Conformational analysis of n-perfluoroalkanes: n-C4F10 and n-C6F14

Journal of Molecular

Structure,

26 (1975)

421-428

421

OElsevier Scientific Publishing Company, Amsterdam -

CONFORMATIONAL AND n-C, F14

P. PIAGGIO Istituto

ANALYSIS

Printed in The Netherlands

OF n-PERFLUOROALKANES:

n-C4 Flo

and P. G. FRANCESE

di Chimica Industriale,

Uniuersit& di Genova,

Via Pastore,

3-

16132 Genoa

(Italy) G. MASE’M’I Istituto

and G. DELLEPIANE

di Chimica delle Macromolecole

de1 CNR.

Via A. Corti, 12 -

20133 Milan (Italy)

(Received 27 August 1974)

ABSTRACT The vibrational spectra of n-C, F,, and n-C, F,, were investigated in the vapour, liquid and solid phases and the equilibrium conformations of these molecules were determined. Different solid-phase spectra were recorded by changing deposition and/or annealing conditions. INTRODUCTION

The study of the structure and of the vibrational spectra of a polymer can be carried out on the basis of the available calculation techniques, either for ideal-chain models or for disordered-chain models [l] . The validity of

the results so obtained is strictly dependent on the validity and generality of the force fields employed. Therefore the problem of the determination of force fields valid for polymeric chains is of great interest, either from a purely spectroscopic point of view, or for the study of the properties of polymers. So far a general force field has been derived only for polyethylene and for polyoxymethylene by refining the vibrational frequencies of a large number of oligomers [Z-4] . Various vibrational studies on PTFE [5] have been carried out leading to different or partial results. The solution of this vibrational problem can be obtained through a force field whose validity can be checked on the perfluoroalkane series. Moreover, a vibrational study of perfluoroalkanes is interesting in order to understand the particular properties of these compounds which have received, so far, very little attention [6--81. This first paper is mainly addressed to the understanding of the problem of rotational isomerism in n-C, F,Oand n-C6 F14,which is indispensable in order to interpret the vibrational spectra. In the course of this-work we have observed clear evidence of the presence of various rotational isomers in both liquid and gaseous phases and have noted the great differences appearing in

422

the solid spectra of each compound by changing deposition and/or annealing conditions. The investigation of this problem should moreover allow the evaluation of the force constants relative to the different conformations of the CF, groups in the molecule. These interaction force constants are of great interest because, even in highly crystalline PTFE, amorphous zones exist in which the chain takes on conformational sequences different from the more stable one. EXPERIMENTAL

The samples of high purity n-C4Ftofrom Peninsular Chemresearch Inc., Gainesville, Fla., and n-C,FX4 from I.C.L., Milan, were used without further pupation. The infrared spectra were recorded with Perk&Elmer 225 (4000-400 cm-‘) and Perkin-Elmer FIS-3 (400-50 em-‘) spectrometers. Conventional cells fitted with KBr or polyethylene windows were used for gaseous and liquid samples. The low-temperature spectra were recorded with conventional low-temperature cells, equipped with KBr, CsI and polyethylene windows. A large number of low-temperature spectra were obtained by varying both the temperature of deposition of the sample and the annealing conditions. The following results were obtained: n-C4Flo. Two kinds of low-temperat~e spectra were obtained. The first was measured after depositing the sample at a temperature below -120 ‘C. The second was observed after annealing these samples at a temperature of - -120°C. In the first kind of spectrum a few bands show very small differences in relative intensities with respect to the second one. n-C&F,,. Four kinds of spectra were measured. The first (I) was obtained by depositing the sample in the temperature range -170” to -150°C. The second (II) was measured either by annealing the preceding samples at -130°C or by depositing them in the temperature range -145” to -120” C. The third (III) was measured by annealing the samples which give spectrum II at a temperature near to melting point (- -86°C). The fourth (IV) was measured by depositing the sample in the temperature range -130” to-120” Further annealing of the samples which give rise to these four kinds of spectra does not lead to any appreciable variation in the infrared spectra. The Raman spectra of liquid samples were recorded with a Jarrel-Ash Spectrometer equipped with a Ar-Kr laser source. For C6 F,, a “double pass” cell was used and depolarization measurements were carried out by the Edsal-Wilson method. The Raman spectra of solid samples were recorded with a Gary Model 81 (&Kr laser source) using a low-temperature cell. No attempts were made to separate the various possible crystalline mod~ications.

423 RESULTS

The rotational isomers taken into consideration, in this paper, for each molecule, are those which are not hindered by strong steric repulsions. To define the relative position between the C-C bonds of the molecular skeleton of the different isomers we shall use the following notation: T(trans), G(gauche), G’, H(helix), H’ corresponding to the internal rotation angles of 180, 60, 300,165 and 195” respectively, which correspond to the minima of the potential energy curves [9] . The obtained results are here reported, separately, for each molecule:

The spectra of n-C,F,, in the liquid phase could not be registered because of the low melting point. The symmetry group, the structure of the irreducible representations and the spectral activity of the possible rotational isomer (T, G and H) of n-C4 F10are reported in Table 1.

TABLE 1 Structure of the irreducible representations and spectral activity for a few possible

modelsof n-C,F,, Model

Point group

Structure of the irreducible representations

IR-active modes

Ramanactive mode*

T WI

c zh C,

llAg + Sq, 19A + 1’7B

18 36

18(11p 36(19p

+ 7B, + lOB,

+ 7d) + 17d)

ap = polarized; d = depolarized.

The Raman and infrared spectra of the crystal phase are schem&ically reported in Figs. la and lb respectively. The number of bands agrees only with those expected for the conformational model T which belongs to the symmetry group &h. This hypothesis is further supported by the mutual exclusion of several infrared and Raman lines which suggests the existence of a centre of inversion in the molecule and which is actually found in the T model. The infrared spectrum of the gas is schematically reported in Fig. lc. Four bands, two of medium intensity at 1072 and 960 cm-‘, two weak ones at 768 and - 655 cm-‘, appear only in the gas phase spectrum. These bands should therefore be assigned to the rotational isomers G or H.

424

Ill I I I

I

I I

_-. _-_c~_

I.

1_.

__)

II I I

1400 1300 Ii00

I

Ll~.~. *I. ._I

-

llb0

I&O

f___-f~

900

800

x._L

I;

,i

i

I. I..._.__,_ .. .I I -t-

1.~ 700 6&J

&I

do-

300

la

i

I

I

I

I

I

2&

lb

_lc

-cm-~

Fig. 1. Comparison between infrared and Raman spectra of n-C=F,,. la, Raman spectrum of crystal; lb, infrared spectrum of crystal; lc, infrared spectrum of gas.

n-G 3’~ In the following discussion, solid IV will be considered as the more crystalline form of n-C6F,4. The infrared spectrum of this solid does indeed show very sharp bands with crystal splittings. These splittings are evidence of a considerable interaction between the molecular chains. This intermolecular interaction has not been found in polytetrafluoroethylene. The vibrational frequencies of n-&F, are reported in Table 2 and the predicted spectra for its various possible isomers in Table 3. The infrared and Raman spectra of the crystal phase are schematically reported in Fig. 2a and 2b respectively. The following observations can be made : (a) Seven Raman bands at 1382 (ms), 977 (mw), 757 (vs), 749 (w), 716 (ms), 388 (ms) and 172 (mw) cm -’ do not appear in the infrared spectrum. (b) 14 infrared bands at 1352 (mw), 1339 (mw), 1180 (m), 1166 (WV), 1149 (s), 1105 (ms), 797 (mw), 711 (m), 570 (m), 556 (m), 502 (w), 348 (w), 224 (w),204(w) and 162 (VW) cm-’ do not appear in the Raman spectrum. The two IR bands at 723 and 658 cm-’ are of uncertain Raman activity. (c)Eightweakbands at 1322,620,612,603, 540,375,316 and 146 cm-’ appear both in the infrared and in the Raman spectra. (d) Because of the very large number of bands in the 1200-1270 cm-’ region, no comparison between infrared and Raman data can be made in this region. These results point towards a conformational model with a centre of inversion like the more stable in the crystalline phase. We can therefore disregard all the models which do not belong to the point groups CZhand Ci* To make a further distinction between these two models it is necessary to use depolarization data. Since these data are not available for the crystal, we will use depolarization data obtained from the Raman spectrum of the liquid. In order to do this, we must, first, look for the bands due to the less stable isomers (comparison between liquid and crystal) and secondly, make the

425

TABLE 2 Vibrational

frequencies

of n-C, F,,

Raman liq.

Raman cry&.

1380 1368

1382

P, m D, VW

1322 1314

IR cry&.

IR liq;

IR vap.

1339s

1338 s

ms

mw

1352 1339 1322

mw mw w

D, mw

1272 w 1240

D, VW

1211

D, w

1242 1231 1209 1200

w w VW w, db

1270 s 1259 m

1232-1200

1253 1240

sh vs

vs

1254

s

1219

s

1200 s 1190

1183

1180 1166 1149 1105

D, VW

ms, db VW s ms, db

1072 P, VW 991 P, VW 977 P, w

977

mw

818 P, VW 797 mw, db

757 P, vs

757 vs

731 P, w

749 w

714 P, s 655 P, w

716 ms

609

D,

mw

603 1 broad 578

620 w 612 mw

603 mw

620 VW 612 vw 603 w

540 ?

570 m, db 556 m 537 VW

D, ww

540 P, mw

723 m 711 m 658 m, db

1185

sh

1149 1100 1071 1045 989 980 927 870 855 817 794 758 740 730

s ms m m ms VW mw w ms ms ms m m ms

357 D, VW ?

992 973 927 875 856 817 798 758 741 731

w sh w sh m m m sh sh m

716 ms 656 m 634 w

611 w 603 sh 581 m

612 VW

572 w

529 m

442 ? 388 ms 375 w

ms m w

711 ms 655 ms 632 w

526 mw 502 w

440 D, VW 383 P, vs

1151 1102 1075

437 VW

499 VW 478 VW 440 vw

376 m

374

m

530 mw

426 TABLE 2 (continued) Raman liq.

Baman crsyt.

IR cry&

IR liq;

IR vap

348 VW

342 I?,VW 312I),vs 293 F, s 274 P, VW 274 P, VW

316 w

315 ms, db 296 m

290 s

290 s

265 m 256 ? 232 ? 224 w 204 s

184 D, vw 172 P, vs 145 P, ms

172 mm 146 w

226 sh 208 s

162 VW 146 ?

TABLE: 3 Structure of the irreducible Representations and Spectral Activities for a few passibIe models of n-C, I?,., Model

Point group

Structure of the irreducible representations

c zh

16A,

c,

X.&active modes

Raman active modesa

27

27(16P + lid)

28A + 26B

54

54(28p

Ci

27%

27

27W’p)

C,

54A

54

54(54p)

+ 12q,

-s- 27An

+ llBg

+ 15B,

-t 26d)

ap = polarized; d = depolarized.

assumption that the bands which do not disappear are due only to the isomer which is stable in the crystalline phase. On this second assumption, one can assign to the bands of the cry&4 the depolarization factor measured in the corresponding bands of the liquid. The comparison between the Raman spectra of liquid (Fig. 2~) and crystal (Fig_ Zb), the comparison between the infrared and Raman spectra of the liquid phase, reported in Figs. 2d and 2c respectively and the depolarization data reported in Table 2, give the following results:

427

I

1400

1300

I

1200

1100

1000

2d

1

I

900

800

700

600

500

400

300

200

100

cm-l

Fig. 2. Comparison between infrared and Raman spectra of n-C,F,,. 2a, infrared spectrum of crystal; 2b, Raman spectrum of crystal; 2c, Raman spectrum of liquid; 2d, infrared spectrum of liquid

(a) There are nine polarized Raman bands in the liquid phase at 1380, 977, 757, 714, 540, 383, 293,172 and 145 cm-’ which also appear in the crystalline phase. (b) There are five. depolarized bands in the Raman spectrum of the liquid at 1240,1211, 608,440 and 312 cm-’ which also appear in the spectrum of the crystal. (c) All the other bands reported in Table 2 and in Fig. 2, which appear only in the liquid phase may be considered “conformational bands” due to less stable isomers. It is interesting to notice the great difference between the infrared spectra of liquid and crystal. For instance, the bands which appear in the infrared spectrum of liquid and solid I between 800 cm-’ and 1070 cm-’ are absent from the spectrum of the other solids. This variation is also observed in the spectra of CqFlo. Another interesting region can be found around 600 cm-‘. In this region, in which great variations are observed in the infrared spectra of the four different solid phases, strong absorptions in the infrared and Raman spectra of the liquid occur. As far as the crystal is concerned, it can then be concluded that the number of observed bands and the depolarization data agree with the expectations for the structure TTT, belonging to the symmetry group &h. A great number of conformational bands appear in the spectrum of the liquid. Finally the examination of the spectrum of solid I shows that its spectral behaviour is analogous to that of the liquid_ By remembering what has been observed in the experimental part on solids I, II, III, IV, we can also conclude that n-C6 F,, shows different solid phases in which different molecular conformations are present, probably as mixtures.

428 CONCLUSIONS

AND DEVELOPMENT

LINES

In this paper we have shown that the planar zig-zag is the more stable isomer of n-&F*,, and n-C, F14. A large number of “‘conformational bands” has been identified. Of particular interest is the spectral region between 1100 and 800 cm-‘. Here we have found that various bands, which are only active in the liquid and gaseous phases, occur. A study of the variation of the band intensity with the temperature should be quite interesting. Such a study should allow the evaluation of the relative stability of the various isomers and should also allow their identification with the help of an approximate calculation of the energies of each isomer. Unluckily few thermodynamic data are known, for these compounds. Particularly no data on the range of “plasticity” of the crystal phase are available. The difference between the melting point and the temperature of order/disorder transition is -30” for CsFs, -270” for C!16F30and -3OO* for PTFE. It is therefore reasonable to expect that the range of “plasticity” of the molecules studied in this paper is quite large. It coincides with the temperature range in which the spectra of the solid phases are recorded, It should then be interesting to study the variations found in the various kinds of crystal spectra. Of great interest is the spectral region 1300-1200 cm-‘, in which the bands of the C-F stretch appear A particular phenomenon has been observed in the crystal-phase spectra of all compounds: the bands are not well resolved and their shape, which is symmetric in the gaseous and liquid phases, becomes asymmetric, with the absorption maximum toward the lower frequencies. Since the normal modes, which give rise to these absorptions, are strongly dependent on both the intra- and inter-molecuIar interactions, an analysis of the spectra, obtained with the matrix-isolation technique, should throw some light on this probIem.

REFERENCES 1 2 3 4 5 6 7 8 9

G. Zerbi,PureAppLChem., 26 (1971) 499. J. H. Schachtschneider and R. G. Snyder, Spectrochim. Acta, 19 (1963) 85,117; ibid, 20 (1964) 853. R. G. Snyder, J. Mol. Spectrosc,, 23 (1967) 224; J. Chem. Phys., 47 (1967) 1316. R. G. Snyder and G. Zerbi, Spectrochim. Acta, Part A, 23 (1967) 391. G, Zerbi and M. Sacchi, Macromolecules, 6 (1973) 692; G. Masetti, F. Cabassi, G. Morelli and G. Zerbi, Macromolecules, 6 (1973) 700, see aiso references quoted. G. J. Szasz, J. Chem. Phys., 18 (1950) 1417. J. R. Nielsen and C. W. Gullikson, J_ Chem. Phys., 21 (1953) 1416. E. L. Pace, A. C. Plaush and M. V. Samuelson, Spectrochim. Acta, 22 (1953) 1416. P. De Santis, E. Giglio, A. M. Liquori and A_ Ripamonti, J, Polym. Sci., Part A-l (1963) 1383.