Direct visualization of the H–Xe bond in xenon hydrides: Xenon isotopic shift in the IR spectra

The noble-gas hydrides of HNgY type (Ng is a noble gas atom and Y is an electronegative fragment) were discovered in 1995.^1,2 The fingerprints of these unusual molecules are strong characteristic absorptions related with H–Ng stretching,^3,4 resulting from extremely large transition dipole moment of the corresponding vibrations. In addition to chemical arguments, assignment of the IR features was supported by deuterium isotopic effect, which actually demonstrated H bonding to a heavy atom. Meanwhile, only in the case of HArF,⁵ the involvement of a noble-gas atom has been proved directly by argon isotopic shift. Heavier noble-gas atoms (Kr and Xe) could not be studied in this way in view of very large mass difference between H and Ng. It is worth noting that there is no chance to get such information from the studies in natural xenon, because the observed absorption half-width is substantially larger than the expected isotopic shifts, so it is not determined by isotopic profile or instrument resolution. In this communication we report the first observation of xenon isotopic shift in the IR spectra of xenon hydrides from the experiments with selected xenon isotopes, which were applied recently in our laboratory for EPR studies.^6,7

A monoisotopic “heavy xenon” sample (99.4% of $X136e$⁠) and highly enriched “light xenon” sample (86.2% of $X129e$⁠, 13.6% of $X128e$⁠) were obtained from Russian Research Center Kurchatov Institute (these samples will be further referred as $X136e$ and $X129e$⁠, respectively). Acetylene was used as a precursor compound for preparation of different-type xenon hydrides (HXeCCH, HXeH, HXeCC, and HXeCCXeH).^8–10 A common method of synthesis of HXeY includes homolytic dissociation of HY or another possible precursor molecule in a solid xenon matrix at low temperature followed by thermally induced mobility of H atoms at $∼40K$⁠.^11,12 In the case of acetylene, dissociation is a two-step process, and annealing results in the formation of different xenon hydrides, depending on the initial concentration of acetylene and dissociation degree. The whole scheme may be represented as follows:^9,10,13,14

Dissociation processes (1) and (2) occur upon photolysis or radiolysis of the deposited matrix samples $(Xe/C2H2)$ below 20 K, whereas reactions (3)–(6) result from mobilization of trapped H atoms upon annealing at 40–45 K. The modification of dissociation method used in this work is based on x-ray irradiation. From mechanistic viewpoint, it is qualitatively similar to fast electron irradiation applied in our previous studies,^10,11,14 because in both cases the radiation energy is mostly absorbed by the matrix and isolated molecules are affected due to transfer from the host.

The gaseous mixtures of $C2H2/X136e$ and $C2H2/X129e$ $(∼1/1500)$ were deposited onto a cooled KBr window in an original continuous-flow helium cryostat¹⁵ at 30 K and irradiated with x-rays at 16 K using a 5-BKhV-6 tube with a tungsten anode operating at 40 kV and 80 mA. IR spectra were measured at 16 K with a Perkin Elmer 1720 X spectrometer (MCT detector, wavenumber accuracy of $0.01cm−1$⁠, resolution of $0.5cm−1$⁠, typically 200 scans). After 45–60 min deposition the deposited layer thickness was between 40 and $65μm$⁠, as controlled by the interference picture. Special attention was paid to the similarity of conditions of matrix deposition and annealing in the experiments with different isotopes in order to avoid any possible effect of matrix morphology.

A 90–120 min irradiation resulted in decomposition of about 44%–49% of initial acetylene. In agreement with the previous studies, subsequent annealing at 42 K led to formation of new bands assigned to xenon hydrides [HXeH,² HXeCCH,^9,10 HXeCC,¹⁰ and HXeCCXeH (Ref. 10)] and vinyl $(C2H3)$ radical.^16,17 In our experiments the absorptions attributed to HXeCCXeH (triplet centered at $∼1301cm−1$⁠) were rather weak and broad, whereas other bands were well defined and suitable for precise determination of the maxima positions. Here we will focus on comparison of the data for “heavy” and “light” xenon. One can see that the maxima of all absorptions ascribed to xenon hydrides, which should contain $X136e$ isotope, are slightly red shifted in comparison to the corresponding features observed in a $X129e$ matrix (an example is shown in Fig. 1). The shift magnitude is between 0.17 and $0.38cm−1$ or 0.015%–0.03% (Table I). Regarding the reliability of these results, one might notice that the values reported by different groups sometimes differ by up to $0.5cm−1$ due to a number of experimental reasons. However, the reproducibility of maximum positions of absorptions from a stable component of HXeH observed in different experiments in a $X136e$ matrix in our studies was within $±0.01cm−1$⁠. The values for HXeCCH and HXeCC are somewhat less reliable because of the effect of atmospheric water absorbing in the same region, but, in any case, this uncertainty is much smaller than the observed difference in the experiments with different xenon isotopes. Furthermore, no measurable shift was found for the absorption of vinyl radical $(890.97cm−1)$ obtained in $X129e$ and $X136e$ matrices in the same experiments. Thus, we may conclude that the observed shifts are of fundamental nature and they reflect the xenon isotopic effect on the H–Xe stretching frequency.

The simplest way to estimate the expected shift for H–Xe stretching is through classical mechanical approach, assuming that $ν(H–Xe)$ is proportional to $[(mH+mXe)/(mHmXe)]1/2$⁠, which yields $ν(X129e–H)/ν(X136e–H)≈1.0002$⁠, in qualitative agreement with experimental results. A more detailed view can be obtained from quantum-chemical calculations, which were performed using an ab initio MP2 method with the DEF2-TZVPP triple-zeta basis. The GAMESS suite of programs was used in the calculations.¹⁸ 

The geometric structures of HXeCCH and HXeH molecules were optimized, and for each isotope (⁠$X129e$ and $X136e$⁠) the harmonic frequencies were calculated. In the case of HXeH, anharmonic effects were also included in a separate VSCF and VCI-VSCF calculation.^19,20 The absolute values of the H–Xe stretching frequencies are typically somewhat overestimated in the MP2 calculations,^2,3 so we are focusing here only on the isotopic shifts. One can see (Table I) that the computed shifts agree well enough with the experimental values. It is worth noting that the calculations also predict appreciable xenon isotopic shifts for HCC–XeH stretching mode in HXeCCH and bending mode in HXeH. However, the former absorption should appear well below $400cm−1$ (i.e., beyond the range accessible for our instrument) and the latter one (around $700cm−1$⁠) was too weak to be examined in detail in our experiments. No attempts were made to calculate the xenon isotopic shift for the HXeCC radical with the same computational method since it is known that the MP2 calculations are unsuitable for the open-shell systems of this kind.⁹ 

In summary, we may conclude that the studies using highly isotopically enriched xenon matrices allowed us to “visualize” the H–Xe bonds in a number of xenon hydrides. In our view, this is an important point for final proof of the chemical nature of these species. The xenon isotopic studies may be especially helpful for assignment of the absorptions in complicated situations (e.g., in preparation of xenon hydrides from complex organic molecules).

This work was supported by INTAS (Project No. 05-1000008-8017) and Division of Chemistry and Material Science of the Russian Academy of Sciences (programme no. 1). Technical assistance from I. V. Tyulpina is acknowledged.