Reports of the Academy of Sciences of the USSR

1. Volume 160, No. 6

CHEMISTRY

Yu. N. SHEINKER, L. B. SENYAVINA, V. N. ZHELTOVA

POSITION AND INTENSITY OF THE ABSORPTION BAND OF THE ANTISYMMETRIC STRETCHING VIBRATION OF THE $\mathrm{N}_3$ GROUP IN THE IR SPECTRA OF ORGANIC AZIDES

(Presented by Academician M. M. Shemyakin on 29 VII 1964)

In the literature ($^{1-3}$) it is still considered that the position of the band of the antisymmetric stretching vibration of the azido group, $\nu_{(as)}\mathrm{N}_3$,* in the IR spectra of organic azides is localized in the range 2120–2160 cm$^{-1}$.

At the same time, consideration of the spectra of azides available to us, both in the present study and in an earlier study ($^4$), showed that the absorption range of $\nu_{(as)}\mathrm{N}_3$ is broader than that given in the cited works ($^{1-3}$) and is approximately confined within 2100–2170 cm$^{-1}$. The position of $\nu_{(as)}\mathrm{N}_3$ within this range is determined by the nature of the group directly adjacent to the azido group. In aliphatic azides, $\nu_{(as)}\mathrm{N}_3$ is located in the range 2100–2110 cm$^{-1}$. Conjugation of the azido group with $\mathrm{C{=}O}$, $\mathrm{C{=}C}$, $\mathrm{N{=}C}$, and $\mathrm{SO}_2$ raises the frequency of the absorption band to 2130–2150 cm$^{-1}$; conjugation with $\mathrm{P{=}O}$ and $\mathrm{P{=}S}$, or the presence of an electron acceptor at the $\alpha$-C atom relative to the $\mathrm{N}_3$ group, raises $\nu_{(as)}$ to 2163–2166 cm$^{-1}$. In ferrocenyl azide, $\nu_{(as)}$ occupies a position only slightly higher (2113 cm$^{-1}$) than in aliphatic azides (Table 1).

There are no data in the literature on the integral intensity of the $\nu_{(as)}\mathrm{N}_3$ band. We measured these characteristics in the spectra of a number of organic azides of various classes. The calculation of integral intensity was carried out without taking into account wing corrections (over a broad integration range). When the band was split (a rather frequent occurrence for $\nu_{(as)}\mathrm{N}_3$), the total intensity was taken.

Our measurements showed that the integral intensity of the $\nu_{(as)}\mathrm{N}_3$ band is considerably more sensitive to the nature of the radical bonded to the azido group than is the frequency. In aliphatic azides, the intensity of $\nu_{(as)}$ has a value of $4.4$–$4.5 \cdot 10^4$ m$^{-1}\cdot$l·cm$^{-2}$. Upon conjugation of the azido group with double bonds, the intensity $A_{(as)}\mathrm{N}_3$ changes noticeably. This change may be either toward an increase or toward a decrease in intensity, depending on the electronegativity of the substituents, whereas the frequency upon conjugation with any double bonds, as noted earlier, only increases. In aromatic azides the intensity increases by $\sim 50\%$. Lengthening of the conjugation chain leads to a further increase in intensity. Introduction into the phenyl ring of an electron-acceptor substituent—the nitro group (compound 3)—lowers the intensity. The presence of an electron acceptor at the $\alpha$-C atom, as well as conjugation with a carbonyl group, causes a decrease in intensity. In compounds in which the azido group is conjugated with a carbonyl, alongside a drop in $A_{(as)}\mathrm{N}_3$ there is observed some increase in the intensity of the $\nu_{\mathrm{C}=0}$ band (by 0.2 units in comparison with compounds in which, instead of the group

* The designation of vibrations in the form $\nu_{(as)}\mathrm{N}_3$ and $\nu_{(s)}\mathrm{N}_3$ is strictly correct only for a group with symmetry $D_{\infty h}$; however, we retain these symbols for designating the characteristic vibrations of the $\mathrm{N}_3$ group in organic azides, where this symmetry no longer exists.

Table 1*

* All investigated substances had constants corresponding to literature data; \(a\)—splitting of \(\nu_{as}\mathrm{N}_3\) is observed; values of \(\nu_{(as)}\mathrm{N}_3\) are given for the most intense band; \(b\)—data for solution in CCl\(_4\).

N$_3$ there is a CH$_3$ group). Thus, in these compounds the azido group manifests itself as a weak electron donor. When an additional strong electron donor is present in the compound (even with conjugation of the azide group with a carbonyl), it acts as an acceptor and the intensity of its band increases. Thus, in the presence of mercury in chloromercuribenzoic acid azide (compound 19), the intensity of the azido-group band increases markedly (to $6.4 \cdot 10^4\ \mathrm{m}^{-1}\cdot\mathrm{l}\cdot\mathrm{cm}^{-2}$, as compared with $4.1 \cdot 10^4\ \mathrm{m}^{-1}\cdot\mathrm{l}\cdot\mathrm{cm}^{-2}$ for benzoyl azide), while the intensity of the carbo-

Table 2*

* The position of the most intense band is given first; the remaining bands are weaker or are shoulders.

nyl increases by 0.5 unit. In another derivative of mercuribenzoic acid (compound 20), where the donor properties of mercury are lowered by the presence of an $\mathrm{HO_2C{-}CH_2{-}S}$ group, the intensity of the azide-group band decreases to $5.2 \cdot 10^4\ \mathrm{m}^{-1}\cdot\mathrm{l}\cdot\mathrm{cm}^{-2}$, but it is still higher than in benzoyl azide. From the intensities in compounds 29 and 30, where the N$_3$ group is attached to phosphorus, it is seen that the $\mathrm{P{=}O}$ and $\mathrm{P{=}S}$ bonds are electron donors with respect to the azide group. The value of the intensity of the band under consideration in ferrocene azide increases only slightly (analogously to the increase in frequency) in comparison with aliphatic azides, although a greater rise in intensity might have been expected, since the ferrocene nucleus is usually regarded as a considerably stronger donor than the benzene ring. Apparently this is due to the closeness of ferrocene and the N$_3$ group in electron-donor ability.

Thus, from the intensity data it follows that the azide group is a weak electron-donor group which, however, is capable of exhibiting noticeable acceptor properties when attached to a sufficiently strong donor.

In considering the IR spectra of azides, splitting of the $\nu_{(as)}\mathrm{N}_3$ band is often observed (see Table 2) and ($^4$, $^5$). An attempt was once made to explain this splitting in the case of methyl carbonic acid ester by rotational isomerism ($^4$). Lieber in 1963 came to the conclusion ($^3$) that

splitting of the \(\nu_{(as)}\mathrm{N}_3\) band can be explained by Fermi resonance of \(\nu_{(as)}\mathrm{N}_3\) with combination tones involving symmetric and C—N stretching vibrations and other lower frequencies. Indeed, in all the cases considered by him, when splitting was present the spectra contained intense bands capable of forming a combination tone with a frequency of \(\sim 2100\ \mathrm{cm}^{-1}\); conversely, in the absence of such bands in the spectrum he did not observe splitting of the \(\nu_{(as)}\mathrm{N}_3\) band \((\mathrm{C}_6\mathrm{H}_5\mathrm{CH}_2\mathrm{CON}_3,\ (\mathrm{CH}_3)_2\mathrm{NCH}_2\mathrm{N}_3)\). A similar explanation of the splitting of \(\nu_{(as)}\mathrm{N}_3\) in the case of methyl azide was given in work \((^6)\). From this point of view we examined the spectra of azides obtained by us, some of which are presented in Table 2. Consideration of the data in this table shows that in a number of cases the explanation given by Lieber is confirmed. Thus, in compound 3 no splitting is observed, and its spectrum lacks bands that could give combination tones with a frequency of about \(2100\ \mathrm{cm}^{-1}\). At the same time, in compounds 18, 16, and 12 splitting is observed (especially strong in 16), and the spectra contain possibilities for the formation of combination tones (see Table 2). However, analysis of the data in Table 2 shows that the spectra of compounds 5, 30, and 27 contain absorption bands that could form a combination tone and produce Fermi interaction, whereas the antisymmetric stretching vibration of the azide group in these cases appears as an individual symmetric band. This is not in agreement with Lieber’s point of view. Probably the presence in the spectrum of bands which, in sum, give combination tones corresponding in frequency, or which have corresponding overtones, is not a sufficient condition for splitting of \(\nu_{(as)}\mathrm{N}_3\). Evidently, these frequencies must correspond to vibrations capable of interacting with the vibrations of the azide group. Thus, in the spectrum of ferrocenyl azide there are bands that could form a tone with frequencies 2114 and \(2111\ \mathrm{cm}^{-1}\), close to \(\nu_{(as)}\mathrm{N}_3\). However, these bands are associated with vibrations of the unsubstituted cyclopentadienyl ring, and because of this their interaction with the vibration of the azide group cannot be transmitted through the Fe atom, and no splitting of the band is observed. In other cases, the barrier to interaction of vibrations apparently is phosphorus (compounds 29 and 30) or sulfur (compound 27). In these compounds the \(\nu_{(as)}\mathrm{N}_3\) band is not split, although possibilities exist for the formation of combination tones. Thus, a condition for splitting of azide bands is not only the presence of vibrations capable of giving the corresponding combination frequencies in the spectrum, but also the absence of obstacles to interaction between \(\nu_{(as)}\mathrm{N}_3\) and these combination frequencies; such obstacles may be decoupling of vibrations because of their remoteness from one another in the molecule, separation of groups by a heavy atom, and also differences in the symmetry of these vibrations.

The authors express their deep gratitude to V. A. Gilyarov, M. I. Rybinskaya, L. A. Neiman, L. F. Matyash, I. S. Berdinskii, A. L. Zhuzo, I. G. Bolesov, and L. I. Maimind for the synthesis of the compounds investigated.

Institute of Chemistry of Natural Compounds
Academy of Sciences of the USSR

Received
25 VII 1964

REFERENCES CITED

1. L. Bellamy, Infrared Spectra of Complex Molecules, 1963, p. 378.
2. E. Lieber, C. N. R. Rao, Anal. Chem., 29, 916 (1957).
3. E. Lieber, C. N. R. Rao, A. Thomas, Spectrochim. Acta, 19, 1135 (1963).
4. Yu. N. Sheinker, Ya. K. Syrkin, Izv. AN SSSR, Ser. Fiz., 14, 478 (1950).
5. E. Lieber, E. Oftedahl, J. Org. Chem., 24, 1014 (1959).
6. E. Mantica, G. Zerbi, Gazz. Chim. Ital., 90, 53 (1960).