Far-Infrared Synchrotron Spectroscopy and Quantum Chemical

Subscriber access provided by TULANE UNIVERSITY

A: Spectroscopy, Molecular Structure, and Quantum Chemistry

Far-Infrared Synchrotron Spectroscopy and Quantum Chemical Calculations of the Potentially Important Interstellar Molecule, 2-Chloroethanol Rebekah M. Soliday, Hayley Bunn, Isaiah Sumner, and Paul L Raston J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.8b11333 • Publication Date (Web): 16 Jan 2019 Downloaded from http://pubs.acs.org on January 17, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Far-Infrared Synchrotron Spectroscopy and Quantum Chemical Calculations of the Potentially Important Interstellar Molecule, 2Chloroethanol Rebekah M. Soliday1, Hayley Bunn2,a, Isaiah Sumner1, and Paul L. Raston1,2,* 1Department

of Chemistry and Biochemistry, James Madison University, Harrisonburg, Virginia 22807, USA

2Department

a)Present

of Chemistry, University of Adelaide, SA 5005, Australia

address: Department of Chemistry, Emory University, Atlanta, Georgia, 30322, USA

*Author

to whom correspondence should be addressed. Electronic mail: [email protected]

Abstract The high brightness of the Australian synchrotron allowed for detailed spectra to be collected at high resolution (0.00096 cm-1) in the vicinity of the a/b/c-type 19 band of 2-chloroethanol, which involves O-H torsional motion about the C-O bond. A rovibrational analysis was performed for both chlorine isotopologues in the 19 fundamental (centred at ~344 cm-1) which involved the assignment of 7153 lines (J ≤ 90, Ka ≤ 41). A global fit to these lines in addition to 119 microwave lines (J ≤ 29, Ka ≤ 11) led to the determination of spectroscopic constants up to the sextic level in both the ground and excited states using Watson’s Areduction Hamiltonian. The constants agree well with those calculated at the anharmonic MP2/cc-pVTZ level and allow for spectroscopically accurate predictions of rotational transitions in the ground vibrational state to be made over a broad range of rotational energies (TR100 K),14 or escape following energetic processing.15 These routes could foreseeably lead to the production of gas phase chlorinated molecules (which have low volatility). While only five neutral chlorine containing molecules have been detected in molecular clouds thus far (HCl9, CH3Cl16, AlCl, KCl, and NaCl17), several others have been (unsuccessfully) searched for, including ClO, CCl18, NH2Cl19, and ClCN20 (note that several interstellar lines have very tentatively been assigned to others21). Although 2CE is larger than these species (and so has a less favourable partition function), the intramolecular Cl···HO bond has a stabilizing effect, which could reduce its relative reactivity.

While quantitatively estimating the abundance of interstellar 2CE would be quite involved, we can make a qualitative estimate based on comparison to other species. If we assume that the methyl chloride/2CE abundance ratio is the same as that for methanol/ethylene glycol in various sources, then prospects for detecting 2CE are somewhat promising. For example, the methanol/ethylene glycol ratio is ~100 towards the protostellar binary, IRAS 16293-2422,22 where CH3Cl was detected,16 and while it is much lower in the “tail” of comet 67P/Churyumov-Gerasimenko23 (where methyl chloride was also detected), it has been observed in others such as Hale-Bopp with a much more favorable methanol/ethylene glycol abundance ratio (~10; see Ref. 24). Naturally, one would only expect similar ratios if each pair is connected by similar chemical transformations in astrophysical environments. On this end, Fedoseev et al. suggested that ethylene glycol can be efficiently produced from the association of two HCO radicals which is followed by sequential H atom addition reactions.25 In their scheme, methanol is also produced following sequential H atom addition to HCO. We speculate that an analogous scheme could be of relevance here that involves HCO and HCCl, which would result in the formation of both 2CE (from association of HCO and HCCl followed by H atom addition reactions) and methyl chloride (following H atom additions to HCCl). In order to accurately predict line positions that are required for searches of 2CE in interstellar molecular clouds to be made, accurate spectroscopic constants are required, and by analysing

-3-

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the torsional spectrum collected here, we are able to significantly improve on the number and accuracy of the constants for 2CE. Torsional bands are well suited for investigation by synchrotron spectroscopy since they have relatively low energies that fall within well suited regions of the power spectra of far-infrared synchrotron sources. The analysis of the farinfrared synchrotron spectra of chlorine containing molecules presents a challenge due to the increased number of bands that can lead to somewhat congested spectra;26-35 this arises from the two naturally occurring isotopes of chlorine (35Cl/37Cl≈3). In the following we analyse the 19 fundamental for both isotopologues of 2CE, and complement the results with ab initio

calculations. 2. Experimental 2.1. Spectroscopy The high resolution Fourier transform infrared (FTIR) spectrum of 2CE was collected at the far-infrared beamline of the Australian Synchrotron. The far-infrared radiation generated from deflecting the (200 mA) electron beam is directed into a Bruker IFS 125 HR FTIR spectrometer that houses a room temperature glass sample cell containing a White cell (path length at 12 passes = 6.6 m).36 Similar to what we previously used,37-39 the instrument was equipped with a Mylar beamsplitter, polyethylene windows, and liquid helium cooled Si bolometer detector. This set up allowed for high spectral brightness around 350 cm–1, wherein lies the 19 band of 2CE. Individual spectra were acquired at maximum resolution (unapodized resolution = 0.00096 cm–1), and calibrated with suitable residual H2O lines in the HITRAN database.40 The reported absorption spectrum was ratioed against a background spectrum that was collected at medium resolution. The 2CE (99%) was obtained from Sigma Aldrich, and the volatile impurities were removed by pumping on each sample for ~10 min prior to use. While 2CE has a vapour pressure of 5 Torr at 20oC,41 we recorded the high resolution spectra at a reduced pressure (~0.5 Torr) to avoid saturation effects. The lower resolution spectra (Figure 1) were recorded at a somewhat higher pressure. 2.2. Calculations

-4-


Page 4 of 29

Page 5 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


We used the Gaussian 09 package for the ab initio calculations.42 They were done in two parts: 1) Spectroscopic parameters: The geometry optimization of the lowest energy rotamer of 2CE was done at the CCSD(T) level with a triple ζ, cc-pVTZ basis set. Anharmonic frequency calculations (for both isotopologues) were then performed at the MP2/cc-pVTZ level. 2) Reaction coordinates: Geometry optimizations were performed using both cc-pVTZ and aug-cc-pVTZ basis sets at the MP2 level. The minimum energy pathways (MEPs) were calculated at the MP2/cc-pVTZ level by following the intrinsic reaction coordinate (IRC).43-44 The calculations that included a water-like solvent were performed using the SMD implicit solvation method.45 3. Results and discussion 3.1. Spectroscopic analysis 2CE is a near prolate top (asymmetry parameter  = (2B-A-C)/(A-C) = -0.896) for the normal isotopologue), and its ground state conformer belongs to the trivial C1 point group. We label the normal modes according to C1, and wish to point out that previously the Cs point group was used for labelling the vibrational bands for convenience.6 2CE has 21 vibrational modes, ~1/4 of which fall in the far-infrared region as shown in Figure 1; the fundamental bands from low to high energy are C-C torsion (21), CCl bend (20), C-O torsion (19), and CCO bend (18), all of which show some rotationally resolved substructure. In addition to the band assignments previously reported by Durig et al.,6 we made some new assignments through comparison with the anharmonic frequency calculations reported here. Most notably, we were able to identify the 18 and 17 bands of the minor isotopologue of Cl (37ClCH2CH2OH), and assign vibrational quantum numbers to a Fermi resonance peak (for the normal isotopologue) in the mid-infrared (see Table S1 in the Supplementary Material (SM)). A comparison of experimental and calculated band centers in the far-infrared are listed in Table 1 and in the mid-infrared are listed in Table S1 of the SM. It is interesting to note that the anharmonic calculations predict the O-H stretch and C-O torsion very well (3649.01 cm-1 and 340.875 cm1),

while scaled harmonic values (using the recommended scaling factor of 0.953) show good

-5-



agreement for the stretch and poor agreement for torsion (3649.99 cm-1 and 365.952 cm-1), which is naturally a consequence of the way typical scaling factors are determined.46 Out of the observed bands, we found that the 19 fundamental has the most detailed band profile with obvious spectral patterns, and is well separated from associated hot bands. For these reasons, and so we could improve upon the ground state constants, we focus on 19 in this work. The 19 vibration is best described as the local mode torsion of the O-H group about the C-O bond (see inset in Figure 1), with a strong mixed a/b/c-type character. This, in combination with the lack of rotational symmetry and multiple Cl isotopes makes for a fairly rich spectrum, containing many blended lines. Fortunately, Azrak and Wilson performed a detailed microwave study on this molecule, where they reported rotational constants and some tau distortion constants.8 Their corresponding line positions47 were used to simulate the spectrum and perform an initial line assignment in the 19 fundamental which allowed for rough determination of the band center (see Figure S1 in the SM). This was done using the computer program, PGOPHER,48 which we used throughout this work to fit and simulate the spectra. After roughly determining the band center, we were able to perform additional low Ka assignments, and through an iterative procedure whereby we assigned peaks in progressively higher Ka = ±1 subbands, followed by high J peaks in lower Ka = ±1 subbands (while continuously fitting), we were able to steadily improve upon the fit. We included blended lines in the fit, and assigned them to reasonably strong transitions that did not result in residuals greater than the unapodized resolution. Figure 2 shows a Loomis-Wood type plot in the vicinity of the band center where we initially focused, and highlights series associated with Ka = ±1 transitions. Along the way, we systematically removed assignments from the fit with residuals greater than the unapodized resolution (0.00096 cm-1), and in the final fit we reduced this cut-off to the apodized resolution (0.00048 cm-1). Following calibration with water lines,40 we estimate the line position accuracy to be better than 0.0001 cm-1. We included previous microwave lines in the fit and gave them a weighting of 143.9 times that of the infrared lines (according to the estimated measurement accuracies, i.e. 100 kHz for microwave and 0.00048 cm-1 for infrared).

-6-


Page 6 of 29

Page 7 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ultimately, to perform a good quality fit to each isotopologue of 2CE, we found it necessary to use the following A-reduced Watson-type Hamiltonian49 in the Ir representation,50 ~ H red  [ A  12 ( B  C )]Jâ2  12 ( B  C ) Jˆ 2  12 ( B  C ) Jˆbc2   K Jâ4   JK Jˆ 2 Jâ2   J Jˆ 4   K ( Jâ2 Jˆbc2  Jˆbc2 Jâ2 )  2 J Jˆ 2 Jˆbc2   K Jâ6   KJ Jˆ 2 Jâ4   JK Jˆ 4 Jâ2 ,

~ where Jˆbc2  Jˆb2  Jˆ c2 . Included in H red is the total angular momentum operator ( Jˆ ) and its

components along the principal inertial axes ( Jˆ a , Jˆb , Jˆ c ), the rotational constants (A, B, C), quartic centrifugal distortion constants (  K ,  JK ,  J ,  K ,  J ), and select sextic distortion constants (  K ,  KJ ,  JK ). Inclusion of the full set of sextic constants did not significantly improve the fit and resulted in values that were in disagreement with the calculated ones, so we did not include all of them (we excluded  J , K , JK , and  J ). The fitted and calculated constants (given in Table 2) show good agreement for both isotopologues of 2CE, indicating that we have converged upon physically reasonable values. A comparison of the high resolution spectrum and fitted spectrum is given in Figs. 3 & 4: Figure 3 shows a breakdown of the a-,b-, and c-type bands that contribute to the overall band intensity for the normal isotopologue, and Figure 4 shows a close-up of the rQ0(J)/ pQ1(J) branch substructure showing the need to include each isotopologue in the fit. Observed infrared line positions and their residuals from the fit are included in the SM. 3.2. Potential astrophysical importance It is interesting to note that a common synthetic route to 2CE involves reacting ethylene glycol and hydrochloric acid at warm temperatures (~160oC), and that both reactants have been observed in the interstellar medium (see Ref. 9 for HCl and Ref. 13 for HOCH2CH2OH). Based on the available standard enthalpies of formation, the SN2 reaction between hydrogen chloride and ethylene glycol: HCl + HOCH2CH2OH → ClCH2CH2OH + H2O

(R1)

is slightly exothermic (by ~2 kcal/mol, using a scaled Hf for 2CE).51-52 While this is encouraging with respect to it being an important astrophysical reaction with respect to 2CE synthesis, there is a rather large uncertainty in the magnitude of the enthalpy change. In order to obtain further insight into the energetics, we have performed calculations of the various species in R1.

-7-



In figure 5 (top), we plot the relative energies of the stationary points in R1, with the inclusion of zero-point vibrational energy (ZPVE) corrections. It predicts this substitution reaction to be exothermic by 7.3 kcal/mol (MP2/aug-cc-pVTZ), and features a relatively broad barrier (see figure 6 for an IRC) with a large activation energy of 52 kcal/mol. Not surprisingly, these features are similar to what has previously been calculated for the methanol + hydrogen chloride SN2 reaction; it is also slightly exothermic (by ~5 kcal/mol) with a high barrier (~60 kcal/mol).53 Note that similar high barriers exist on certain pathways to the formation of known interstellar molecules, such as glycine (~40 kcal/mol; also starting from known interstellar molecules).54 While reactions with large barriers are unlikely to occur between “gas phase” molecules in cold molecular clouds where the average translational kinetic energies are low, they could feasibly occur in astrophysical ices that are bombarded with UV photons or high energy cosmic rays.55 Further, the excess energy (following reaction) could lead to vaporization of the newly formed molecules, potentially allowing for their spectroscopic detection (e.g. see Ref. 3). In addition to R1, we consider the following addition reactions to be potential sources for interstellar 2CE: HCl + CH3CHO → ClCH2CH2OH

(R2)

HCl + CH2CHOH → ClCH2CH2OH

(R3)

HCl + CH2OCH2 → ClCH2CH2OH

(R4)

where the second reactant is acetaldehyde (R2), vinyl alcohol (R3), or oxirane (R4). Two of these reactions have been investigated computationally (by Osman et al.56), where they found R2 to be nearly thermoneutral (H = -0.4 kcal/mol) with a barrier of 60 kcal/mol, and R3 to be exothermic (H = -12 and -13 kcal/mol for syn- and anti-vinyl alcohol) with barriers of ~48 kcal/mol. Based on the standard enthalpies of formation, R4 should be the most exothermic of the four reactions considered (H = -29 kcal/mol at 298.15 K), and have the smallest barrier. Experimental evidence that the barrier is small comes from the observation that it occurs rapidly in the gas phase at room temperature.57 In order to better characterize the energetics of this reaction, we have also performed calculations on it. Figure 5 (bottom) shows the calculated energies of the stationary points in R4, which reveals this reaction to have a lower activation energy than R1 (by 12 kcal/mol), and a much larger exothermicity (28 kcal/mol). Because high activation barriers can be significantly lowered in

-8-


Page 8 of 29

Page 9 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


solid ices, we decided to do some calculations that include an implicit solvation model to mimic the effect of a water environment. Although the ability of an implicit solvation model to quantitatively model solid ice is doubtful, it may provide a qualitative description of how the barrier may change in a polar environment. We chose a water solvent since it is considered to be the predominant component of interstellar ices (see 58 and Refs. therein). The solvated calculations predict a slightly increased exothermicity (29 kcal/mol), with a significant reduction in the barrier height to 30 kcal/mol. We note this is a slightly larger overall reduction in the activation energy previously reported for the methanol + HCl reaction in going from gas phase to a simulated water environment.53 Naturally, this is a consequence of the transition state being more stabilized by the solvent than the reactants, which is thought to relate to the transition state dipole moment (7.4 D for unsolvated) being much larger than that of the reactants (1.2 D for HCl and 2.3 D for oxirane (calc.)). A comparison of the IRC diagrams (figure 7) reveals that while the barrier is lower in the water solvent, it is wider, the former [latter] of which will facilitate [impede] quantum mechanical tunnelling through the barrier. In comparison to R1, R4 is more favorable, with a much greater exothermicity and greatly reduced barrier. The smaller barrier can be explained by the pre-reactive complex more-so resembling the transition state in R4 (see Ref. 59 and 60 for microwave structure) than R1, and the greater exothermicity is related to the large amount of energy stored in the three membered epoxide ring. It is certainly worth mentioning that the calculated barriers should be significantly lowered in real astrophysical ices owing to the solvent induced dissociation of HCl into ion pairs (e.g. see Ref. 61 and 62). This will naturally enhance the reactivity, both classically (over the barrier) and quantum mechanically (through the barrier), especially for the more highly exothermic HCl + oxirane reaction. Consideration of this in our calculations would be ideal, however, full consideration of explicit water molecules is beyond the scope of this work. As mentioned above, pathways leading to the astrophysical production of 2CE involve the cosmic ray processing of HCl/C2H4O or HCl/HOCH2CH2OH containing ices. These ideas are natural extension of studies that have shown the production of acetaldehyde,11 oxirane,11 vinyl alcohol,11, 63 and ethylene glycol,64 from the simulated cosmic ray bombardment of different ice mixtures. Detection of gas phase 2CE would naturally require its desorption following formation, and on this end, reactions that occur on the icy mantles of interstellar dust grains -9-



would be the most important (because the excess energy following reaction could result in evaporation). There must be numerous pathways that could result in the build-up of gas phase 2CE that are related to R1-R4, and most probably involve HCl since it is expected to be the dominant reservoir species of chlorine in the icy mantle on grain surfaces in dark clouds (~90% of chlorine is expected to be in the form of HCl).12 Note that several simple reactions occurring on the surface of interstellar grains have recently been proposed to explain the production of gas phase methyl chloride.65 The authors also considered ion-molecule reactions that we overlook here, but we note they could also contribute to the gas phase accretion of 2CE; e.g., H2Cl+ + C2H4O → ClCH2CH2OH2+, followed by ClCH2CH2OH2+ + e→ 2CE + H. Not surprisingly, promising gas phase 2CE lines, as predicted from our refined constants are not evident in the survey spectra of Sagittarius B2(N) reported in Ref. 66. Considering the large partition function for the 9 atom asymmetric top molecule studied here, in addition to its low volatility (compared to most detected interstellar molecules), we suspect extensive averaging will be required for detection. Nonetheless, we expect molecules like 2CE have important roles in the chemistry that occurs on icy grain mantles, and suggest that 2CE might be a promising candidate for future observations, especially at millimetre wave frequencies where the most intense lines should fall. At this end, Figure 8 shows the simulated spectra at TR=20 K (a typical cold molecular cloud rotational temperature) which maximizes in intensity around 100 GHz (see also Figure S4 (SM) which shows the simulated spectra at 200 K, i.e. the rotational temperature of a typical hot molecular core). The uncertainty in the predicted rotational line positions at 100 GHz (3 mm) for low J values (J

Far-Infrared Synchrotron Spectroscopy and Quantum Chemical

Recommend Documents