J. Phys. Chem. A 2010, 114, 7311–7314
7311
Geometric Isomerism in the OCS-CS2 Complex: Observation of a Cross-Shaped Isomer J. Norooz Oliaee, F. Mivehvar, M. Dehghany, and N. Moazzen-Ahmadi* Department of Physics and Astronomy, UniVersity of Calgary, 2500 UniVersity DriVe North West, Calgary, Alberta T2N 1N4, Canada ReceiVed: May 11, 2010
Infrared spectra of the OCS-CS2 van der Waals complex were studied in a pulsed supersonic slit-jet using a tunable diode laser probe. Spectra were recorded in the region of ν1 fundamental of OCS. Two bands were observed and analyzed, one band corresponding to a previously observed planar isomer and another due to a new isomer which has a nonplanar cross-shaped structure. The intermolecular (center of mass) separation of the planar isomer is 3.87017(2) Å. The structure of this isomer has been determined previously from its rotational spectrum. The cross-shaped isomer was observed here for the first time, and its structure was determined with the help of isotopic substitution. Two structural parameters, the intermolecular distance (R) and an angle (φ), are necessary to completely define the structure. These were determined to be R 3.5553(8) Å and φ ) 104.82(22)° which are in fair agreement with the theoretical predictions. Introduction Weakly bound complexes provide a convenient starting point for a detailed understanding of different pathways that can be taken between the gas and condensed phases of matter. In this regard, it is of considerable interest to determine if and how geometrical choices made in the early stages of condensation influence the growth of larger clusters. Although it is expected that the number of isomers grows rapidly with cluster size, in many cases only a single isomer is observed experimentally. Binary and ternary complexes formed from linear molecules are among the simplest systems for which geometrical isomerism can be studied experimentally, since this minimizes the number of intermolecular parameters required to determine the relative orientation of the monomers. Peebles and Kuczkowski have reported a number of trimer structures deduced from rotational spectra1–8 in which they conclude that the trimer faces are often composed of sets of dimer-like geometries, with subtle differences in intermolecular angles and/or distances. These findings have been largely confirmed by Valdes and Sordo in a series of high level quantum chemical calculations.9–12 For example, ab initio and density functional theory calculations were carried out to study the OCS-(CO2)2 complex.12 Three isomers, two barrel-like structures (with C1 symmetry) and one with a cyclic (pinwheel) structure were located and characterized. Thus far, only the lowest energy barrel-like isomer of the complex has been observed experimentally.2 The geometrical disposition of the three dimer faces of this isomer is quite similar to that of the slipped CO2 dimer,13–15 the lowest energy form of OCS-CO1 (isomer a), also observed and analyzed in the microwave region,16 and a higher energy form of OCS-CO2 (isomer b), only recently observed by our group in the infrared region.17 Therefore, the rationalization of structures of larger size clusters (at least in the case of clusters with linear triatomic monomers) seems to rely on a full knowledge of the number of minima on the potential energy surface for the clusters of smaller size. * To whom correspondence phas.ucalgary.ca.
should
be
addressed,
ahmadi@
In the past few years, observation of second isomers of the binary complexes formed from OCS, CO2, and N2O have become more common. The first higher energy isomer observed was the polar form of (OCS)2.18 Subsequently, the polar isomer of (N2O)219 and the higher energy isomers of OCS-CO217 and OCS-N2O20 were reported. In all cases the detection of the second isomer was made using helium as a gas carrier instead of the heavier rare gases such as neon or argon. In this paper, we report the first infrared observation of the OCS-CS2 complex, including the discovery of a new isomeric form. The lowest energy isomer of this complex was first studied by Newby et al.21 who used Fourier transform microwave spectroscopy to observe its rotational spectrum. The deduced rotational constants and the dipole moment components were consistent with a planar structure in which the CS2 and OCS monomers were nearly parallel to one another and had a center of mass separation of R ) 3.8017(2) Å. In the same work, Newby et al. reported three calculated minimum energy structures which they obtained using a semiempirical modeling program, ORIENT.22 Two of the isomers (I and III) were planar with OCS and CS2 monomers almost parallel, whereas isomer II had a cross-shaped structure, reminiscent of the experimental structure obtained for CO2-CS223 with a dihedral angle of 90°. Here we report the experimental confirmation of isomer II. Initially, two bands were recorded in the region of the ν1 fundamental band of OCS. A prominent band around 2058 cm-1 was assigned to isomer I and a second weaker band around 2060 cm-1 was assigned to isomer II. Preliminary analysis of the higher frequency band indicated that the dihedral angle in the complex was not exactly 90°. This meant that in addition to the intermolecular distance R the tilt angle of the OCS monomer axis with respect to the vector R connecting the OCS and CS2 centers of mass (φ) was required for an unambiguous structural determination. By using 18O-substituted OCS in our gas mix and observing the corresponding 18OCS-CS2 band, it was then straightforward to determine the required structural parameters. These were in fair agreement with those predicted by Newby et al.21
10.1021/jp104305r 2010 American Chemical Society Published on Web 06/22/2010
7312
J. Phys. Chem. A, Vol. 114, No. 27, 2010
Norooz Oliaee et al. TABLE 1: Molecular Parameters for the Planar Isomer of the 16OCS-CS2 Complex in the Region of OCS ν1 Fundamental (in cm-1)a
Figure 1. Observed and simulated (3.2 K) infrared spectra of the planar isomer of the 16OCS-CS2 complexes in the region of the OCS ν1 fundamental. An assumed Gaussian line width of 0.0017 cm-1 was used for the simulated spectrum. The blank regions in the experimental spectra are obscured by absorption due to 16OCS monomer or He-16OCS complex.
Experimental Methods
ν0 A′ B′ C′ 107∆K′ 109∆JK′ 108∆J′ 107δK′ 108δJ′ A′′ B′′ C′′ 107∆K′′ 109∆JK′′ 108∆J′′ 107δK′′ 108δJ′′
present work
ref 21
2057.911607(27) 0.0789058(20) 0.0331977(14) 0.02336604(75) 3.23(26) 5.44b 4.72(40) 1.392b 1.61(23) 0.079044479(27) 0.033171173(18) 0.0233666217(91) 2.484(22) 5.44(99) 4.921(25) 1.392(30) 1.617(11)
0.079044490(30) 0.033171171(20) 0.023366622(10) 2.492(23) 5.6(11) 4.920(30) 1.388(33) 1.618(13)
a Uncertainties in parentheses are 1σ from the least-squares fits in units of the last quoted digit. b These parameters were fixed at their ground state values.
The experiments were carried out at the University of Calgary using our previously described tunable diode laser in combination with a pulsed supersonic jet apparatus.24 Spectra were calibrated by using a fixed etalon (free spectral range, 0.00997 cm-1) and a reference gas cell containing OCS. A dilute mixture of CS2 (∼0.30%) and OCS (∼0.15%) in He with backing pressure of about 8 atm was used for the supersonic expansion through a slit-shaped nozzle. Isotopically enriched 18OCS was prepared from the reaction between sulfur and C18O.25 Sulfur and C18O were heated together in a 1 L glass bulb at a temperature of 300 °C and were kept at this temperature for approximately 60 h. Subsequent low resolution (2 cm-1) infrared analysis of the gas from the reaction showed that the purity of 18OCS sample was higher than 90%, with carbon monoxide as the major impurity. Since the latter has a much higher vapor pressure than OCS at liquid nitrogen temperature, it could be removed by freezing the content of the storage vessel to 77 K and opening it to the pump. The isotopically enriched sample of C18O was obtained from Cambridge Isotopes, with a stated purity of 95%.
Figure 2. Observed and simulated (3.2 K) infrared spectra of the crossshaped isomer of the 16OCS-CS2 complexes in the region of the OCS ν1 fundamental. An assumed Gaussian line width of 0.0017 cm-1 was used for the simulated spectrum. The blank regions in the experimental spectra are obscured by absorption due to the 16OCS monomer or He-16OCS complex.
Results and Analysis A. Planar Isomer in the Region of the 16OCS ν1 Fundamental. A search to the lower side of the 16OCS monomer band origin quickly resulted in detection of strong absorption lines which were later assigned to the vibrational fundamental of the planar isomer. The gas mixture of 2:1 in favor of CS2 was chosen to minimize interference from absorptions by OCS clusters. The region between 2056 and 2060 cm-1 was scanned to record the entire band. It was then easy to show that this band could be well accounted for in terms of the planar isomer of the 16OCS-CS2 complex whose lower state parameters were already known from microwave spectroscopy.21 As expected from the microwave structure, this band was predominantly b-type (∆Kb ) (1), but with a weaker a-type component. The observed a-to-b-type intensity ratio was approximately 1:5. Portions of the observed (top trace) and simulated (bottom trace) spectra are shown in Figure 1. For the simulated spectrum, an effective rotational temperature of 3.2 K and an assumed Gaussian line width of 0.0017 cm-1 were used. The strong and broad feature at 2059.909 cm-1 is the central Q- branch due to the weaker a-type component and to the right is the R-branch region. The blank regions in the experimental trace are obscured by strong OCS monomer or He-OCS absorptions. Assignment,
fitting, and simulations were done using the PGOPHER computer program.26 We assigned 327 infrared transitions to 212 observed lines, with values of J and Ka ranging up to 21 and 11, respectively. The rotational transitions reported by Newby et al. were also included in the fit. These were given a weight of 1000 to reflect their higher precision. A fit with an root mean square (rms) deviation 0.00012 cm-1 (with the unweighted deviation of 0.0026 cm-1) for the infrared was obtained. Results of the fit are listed in Table 1. The observed transitions and assignments are given in Table A1 of the Supporting Information. B. Cross-Shaped Isomer in the Region of the OCS ν1 Fundamental. Searching to the higher side of the planar isomer band, we found a new band also requiring both 16OCS and CS2 in the gas mix. This band which was approximately five times weaker than the band due to the planar isomer had its central Q-branch at 2060.06 cm-1. A rotational analysis showed that it was a c-type band, with rotational parameters rather similar to those calculated for structure II (the higher energy cross-shaped isomer) of the OCS-CS2 complex by Newby et al.21 Figure 2 shows a portion of the spectrum including the characteristic c-type Q-branch (left) and the R-branch (right) for this band. To obtain a better match for the central Q-branch, it was
Geometric Isomerism in the OCS-CS2 Complex
J. Phys. Chem. A, Vol. 114, No. 27, 2010 7313
TABLE 2: Molecular Parameters for the Cross-Shaped Isomer of the OCS-CS2 Complex in the Region of OCS ν1 Fundamental (in cm-1)a present work 16
ν0 A′ B′ C′ A′′ B′′ C′′
12
32
O C S-CS2
2060.066306(41) 0.0729047(43) 0.0332888(46) 0.0288672(55) 0.0730581(45) 0.0333086(54) 0.0288564(58)
18
O12C32S-CS2
2024.060732(41) 0.0711220(54) 0.0326148(27) 0.0286301(45) 0.0712925(64) 0.0326411(30) 0.0286340(46)
semiempirical theory21
0.071332 0.034324 0.029954
a
Uncertainties in parentheses are 1σ from the least-squares fits in units of the last quoted digit.
Figure 4. Illustration of the experimentally determined structures of the isomers of the OCS-CS2 complex. The structure of isomer I was determined from MW spectroscopy ref 21.
TABLE 3: Structural Parameters for the Cross-Shaped Isomer of the OCS-CS2 Complex R/Å φ/dega
present work
ref 21
3.5553(8) 104.82(22)
3.490 95.2
a φ is the angle formed by the carbon atom in CS2 and the center of mass and the sulfur atom of OCS.
shown in Figure 4. This structure is very similar to that observed for the CO2-CS2 complex.23 The rotational constants listed in Table 2 are consistent with a nonplanar geometry because of the large inertial defect
Ic - Ia - Ib ) 584.456 - 230.774 - 506.306 ) Figure 3. Observed and simulated (3.2 K) infrared spectrum of crossshaped isomer of the 18OCS-CS2 complex in the region of OCS ν1 fundamental. An assumed Gaussian line width of 0.0017 cm-1 was used for the simulated spectrum. The blank regions in the upper trace are obscured by absorption due to 18OCS monomer or He-18OCS complex.
necessary to include a weak contribution from a-type transitions. The observed a- to c-type intensity ratio was approximately 1:13. We were able to assign 193 c-type transitions to 148 observed lines. The maximum values for J and Ka were 13 and 10, respectively. Results of the fit are shown in Table 2, where the ground state parameters obtained from the semiempirical modeling program, ORIENT, are also given for comparison.21 The weighted rms deviation and the associated unweighted rms were 0.00016 and 0.00038 cm-1, respectively. The assigned transitions can be found in Table A2 of the Supporting Information. We also recorded this band for the 18OCS-CS2 isotopomer to determine the structural parameters of the cross-shaped isomer (see below). Parts of the Q- and R-branch of this band along with the simulated spectrum are shown in Figure 3. The number of transitions, maximum J, and maximum Ka values for this band were 155 (122 frequencies), 12, and 9, respectively. The observed transitions and assignments for this isotopomer are given in Table A3 of the Supporting Information. Results of the fit whose weighted and unweighted rms were 0.00018 and 0.00034 cm-1, respectively, are listed in Table 2. In the fits to the normal and 18O forms of cross-shaped OCS-CS2, we found that allowing centrifugal distortion parameters to vary gave only a slight improvement in the standard deviations, while also giving worse agreement for the unresolved central Q-branch region. Thus all distortion parameters were fixed to zero in the final analyses reported here. C. Structural Analysis of the Cross-Shaped Isomer. The band centered at 2060.06 cm-1 was analyzed in terms a complex having CS symmetry. The CS symmetry and observation of a predominantly c-type band suggest a cross-shaped structure as
- 152.624 amu Å2 Furthermore, an orthogonal orientation of the CS2 to the plane of symmetry can be substantiated by calculating
2Pb ) Ic + Ia - Ib ) 2
∑ mibi2 i
As expected, the experimental value of Pb ) 154.462 amu Å2 is nearly the same as the value of ICS2 ) 154.022 amu Å2 with a small discrepancy of ∆ac ) 0.440 amu Å2 that would be equivalent to the inertial defect in a planar molecule. However
2Pc ) Ia + Ib - Ic ) 2
∑ mici2 i
gives an experimental value of Pc ) 76.312 amu Å2, which is measurably different from IOCS ) 83.098 amu Å2, indicating the OCS monomer is not exactly orthogonal to the intermolecular axis. Therefore, two parameters were required to specify the structure completely (assuming the orthogonal orientation of the CS2 and unchanged monomer geometries). We used the STRFIT of Kisiel,27 with input data being the six observed rotational parameters of the two isotopomers as listed in Table 2. The results of the fit are given in Table 3 and shown graphically in Figure 4. We used the same monomer geometries for CS2 and OCS as in ref 21: rCS ) 1.552 Å in CS2, and rOC ) 1.160 Å, and rCS ) 1.560 Å in OCS. Discussion and Conclusion Since our spectra do not offer the accuracy that is possible using microwave spectroscopy, the new information for the planar isomer is limited to vibrational shift and the changes in rotational parameters A,B, and C between the ground and excited vibrational states. The observed vibrational shift (the shift in
7314
J. Phys. Chem. A, Vol. 114, No. 27, 2010
the dimer band center from the OCS monomer band center) was ∆ν ) -4.289(2) cm-1, and as can be seen from Table 1, the observed changes for the rotational parameters between the ground and excited vibrational state are very small (
