Microwave Spectra, Planarity, and Conformational Preferences of cis

Nov 16, 2012 - Raphaela Kannengießer , Sebastian Klahm , Ha Vinh Lam Nguyen , Arne Lüchow , Wolfgang Stahl. The Journal of Chemical Physics 2014 141 ...
0 downloads 0 Views 495KB Size
Download PDF
Article pubs.acs.org/JPCA

Microwave Spectra, Planarity, and Conformational Preferences of cisand trans-N‑Vinylformamide Harald Møllendal* and Svein Samdal Centre for Theoretical and Computational Chemistry (CTCC), Department of Chemistry, University of Oslo, P.O. Box 1033 Blindern, NO-0315 Oslo, Norway S Supporting Information *

ABSTRACT: The microwave spectra of a mixture of cis- and trans-HNCO forms of N-vinylformamide, (H2C CHNHC(O)H), have been measured at room temperature in the 18−75 GHz spectral range. The spectra of two forms were assigned. The first of these forms has a cis arrangement for the HNCO chain of atoms, whereas the second form has a trans arrangement. The C−C−N−C chain of atoms is antiperiplanar (180°) in both forms. The inertial defect of the ground vibrational state of cis is −0.142(5) × 10−20 u m2, whereas this parameter is −0.087098(26) × 10−20 u m2 for trans. It is concluded that the equilibrium structures of both cis and trans are completely planar. The dipole moment determined from Stark effect measurements is μa = 9.96(8), μb = 2.22(3), μc = 0 (by symmetry), and μtot = 10.20(8) × 10−30 C m [3.06(2) D], for cis, and μa = 7.64(16), μb = 9.24(10), μc = 0 (by symmetry), and μtot = 12.0(2) × 10−30 C m [3.59(5) D] for trans. The spectrum of one vibrationally excited state, presumably the first excited state of the torsion about the C−N bond of cis, was assigned and the frequency of this state was determined to be 76(15) cm−1 by relative intensity measurements. The spectra of two vibrationally excited states of trans were assigned. These states are assumed to be the first excited state of the torsion about the CN bond, and a low bending vibration. Relative intensity measurements yielded 101(20) and ca. 300 cm−1, respectively, for the frequencies of these normal vibrations. Accurate values of the quartic centrifugal distortion constants, the dipole moments, and the vibration−rotation constants have been obtained for both cis and trans. The experimental work has been augmented by high-level quantum chemical calculations at the B3LYP/ccpVTZ and CCSD(T)/cc-pVTZ levels of theory. The theoretical calculation performed without symmetry restrictions correctly predict that cis and trans are both planar. The CCSD(T) rotational constants are in excellent agreement with their experimental counterparts, whereas the B3LYP quartic centrifugal distortion constants and the vibration−rotation constants are in fairly good agreement with experiments. The CCSD(T) dipole moments deviate more than expected from the experimental dipole moments. It is estimated that further conformers of cis and trans must be at least 4 kJ/mol higher in energy.

■

INTRODUCTION The peptide linkage, −C(O)NH−, found in amides is an essential feature of proteins. The properties of amides are therefore of great biological as well as chemical interest.1−3 Amides are crystalline substances with comparatively low sublimation pressures, or liquids of low volatility at room temperature. Intermolecular hydrogen bonding is prevalent in condensed phases of N-unsubstituted and N-monosubstituted amides due to the proton accepting oxygen atom and the hydrogen atom(s) attached to the nitrogen atom. Many of these substances therefore form hydrogen-bonded networks in the crystalline state.2 Extensive hydrogen bonding and crystal effects may therefore obscure the genuine structural and conformational properties of this class of compounds. X-ray crystallography may provide excellent knowledge of crystal structures and conformations of solid amides, but the gas-phase properties may, for reasons already stated, be significantly different from the condensed-phase properties. Experimental studies of amides in the gas phase at low pressures augmented with modern, high-level quantum chemical calculations are therefore preferred to obtain the © 2012 American Chemical Society

best possible insight in the true, unperturbed structural and conformational preferences of this class of compound. Gas electron diffraction (GED) and microwave (MW) spectroscopy are the experimental methods of choice for gasphase investigations due to their high resolution. Unfortunately, the low vapor pressures of amides and their tendency to decompose when heated have made gas-phase studies of amides less accessible than for compounds containing other functional groups resulting in a relatively limited literature about this class of compounds. The relative scarcity of studies of gaseous amides and their great biological and chemical interest motivated our laboratory to undertake investigations of them using the GED method and MW spectroscopy in combination with quantum chemical calculations. Our previous studies include acetamide (CH3CONH2),4 2-fluoroacetamide (CH2FCONH2),5,6 2-chloroacetamide (CH2ClCONH2),6,7 2iodoacetamide (CH 2 ICONH 2 ), 8 2,2-difluoroacetamide Received: September 26, 2012 Revised: November 12, 2012 Published: November 16, 2012 12073

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

Article

(CF2HCONH2)9 2,2-dichloroacetamide (CHCl2CONH2),10 2chloro-2,2-difluoroacetamide (CF2ClCONH2),11 2,2,2-trifluoroacetamide (CF 3 CONH 2 ), 12 2,2,2-trichloroacetamide (CCl3CONH2),13 propionamide (CH3CH2CONH2),14 formic hydrazide (H2NNHCHO),15 acrylamide (H2C CHCONH2),16 methyl carbamate (CH3OCONH2),17,18 methoxyacetamide (CH3OCH2CONH2),19 and 2-azetidinone.20,21 Experimental contributions from other laboratories include, for example, the prototype formamide (HCONH2),22−24 urea (H2NCONH2),25 H−N−C−O-cis- and trans-N-methylformamide (HCONHCH3),26 N,N-dimethylformamide (HCON(CH3)2),27 H−N−C−O-trans-N-ethylformamide (HCONHCH2CH3),28 H−N−C−O-cis-methoxyformamide ( H C O N H O C H 3 ), 2 9 H −N −C − O - t r a n s - f o r m a n il i d e (C6H5NHCHO),30,31 acetamide,32−35 H−N−C−O-trans-Nmethylacetamide (CH3CONHCH3),36 H−N−C−O-trans-Nmethylpropionamide (CH3CONHCH2CH3)37 H−N−C−Otrans-acetanilide (CH 3 CONHC 6 H 5 ), 38 and alaninamide (H2NCH2CONH2).39 These studies have revealed a number of interesting, and in some cases, unexpected findings. It has, for example, often been assumed that the amide group as a rule is completely planar, which is consistent with a significant weight of the −OCN+ resonance structure. However, a study based on accurate MW data and very high-level quantum chemical calculations40 of several small amides concluded that only one of the investigated compounds, namely formamide, is planar, whereas all the other amides in this study were found to have nonplanar amide groups in the free state. It therefore appears that a planar amide group in gaseous amides is the exception rather than the rule. The planarity problem is not the only interesting feature of amides. Recent studies have revealed a number of unexpected properties. Acetamide is one example of an amide displaying an unusual behavior. First, the conformation of the methyl group is exceptional in that one of the C−H bonds is almost perpendicular to the heavy atom framework according to very high-level ab initio calculations,4 which also found that its amide group is nonplanar. Interestingly, the barrier to internal rotation of this group is only 0.30466413(14) kJ/mol,35 much lower than, for example, the methyl barrier in the isoelectronic compound acetic acid (2.01258(20) kJ/mol).41 There are also significant changes in the barrier heights of the methyl group when the amide group hydrogen atoms are substituted with deuterium atoms due to coupling of the methyl torsion with the amide group inversion.35 Remarkably, the small perturbation caused by deuterium substitution has a substantial influence on the electronic structure of the peptide linkage.35 Moreover, although conformational mixtures in crystalline amides are very rare, this phenomenon is expected to be much more prevalent in the gas phase. However, so far acrylamide appears to be the only amide for which more than one rotameric form has been safely assigned by MW spectroscopy.16 In this work, our amide investigations are extended to include the first MW study, augmented with high-level quantum chemical calculations, of cis- and trans-N-vinylformamide (H2CCHNHC(O)H), where cis and trans refer to the configuration of the HNCO chain of atoms. In both the two title compounds, the two π-electron systems of the amide and vinyl groups, respectively, are conjugated. Moreover, the CN bond connecting the amide and vinyl groups has presumably a substantial σ-character. This opens up for rotational isomerism about the CN bond. There appears to be no experimental gas-phase structural work reported for this

kind of vinylamide rotational isomerism, but the MW spectra of the trans forms of formanilide30,31 and acetanilide,38 which have some conformational resemblance to the title compounds, are known. The π-electron conjugation implies that planar, or nearplanar forms can be expected for both cis and trans. Four representative forms, denoted I−IV, are depicted in Figure 1,

Figure 1. Models of four forms (I−IV) of N-vinylformamide found to be minima on the potential energy hypersurface in the B3LYP/ccpVTZ calculations. Atom numbering is indicated on I. Note that the H7N6C8O9 chain of atoms is cis in I and II, and trans in III and IV. The C1C2N6C8 link of atoms is antiperiplanar (dihedral angle =180°) in I and in III, and synperiplanar in II (∼20°) and in IV (0°). The MW spectra of I and III are reported herein.

with atom numbering indicated on I. The H7N6C8O9 chain of atoms is locked in a cis (forms I and II) or in a trans (III and IV) configuration due to restricted rotation about the amide N6C8 bond. Rotation about the C2N6 bond allows for rotational isomerism in both cis and trans. Four representative conformers, namely, I and II of cis, and III and IV of trans, are indicated in Figure 1. The C1C2N6C8 chain of atoms is exactly antiperiplanar in I and III, nearly synperiplanar in II, and exactly synperiplanar in IV. Although no GED or MW studies have been reported for the two N-vinylformamides, a recent quantum chemical investigation exists.42 Using HF and B3LYP calculations with the 6311++G** basis set, it was predicted that I and III are both planar and have approximately the same energy, whereas II and IV were found to be several kJ/mol less stable.42 II was predicted to be slightly nonplanar, whereas IV was found to be completely planar in these calculations. The interesting structural and conformational problems presented by the two N-vinylamides motivated the present MW research, which has been augmented with quantum chemical calculations performed at a much higher level of theory than previously employed42 to obtain information that could be useful for the assignment of the MW spectra and for investigating properties of the potential energy hypersurface. 12074

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

■ ■

Article

RESULTS AND DISCUSSION

predicted to be 8.9 kJ/mol more stable than IV. The energy differences corrected for harmonic zero-point-vibrational interaction are 12.2 for the I−II pair, 9.7 for the III−IV pair, and 0.3 kJ/mol for the difference between cis represented by I and trans III, with cis as the more stable form. The transition states of cis and trans were also explored. Two transition states, at 0° and 99.5° of the C1C2N6C8 dihedral angle were located for cis. The first of these transition states at 0° has an energy of 12.4 kJ/mol above the energy of I, or only 0.2 kJ/mol above the energy of II. The energy of the second transition state at 99.5° is 31.5 kJ/mol higher in energy than the energy of I. The trans form has only one transition state at 90.9° of the said dihedral angle. The energy of this state is 34.6 kJ/mol above the energy of III. The comparatively small energies of the cis and trans transition states indicates that there is little double bond character in the C2N6 bond. The data above allow potential functions for rotation about the C2N6 bond to be drawn for cis and trans, as shown in Figures 2 and 3.

EXPERIMENTAL SECTION Compound and MW Experiment. A commercial sample of N-vinylformamide was used as received. The sample contained roughly 50% of both cis and trans as judged by the intensities of the MW transitions. No impurities were noted. The MW spectra were recorded at room temperature. NVinylformamide is a liquid at room temperature and has a vapor pressure of about 30 Pa at this temperature. The MW spectrum was studied in the 18−75 GHz frequency interval by Stark-modulation spectroscopy using the microwave spectrometer of the University of Oslo. Details of the construction and operation of this device have been given elsewhere.43−45 This spectrometer has a resolution of about 0.5 MHz and measures the frequency of isolated transitions with an estimated accuracy of ≈0.10 MHz. Radio-frequency microwave double-resonance experiments (RFMWDR), similar to those performed by Wodarczyk and Wilson,46 were also conducted to unambiguously assign particular transitions, using the equipment described elsewhere.43 The spectra were measured at room temperature at a pressure of roughly 7 Pa. Quantum Chemical Methods. The present ab initio calculations were performed employing the Gaussian 0947 and Molpro48 programs, running on the Titan cluster in Oslo. Becke’s three-parameter hybrid functional employing the Lee, Yang, and Parr correlation functional (B3LYP)49 was employed in the density functional theory (DFT) calculations. Coupledcluster calculations with singlet and doublet excitations including noniterative triplet excitations, CCSD(T),50 were also performed. Peterson and Dunning’s51 correlation-consistent cc-pVTZ basis set, which is of triple-ζ quality was used in all the calculations. Quantum Chemical Calculations. The conformational properties of the O9C8N6H7 cis and trans forms were first explored. The variation of the energy with the C1C2N6C8 dihedral angle can conveniently be used for this purpose. The energies were calculated while the dihedral angle was stepped in intervals of 10° allowing all remaining structural parameters to vary freely. The B3LYP method was employed for these potential function calculations because CCSD(T) calculations would have been far too expensive. The first calculations indicate that cis has minima at 180° (I) and at about 20° (II), whereas trans has minima at 180° (III) and 0° (IV) of the C1C2N6C8 dihedral angle. Calculations of the structures, dipole moments, vibrational frequencies, quartic and sextic Watson A-reduction centrifugal distortion constants, and vibration−rotation constants (the α's), were then undertaken for the corresponding four conformers. Due precautions were observed in the calculations of the vibration−rotation interactions.52 All structural parameters were varied freely in these calculations with no symmetry restrictions. No imaginary vibrational frequencies were found for any of these four forms, which indicate that they are minima on the potential energy hypersurface. Selected results of these calculations are listed in Tables 1S−4S in the Supporting Information. It is seen from these tables that I, III, and IV are predicted to be exactly planar, whereas II is nonplanar with a C1C2N6C8 dihedral angle of −20.6°. All four forms were predicted to have completely planar amide groups. The electronic energy difference between I and II is calculated to be rather large (12.2 kJ/ mol) with I as the more stable conformer. Likewise, III is

Figure 2. B3LYP/cc-pVTZ potential function for rotation about the C2N6 bond for the H7N6C8O9 cis configuration. This function has minima at 20.6° of the C1C2N6C8 dihedral angle (form II) and at 180° (I) and transitions states at 0° and 99.5°. The B3LYP electronic energy difference between II and I is 12.2 kJ/mol, with I as the more stable conformer. The energy of the transition state at 0° is 12.4 kJ/ mol, and the energy of the transition state at 99.5° is 31.0 kJ/mol higher than the energy of I.

Much more comprehensive CCSD(T)/cc-pVTZ calculations of optimized structures, dipole moments, nuclear quadupole coupling constants, and electronic energies of the four forms I− IV were finally undertaken. It was important to investigate whether these extensive calculations predict a planar amide group. All starting conformations were therefore chosen to be nonplanar. The optimizations were done without symmetry restrictions employing the default convergence criteria of Molpro. All four forms were predicted to have completely planar amide groups. Moreover, conformers I, III, and VI were calculated to be exactly planar in this case, whereas II was found to be nonplanar with a C1C2N6C8 dihedral angle of −26.7°. The resulting structures are listed in Table 1. The rotational constants calculated from these structures are shown in Table 2. The dipole moment components of the Molpro calculations were transferred to the principal-axes dipole moment components using Bailey’s program Axis53 with the results listed in Table 2. The CCSD(T) electric field gradients were 12075

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

Article

Table 1. CCSD(T)/cc-pVTZ Structures of Four Formsa of N-Vinylformamide I C1−C2 C1−H3 C1−H4 C2−H5 C2−N6 N6−H7 N6−C8 C8−O9 C8−H10 C2−C1−H3 C2−C1−H4 H3−C1−H4 C1−C2−H5 C1−C2−N6 H5−C2−N6 C2−N6−H7 C2−N6−C8 H7−N6−C8 N6−C8−O9 N6−C8−H10 O9−C8−H10

Figure 3. B3LYP/cc-pVTZ potential function for rotation about the C2N6 bond for the H7N6C8O9 trans configuration. This function has minima at 0° (form III) and at 180° (IV) and a transition state at 90.9°. The B3LYP energy difference between IV and III is 8.9 kJ/mol, with III as the more stable conformer. The energy of the transition state is 34.6 kJ/mol above the energy of III.

calculated with Gaussian 09 using the structures in Table 2. Results of these calculations were employed to calculate the principal axis nuclear quadrupole coupling constants of the 14N nucleus shown in Table 2 using the program Nqc.53 Calculation of the centrifugal distortion constants by the CCSD(T) method is beyond our computational resources. The B3LYP quartic centrifugal distortion constants in the A-reduction form are therefore included in Table 2. The results in Tables 1 and 2 warrant further comments. The nonplanarity of II may result from nonbonded repulsion between the H4 and H10 atoms. This CCSD(T) distance (not given in Table 1) is 228 pm, compared to 240 pm, which is twice the Pauling van der Waals radius of hydrogen.3 A planar conformation of II would have brought these two hydrogen atoms into considerable closer proximity with increased repulsion as a result. The effect of a maximal conjugation stabilization of a completely planar II therefore appears to be insufficient to offset the ensuing nonbonded repulsion in this case. The situation in the planar IV rotamer is different. There is a short contact between H4 and O9 atoms in this case. The CCSD(T) nonbonded distance is as short as 231 pm compared to the Pauling van der Waals3 sum (260 pm) of hydrogen (120 pm) and oxygen (140 pm). It is possible that the overall effect of this close contact results in stabilization of the planar form due to the electronegative character of oxygen and the electropositive property of hydrogen. Whether this contact should be called an intramolecular hydrogen bond is a matter of semantics. The corresponding bond lengths (Table 1) of all forms are remarkably similar, varying by less than 0.5 pm. Partial loss of π-electron conjugation in II relative to the three other forms (I, III, and IV) therefore seems to influence its C2N6 bond length, as well as other bond lengths, to a minor degree. The N6C8 bond lengths of I−IV vary between 137.0 and 137.3 pm (Table 1), compared to 135.47 pm found for the equilibrium NC bond length in formamide.40 A lengthening of this bond in the title compound compared to the corresponding bond in formamide could perhaps be due to

H3−C1−C2−H5 H3−C1−C2−N6 H4−C1−C2−H5 H4−C1−C2−N6 C1−C2−N6−H7 C1−C2−N6−C8 H5−C2−N6−H7 H5−C2−N6−C8 C2−N6−C8−O9 C2−N6−C8−H10 H7−N6−C8−O9 H7−N6−C8−H10

II

Bond Length (pm) 133.9 134.0 108. 0 108.0 108.3 108.2 108.4 108.3 139.6 140.4 100.9 100.8 137.1 137.3 121.3 121.4 110.2 109.9 Angle (deg) 119.8 119.3 121.9 122.7 118.3 118.1 121.5 120.5 124.9 125.8 113.6 113.7 119.6 118.9 124.0 125.6 116.3 115.3 124.4 123.9 112.2 112.6 123.4 123.5 Dihedral Angle (deg) 0.0 −2.8 180.0 176.3 180.0 177.2 0.0 −3.8 0.0 156.7 180.0 −26.7 180.0 −24.3 0.0 152.4 180.0 177.7 0.0 −3.5 0.0 −5.5 180.0 173.2

III

IV

133.9 108.0 108.4 108.1 140.0 100.7 137.2 121.4 110.1

134.0 108.1 107.8 108.3 140.7 100.5 137.0 121.4 110.2

119.9 122.0 118.1 123.2 124.0 112.8 119.1 123.0 118.0 124.9 111.9 123.2

118.0 122.3 119.7 120.7 127.0 112.3 117.0 127.1 115.9 125.9 111.3 122.8

0.0 180.0 180.0 0.0 0.0 180.0 180.0 0.0 0.0 180.0 180.0 0.0

0.0 180.0 180.0 0.0 180.0 0.0 0.0 180.0 0.0 180.0 180.0 0.0

a

The four forms are depicted in Figure 1. The parameters of the forms whose MW spectra were assigned, are in boldface.

delocalization of π-electron density of the amide group into the vinyl part with an increased bond length as a result. The C8O9 bonds of I−IV should also be affected by conjugation, but their lengths vary little and are about 0.4 pm longer than the corresponding equilibrium bond length in formamide.40 Conjugation should also influence the C1C2 bond length in the title compound. There is a report of the rz value [133.91(13) pm] of the CC bond length in ethylene,54 which is the same as 133.9−134.0 pm calculated for the C1C2 bond lengths of the four conformers in Table 1. It is noted that most corresponding bond angles in Table 1 are quite similar. The largest differences are seen for the C1C2N6 and C2N6C8 angles of the III−IV pair, where the larger angles (by 3−4°) are found in IV. The short nonbonded contact between H4 and O9 mentioned above may have resulted in a opening up of this angle due to a repulsive interaction, which is effectively countered by electron conjugation and hydrogen-bond like stabilization. The energy differences of the rotamer pairs are interesting. The CCSD(T) electronic energy of I is lower than that of II by 12076

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

Article

Table 2. Theoretical Spectroscopic Contants of Four Formsa of N-Vinylformamide I A B C

ΔJ ΔJK ΔK δJ δK μa μb μc μtot χaa χbb χcc ΔE

II

III

moment component is μa, predicted to be as large as (10−11) × 10−30 C m (Table 2). A relatively intense series a-type Rbranch pile-ups of MW transitions separated by almost exactly the sum of the B and C rotational constants was predicted for its spectrum. These predictions turned out to be correct, and the spectrum of the ground vibrational state, which is shown in Table 5S in the Supporting Information, was readily assigned. The assignments of several of its transitions were confirmed by RFMWDR experiments. Only aR-transitions with J values up to 15 with K−1 between 0 and 4 were used to determine the spectroscopic constants listed in Table 1 because other K−1 transitions were frequently overlapped. b-type lines were searched for, but not definitely identified presumably because they are comparatively weak, which is consistent the fact that μa is almost than 5 times larger than μb (see dipole moment section below). The quadrupole hyperfine structure of the assigned lines caused by the nitrogen nucleus were predicted using the program MB0956 employing the nuclear quadrupole coupling constants listed in Table 2. However, the splittings were predicted to be too small to be resolved and this effect was therefore not taken into consideration. A total of 57 transitions were assigned and least-squares fitted to Watson’s A-reduction Hamiltonian in the Ir-representation using Sørensen’s program Rotfit.57 It was only possible to determine two quartic centrifugal distortion constans ΔJ and ΔJK. The remaining centrifugal distortion constants were preset at zero in the leastsquares fit with the resulting spectroscopic constants listed in Table 3.

IV

Rotational Constantsb (MHz) 37516.9 19598.4 19817.0 10220.2 2407.3 2887.8 2960.8 4491.5 2262.2 2542.8 2575.9 3120.3 Inertial Defectc (10−20 u m2) 0.0 −2.00 0.0 0.0 Quartic Centrifugal Distortion Constantsd (kHz) 0.190 0.686 0.664 2.62 −3.81 −11.5 −9.10 −6.93 280 165 108 12.6 0.0162 0.0983 0.133 0.918 1.11 4.05 2.60 3.89 Dipole Momente (10−30 C m) 10.56 11.67 6.11 3.21 3.15 2.86 10.65 13.50 0.00 1.16 0.00 0.00 11.02 12.07 12.28 13.87 Principal Axis Quadrupole Coupling Constantsb (MHz) 2.05 2.27 1.96 2.04 1.98 1.63 2.14 2.03 −4.03 −3.90 −4.10 −4.07 Energy Difference (kJ/mol) 0.0f 9.6g 0.0h 2.4i

a

The four forms are depicted in Figure 1. The parameters of the forms whose MW spectra were assigned, are in boldface. bCalculated from the CCSD(T) structures given in Table 1. cConversion factor: 505379.0 × 10−20 MHz u m2. dB3LYP/cc-pVTZ values. eCCSD(T) principal-axes dipole moment components and total dipole moment. f Total CCSD(T) electronic energy: −648187.60 kJ/mol. gRelative to I. hTotal CCSD(T) electronic energy: −648206.31 kJ/mol. iRelative to III.

Table 3. Spectroscopic Constantsa of Form I of cis-NVinylformamide A (MHz) B (MHz) C (MHz) Δc (10−20 u m2) ΔJ (kHz) ΔJKd (kHz) rmse Nf

9.6 kJ/mol (Table 2) compared to 12.2 kJ/mol obtained in the B3LYP calculations above. A difference of this order of magnitude found in the two methods is to be expected. A much larger variance is found for the III−IV pair. The CCSD(T) method yields 2.4 kJ/mol, whereas the B3LYP method predicts 9.7 kJ/mol. A smaller energy difference was expected for the III−IV pair than for the I−II pair because the two conformers of the first pair are both planar and the full effect of π-electron conjugation is preserved for both forms. The nonbonded interaction between the oxygen atom O9 and the hydrogen atom H4 could also be of importance for the reduced energy difference for this pair compared to the I−II pair. Finally, the energy difference between trans and cis is also of interest. The CCSD(T) electronic energy difference between cis and trans represented by I and III is 2.8 kJ/mol, with I as the global minimum, whereas the corresponding B3LYP difference given above is 0.3 kJ/mol. MW Spectrum and Assignment of the MW Spectrum of Form I. The MW spectrum of the two N-vinylformamides was found to be dense with absorption lines occurring every few megahertz throughout the MW region. A high spectral density was expected because both cis and trans forms contribute to the spectrum. The spectrum of I was first searched for. This rotamer was predicted to be the preferred form of cis in the quantum chemical calculations above. This conformer is very prolate with Ray’s asymmetry parameter55 κ ≈ −0.99. Its major dipole

ground

first ex tors stateb

37001(14) 2419.1292(38) 2272.1181(40) −0.142(5) 0.2042(75) −3.39(23) 1.078 57

35540(11) 2424.3806(45) 2280.3965(48) −1.058(5) 0.2057(81) −2.79(18) 1.226 53

a

A-reduction Ir-representation.64 bFirst excited state of the torsion about the C2N6 bond. cInertial defect. Conversion factor: 505379.01 × 10−20 MHz u m2. dFurther quartic constants preset at zero in the least-squares fit. eRoot-mean-square deviation of a weighted fit defined by rms2 = ∑[(νobs − νcalc)/u]2/(N − P), where νobs and νcalc are the observed and calculated frequencies, u is the uncertainty of the observed frequency, N is the number of transitions used in the leastsquares fit, and P is the number of spectroscopic constants used in the fit. fNumber of transitions used in the least-squares fit.

Comparison of the experimental (Table 3) and CCSD(T) (Table 2) rotational constants show that there is very good agreement between the two sets of constants. The B and C rotational constants differ by 0.5%, whereas the A rotational constant differs by 1.4%. The ef fective rotational constants in Table 3 are defined differently from the approximate equilibrium rotational constants in Table 2. The differences between the two sets of constants are expected to be relatively small, similar to what has been found in the present case. This indicates that the structure in Table 1 is very close to the equilibrium structure. Interestingly, the inertial defect is 12077

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

Article

−0.142(5) × 10−20 u m2 (Table 3). The inertial defect of a completely planar molecule is zero. It is also noted that the experimental values of the ΔJ and ΔJK quartic centrifugal distortion constants (Table 3) are quite similar to their B3LYP counterparts (Table 2). There could be two main reasons for the observed, relatively small inertial defect, zero-point vibrations and/or a nonplanar amide group. The fact that the CCSD(T) calculations predict a completely planar structure for I, including the amide group, is a strong indication that this is indeed the case. The observed inertial defect of −0.142(5) × 10−20 u m2 is therefore thought to arise mainly from low-frequency out-of-plane vibration(s). This inertial defect is similar to that of acrylamide (−0.131300(34) × 10−20 u m2),16 which is isomeric with Nvinylformamide. Acrylamide is planar, or very nearly planar.16 The inertial defect of the planar formamide molecule is +0.00862 × 10−20 u m2,22 whereas −0.60 × 10−20 u m2 has been reported for formanilide, which is assumed to be planar.30,31 Vibrationally Excited State of I. The ground-state spectrum was accompanied by several satellite lines presumably originating from vibrationally excited states. One such state, whose spectrum is shown in Table 6S of the Supporting Information, was assigned in the same manner as described for the ground state. Maximum value of J is 15 and maximum value of K−1 is 4. The spectroscopic constants are listed in Table 3. Attempts were made to assign further vibrationally excited states, but complete unambiguous assignments were not achieved. Relative intensity measurements performed largely as described by Esbitt and Wilson58 yielded 76(15) cm−1 for the torsional vibration. It is seen from Table 3 that the inertial defect is −1.058(5) × 10−20 u m2 for this state. The increase in the absolute value of this quantity from the ground state (−0.142(5) × 10−20 u m2) indicates that this state is an out-ofplane vibration,59 presumably the first excited state of the torsion about the C2N6 bond. The B3LYP calculations yield 116 cm−1 (anharmonic value; Table 1S, Supporting Information) for this mode, significantly higher than the experimental value. It is possible to compare the experimental and theoretical vibration−rotation constants defined by αex = X0 − Xex, where X0 is the rotational constants of the ground state and Xex are the rotational constants of the excited state under consideration. The experimental vibration−rotation constants derived from the entries in Table 3 are αA = +1461(18), αB = −5.25(1), and αC = −8.28(1) MHz, compared to +1644, −4.74, and −7.27 MHz, respectively (Table 1S, Supporting Information). The agreement between the two sets of vibration−rotation constants is satisfactory. Planarity of I. The three lowest normal vibrations have anharmonic frequencies of 116, 222, and 241 cm−1 according to the B3LYP calculations shown in Table 1S (Supporting Information). Two semiempirical equations by Oka60 and by Hanyu et al.61 can be employed to derive a fairly accurate value for the lowest torsional vibration provided it is well isolated from other vibrational modes. The B3LYP torsional vibration at 116 cm−1 is well separated from other vibrational modes. According to Oka,60 the dependency of the inertial defect, Δ0, of the ground vibrational state of planar molecules on the torsional frequency, ν1, and the largest principal moment of inertia, Icc, of the molecule is given by Δ0 = (−33.715/ν1) + 0.0186Icc1/2 in units of u Å2. Insertion of the inertial defect (−0.142(5) × 10−20 u m2; Table 3) of the ground state and Icc

derived from the C rotational constant in the same table gives ν1 = 80 cm−1, in good agreement with the experimental torsional frequency of 76(15) cm−1. An alternative equation by Hanyu et al.,61 ν1 is given by ν1 = −67.4/Δ1, where Δ1 is the change in the inertial defect (in u Å2 units) upon excitation of an out-of-plane vibration. Using the inertial defects of the ground and of the first excited state, one gets a torsional frequency of 74 cm−1 in this manner. This value is smaller than 80 cm−1 derived from the Oka60 formula. A smaller value was expected because the equation by Hanyu et al.61 neglects interaction terms.60 It is concluded that the low torsional frequency about the C2N6 bond is the major contributor to the inertial defect of I, whose equilibrium structure is completely planar. This conclusion is strongly supported by the advanced CCSD(T) calculations. Dipole Moment of I. The second-order Stark coefficients of the ground vibrational state shown in Table 4 were used to Table 4. Second-Order Stark Coefficientsa and Dipole Moment of Form I of cis-N-Vinylformamide ΔνE−2/10−6 MHz V−2 cm2 transition 41,3 50,5 51,4 51,4 51,4 51,5 51,5 51,5

← ← ← ← ← ← ← ←

31,2 40,4 41,3 41,3 41,3 41,4 41,4 41,4

μa = 9.96(8) a

obs = |1| = |1| =0 = |1| = |2| =0 = |1| = |2| Dipole Moment μb = 2.22(3) M M M M M M M M

calc

−15.8(4) −2.10(4) −1.57(3) −3.90(3) −10.9(5) −0.548(10) 2.48(4) 12.4(4) (10−30 C m) μc = 0.0b

−16.3 −1.97 −1.59 −3.81 −10.6 −0.554 2.63 12.1 μtot = 10.20(8)c b

Uncertainties represent one standard deviation. By symmetry; see text. cIn debye units: 3.06(2) D. Conversion factor 1 D = 3.33564 × 10−30 C m.

determine the dipole moment of I. The cell was calibrated using OCS whose dipole moment was taken to be 2.38568(66) × 10−30 C m.62 The theoretical second-order Stark coefficients were calculated using the Golden and Wilson formalism,63 which is implemented in the program MB04.56 The out-ofplane component of the dipole moment, μc, was preset at zero in the least-squares fit with the results shown in Table 4. The experimental dipole moment components of this table, μa = 9.96(8), μb = 2.22(3), and μtot = 10.20(8) × 10−30 C m, deviate more than expected from the CCSD(T) predictions (10.56, 3.15, and 11.02 × 10−30 C m, respectively; Table 1). It is somewhat surprising that the very advanced CCSD(T) calculations are not able to reproduce the dipole moment components more accurately. Assignment of the MW Spectrum of Form III. This conformer has a sizable μa component of about 6 × 10−30 C m, according to the CCSD(T) calculations (Table 2) and a value of Ray’s asymmetry parameter (κ = −0.95) that is not much different from the corresponding asymmetry parameter of I. A typical, quite strong aR-pile-up series was therefore expected for this form and searches for this series were first undertaken with immediate success. Several assignments were confirmed by MWRFDR experiments. The a-type lines assigned in this manner were used to predict bQ-branch lines, which were soon identified because most of them are among the strongest lines 12078

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

Article

of the spectrum caused by a relatively large μb of approximately 11 × 10−30 C m (CCSD(T) value; Table 2). The assignments of the strongest bR-branch lines were now straightforward. Additional P- and R-branch lines were gradually assigned and included in the least-squares fit. In this manner, 349 transitions with maximum value of J = 71 and K−1 = 14 were assigned. All quartic and one sextic (ΦKJ) centrifugal distortion constants were fitted using Rotfit,57 with the remaining sextic constants preset at zero. The spectrum is listed in Table 7S (Supporting Information), and the resulting spectroscopic constants are shown in Table 5.

as stated above for I. A total of 141 transitions with Jmax = 30 and K−1max = 4 were used to derive the spectroscopic constants. Relative intensity measurements yielded 101(20) cm−1 for this vibration, compared to 131 cm−1 predicted by the B3LYP method (Table 3S, Supporting Information). The experimental vibration−rotation constants calculated from the entries in Table 5 are in this case αA = +364.38 (2), αB = −7.08(1), and αC = −9.77(1) MHz, compared to +403.4, −6.26, and −8.59 MHz (Table 3S, Supporting Information). The agreement between the two sets of vibration−rotation constants is again satisfactory. The second excited state whose spectroscopic constants are found in Table 5 is presumably an out-of-plane vibration because its inertial defect (−0.1996(1) × 10−20 u m2) is in absolute terms larger than that of the ground state (−0.087098(27) × 10−20 u m2). A total of 65 transitions of this form with Jmax = 31 and K−1max = 2 were employed to obtain the spectroscopic constants of this excited state. Relative intensity measurement yielded ca. 300 cm−1, compared with the B3LYP frequency of 296 cm−1 (Table 3S, Supporting Information). The experimental vibration−rotation constants are αA = +21.90(3), αB = −2.32(1), and αC = −2.87(1) MHz, compared to +26.89, −1.49, and −2.58 MHz, respectively (Table 3S, Supporting Information). The agreement between the two sets of vibration−rotation constants is satisfactory in this case too. Planarity of III. The inertial defect of the ground vibrational state (Table 5) yields 97 cm−1 for the torsion about the C2N6 bond using the Oka formalism.60 This value agrees with 101(20) cm−1 found by relative intensity measurements. The Hanyu et al. equation61 yields a lower value of 83 cm−1, compared to the result using Oka’s formula,60 as expected.. The torsional fundamental is well separated from the other vibrational modes according to the B3LYP calculations (Table 3S, Supporting Information), just as in the case of I. All this, combined with the CCSD(T) result, is taken as evidence that III has a completely planar equilibrium structure. Dipole Moment of III. The determination of the dipole moment of III was made in the same manner as reported above for I. The results are presented in Table 6. The experimental dipole moment components of this table, μa = 7.64(16), μb = 9.24(10), and μtot = 12.0(2) × 10−30 C m, are again found to

Table 5. Spectroscopic Constantsa of Form III of trans-NVinylformamide A (MHz) B (MHz) C (MHz) Δd (10−20 u m2) ΔJ (kHz) ΔJK (kHz) ΔK (kHz) δJ (kHz) δK (kHz) ΦKJe (Hz) rmsg Nh

ground

first ex tors stateb

in-plane bend.c

19723.2338(55) 2976.65651(78) 2587.47826(79) −0.087098(27)

19358.851(19) 2983.7425(17) 2597.2452(18) −0.900684(74)

19701.332(33) 2978.9758(58) 2590.3460(58) −0.1996(1)

0.6874(10) −8.866(15) 103.162(69) 0.13766(23) 2.754(36) −0.1342(89) 1.336 349

0.7269(28) −9.193(65) 91.6(11) 0.14229(80) 2.292(99) −4.5(32) 1.702 141

0.633(26) −9.17(20) 92.5(73) 0.1420(15) 2.96(25) −f 1.773 65

a

A-reduction Ir-representation.64 bFirst excited state of the torsion about the C2N6 bond. cIn-plane bending vibration. dInertial defect. Conversion factor: 505379.01 × 10−20 MHz u m2. eFurther sextic centrifugal distortion constants preset at zero in the least-squares fit. f Preset at zero in the least-squares fit. gRoot-mean-square deviation of a weighted fit, as defined in the footnote of Table 3. hNumber of transitions used in the least-squares fit.

Attempts to get the nuclear quadrupole coupling constants of the nitrogen nucleus failed for III, just as for I, because no wellresolved quadrupole hyperfine structures suitable for exploitation were observed for the b-type transitions, which were explored in an attempt to derive these constants. However, the shapes of several of these transitions were unsymmetrical and definitely perturbed by this effect. The experimental (Table 5) and CCSD(T) rotational constants (Table 2) all agree to within about 0.5%, which is again taken as an indication that the CCSD(T) structure is close to the equilibrium structure. The inertial defect of III is −0.087098(27) × 10−20 u m2, even smaller, in absolute terms, than that of I (−0.142(5) × 10−20 u m2). The experimental and B3LYP quartic centrifugal distortion constants are in quite good agreement (Tables 2 and 5). The sextic centrifugal distortion constant ΦKJ = −0.1342(89) Hz (Table 4) should be compared with the B3LYP value (−0.219 Hz), shown in the Supporting Information, Table 3S. Vibrationally Excited State of III. The spectra of two vibrationally excited states were assigned for this rotamer in the same manner as described for the ground state. The spectra consisting of 141 and 65 transitions, respectively, are listed in Tables 8S and 9S of the Supporting Information, and the spectroscopic constants are displayed in Table 5. The strongest of these spectra, which has an inertial defect of −0.900684(74) × 10−20 u m2 (Table 5) is presumed to belong to the first excited state of the C2N6 torsion for a similar reason

Table 6. Second-Order Stark Coefficientsa and Dipole Moment of Form III of trans-N-Vinylformamide ΔνE−2/10−4 MHz V−2 cm2 transition 11,1 11,0 21,1 31,2 41,3 41,3 51,4 51,4 51,4

← ← ← ← ← ← ← ← ←

00,0 10,1 20,2 30,3 40,4 40,4 50,5 50,5 50,5

μa = 7.64(16)

obs M=0 1.24(2) M = |1| 9.36(13) M = |2| 1.95(8) M = |3| 1.18(3) M = |4| 0.918(15) M = |3| 0.349(5) M = |5| 0.758(10) M = |4| 0.406(5) M = |3| 0.149(2) Dipole Moment (10−30 C m) μb = 9.24(10) μc = 0.0b

calc 1.31 9.22 2.00 1.10 0.855 0.353 0.732 0.408 0.156 μtot = 12.0(2)c

a

Uncertainties represent one standard deviation. bBy symmetry; see text. cIn debye units: 3.59(5) D. Conversion factor 1 D = 3.33564 × 10−30 C m.

12079

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

■

deviate significantly from the CCSD(T) predictions (6.11, 10.65, and 12.28 × 10−30 C m; Table 1). Searches for the Hypothetical Forms II and IV. The theoretical calculations predict that there should be an energy difference of 7−13 kJ/mol between I and II, and the latter form was therefore expected to be present in small concentration, if it exists at all as a stable form of cis-N-vinylformamide. This conformer was calculated to have μa as the major dipole moment component (Table 2). Searches were therefore undertaken for the a-type R-branch transitions of this compound using the spectroscopic constants of Table 2 to predict their approximate frequencies. However, no assignments could be made. The B3LYP and CCSD(T) methods predict rather different (9.7 and 2.4 kJ/mol) energy differences between III and IV, as discussed above. IV is predicted to have a comparatively large μb of about 13 × 10−30 C m. Extensive searches for the spectrum of IV were undertaken by employing the spectroscopic constants of Table 2, but no assignments could be made. It is very unlikely that the spectrum of IV would have been overlooked if the energy difference were as small as 2.4 kJ/mol. It is concluded that IV must have a higher energy relative to III than 2.4 kJ/mol. A conservative estimate is that the hypothetical form IV is at least 4 kJ/mol less stable than III.

■

CONCLUSIONS

■

ASSOCIATED CONTENT

Article

AUTHOR INFORMATION

Corresponding Author

*Tel: +47 2285 5674. Fax: +47 2285 5441. E-mail: harald. [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Anne Horn for her skillful assistance. The Research Council of Norway (Program for Supercomputing) is thanked for a grant of computer time.

■

REFERENCES

(1) Glover, S. A. Tetrahedron 1998, 54, 7229−7271. (2) Schweitzer-Stenner, R. Vibr. Spectrosc. 2006, 42, 98−117. (3) Pauling, L., The Nature of the Chemical Bond: Cornell University Press: Ithaca, NY, 1960; p 260. (4) Samdal, S. J. Mol. Struct. 1998, 440 (1−3), 165−174. (5) Marstokk, K.-M.; Møllendal, H. J. Mol. Struct. 1974, 22 (2), 287− 300. (6) Samdal, S.; Seip, R. J. Mol. Struct. 1979, 52 (2), 195−210. (7) Møllendal, H.; Samdal, S. J. Phys. Chem. A 2006, 110 (6), 2139− 2146. (8) Samdal, S.; Seip, R. J. Mol. Struct. 1980, 62, 131−141. (9) Gundersen, S.; Samdal, S.; Seip, R.; Shorokhov, D. J. J. Mol. Struct. 1999, 477 (1−3), 225−240. (10) Gundersen, S.; Samdal, S.; Seip, R.; Strand, T. G. J. Mol. Struct. 2004, 691 (1−3), 149−158. (11) Gundersen, S.; Novikov, V. P.; Samdal, S.; Seip, R.; Shorokhov, D. J.; Sipachev, V. A. J. Mol. Struct. 1999, 485−486, 97−114. (12) Gundersen, S.; Samdal, S.; Seip, R.; Shorokhov, D. J.; Strand, T. G. J. Mol. Struct. 1998, 445 (1−3), 229−242. (13) Samdal, S.; Seip, R. J. Mol. Struct. 1997, 413−414, 423−439. (14) Marstokk, K. M.; Møllendal, H.; Samdal, S. J. Mol. Struct. 1996, 376, 11−24. (15) Samdal, S.; Møllendal, H. J. Phys. Chem. A 2003, 107 (42), 8845−8850. (16) Marstokk, K.-M.; Møllendal, H.; Samdal, S. J. Mol. Struct. 2000, 524, 69−85. (17) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. 1999, 53 (2), 79−84. (18) Bakri, B.; Demaison, J.; Kleiner, I.; Margulès, L.; Møllendal, H.; Petitprez, D.; Wlodarczak, G. J. Mol. Spectrosc. 2002, 215 (2), 312− 316. (19) Marstokk, K.-M.; Møllendal, H.; Samdal, S. Acta Chem. Scand. 1996, 50 (9), 845−847. (20) Marstokk, K.-M.; Møllendal, H.; Samdal, S.; Uggerud, E. Acta Chem. Scand. 1989, 43 (4), 351−363. (21) Demyk, K.; Petitprez, D.; Demaison, J.; Møllendal, H.; Wlodarczak, G. Phys. Chem. Chem. Phys. 2003, 5, 5038−5043. (22) Hirota, E.; Sugisaki, R.; Nielsen, C. J.; Sørensen, G. O. J. Mol. Spectrosc. 1974, 49, 251−267. (23) Brown, R. D.; Godfrey, P. D.; Kleibömer, B. J. Mol. Spectrosc. 1987, 124 (1), 34−95. (24) McNaughton, D.; Evans, C. J.; Lane, S.; Nielsen, C. J. J. Mol. Spectrosc. 1999, 193, 104−117. (25) Brown, R. D.; Godfrey, P. D.; Storey, J. J. Mol. Spectrosc. 1975, 58, 445−50. (26) Kawashima, Y.; Usami, T.; Suenram, R. D.; Golubiatnikov, G. Y.; Hirota, E. J. Mol. Spectrosc. 2010, 263, 11−20. (27) Schultz, G.; Hargittai, I. J. Phys. Chem. 1993, 97, 4966−4969. (28) Ohba, K.; Usami, T.; Kawashima, Y.; Hirota, E. J. Mol. Struct. 2005, 744−747, 815−819. (29) Styger, C.; Caminati, W.; Ha, T. K.; Bauder, A. J. Mol. Spectrosc. 1991, 148, 494−505. (30) Ottaviani, P.; Melandri, S.; Maris, A.; Favero, P. G.; Caminati, W. J. Mol. Spectrosc. 2001, 205, 173−176.

The MW spectra of a mixture of cis- and trans-HNCO forms of N-vinylformamide, (H2CCHNHC(O)H) have been measured and analyzed. Rotational isomerism is possible for both cis and trans. However, the MW spectra of only one form of cis and one form of trans were assigned. Both these forms have the CCNC chain of atoms in an antiperiplanar conformation. The behavior of the inertial defects, as well as CCSD(T)/ccpVTZ calculations, of the two forms strongly indicates that the equilibrium structures of both forms are exactly planar. The experimental work has been augmented with quantum chemical calculations at the B3LYP/cc-pVTZ and the very high CCSD(T)/cc-pVTZ level of theory. The two theoretical methods predict that the identified forms are both planar. The CCSD(T) rotational constants are in excellent agreement with their experimental counterparts. The B3LYP quartic centrifugal distortion constants and the vibration−rotation constants are in fair agreement with experiments, whereas larger discrepancies are seen for the CCSD(T) dipole moments. The theoretical calculations predict that an additional rotameric form exists for each of cis and trans with an energy that is several kJ/mol higher than those of the identified forms. However, searches for the spectra of these two hypothetical conformers were unsuccessful. It is estimated that the additional rotamers must be at least 4 kJ/mol less stable than the identified forms, if they exist at all.

S Supporting Information *

Results of the theoretical calculations, including molecular structures; geometrical parameters; dipole moments; harmonic frequencies, IR intensities, Raman scattering activities, depolarization ratios, reduced masses, and force constants; and vibrational energies and rotational constants. Microwave spectra. This material is available free of charge via the Internet at http://pubs.acs.org. 12080

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081

The Journal of Physical Chemistry A

Article

(31) Aviles, M. J. R.; Huet, T. R.; Petitprez, D. J. Mol. Struct. 2006, 780−781, 234−237. (32) Kitano, M.; Kuchitsu, K. Bull. Chem. Soc. Jpn. 1973, 46, 3048− 3051. (33) Yamaguchi, A.; Hagiwara, S.; Odashima, H.; Takagi, K.; Tsunekawa, S. J. Mol. Spectrosc. 2002, 215 (1), 144−154. (34) Ilyushin, V. V.; Alekseev, E. A.; Dyubko, S. F.; Kleiner, I.; Hougen, J. T. J. Mol. Spectrosc. 2004, 227, 115−139. (35) Hirota, E.; Kawashima, Y.; Usami, T.; Seto, K. J. Mol. Spectrosc. 2010, 260, 30−35. (36) Kitano, M.; Fukuyama, T.; Kuchitsu, K. Bull. Chem. Soc. Jpn. 1973, 46, 384−387. (37) Kawashima, Y.; Suenram, R. D.; Hirota, E. J. Mol. Spectrosc. 2003, 219, 105−118. (38) Caminati, W.; Maris, A.; Millemaggi, A. New J. Chem. 2000, 24, 821−824. (39) Lavrich, R. J.; Farrar, J. O.; Tubergen, M. J. J. Phys. Chem. A 1999, 103, 4659−4663. (40) Demaison, J.; Csaszar, A. G.; Kleiner, I.; Møllendal, H. J. Phys. Chem. A 2007, 111 (13), 2574−2586. (41) Demaison, J.; Dubrulle, A.; Boucher, D.; Burie, J. J. Mol. Spectrosc. 1982, 94, 211−214. (42) Aparicio-Martinez, S.; Hall, K. R.; Balbuena, P. B. J. Phys. Chem. A 2006, 110, 9183−9193. (43) Møllendal, H.; Leonov, A.; de Meijere, A. J. Phys. Chem. A 2005, 109 (28), 6344−6350. (44) Møllendal, H.; Cole, G. C.; Guillemin, J.-C. J. Phys. Chem. A 2006, 110 (3), 921−925. (45) Samdal, S.; Møllendal, H.; Hnyk, D.; Holub, J. J. Phys. Chem. A 2011, 115, 3380−3385. (46) Wodarczyk, F. J.; Wilson, E. B., Jr. J. Mol. Spectrosc. 1971, 37 (3), 445−463. (47) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, Revision B.01, Gaussian, Inc: Wallingford, CT, 2010. (48) Werner, H.-J.; Knowles, P. J.; Knizia, G.; Manby, F. R.; Schütz, M.; et al. MOLPRO, version 2010.1, a package of ab initio programs, 2010; http://www.molpro.net/. (49) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37 (2), 785− 789. (50) Deegan, M. J. O.; Knowles, P. J. Chem. Phys. Lett. 1994, 227 (3), 321−326. (51) Peterson, K. A.; Dunning, T. H., Jr. J. Chem. Phys. 2002, 117 (23), 10548−10560. (52) McKean, D. C.; Craig, N. C.; Law, M. M. J. Phys. Chem. A 2008, 112 (29), 6760−6771. (53) Bailey, W. C. Calculation of Nuclear Quadrupole Coupling Constants in Gaseous State Molecules. http://nqcc.wcbailey.net/ (Accessed October 2011). (54) Hirota, E.; Endo, Y.; Saito, S.; Yoshida, K.; Yamaguchi, I.; Machida, K. J. Mol. Spectrosc. 1981, 89 (1), 223−231. (55) Ray, B. S. Z. Phys 1932, 78, 74−91. (56) Marstokk, K.-M.; Møllendal, H. J. Mol. Struct. 1969, 4, 470−472. (57) Sørensen, G. O. J. Mol. Spectrosc. 1967, 22, 325−346. (58) Esbitt, A. S.; Wilson, E. B. Rev. Sci. Instrum. 1963, 34 (1963), 901−907. (59) Herschbach, D. R.; Laurie, V. W. J. Chem. Phys. 1964, 40 (11), 3142−3153. (60) Oka, T. J. Mol. Struct. 1995, 352/353, 225−233. (61) Hanyu, Y.; Britt, C. O.; Boggs, J. E. J. Chem. Phys. 1966, 45 (12), 4725−4728. (62) Muenter, J. S. J. Chem. Phys. 1968, 48 (10), 4544−4547. (63) Golden, S.; Wilson, E. B., Jr. J. Chem. Phys. 1948, 16, 669−685. (64) Watson, J. K. G. Vibrational Spectra and Structure; Elsevier: Amsterdam, 1977; Vol. 6, pp 1−89.

12081

dx.doi.org/10.1021/jp309552m | J. Phys. Chem. A 2012, 116, 12073−12081