1321. l3 T. G. Clune, W. E. Waghorne and B. G. Cox, J. Chem. SOC., Faraday Trans. I , 1976, 72, 1294. l4 Handbook ofchemistry and Physics, ed. C. R. Weast (Chemical Rubber Co., Cleveland, Ohio, 1968).W. E. WAGHORNE A N D H. RUBALCAVA 1207 l5 G. Durgaprasad, D. N. Sathyanarayana and C. C. Patel, Bull. Chem. Soc. Jpn, 1971, 44, 316. l6 K. L. Dorris, T. H. Siddal, W. E. Stewart and M. L. Good, Spectrochim. Acta, Part A, 1967,23,1657. L. Pauling, The Nature ofthe Chemical Bond (Oxford University Press, London, 2nd edn, 1952), (a) p. 207, (b) p. 343. D. N. Waters, Z. Kantaret and N. N. Rhamna, J. Raman Spectrosc., 1978, 7 , 288. l9 P. A. Temussi and F. Quadrifoglio, Chem. Commun., 1968, 844. 2o P. A. Temussi, T. Tancredi and F. Quadrifoglio, J. Phys. Chem., 1969, 73, 4227. 21 B. G. Cox, J. Chem. Soc. B, 1970, 1780. 22 F. A. Cotton and G. Wiikinson, Advanced Inorganic Chemistry (Interscience, London, 1962). (PAPER 1/855)