Far-Infrared Synchrotron Spectroscopy and Torsional Analysis of the

Feb 13, 2017 - We report far-infrared spectra (100–600 cm–1) covering the torsional bands of the important interstellar molecule, vinyl alcohol. W...
1 downloads 4 Views 3MB Size
Article http://pubs.acs.org/journal/aesccq

Far-Infrared Synchrotron Spectroscopy and Torsional Analysis of the Important Interstellar Molecule, Vinyl Alcohol Hayley Bunn,† Rohan J. Hudson,† Alexander S. Gentleman,†,‡ and Paul L. Raston*,†,§ †

Department of Chemistry, The University of Adelaide, Adelaide, South Australia 5005, Australia Department of Chemistry and Biochemistry, James Madison University, Harrisonburg, Virginia 22807, United States

§

S Supporting Information *

ABSTRACT: We report far-infrared spectra (100−600 cm−1) covering the torsional bands of the important interstellar molecule, vinyl alcohol. We observed the fundamental and first two hot bands of syn-vinyl alcohol and the fundamental and first hot band of anti-vinyl alcohol, whose infrared spectrum has not been previously reported. The hot bands for syn-vinyl alcohol were incrementally shifted by ∼40 cm−1 from the fundamental, implying a high degree of anharmonicity in the torsional potential in the vicinity of the global minimum. The opposite holds true for anti-vinyl alcohol, as evidenced by the relatively small red shift (∼2 cm−1) of the hot band with respect to the fundamental. These characteristics were captured by ab initio calculations of the torsional potential, which we fit a seven-term Fourier series expansion to, and used along with the calculated torsional inertias to solve for the energy levels and wave functions. The potential was empirically refined by fitting the three leading terms in the expansion to the observed torsional frequencies, which resulted in a 3-fold decrease in the root-mean-square deviation between theory and experiment. The resulting energy difference between the ground states of the two rotamers is 4.0 kJ/mol, which is approximately consistent with the measured intensity ratio of the two ground-state torsional bands and with microwave spectroscopic measurements (4.5 ± 0.6 kJ/mol) [Rodler, M. Microwave Spectrum, Dipole Moment, and Structure of anti-Vinyl Alcohol. J. Mol. Spectrosc. 1985, 114, 23−30, DOI: 10.1016/0022-2852(85)90332-7]. Using the empirically refined potential and reported column densities of syn- and anti-vinyl alcohol in Sagittarius B2(N) (Turner, B. E.; Apponi, A. J. Microwave Detection of Interstellar Vinyl Alcohol, CH2CHOH. Astrophys. J. 2001, 561, L207−L210, DOI: 10.1086/324762), we determined the conformer temperature of vinyl alcohol to be ∼230 K. Important parameters in determining the column densities are the dipole moments, which were found to undergo a relatively large (small) amount of torsional averaging for anti(syn-) vinyl alcohol. KEYWORDS: astrochemistry, internal rotation, ISM, potential energy surface, CH2CHOH

1. INTRODUCTION It was proposed by Erlenmeyer over a century ago that alcohols that contain a CC bond adjacent to an O−H group (i.e., enols) immediately tautomerize to the corresponding aldehyde.1 Vinyl alcohol is the simplest of the enols (also known as ethenol) and was first observed by nuclear magnetic resonance (NMR) spectroscopy following the photolysis of acetaldehyde or acetoin in a variety of nonpolar solvents.2 A few years later, Saito observed the microwave spectrum of vinyl alcohol, which was produced from the pyrolysis of ethylene glycol.3 He found that it is stable in the gas phase for extended periods, with a half-life of ∼30 min in a Pyrex container at room temperature.3 In that study, it was proposed that vinyl alcohol should be observable in the interstellar medium (ISM).3 Although predicted to be present in the ISM some 25 years earlier,3 vinyl alcohol was not observed until 2001 by Turner and Apponi, who detected its microwave transitions in emission toward the dense molecular cloud Sagittarius B2(N).4 However, its more thermodynamically and kinetically stable © 2017 American Chemical Society

tautomers, acetaldehyde and oxirane, respectively, have been known to exist in the ISM for some time.5,6 It has been pointed out that the more stable tautomers of vinyl alcohol (acetaldehyde and oxirane) are commonly observed in hotter regions of the ISM, perhaps implying a significant energy barrier for the synthesis of these species and/or conversion from vinyl alcohol.4 All three are of considerable interest particularly as a result of their astrobiological importance7,8 and may play an important role in understanding the synthesis and evolution of more complex organics observed on Earth.9 The formation of vinyl alcohol and its tautomers in the ISM has been extensively studied and is of considerable discussion.9 Calculations have shown that vinyl alcohol could be synthesized in gas-phase reactions from known interstellar molecules;10 Received: Revised: Accepted: Published: 70

November 25, 2016 February 8, 2017 February 13, 2017 February 13, 2017 DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry

Figure 1. Medium-resolution (0.01 cm−1) far-infrared spectrum of vinyl alcohol following the pyrolysis of 2-chloroethanol at ∼950 °C.

2. EXPERIMENTAL SECTION The high-resolution Fourier transform infrared (FTIR) spectrum of vinyl alcohol was obtained at the far-infrared beamline of the Australian Synchrotron. The sample was produced from the pyrolysis of 2-chloroethanol at 950 °C. As indicated by Joo et al., the use of this precursor minimizes the formation of interfering hydrocarbon absorptions.19 The gaseous sample was introduced into a multiple reflection gas cell (White type20), which was held at room temperature and a pressure of ∼0.5 or 2 Torr. We found it necessary to continuously flow the sample through the cell to minimize interfering absorptions, which presumably result from the back reaction between vinyl alcohol and hydrogen chloride. The setup of the spectrometer was similar to what we used previously.21 Briefly, the synchrotron radiation was coupled to a Bruker IFS 125 HR FTIR spectrometer equipped with a Mylar beam splitter and polyethylene windows. After the beam passed through the interferometer, it was directed into the gas cell and detected with a liquid-helium-cooled silicon bolometer. This setup allowed for maximum light intensity in the far-infrared region, particularly around 250 cm−1, wherein lies bands of antivinyl alcohol. The spectra were recorded at the highest resolution possible (0.000 96 cm−1), with a total path length of 10 m (16 passes).

however, they require the use of reactants in high abundance and some intermediates that are yet to be observed in the ISM.9 Hudson and Moore showed that vinyl alcohol is formed from the simulated cosmic ray bombardment or far-ultraviolet irradiation of synthetic ice composed of water and acetylene.11 The excess energy following reaction could lead to desorption of vinyl alcohol into the gas phase. The dominant mechanism(s) leading to the formation of vinyl alcohol, however, seems to be unclear, because experiments are as of yet unable to reproduce the observed relative abundances of C2H4O.9 Vinyl alcohol can exist in one of two conformations in relation to the orientation of the hydroxyl group with respect to the rest of the molecule; these conformers or more specifically rotamers are referred to as syn- and anti-vinyl alcohol, where the CC−O−H dihedral angles are 0° and 180°, respectively (see inset of Figure 1). In the original microwave experiments, it was the lower energy syn rotamer that was observed.3 Extended microwave measurements covering more transitions and additional isotopologues yielded a complete set of structural parameters12 that were further refined with the help of theory.13 Rodler first observed the microwave spectrum of anti-vinyl alcohol,14 which was predicted to be the higher energy rotamer of vinyl alcohol by ab initio molecular orbital theory.15 Infrared spectra of the syn rotamer were first observed by rare gas matrix isolation experiments16,17 and then in the gas phase at mediumand then high-resolution in the mid-infrared.18,19 These investigations led to the identification of 9 of the 15 fundamental vibrational bands between 700 and 3700 cm−1. In the far-infrared spectral region investigated here, the two lowest frequency vibrational bands, ν15 (OH torsion) and ν11 (CCO bend), are located. This paper reports the first far-infrared spectrum of vinyl alcohol in the gas phase, where absorption as a result of both the syn- and anti-vinyl alcohol rotamers has been detected. This is the first reported observation of anti-vinyl alcohol by infrared spectroscopy, and the observed torsional bands are used to refine ab initio calculations of the torsional potential. This torsional potential can be used along with observational data to determine the conformational temperature of vinyl alcohol in the ISM. This is important because the conformational temperature may provide important clues as to the astrophysical synthesis of vinyl alcohol.

3. CALCULATIONS 3.1. Ab Initio Anharmonic Calculations. The equilibrium structures of both syn- and anti-vinyl alcohol were obtained from geometry optimization at the coupled cluster level of theory incorporating single, double, and perturbative triple excitation contributions [CCSD(T)] coupled with a correlation consistent quadruple ζ basis set (cc-pVQZ). To conveniently introduce second-order vibrational perturbation theory (VPT2) for anharmonic frequency calculations, a second-order Møller− Plesset (MP2) level of theory was used. Because the Gaussian 09 electronic structure package22 only provided anharmonic fundamental and first hot band frequencies, we performed additional calculations to obtain torsional state energies to predict warmer hot bands (see the following). 3.2. Internal Rotation Hamiltonian. Quantum chemical optimization calculations were performed on vinyl alcohol at every 5° about the CC−O−H dihedral (torsional) coordinate, also using Gaussian 09.22 They were initially 71

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry

Figure 2. Medium-resolution (0.01 cm−1) spectra of (top) syn-vinyl alcohol and (bottom) anti-vinyl alcohol, highlighting the fundamental and hot bands of each rotamer. Several p/rQ bandheads are labeled for both the fundamental (black arrows) and first hot band (red arrows) that were used to help with the band assignments; they were identified with the help of the ground-state rotational constants.3,14

spectrum shown in Figure 1. This is easily assignable to synvinyl alcohol based on a comparison to the anharmonic frequency calculations (which predict 406.2 cm−1) and matrix isolation experiments (∼413 cm−1 in solid argon16,17). The formation of vinyl alcohol is accompanied by the production of HCl, and its pure rotational lines are evident below 300 cm−1. In addition to the ν15 fundamentals, we observed the very weak ν11 fundamental of syn-vinyl alcohol, which is an a′ hybrid a/b-type band corresponding to the CCO bending vibration. Its band center is at ∼488.7 cm−1, which is relatively close to the calculated anharmonic value (495.4 cm−1). We also observed the ν10 band of acetaldehyde near 509 cm−1,29 indicating that vinyl alcohol somewhat tautomerizes, and we suspect that this most likely occurs in the furnace. We note that absorptions corresponding to the precursor, 2chloroethanol, are not evident, indicating near complete pyrolysis. One of the most interesting features of this work is the infrared observation of the anti rotamer of vinyl alcohol. Similar to the syn rotamer, it is also observed via its ν15 torsional band, with a prominent central Q-branch peak at 261.77 cm−1, which compares reasonably well to the calculated anharmonic value (248.5 cm−1). It is accompanied by a weaker band slightly to the red that we assign to the torsional hot band. It is situated at 259.41 cm−1, which is also in fair agreement with the calculated value (266.7 cm−1). Similarly, we observed hot bands of the syn rotamer, the first band of which lies at 368.53 cm−1, which is very close to the calculated value (371.9 cm−1), and the second band is shifted by an equally large amount to the red at 323.32 cm−1. While the anharmonic frequency calculations do not provide a prediction for the second hot band, potential energy function calculations do (vide inf ra) and are in excellent agreement all around. Further support of the band assignments to vinyl alcohol comes from the reasonable spacings between the ΔK = ±1 sub-branches for syn- and anti-vinyl alcohol (see Figure 2), comparable to 2(A − B̅ ). To confirm our assignments, we analyzed the high-resolution spectra to make sure that the rotational substructure was in agreement with what is expected for vinyl alcohol.3,12,14,19,30 High-resolution spectra of all five ν15 torsional bands, i.e., the fundamental and first two hot bands of syn-vinyl alcohol along with the fundamental and first hot band of anti-vinyl alcohol,

performed at a lower level of theory (MP2) and then continued at a higher level [CCSD(T)] to save on computational time. Both methods were coupled with a correlation consistent triple ζ basis set (cc-pVTZ). Various molecular parameters obtained from the relaxed scan are provided in Table S1 of the Supporting Information. The following Fourier expansion was then fit to the potential energy as a function of the dihedral angle: V (ϕ) =

1 2

7

∑ Vn(1 − cos nϕ) n=1

(1)

The atomic masses and optimized coordinates as a function of the dihedral angle were used to calculate the Pitzer torsional inertias as previously outlined,23 and the following Fourier expansion was then fit to the associated internal rotation constants: 7

F(ϕ) =

∑ Fn cos nϕ n=0

(2)

The resulting potential energy and internal rotation functions were then inserted into the internal rotation Hamiltonian, which is given by H=−

d d + V (ϕ) F(ϕ) dϕ dϕ

(3)

We included 100 even/odd cosine/sine functions to solve the associated Schrödinger equation (HΨ = λΨ) using a modified version of an existing script.24 Further details of the methodology can be found elsewhere.25,26 We note that the results determined using the renormalized Numerov−Cooley technique27,28 are similar.

4. RESULTS AND DISCUSSION 4.1. Far-Infrared Spectroscopy. Vinyl alcohol is a planar asymmetric top with Cs symmetry and 15 vibrational modes, 11 of which are symmetric (a′) and 4 are asymmetric (a″). The ν15 fundamental that we focus on here is a relatively intense a″ (ctype) band corresponding to the OH torsional vibration. This band that features a strong central Q-branch peak is one of the most prominent features (at 407.16 cm−1) in the survey 72

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry

Figure 3. High-resolution (∼0.001 cm−1) spectra showing the central Q-branch region of the five torsional bands. The inverted traces correspond to the fitted spectra, for which we held the ground-state constants to their literature values and the upper state higher order distortion constants to their ground-state values.19 The simulated 3−2 band is shown as a dashed line because it is a very rough preliminary fit.

are shown in Figure 3 (covering a 0.8 cm−1 range centered around each respective band origin). We performed a rough fit of the band origins and excited-state rotational constants (A, B, and C) to the spectra, focusing on the displayed regions (Figure 3; see caption for further details). The fits were performed using the computer program PGOPHER31 and are of good quality, except for the 3−2 band, which could be due to unaccounted for Coriolis and/or Fermi interactions involving 3ν15 (also, see below). The corresponding rotational constants are plotted with respect to the number of vibrational quanta in Figure 4, which reveals a smooth progression that qualitatively agrees with theory, thus supporting the assignments. A detailed analysis of these bands will be presented elsewhere. It is worth noting that we considered the assignment of the non-fundamental bands to hot bands of ν15 originating from different vibrationally excited states. What largely argues against this for syn-vinyl alcohol is that the observed red shifts are incrementally about 40 cm−1 from the fundamental and our anharmonic calculations only predict such a large red shift for

2ν15−ν15 (34.3 cm−1). For comparison, while ν11 and ν15 are expected to have similar Boltzmann populations because they are close in energy, our calculations predict a 0.2 cm−1 shift for (ν11 + ν15)−ν11. For the anti rotamer, the observed red shift is only 2.28 cm−1, and therefore, we cannot confirm the 2ν15−ν15 assignment based on shift alone. Similar to syn-vinyl alcohol, the lowest excited state for anti-vinyl alcohol is ν15, which is followed by ν11. Unlike syn-vinyl alcohol, however, the ν11 state for anti-vinyl alcohol is very different in energy, being ∼2× that of ν15. On the basis of Boltzmann populations alone, we expect (ν11 + ν15)−ν11 to be ∼1/10 the intensity of the fundamental and 2ν15−ν15 to be ∼1/3 of the fundamental, the latter of which agrees with the experiment. Additional evidence that the 259.27 cm−1 band corresponds to 2ν15−ν15 comes from the near linear variation of the rotational constants in going up the torsional ladder (see Figure 4), in addition to the good quality fit we achieve using the ν15 constants for its lower state (see Figure 3). The band origins for the ν15 fundamentals and hot bands of both syn- and anti-vinyl alcohol are presented in Table 1. There, 73

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry

Figure 4. Fitted (black squares) and calculated (red or blue diamonds) rotational constants as a function of the number (n) of vibrational quanta (see the main text for details).

Table 1. Observed ν15 Vibrational Frequencies (cm−1) of syn- and anti-Vinyl Alcohol in the Far-Infrared, with Comparison to Theoretical Frequencies rotamer

ν′−ν″

observeda

ab initio functionb

empirically refined functiona

anharmonicc

harmonicc

syn

1−0 2−1 3−2 1−0 2−1 3−2

407.17 368.49 323.25 261.55 259.27

401.9 369.4 329.5 259.3 254.1 238.4

405.1 370.9 323.2 261.7 258.9 243.2

406.2 371.9

439.2 439.2 439.2 251.9 251.9 251.9

anti

248.5 266.7

a Band origins determined in the preliminary fits (see Figure 3). bDifference frequencies determined from the calculated vibrational levels using the appropriate double-well potential function (see the text). cComputationally determined frequencies at the MP2 level (see the text).

respectively. Several previous theoretical investigations have focused on mapping out the torsional potential of vinyl alcohol. Samdal and Seip reported an energy difference between the syn and anti minima of 8 kJ/mol and a barrier to interconversion (syn → anti) of 18 kJ/mol;32 Nobes et al. reported a range of 7−8 kJ/mol between the minima and a barrier of 18 kJ/mol, where higher levels of computational theory were found to lower the energy difference.33 More recently, there have been several computational investigations,34−36 the most sophisticated of which involved a focal point analysis (FPA) that predicts the minima to be 5.3 kJ/mol different with a barrier to interconversion of 20 kJ/mol.36 Here, we obtained a well

a comparison is made to the calculated harmonic and anharmonic frequencies, in addition to the predictions using the internal rotation Hamiltonian mentioned earlier (eq 3). The best all-around agreement is apparent between the observed and calculated frequencies using the internal rotation Hamiltonian (vide inf ra). Table 2 provides calculated rotational constants and percentage Boltzmann population factors at 300 K. 4.2. Torsional Analysis. Figure 5 shows the relaxed torsional potential (dots) and fitted Fourier series function (curve). The corresponding Vn coefficients from eq 1 are presented in Table 3. As expected, the potential shows minima at 0° and 180° corresponding to syn- and anti-vinyl alcohol, 74

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry

Table 2. Rotational Constants (A, B, and C, in cm−1), State Energies (E, in cm−1), and Corresponding Percentage Boltzmann Population Factors (at 300 K)a rotamer

stateb

A

B

C

Ec

population (%)

syn syn anti syn anti syn syn syn syn anti syn anti syn syn

GS ν15 GS ν11 ν15 ν14 2ν15 ν13 ν11 + ν15 ν11 ν10 2ν15 ν12 2ν11

1.98869 1.97882 2.09481 1.98812 2.06647 1.98899 1.96896 1.96473 1.97826 2.09069 2.00738 2.03813 1.98121 1.98756

0.35043 0.34849 0.34688 0.35125 0.34641 0.34990 0.34654 0.35000 0.34930 0.34712 0.34799 0.34594 0.35140 0.35206

0.29766 0.29782 0.29759 0.29722 0.29834 0.29778 0.29798 0.29801 0.29738 0.29728 0.29698 0.29910 0.29785 0.29678

0 406.21 444.10 495.41 692.65 718.71 778.15 836.23 901.42 919.41 951.13 959.34 984.02 993.99

63.32 9.03 7.53 5.88 2.28 2.02 1.52 1.15 0.84 0.77 0.66 0.64 0.56 0.54

a

The entries are determined from anharmonic MP2 calculations and are ordered from most populated to least populated for energies up to 1000 cm−1; for a complete list, see Table S3 of the Supporting Information. bGS = ground state. cE is the energy relative to the GS energy of the syn rotamer.

difference of 6.4 kJ/mol and barrier height of 21 kJ/mol, both of which fall within the range of those studies. The “rigid” and “relaxed” internal rotation constants are given in Figure 6, and the corresponding Fn coefficients from a

Figure 5. Ab initio potential function for vinyl alcohol (see Table 3 for Fourier series coefficients). The energy levels and probability distributions are shown for even (red) and odd (blue) symmetry states and were determined using the method of Laane and co-workers (see the text for further details).25,26 Observed torsional bands are shown by arrows. Figure 6. Internal rotation constants as a function of the dihedral angle for vinyl alcohol. The inset shows Frigid − Frelaxed (blue diamonds) and the O−H bond length, rO−H (green circles).

Table 3. Coefficients Extracted from the Fourier Series Fits to the Potential Energy and Internal Rotation Constants (in cm−1)

fit of eq 2 to the data are presented in Table 3. The plot reveals a decrease by 8 and 7%, respectively, in going from 0° (syn) to 180° (anti). Because these values are similar, we can be confident that structural relaxation upon internal rotation is not responsible for the large decreases. Rather, it is related to the decrease (increase) in alignment of the internal rotation axis with the b-axis (a-axis), which has a reducing effect on the “F number”.37,38 The relaxed internal rotation constants diverge from the rigid internal rotation constants in going from, in particular, ∼120° to a maximum difference of 1% at 180°. While structural relaxation must be responsible for this difference, after examining how the structural parameters vary with the CC−O−H dihedral angle (see Table S1 of the Supporting Information), it is apparent that a concerted variation of at least more than one of them is responsible.

potential energy parameters

V1 V2 V3 V4 V5 V6 V7 a

ab initio

empirically refined

317.360 1452.178 221.193 −35.155 −0.371 −2.072 0.431

168.4 1471.8 246.9 −35.155a −0.371a −2.072a 0.431a

internal rotation parameters F0 F1 F2 F3 F4 F5 F6 F7

23.545 0.955 0.297 −0.0608 −0.00878 −0.00578 0.00284 0.00170

Constrained to ab initio values.

75

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry

to the ν15 = 3−2 band. We do not focus on this phenomenon here though, because we would like to observe more bands, including those for other isotopologues (such as CH2CHOD) to help further refine the torsional potential.42 After solving the one-dimensional (1D) vibrational Schrödinger equation, we obtain a zero-point energy difference of 5.5 kJ/mol using the ab initio potential and 4.0 kJ/mol using the empirically refined potential. The empirically refined zero-point energy difference is in good agreement with the zero-point corrected FPA value (4.4 kJ/mol)36 and on the low end of the uncertainty range determined from microwave spectroscopy (4.5 ± 0.6 kJ/mol).14 The zero-point energy difference from microwave spectroscopy was determined using a modified Boltzmann-type analysis that includes the ratio of the fractional populations in the ground state of anti- and syn-vinyl alcohol.14,43 The OH torsional frequency of anti-vinyl alcohol was noted to cause the largest uncertainty in this fraction,14 because it is the lowest frequency vibration by a factor of ∼2. The measured OH torsional frequency for anti-vinyl alcohol reported here should allow for a more accurate determination of the energy difference by this method in the future (note that we cannot input it into the published calculation because the required line intensities and widths were not reported). The column density of anti-vinyl alcohol (Nanti = 2.4 × 1013 cm−2)4 and syn-vinyl alcohol (Nsyn = 2.0 × 1014 cm−2)4 can be used together with the zero-point energy difference (Δν)̅ to determine the conformational temperature of vinyl alcohol, using Nanti/Nsyn = e−hcΔν̅/kBT. Using the zero-point energy difference reported by Rodler,14 we obtain a conformer temperature of 255 K, and using the difference reported here, we obtain 227 K, both with relatively large uncertainties, which can be traced back to the large uncertainty in Nsyn.4 In either case, this is considerably warmer than the rotational temperature of vinyl alcohol in Sagittarius B2(N) (Trot ≈ 12 K), indicating that kinetic effects are important to consider in its formation. Perhaps the most likely one involves the synthesis of vinyl alcohol on the icy mantle of dust grains,4 followed by the non-thermal desorption into the gas phase. Several other conformational isomer pairs have been observed in the ISM, including ethanol (Orion KL),44 methyl formate [Sagittarius B2(N)],45 and ethyl formate (Orion KL).46 Of these, methyl formate has the largest difference between conformer temperature and kinetic temperature (usually taken as Trot), suggesting kinetic effects are also important to consider in its formation.45 Important parameters in determining column densities are the dipole moment components. We plot the evolution of the a- and b-axis dipole moment components (μa and μb) as a function of the torsional angle in Figure 7. At a CC−O−H dihedral angle of 180°, where the anti-vinyl alcohol minimum is located, the b-axis dipole moment is at 1.83 D, which is fairly consistent with the reported value of 1.702 D by Rodler.14 The a-axis dipole moment, however, has a magnitude that is ∼30% too high (0.79 D) in comparison to the experimental value (0.547 D14). The minimum corresponding to syn-vinyl alcohol, at a CC−O−H dihedral angle of 0°, has an a-axis dipole moment of 0.60 D and a b-type magnitude of 0.77 D, which are in reasonable agreement with the experimental values of 0.616 and 0.807 D, respectively.3 To gauge the effects of torsional averaging on the dipole moment components, we calculated their expectation values (⟨μa⟩ψ and ⟨μb⟩ψ) using the respective ground-state wave functions. This gave an overall better agreement between experiment and theory (μa and μb for anti = 0.72 and 1.77 and

We do, however, notice a rough correlation between the F number difference (Frigid − Frelaxed) and the O−H bond length (rO−H), as shown in the inset of Figure 6. Quite reasonably, this indicates that the turnaround and increase in Frelaxed in approaching the dihedral angle of 180° is related to a shortening of rO−H. The potential energy (Vn) and internal rotation (Fn) coefficients mentioned above were then used to numerically solve for eigenvalues on this double-well potential.25 Although the resulting ab initio transition frequencies (see Table 1) show good agreement with experiment, having a root-mean-square (RMS) deviation of 4.47 cm−1, we decided to try and come into better agreement by scaling the potential. However, we did not see significant improvement when applying a common scaling factor to the potential, as has been done for many systems, such as HOOO.39 Instead, we performed a fit of the three leading coefficients in the potential energy expansion (V1, V2, and V3).40 This resulted in a significant reduction of the RMS deviation (to 1.43 cm−1), and we label this potential as “empirically refined”. Note that, after refining this potential, it comes into much better agreement with the FPA values (calculated at 180°, ∼273°, and 360°; see Figure 7).36 In

Figure 7. Empirically refined (black squares) and ab initio (green diamonds) potential functions plotted with the a- and b-axis dipole moment components (in red and blue, respectively) as a function of the torsional angle. They were calculated at the CCSD(T) level of theory. The analogous values obtained at MP2 (same basis set) are not significantly different. The ground-state probability distribution for each rotamer (using the empirically refined potential) is also shown. The pink triangles correspond to previously reported ab initio energies extrapolated using the FPA method.36

particular, the V1 coefficient was found to be ∼2× too high in our calculations, which means that the difference in energy between syn- and anti-vinyl alcohol is overestimated at the level of theory used [CCSD(T)/cc-pVTZ]. We suspect this is because our basis set is not large enough to capture the finer details of the relatively weak intramolecular interactions, which is in line with computations of Allen and co-workers, who found a significant decrease in the energy difference with increasing basis set size.36 It is interesting to point out that, using the empirically refined potential, the ν15 = 3 states for anti- and syn-vinyl alcohol are within 10 cm−1 of each other, and this results in a large degree of wave function mixing (see Figure S3 of the Supporting Information). This has been reported for other systems, such as ethyl phosphine,41 and could result in anomalous ν15 = 3 rotational constants, which could explain the difficulties encountered when trying to do a preliminary fit 76

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry Table 4. Comparison of Experimental and Theoretical a- and b-Axis Dipole Moment Components (in Debye) equilibrium rotamer syn

anti

a

method

μa

μb

0.514 0.602

0.845 0.773

3

experimental MP2/cc-pVQZb CCSD(T)/cc-pVTZ experimental14 MP2/cc-pVQZb CCSD(T)/cc-pVTZ

0.679 0.791

1.713 1.833

vibrationally averaged

torsionally averageda

μa

μb

μa

μb

0.616 0.516

0.807 0.828 0.605

0.759

0.547 0.629

1.702 1.626 0.723

1.768

Determined using the empirically refined potential energy function (section 3.2). bFrom anharmonic frequency calculations (section 3.1).

resolution (0.000 96 cm−1), for which an in-depth analysis of the ν15 bands (which is currently underway) will provide rovibrational constants that could be used for future observations of torsionally warm vinyl alcohol in the ISM and planetary atmospheres.

μa and μb for syn = 0.61 and 0.76), although we note that there is room for improvement (see Table 4). Our results show that the effects of torsional averaging are small for syn-vinyl alcohol, as evidenced by the relatively flat variation in the dipole moment components over the largest amplitude parts of the ground-state probability distribution (see Figure 7). This is in contrast to anti-vinyl alcohol, for which there is a large variation in the dipole moment components over the extent of the probability distribution, which leads to large torsional averaging effects. It is interesting to note that complete vibrational averaging of the dipole moment components (over the 15 vibrational modes) has a similar effect to just torsional averaging them (over ν15), indicating that motion along the CC−O−H dihedral coordinate has by far the greatest effect on the dipole moment components (see Table 4). Vibrational averaging of the dipole moment is also known to be important in the related molecules, HCOOH,47 which has been detected in the ISM,48,49 and HOOO,24 which has been searched for in the ISM.50



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsearthspacechem.6b00008. Calculated parameters (Table S1) and atom-numbering scheme (Figure S1) for the relaxed scan, fitted preliminary rovibrational constants (Table S2), parameters from the anharmonic calculations (Table S3), plotted empirically refined potential function with three leading terms (Figure S2), and plotted empirically refined potential with eigenstates and probability distributions (Figure S3) (PDF)



5. SUMMARY AND OUTLOOK Here, we have reported the far-infrared spectrum of vinyl alcohol, which is dominated by the high infrared intensity torsional bands (ν15). For the syn rotamer, the ν15 fundamental and first two hot bands were found at 407.17, 368.49, and 323.25 cm−1, respectively. The large red shifting with increasing vibrational quantum number indicates a relatively strong anharmonicity in the well corresponding to the global minimum of the torsional potential. For the anti rotamer, the ν15 fundamental and first hot band were found at 261.55 and 259.27 cm−1, respectively, suggesting a harmonic-like potential in the well corresponding to the local minimum of the potential. The clamped coordinate (relaxed scan) approach was used to calculate an ab initio torsional potential and associated inertias that were fit to truncated Fourier series expansions. The parametrized expansions were inserted in the internal rotation Hamiltonian, which was solved using a set of free-rotor basis functions to determine the energy levels and wave functions. While the resulting energy level differences agreed within ∼6 cm−1 of the experiment, we were able to reduce that by more than a factor of 2 by fitting the first three coefficients in the potential energy expansion to the observed band origins. The resulting “empirically refined” potential comes into very good agreement with previously reported stationary-point energies extrapolated using the FPA method,36 and it was used to help determine a 227 K conformer temperature of vinyl alcohol in the ISM. Using the ground-state wave functions, we determined that the a- and b-axis dipole moment components undergo a large (small) amount of torsional averaging for syn- (anti-) vinyl alcohol. The spectra presented were recorded at a high-

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Paul L. Raston: 0000-0003-3717-4154 Present Address ‡

Alexander S. Gentleman: Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QZ, United Kingdom. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Acknowledgement is made to the donors of the American Chemical Society Petroleum Research Fund for partial support of this research (56406-UN16). The experiments reported here were undertaken at the High-Resolution Far-Infrared Beamline at the Australian Synchrotron, Victoria, Australia. The authors are grateful to Dom Appadoo for expert assistance and Greg Metha for helpful advice. This research used high-performance computing services provided by eRSA. We are grateful to the referees for providing constructive comments.



REFERENCES

(1) Erlenmeyer, E. Verhalten der Glycerinsäure und der Weinsäure Gegen Wasserentziehende Substanzen. Ber. Dtsch. Chem. Ges. 1881, 14, 320−323. (2) Blank, B.; Fischer, H. Chemically Induced Dynamic Nuclear Polarization XI. Intermediary Vinyl Alcohol During Photochemical 77

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry Reactions of Acetaldehyde and of Acetoin. Helv. Chim. Acta 1973, 56, 506−510. (3) Saito, S. Microwave Spectroscopic Detection of Vinyl Alcohol, CH2CHOH. Chem. Phys. Lett. 1976, 42, 399−402. (4) Turner, B. E.; Apponi, A. J. Microwave Detection of Interstellar Vinyl Alcohol, CH2CHOH. Astrophys. J. 2001, 561, L207−L210. (5) Dickens, J. E.; Irvine, W. M.; Ohishi, M.; Ikeda, M.; Ishikawa, S.; Nummelin, A.; Hjalmarson, A. Detection of Interstellar Ethylene Oxide (c-C2H4O). Astrophys. J. 1997, 489, 753−757. (6) Matthews, H. E.; Friber, P.; Irvine, W. M. The Detection of Acetaldehyde in Cold Dust Clouds. Astrophys. J. 1985, 290, 609−614. (7) Cleaves, H. J., II The Prebiotic Synthesis of Acrolein. Monatsh. Chem. 2003, 134, 585−593. (8) Hjalmarson, Å.; Bergman, P.; Nummelin, A. Progress in Searches for Prebiotic Interstellar Molecules. Exo-/Astro-Biology 2001, ESA SP496, 263−266. (9) Bennett, C. J.; Osamura, Y.; Lebar, M. D.; Kaiser, R. I. Laboratory Studies on the Formation of Three C2H4O IsomersAcetaldehyde (CH3CHO), Ethylene Oxide (c-C 2H4O), and Vinyl Alcohol (CH2CHOH)In Interstellar and Cometary Ices. Astrophys. J. 2005, 634, 698−711. (10) Basiuk, V. A.; Kobayashi, K. Formation of Interstellar Vinyl Alcohol via Simple Radical Processes: Theoretical Study. Int. J. Quantum Chem. 2004, 97, 713−718. (11) Hudson, R. L.; Moore, M. H. Solid-Phase Formation of Interstellar Vinyl Alcohol. Astrophys. J. 2003, 586, L107−L110. (12) Rodler, M.; Bauder, A. Structure of syn-Vinyl Alcohol Determined by Microwave Spectroscopy. J. Am. Chem. Soc. 1984, 106, 4025−4028. (13) Smith, B. J.; Radom, L. Structure of Vinyl AlcoholA Resolution of the Discrepancy between Theory and Experiment. J. Am. Chem. Soc. 1990, 112, 7525−7528. (14) Rodler, M. Micorwave Spectrum, Dipole Moment, and Structure of anti-Vinyl Alcohol. J. Mol. Spectrosc. 1985, 114, 23−30. (15) Pople, J. A.; Radom, L.; Hehre, W. J. Molecular Orbital Theory of the Electronic Structure of Organic Compounds. VII. Systematic Study of Energies, Conformations, and Bond Interactions. J. Am. Chem. Soc. 1971, 93, 289−300. (16) Hawkins, M.; Andrews, L. Reactions of Atomic Oxygen with Ethene in Solid Argon. The Infrared Spectrum of Vinyl Alcohol. J. Am. Chem. Soc. 1983, 105, 2523−2530. (17) Rodler, M.; Blom, C. E.; Bauder, A. Infrared Spectrum and General Valence Force Field of syn-Vinyl Alcohol. J. Am. Chem. Soc. 1984, 106, 4029−4035. (18) Koga, Y.; Nakanaga, T.; Sugawara, K.; Watanabe, A.; Sugie, M.; Takeo, H.; Kondo, S.; Matsumura, C. Gas Phase Infrared Spectrum of syn-Vinyl Alcohol Produced by Thermal Decomposition of Several Alcohols and Aldehydes. J. Mol. Spectrosc. 1991, 145, 315−322. (19) Joo, D.-L.; Merer, A. J.; Clouthier, D. J. High-Resolution Fourier Transform Infrared Spectroscopy of Vinyl Alcohol: Rotaional Analysis of the ν13 CH2 Wagging Fundamental at 817 cm−1. J. Mol. Spectrosc. 1999, 197, 68−75. (20) White, J. U. Long Optical Paths of Large Aperture. J. Opt. Soc. Am. 1942, 32, 285−288. (21) Bunn, H.; Bennett, T.; Karayilan, A.; Raston, P. L. Far-Infrared Spectroscopy of the H2−O2 van der Waals Complex. Astrophys. J. 2015, 799, 65. (22) 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.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;

Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009. (23) Wong, B. M.; Thom, R. L.; Field, R. W. Accurate Inertias for Large-Amplitude Motions: Improvements on Prevailing Approximations. J. Phys. Chem. A 2006, 110, 7406−7413. (24) Liang, T.; Magers, D. B.; Raston, P. L.; Allen, W. D.; Douberly, G. E. Dipole Moment of the HOOO Radical: Resolution of a Structural Enigma. J. Phys. Chem. Lett. 2013, 4, 3584−3589. (25) Laane, J. One-Dimensional Potential Energy Functions in Vibrational Spectroscopy. Q. Rev., Chem. Soc. 1971, 25, 533−552. (26) Lewis, J. D.; Malloy, T. B. J.; Chao, T. H.; Laane, J. Periodic Potential Functions for Pseudorotation and Internal Rotation. J. Mol. Struct. 1972, 12, 427−449. (27) Blatt, J. M. Practical Points Concerning the Solution of the Schrödinger Equation. J. Comput. Phys. 1967, 1, 382−396. (28) Johnson, B. R. New Numerical Methods Applied to Solving the One-Dimensional Eigenvalue Problem. J. Chem. Phys. 1999, 67, 4086− 4093. (29) Kleiner, I.; Moazzen-Ahmadi, N.; McKellar, A. R. W.; Blake, T. A.; Sams, R. L.; Sharpe, S. W.; Moruzzi, G.; Hougen, J. T. Assignment, Fit, and Theoretical Discussion of the ν10 Band of Acetaldehyde Near 509 cm−1. J. Mol. Spectrosc. 2008, 252, 214−229. (30) Hays, B. M.; Wehres, N.; DePrince, B. A.; Roy, A. A. M.; Laas, J. C.; Widicus Weaver, S. L. Rotational Spectral Studies of O(1D) Insertion Reactions with Methane and Ethylene: Methanol and Vinyl Alcohol in a Supersonic Expansion. Chem. Phys. Lett. 2015, 630, 18− 26. (31) Western, C. M. PGOPHER: A Program for Simulating Rotational, Vibrational and Electronic Spectra. J. Quant. Spectrosc. Radiat. Transfer 2017, 186, 221−242. (32) Samdal, S.; Seip, H. M. Potential for Rotation about C(sp2)−O and C(sp2)−S Bonds: Electron Diffraction Results for CH2CH− OCH3 and CH2CH−SCH3 and Ab-Initio Calculations on CH2 CH−OH and CH2CH−SH. J. Mol. Struct. 1975, 28, 193−203. (33) Nobes, R. H.; Radom, L.; Allinger, N. L. Equilibrium Conformations of Higher-Energy Rotational Isomers of Vinyl Alcohol and Methyl Vinyl Ether. J. Mol. Struct.: THEOCHEM 1981, 85, 185− 194. (34) da Silva, G.; Kim, C.-H.; Bozzelli, J. W. Thermodynamic Properties (Enthalpy, Bond Energy, Entropy, and Heat Capacity) and Internal Rotor Potentials of Vinyl Alcohol, Methyl Vinyl Ether, and Their Corresponding Radicals. J. Phys. Chem. A 2006, 110, 7925− 7934. (35) Tishchenko, O.; Ilieva, S.; Truhlar, D. G. Communication: Energetics of Reaction Pathways for Reactions of Ethenol with the Hydroxyl Radical: The Importance of Internal Hydrogen Bonding at the Transition State. J. Chem. Phys. 2010, 133, 021102. (36) Schreiner, P. R.; Reisenauer, H. P.; Ley, D.; Gerbig, D.; Wu, C. H.; Allen, W. D. Methylhydroxycarbene: Tunneling Control of a Chemical Reaction. Science 2011, 332, 1300−1303. (37) This can be understood by considering the conventional relation, F = ℏ2/2rIα, where the so-called reducing factor r = 1 − ∑gλg2Iα/Ig, ℏ is Dirac’s constant, λg is the direction cosine between the torsional axis and the principal inertial axes (g = a, b, or c), and Iα and Ig are the moments of inertia of the internal rotor and whole molecule, respectively. (38) Gordy, W.; Cook, R. L. Microwave Molecular Spectra; Wiley: New York, 1984; Chapter 12. (39) Beames, J. M.; Lester, M. I.; Murray, C.; Varner, M. E.; Stanton, J. F. Analysis of the HOOO Torsional Potential. J. Chem. Phys. 2011, 134, 044304. (40) We attempted to include V4 in the fit; however, we did not see a significant improvement in the RMS deviation. (41) Groner, P.; Johnson, R. D.; Durig, J. R. Reinvestigation of the Asymmetric Torsional Potential Function in Ethylphosphine. J. Mol. Struct. 1986, 142, 363−366. 78

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79

Article

ACS Earth and Space Chemistry (42) Part of the reason why we did not augment the ab initio potential by including zero-point energy corrections over the other 14 normal modes is so that it would be useful for all isotopologues of vinyl alcohol. Note that such corrections should be minor in light of the relatively small difference of the sum of the zero-point vibrational energies for the modes orthogonal to ν15 between the syn and anti anti syn −1 − ∑i 14 at the MP2/ccrotamers [(∑i 14 = 1ω̅ i = 1ω̅ i )/2 = 16.5 cm pVQZ level]. (43) Kroto, H. W. Molecular Rotation Spectra; Wiley: London, U.K., 1975. (44) Pearson, J. C.; Sastry, K. V. L. N.; Herbst, E.; De Lucia, F. C. Gauche Ethyl Alcohol: Laboratory Assignments and Interstellar Identification. Astrophys. J. 1997, 480, 420−431. (45) Neill, J. L.; Muckle, M. T.; Zaleski, D. P.; Steber, A. L.; Pate, B. H.; Lattanzi, V.; Spezzano, S.; McCarthy, M. C.; Remijan, A. J. Laboratory and Tentative Interstellar Detection of Trans-Methyl Formate Using the Publicly Available Green Bank Telescope PRIMOS Survey. Astrophys. J. 2012, 755, 153. (46) Tercero, B.; Kleiner, I.; Cernicharo, J.; Nguyen, H. V. L; López, A.; Caro, G. M. M. Discovery of Methyl Acetate and Gauche Ethyl Formate in Orion. Astrophys. J., Lett. 2013, 770, L13. (47) Paulson, L. O.; Kaminsky, J.; Anderson, D. T.; Bour, P.; Kubelka, J. Theoretical Study of Vibrationally Averaged Dipole Moments for the Ground and Excited CO Stretching States of Trans-Formic Acid. J. Chem. Theory Comput. 2010, 6, 817−827. (48) Zuckerman, B.; Ball, J. A.; Gottlieb, C. A. Microwave Detection of Interstellar Formic Acid. Astrophys. J. 1971, 163, L41−L45. (49) Cuadrado, S.; Goicoechea, J. R.; Roncero, O.; Aguado, A.; Tercero, B.; Cernicharo, J. Trans−Cis Molecular Photoswitching in Interstellar Space. Astron. Astrophys. 2016, 596, L1. (50) Zou, L. Y.; Hays, B. M.; Weaver, S. L. W. Weakly Bound Clusters in Astrochemistry? Millimeter and Submillimeter Spectroscopy of Trans-HO3 and Comparison to Astronomical Observations. J. Phys. Chem. A 2016, 120, 657−667.

79

DOI: 10.1021/acsearthspacechem.6b00008 ACS Earth Space Chem. 2017, 1, 70−79