Infrared and Raman studies of solid formaldehyde

Spectrcchimics Acta, Vol.3OA,pp. 863to868.Pergsmon Prees1974.Printed inNorthern Ireland

Infrared and Raman studies of solid formaldehyde HOSSEIN KHOSHKOO, STUART J. HEMPLE and ETJQENE R. NIXON Department of Chemistry and Laboratory for Research on the Structure of Matter, University of Pennsylvania, Philadelphia, Pennsylvania 19174, U.S.A. (Received11 July 1973; Revised 18 October 1973) Ah&&---The infrared and Raman spectra of the fundamental vibrational regions and the Raman spectra of the lattice mode regions are reported for polyorystalline films of H&O and D&O at 8O’K. It is inferred that the primitive cell contains at least four molecules, that the unit call group has no center of symmetry, and that the intermolecular interactions weaken the CO bonds and strengthen the CR bonds. INTRODUCTION THE INFRARED spectrum of solid formaldehyde

at - 180°C was examined some years ago by SCHNEIDER and BERNSTEIN [l], who noted that the carbonyl stretching fundamental is broad and falls at a lower frequency than the gas value and that both CH stretching modes, on the other hand, are blue-shifted from the gas. They concluded that hydrogen bonding is not involved in molecular association and that the infrared observations can be qualitatively explained in terms of a charge displacement in the carbonyl group with a lone pair orbital of the negatively charged oxygen atom of one molecule directed toward the positively charged carbon atom of another molecule, this type of association repeating in chains running through the crystal. According to their picture, the CO bonds in the solid become weaker and the CH bonds stronger. SCHNEIDER and BERNSTEIN further reported that rapid deposition of thick films from the vapor gave spectra quite different from those of thin films and suggested that the fast deposition produces a mixture of crystalline forms. HARVEY and OGIILVIE [2] re-examined briefly the infrared spectrum of formaldehyde at 77” and 4°K and found some additional fine structure in some of the fundamentals but no evidence for the presence of more than one crystalline form. In this paper, we report a more extensive study of solid H,CO and solid D&O in the Raman as well as the infrared. EXPERIMENTAL

The vapors of H&O and D&O were prepared in the usual way by heating the For the infrared spectra, the vapor was deposited from a Dry paraformaldehydes. Ice reservoir (v.p. of H,CO = 22 torr) onto a CsI window cooled to 80°K in a suitable dewar. The period of deposition varied from less than 1 min to about 6 min, depending upon the film thickness desired. The spectra of the films deposited at 80°K in every case were recorded at 4”, 20”, 40°, 60°, 80’ and 110°K on a PerkinElmer 225 spectrophotometer with spectral slit widths ranging from 0.3-0.5 cm-l. There were no noticeable differences in the spectra at these various temperatures. [l] W.G. SCHNEIDER andH.J. BERNSTEIN, !hans.~amdaySoc.52, [2] K. B. HARVEY and J. F. OGILVIE,Can. J. Chem. 40, 85 (1962). 863 1

13 (1956).

H. KEOSEKOO, S. J. HEMPLX and E. R. NIXON

864

The Raman spectra were all measured on rather thick flms deposited at 80°K onto rhodium plated copper discs, using right angle scattering of either the 4880 or 5146 A line from a Coherent Radiation 52 MG laser. The spectra were recorded by photon counting on a Spex Ramalog system. RESULTS AND DISCUSSION

Lattice modes

The Raman spectra of solid H&O and D&O at 80°K in the lattice mode region are given in Fig. 1 and the frequencies, peak heights and half-widths are listed in Table 1. The correspondences between lattice mode frequencies of the two solids as given in Table 1 were made on the basis of similarity in the intensity patterns and the assignments to either translational or librational type modes made on the basis of frequency ratios. For a pure translational lattice mode, the frequency ratio for the two isotopes would be given by (M@&CO]/MIH,CO])l/a = 1.03. The frequency of a lattice mode involving libration about a single principal axis of the molecule would involve the square root of the moment of inertia about that axis and the isotopic frequency ratio would be 1.42, 1.09 or 1.13, depending upon whether the A, B or C moment were involved [3]. Six of the Raman frequencies below 150 cm-l have ratios sufficiently close to 1.03 to assign them as translational modes. We could not observe the D&O counterparts of the two weak H,CO lines at about 40 and 48 cm-l on the edge of the laser line. Since three of the translational lattice modes of a crystal are acoustic modes, the six, or probably eight, observed optical translational lattice modes require a minimum of three or more probably four molecules in the primitive cell.

I ?m

!

300

I

I

1

250

200 FREQUENCY

150

I

lco

Fig. 1. Raman spectra of lattice mode regions of polycrystalline and D&O

at 80’K.

Excitation

[3] H. BLAU and H. NIELSEN, J. Mol. Spectry.

50

(CM-‘1

films of H&O with 4880 A line from Ar/Kr Iaaer.

1, 124 (1957).

Infrared and Raman studies of solid formaldehyde

866

Table 1. Frequencies (cm-l), relative intensitiesand assignmentsof Raman active lattice modes of solid H&O and D&O at 80’K D&O

H&O Freq. (40) (48) 63 76 85 96 105 135 163 161 178 208 220 316

Av~/a

Rel. int.*

-

V.W.

-

V.W.

4 3 6 (i) (10) (10) (15) (10) 15

20 16 V.W. V.V.W. 10 40 26 60 15 40 30 100

Freq.

62 74 83 95 103 134 141 162 166 (190) 197 236

Rel. int.*

b/n

-

4 4

-

16 20 w.sh. W. W.

rs,

(8)

(12) (10) (G 15

50

50 70 16 (“7”6, 100

Freq. ratio

1.02 1.03 1.02 1.01 1.02 1.01 1.08 1.06 1.07 (1.09) 1.12 1.34

Assign.t TP T? T T T T T T L L L L L L

* Approximate relative intensity given by product of AvI12 and peak height above background. t T = translational, L = librational.

From the frequency ratios, the six highest frequency lattice modes observed in the Reman appear to be largely librational in nature, the ones at 316 and 236 cm-l involving a libration essentially about the A axis. A primitive cell with 4 molecules would have twelve k = 0 librational lattice modes and while we tlpparently observe only six, they are in general broader than the pseudo-translational bands and may possibly consist of some unresolved modes. On the other hand, some of the lattice modes might be forbidden or too weak to be observed in the Raman. Intramolecular

vibrations

The infrared and Raman spectra in the fundamental vibrational regions of solid H,CO and D,CO are shown in Figs. 2 and 3 and in Table 2. The infrared of H,CO agrees generally with that given in Ref. [Z], although we have observed more structure, particularly in the v2 region. Many of the fundamental regions display two or more bands in both the infrared and Ramsn. A few of these are attributed to W molecules, HDCO impurity or a v3 + vg combination (see Table 2). That the remaining components of the vibrations are due to molecular coupling in the crystal was demonstrated by observing the infrared spectra of solid films containing 1% H,CO in D,CO and 1% D,CO in H,CO. In both cases, each fundamental of the impurity molecule gives rise to a single s,bsorption peak ; for example, the fundamentals of H,CO in the D,CO host appear at 2890, 2830, 1713, 1497, 1250 and 1177 cm-‘. The infrared absorption in the v2 (CO stretch) region in both H,CO and D,CO is intense and broad even in the thinnest 6lms examined. It is not possible to determine how many distinct components may be contributing to the v, absorption but other

and E. R. NIXON

H. KZXOSHKOO,S. J. HEMPLE

866

FREQUENCY

(CM-‘)

Fig.

2. Infrared

(upper) and Raman (lower) spectra of polycrystalline H&O at SOoK.

Fig.

3. Infrared

(upper) and Raman (lower) spectra of polycrystalline D&O at SOOK.

es

of

FRECXJENCY (CM-‘)

films of

factors such as the coupling of v2 with lattice phonons may be contributing to the width of the infrared absorption. We of course confirm the observations of SCHNEIDER and BERNSTEIN [l] of the 30-40 cm-l red shift in v2 from the gas to the most intense infrared absorption in the solid and the 40-50 cm-l blue shifts in the CH stretching modes v1 and v4. The present authors [4] have reported similar shifts in v2 and v1 and v4, though smaller in magnitudes, upon going from monomeric formaldehyde to the dimeric species in argon and nitrogen matrices. Again it appears that the intermolecular interactions involve perturbing forces through the carbonyl groups. [4] H. KFIOSHKHOO and E. R. NIXON, Spectrochim. Acta 29, 603 (1973).

Ph.D.

1167.1

1500.0 2843.1 1249.3

1746.1

w w w vw (W)

1179 1177 1174 1166

thesis, University

w s m m w

1505 1490 2886 1250 1245

1720 br sh 1714 1711 Ivs 1705 m sh 1679 w 1674 w (13C)

2831 m

gas*

2782.4

* From S. Hemple,

vs

%

v4

v3

v3

Vl

Region

H&O Infrared solid

Infrared

1970.

w s m m sh w vw (13C) vw w w vw

w s vw vw (13C)

m vs w sh vw

of Pennsylvania,

1505 1491 2887 1250 1249 1246 1241 1237 1180 1178 1175

1712 1701 1686 1676

2831 2825 2821 2811

937.8

1101.3 2160.3 987.6

1700.8

2057.1

vw s vw (W) vw v3 + vs VW w (HDCO) w br sh

and D&O

m m s s sh w sh

942 w

1096 2214 990 988 986

1656 1vs 1663 1641 vw 1632 w (13C) 1624 vw

2120 2095 2082 2971 1691 1688 1682 1670

D&O Infrared solid

regions of solid H&O

Infrared gas*

in the fundamental

Raman solid

Table 2. Infrared and Raman frequencies (cm-l)

Raman solid

m s m m

942 w

1095 2216 992 990

1640 vw 1634 vw

1664 vw 1653 vs

2099 s 2094 vs

at 80’K

3

E V !%

i

E

8

a

i

z

s

& “a

868

H. KHOS~OO, 5. J.

HJWPLE

and E. R. NIXON

In each region of the spectra the most intense Raman peak and the strongest infrared peak are nearly coincident in frequency except for v1 of H&O and ve of H&O and D&O, where the differences are 8-10 cm-l. In the infrared, only v5 and v6 of H&O show reasonably resolved sharp peaks (2-3 cm-l half-widths). The v,, region of H&O is interesting in that the three observed infrared peaks have the same frequencies as the three peaks in the Raman spectrum. As far as we are aware, the crystal structure of formaldehyde has not been determined. From the spectroscopic results presented here, several inferences can reasonably be drawn: the primitive cell contains at least four molecules all in equivalent sites; the unit cell group does not contain a center of symmetry, and the molecular arrangement, as has been pointed out before, is one in which the carbonyl bonds are weakened and the CH bonds strengthened. In the Raman spectrum of a non-centrosymmetric crystal one might expect to observe both the transverse and longitudinal modes of the vibrations, the separation of the modes being proportional to the infrared intensity. Perhaps the polycrystalline nature of our films modifies the effect expected for single crystals and it may be that in some regions, weak peaks on the high frequency side of the strong Raman peaks are really longitudinal modes rather than factor group components but this could not be proved. As a final note, we observe that in the H&O solid just as in the gas [3], vp and v2 + v6 are in strong resonance. The observed frequency of v, in the solid is about 87 cm-l lower than the value calculated from a set of harmonic force constants which reproduce within a maximum difference of 8 cm-l the frequencies of the most intense infrared peak in all six fundamental regions of solid D&O and all the fundamental regions except vp for solid H&O. Consistent with strong Fermi resonance is the fact that the higher frequency member of the dyad (largely v2 + vg) has an integrated infrared intensity of about half that of the lower frequency band (mostly v4) and is at least 10 times more intense than any other combination band observed in the spectrum. wish to acknowledgepartial support of this work by Advanced Projects Agency Contract 15-67-C-0216.

Acknowledgementa-We