Matrix Isolation Spectroscopy: Aqueous p-Toluenesulfonic Acid

Subscriber access provided by READING UNIV

Article

Matrix Isolation Spectroscopy: Aqueous p-Toluenesulfonic Acid Solvation Thien Khuu, David Jay Anick, and Mary Jane Shultz J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b08939 • Publication Date (Web): 02 Jan 2018 Downloaded from http://pubs.acs.org on January 2, 2018

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 free 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 accessible to all readers and 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.

The Journal of Physical Chemistry A 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 35 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

Matrix Isolation Spectroscopy: Aqueous p-Toluenesulfonic Acid Solvation

Authors: Thien Khuu, David Anick, and Mary Jane Shultz* Laboratory for Aqueous and Surface Studies, Pearson Building, Tufts University, Medford, MA 02155

Abstract Interaction between p-toluenesulfonic acid (pTSA) and water is studied at -20°C in a CCl4 matrix. In CCl4 water exists as monomers with restricted rotational motion about its symmetry axis. Additionally, CCl4 is transparent in the hydrogen-bonded region; CCl4 thus constitutes an excellent, ambient thermal energy matrix isolation medium for diagnosing interactions with water. Introducing pTSA-nH2O gives rise to two narrow resonances at 3642 cm-1 and at 2835 cm-1 plus a broad 3000-3550 cm-1 absorption. In addition, negative monomer symmetric and asymmetric stretch features relative to nominally dry CCl4 indicate that fewer water monomers exist in the cooled (-20° C) acid solution than in room-temperature anhydrous CCl4. The negative peaks along with the broad absorption band indicate that water monomers are incorporated into clusters. The 3642 cm-1 resonance is assigned to the OH-π interaction with a cluster containing many water molecules per acid molecule. The 2835 cm-1 resonance is assigned to the (S-)O-H stretch of pTSA-dihydrate. The coexistence of these two species provides insights into interactions in this acid-water CCl4 system.

Introduction

w

ater clusters have been studied in the IR region since the 1970s1-5 and are important in many biological systems,6 organic chemistry mechanisms,7,8 ice9,10 and clathrate

hydrate nucleation.11 Infrared absorption spectroscopy (IR) is perhaps the most useful method to study hydrogen bonded (H-bonded) water systems due to the high sensitivity of OH stretch to the -1-

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

local chemical environment. Unfortunately, broadening in the H-bonded region, which extends hundreds to thousands of wavenumbers, is a major limitation for assigning stretch resonances and diagnosing interactions. Despite the challenges, there have been pioneering theoretical12-16 and experimental16-20 works providing glimpses into H-bonded network structures through IR. A classic IR method consists of isolating water and interactants in a low temperature, usually solid, matrix.21-24 The matrix cages the species of interest and, due to the low temperature, produces narrow resonances that greatly aid in spectral interpretation. Unfortunately, the cage also precludes water molecule exchange among various hydrated clusters and the low temperature can inhibit rearrangement of the cluster, kinetically trapping high-energy configurations. Nozzle beam expansion25,26 retains the cold temperature yielding well defined resonances but avoids caging effects. For example, pioneering work by Johnson and coworkers16-20,24,27 resolved the broad proton band and provided clear and consistent spectral assignments of the H+ proton in the Eigen and Zundel ions as well as in larger hydrated clusters. The Johnson work27 also produced clear evidence that the quantum nature of the proton results in inherently broad spectra. Herein, acid-water structures are isolated in a CCl4 matrix at relatively high temperature: 20 °C, and probed with IR; a technique labeled room temperature-matrix isolation spectroscopy (RT-MIS). RT-MIS retains the small clusters that are characteristic of classic matrix isolation but with ambient thermal energy. This allows free cluster rearrangement as well as exchange or collection of water among clusters. RT-MIS water has a simplified IR spectrum28 since water exists only as monomers in this matrix. The partial negative charge on the water oxygen atom interacts weakly with partial positive charge on the CCl4 molecule resulting in free rotation about the water symmetry axis but quenched rotation about the other two directions. The water IR absorption spectrum in the 3500-4000 cm-1 region consists of the symmetric and asymmetric stretches along -2-


Page 2 of 35

Page 3 of 35 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


with a rotational structure associated with the asymmetric stretch. Interactions between water and numerous different molecules29-31,32,Vu, 2013 #293 have been investigated using this matrix. These studies led to identification of the room temperature dangling OH frequency as well as assigning the OH- ion vibration. Surprisingly the p-toluenesulfonic acid (pTSA, pKa -1.3433) water cluster resonances reported in this work are remarkably sharp despite ambient thermal energy. The narrow width is tentatively attributed to large populations of specific, nonwater-water H-bonding motifs with limited motion. There have been multiple reports of sulfonic acid and its derivatives forming clusters with water34-37 including reports that pTSA induces clathrate hydrate formation at low pressure.38-40 In contrast to these bulk investigations,38-40 RT-MIS vibrational spectra provide insight into molecular interactions between p-TSA and water. Compared with previous RT-MIS studies, this work shows definitive evidence for larger water clusters in the matrix. To the best of our knowledge, IR spectra for this system have not previously been reported. Theoretical modeling and the RT-MIS system are used to assign two major features from the IR spectra of pTSA and water: 2835 cm-1 and 3642 cm-1. The 2835 cm-1 feature is observed in both acid·(H2O)n and acid·(H2O)(D2O)n-1 clusters. This peak is attributed a stretching motion of the sulfonic O-H group of the dihydrate. The feature does not shift upon isotopic substitution due to non-ionization of the acid; i.e. the limited water environment fails to ionize the acid. The feature at 3642 cm-1 is assigned to the O-H stretch of a water molecule in a water cluster that donates a hydrogen bond to the π system. These two resonances provide important insights into the acid hydration as well as information about the acid-water clusters. These observations and assignments are elaborated in the discussion section.

-3-



Experimental Anhydrous carbon tetrachloride (CCl4, Sigma-Aldrich, ≥99.5% anhydrous) is dried further with silica gel for at least two days prior to use. The dried CCl4 is added to a glass cell (25 cm path length, 30 cm diameter) with calcium fluoride IR windows (CaF2, 40 cm diameter). An IR spectrum (a Nicolet Magna-IR 760 spectrometer (DTGS KBr detector)) of dried CCl4 at room temperature serves as the background and is thus subtracted from all spectra. Stock 11 M ptoluenesulfonic acid monohydrate (pTSA·H2O, Sigma-Aldrich, ≥98.5%, A.C.S. reagent) solution is prepared by dissolving 2.6198±0.0001 g of pTSA·H2O in 1.25 mL of Nanopure water (18 MΩ, Barnsted GenPure Pro with UV): monohydrate acid to water mole ratio 1:5. The stock aciddeuterium solution is prepared similarly with deuterium oxide (D2O, Sigma-Aldrich, 99.9 atom % D). Stock pTSA solution is added to CCl4 for an acid concentration of 18.45 mM. The least added water is thus 110.7 mM: more than 20 times neat water saturated concentration. Stock 2.78 M solution sodium p-toluenesulfonate (pTS-Na+, Sigma-Aldich, 95%) is made by dissolving 1.079±0.001 g of pTS-Na+ in 2.0 mL Nanopure water: salt to water mole ratio 1:20. Increased water content samples are prepared by bringing the solution to room temperature, adding water to successively increasing by 2 water molecules per acid, and equilibrating the solution for a minimum of two days prior to spectral acquisition. Cooling CCl4 to -20°C does not change the background spectrum significantly, thus use of the room temperature background does not affect the results. (Spectra of CCl4 at room temperature and at -20 °C are in supplementary information.) The sample compartment is thoroughly purged with dry nitrogen for at least 3 hours to remove atmospheric water and CO2. Each solution spectrum is taken at room temperature prior to cooling the sample in a -20°C freezer for at least two days allowing the system to reach equilibrium. All FTIR spectra are recorded at 64 scans and 1 cm-1 resolution. -4-


Page 4 of 35

Page 5 of 35 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


Theoretical Calculations were carried out on a Parallel Quantum Solutions Linux box using the PQS suite of programs.41 Density functional theory, specifically the hybrid method B3LYP,42,43 has been used for many water cluster studies. When used with a triple-zeta basis that includes both polarization and diffuse functions, B3LYP does a good job predicting energies and spectra for gasphase water clusters (H2O)n as well as for X(H2O)n clusters where X is any of a variety of ions or polar solutes.44-53 When using the PQS programs, diffuse basis functions on the aromatic ring caused overlap problems leading to non-convergence of the self-consistent field. The reason for this glitch was not determined, but diffuse functions are not typically necessary for aromatic optimizations. For this reason we adopted B3LYP with a “mixed basis” consisting of 6311++G(d,p) on the water molecules and on the S and O’s of the sulfonyl group, but consisting of 6-311G** on the aromatic ring and the methyl of pTSA. Herein the mixed basis is denoted as 6311[++]G(d,p). All reported calculations were done via B3LYP/6-311[++]G(d,p). Predicted infrared frequencies were calculated with the harmonic approximation and were scaled by 0.96554 before comparing with experimental results. Free energies at 20 C and 1 atm. were obtained from frequencies via standard formula. This approach does not adequately account for methyl group free rotation, which is treated as a low-frequency vibration. This has minimal effect on the ZPE but potentially a significant effect on the calculated entropy. Full treatment of rotations is complex; the approach is justified by noting that the principal results are free energy differences, for which the errors approximately cancel. Various geometries for gas phase (pTSA)·(H2O)n clusters were optimized for 1  n  12, searching for clusters that have the global minimum free energy or lie close to the global minimum,

-5-



for each n. Because of the immense number of isomers, especially for larger n, only a small subset could be sampled. Correlations between connectivity topologies and cluster energies55,56,58 focused the search toward more likely candidates. For n  5 the lowest energy geometries are dissociated, consisting of the para-toluenesulfonyl anion attached to an H+·(H2O)n cation. Ab initio molecular dynamics (AIMD) studies were carried out for the di- and tri-hydrate clusters, i.e. (pTSA)·(H2O)2 and (pTSA)·(H2O)3. Each run was of the NVE (fixed moles (N), volume (V), and energy (E)) type starting at the global minimum with kinetic energy randomly added to approximate 20C. In this instance, NVE simulates dynamics for a single isolated cluster without solvent, moving under its B3LYP-computed potential conserving the total energy. Instantaneous temperature varies during such a run and is defined as the total kinetic energy divided by kB(3Nat-6)/2, (3Nat-6) is the number of degrees of freedom. The NVE method was a compromise requiring substantially less computation time than a highest-quality simulation including solvent molecules filling out a periodic box along with a thermostat. Runs used a step size of 0.5 fsec. and lasted at least 12 psec. after a 1 psec. equilibration period.

Results Water is an asymmetric rotor hence the gas phase IR spectrum is complex. The spectrum is simplified in CCl4. Isolated monomers are a pseudo prolate top28 having a free rotation around the symmetry axis and quenched rotation about the other two axes. The result is two fundamental vibrations in the free-OH stretching region: the symmetric and asymmetric stretches at 3617 cm-1 and 3709 cm-1, respectively. The asymmetric stretch is accompanied by rotational wings due to rotation about the symmetry axis (pure water spectrum shown in reference 28, these features are -6-


Page 6 of 35

Page 7 of 35 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


also present in Figure 1 indicated with arrows). These features are redshifted relative to water in the gas phase due to the CCl4 dielectric constant (2.2379). Monitoring changes in these features greatly aids in identifying interactions. For example, the presence of a weak hydrogen bond acceptor restricts the symmetry axis rotation and collapses the rotational wings; a strong interaction gives rise to redshifted stretch bands and removes equivalence of the two O-H stretches.30 Interaction with the oxygen atom lone pair enhances the symmetric stretch oscillator strength relative to the asymmetric stretch.25,59-63 The neat water RT-MIS spectrum is altered in the presence of pTSA (Figure 1). The room temperature spectrum uses a stock solution with an acid to water mole ratio of about 1:6. The same concentration ratio is shown at -20°C along with water concentrations of 1:8 and 1:10. “Dry” CCl4 has a trace water concentration. Dissolving and transferring the pTSA·H2O into CCl4 requires a minimum of five moles of water in addition to the water already present in monohydrate, hence the pTSA room temperature spectrum is that of pTSA-(H2O)6. At 1:6, the water concentration is more than 20 times that of saturated water in CCl4. Symmetric and asymmetric stretches for water monomers are clearly visible in the room temperature spectrum as is an additional peak at 3642 cm-1 between the monomer peaks. Cooling the solution changes the spectrum. The water monomer peaks diminish greatly as shown most dramatically by the negative asymmetric stretch peak. Since the background is “dry” CCl4, the negative peak indicates that the low concentration advantageous water in dry CCl4 is collected by pTSA consistent with the dehydrating capacity of sulfuric acids; the solution has less free water than the background, hence the negative peak. The small peak at 3642 cm-1 for the room temperature solution becomes larger at -20 °C and increases further with added water. Additionally, two new features become apparent: a broad hydrogen-bonded feature from 3000 cm-7-



Figure 1. Infrared spectra of pTSA·H2O in CCl4: injected solution mole ratio acid:water 1:6 at room temperature (black line) and -20 °C (red line). Adding 14, and 28 L water and cooling to – 20 °C (acid to water mole ratio 1:6, 1:8, 1:10) (red, blue, and green, respectively; spectra offset for clarity). At 1:6, the injected water is more than 20 times that of saturated neat water. The grey region is inaccessible due to a hydrocarbon impurity in CCl4. The low intensity peak at 2854 cm-1 is also an impurity; it helps scale the growing peak at 2835 cm-1. All spectra are difference spectra subtracting the room temperature, nominally dry CCl4 spectrum. 1

to 3550 cm-1 plus a small peak at 2835 cm-1 that is red of the hydrocarbon impurity peak. This

impurity at 2854 cm-1 appears in all spectra including the CCl4 sample with no acid-water complex and appears in both 99.5% and 99.9% pure CCl4, although it is less intense in the 99.9% sample (see supplemental information). Attempts to identify the impurities by comparison with many known small molecules likely to be in CCl4 yielded no matches. Changes from cooling are entirely reversible; reheating to room temperature and recooling produces the same spectra. Therefore, within the acid:water ratios under investigation in this paper, the 2835 cm-1 and the 3642 cm-1 cannot be distinctly observed at room temperature. -8-


Page 8 of 35

Page 9 of 35 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


To aid assigning the spectral features observed in Figure 1, pTSA·H2O was dissolved in D2O.13,14,64 D2O stretches are red-shifted by √½ relative to ordinary water (Figure 2). Normal OH modes remain due to normal hydrogen in both the acid and the monohydrate water. Peaks blue of 3550 cm-1 appear broader than the corresponding ordinary water resonances. The room temperature, water monomer resonances (black line) are much less intense than the corresponding black line in Figure 1, consistent with replacing H2O with D2O to dissolve the monohydrate. Due

Figure 2. Substitution of D2O for H2O to dissolve the acid monohydrate. In CCl4 at room temperature (black line) and at -20°C with increasing D2O (red, blue, and green, respectively). Spectra are difference spectra subtracting the room temperature, nominally dry CCl4 spectrum. Note that p-TSA is a monohydrate, hence there are three ordinary hydrogen atoms per acid anion. Mole ratios acid ordinary water to heavy water are 1:1:5 (red), 1:1:7 (blue), and 1:1:9 (green). Spectra are offset for clarity.

-9-



to deuterium substitution, intensity in the 3600-3800 cm-1 region is that of the OH resonances of HOD. The corresponding OD resonances of the HOD species appears at 2689 cm-1. Peak assignment is further aided with spectra of the salt pTS-Na+ and water in CCl4. The salt lacks the acid OH. Accordingly, the spectrum shown in Figure 3, contains fewer features. At room temperature, the water monomer peaks are present with lower intensity than in the corresponding acid despite a larger water concentration than in the corresponding acid. Further, the broad hydrogen-bonded resonances are largely absent. It is interesting to note that the impurity peak at 2854 cm-1 is enhanced in the salt. The origin of this increase is likely the increased ionic

Figure 3. Infrared spectra of pTS-Na+ and water in CCl4 at room temperature (black) and at -20°C with 0 and 10 L added water (salt to water mole ratios 1:20, 1:21; (red and blue, respectively). Spectra are difference spectra subtracting the room temperature, nominally dry CCl4 spectrum. Note absence of the 2835 cm-1 peak (the sharp peak is at 2854 cm-1 is the impurity mentioned above).

strength of the solution due to ionization of the salt. The significance of these spectra is that the 2835 cm-1 resonance does not appear in the salt; 2835 cm-1 must thus be associated with the acid. -10-


Page 10 of 35

Page 11 of 35 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


Interpretation of these spectra, peak assignments and implications for pTSA-H2O configurations is contained in the discussion section.

Computational results Low energy structures Table I lists the lowest free energy structure found at 20 C for each n, 1  n  12, as well as others that come within 2 kcal·mol-1 of the lowest. This is given in terms of G253, the free energy of formation from widely separated pTSA and n H2O units. Clusters with the acid intact are named pTSA.WnX, X being a letter; those where the acid proton has transferred to water are named pTS.H+WnX. For isomers having one or more ‘da’ (single-donor-single-acceptor) waters having a free H with two possible orientations, only the lowest energy configuration is listed. Likewise, when several local minima occur with respect to rotation around the C—CH3 or the C— S bonds, only the lowest is included. Table I also shows the calculated electronic energy of formation Eo as well as G0K (defined as Eo + ZPE). Three clusters that do not meet the 2 kcal·mol-1 criterion but are of interest due to occurring in the AIMD or having a H2O molecule that donates to the aromatic ring are also included in Table I. Selected clusters from Table I are illustrated in Figure 4 and all entries therein can be found in xyz format in the supplemental information. Those with the letter “A” are putative global minima. In some cases, greater binding energy is attained by breaking an H-bond. For example, Figure 4(e) (“pTS-H+W7A”) and 4(f) (“pTS-.H+W7D”) are different only in that one H-bond in Figure 4(f) (on the right-hand edge) is broken in Figure 4(e). Per Table I the Figure 4(f) structure -11-



has the lowest electronic energy and lowest free energy at 0 K, but Figure 4(e) dominates at 253 K because the greater entropy from losing that bond lowers the free energy more than the lost enthalpy raises it. As illustrated by Figures Figure 4(g) for n=10 and Figure 4(i) for n=12, global minima for n  8 have an outer surface consisting of the Eigen ion and dda waters, except for daa’s which donate to the negatively charged O’s of the sulfonyl. Such surfaces fit the description of “hydrophobic water”57,65 that tends to repel other such surfaces or free water.

-12-


Page 12 of 35

Page 13 of 35 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


Figure 4: A selection of the optimized (pTSA)(H2O)n clusters listed in Table I. (a) pTSA.W2A (b) pTSA.W3A (c) pTSA.W3B (d) pTS−.H+W4A (e) pTS−.H+W7A (f) pTS−.H+W7D (g) pTS−.H+W10A (h) pTS−.H+W11X (i) pTS−.H+W12A.

-13-



Page 14 of 35

Table I. Selected cluster binding energy (kcal∙mol-1) via B3LYP/6-311[++]G(d,p), including electronic energy, ZPE-corrected electronic energy (denoted G0K), and free energy at 253 K. CLUSTER

∆Eo

pTSA.W1

-11.64

pTSA.W2A

∆G0K

-24.25

-9.10

-19.40

∆G253 -2.04

-5.50

CLUSTER

∆Eo

∆G0K

∆G253

pTS−.H+W7B

-85.89

-66.97

-13.76

pTS−.H+W7C

-87.10

-67.48

-13.57

pTS−.H+W7D

-87.73

-67.87

-13.38

pTS−.H+W7E

-86.33

-66.49

-12.16

pTSA.W3A

-35.70

-28.50

-7.56

pTSA.W3B

-35.93

-28.84

-6.93

pTS−.H+W8A

-100.02

-77.84

-16.16

pTSA.W3C

-35.47

-28.03

-6.30

pTS−.H+W8B

-98.00

-75.97

-14.70

pTSA.W3D

-31.00

-24.55

-4.55

pTS−.H+W3A

-35.14

-28.36

-6.11

pTS−.H+W9A

-112.26

-87.28

-17.66

pTS−.H+W9B

-111.80

-86.72

-17.27

pTS−.H+W4A

-48.70

-38.56

-8.91

pTS−.H+W9C

-112.30

-87.55

-16.75

pTS−.H+W4B

-48.56

-38.66

-8.82

pTS−.H+W9D

-112.74

-87.47

-16.68

pTSA.W4A

-46.60

-37.09

-8.24

pTS−.H+W9E

-110.38

-85.50

-16.19

pTSA.W4B

-45.14

-35.74

-7.80

pTS−.H+W9X

-104.96

-80.79

-12.21

pTSA.W4C

-45.89

-36.22

-7.39 pTS−.H+W10A

-123.48

-95.85

-19.04

pTS−.H+W5A

-59.34

-47.01

-10.49

pTS−.H+W10B

-122.89

-95.34

-18.52

pTS−.H+W5B

-60.51

-47.48

-9.70

pTS−.H+W10C

-123.16

-95.32

-18.29

pTS−.H+W5C

-59.32

-46.24

-8.83

pTS−.H+W10D

-122.37

-94.93

-17.77

pTS−.H+W5D

-61.46

-47.39

-8.74

pTS−.H+W10E

-119.92

-92.95

-17.59

pTSA.W5A

-57.93

-46.14

-9.58

pTSA.W5B

-56.28

-44.76

-8.91

pTS−.H+W11A

-136.05

-105.69

-20.92

pTS−.H+W11B

-135.44

-105.03

-19.98

pTS−.H+W6A

-75.37

-58.21

-11.86

pTS−.H+W11C

-134.85

-104.48

-19.86

pTS−.H+W6B

-70.70

-55.34

-11.82

pTS−.H+W11X

-128.97

-99.38

-15.92

pTS−.H+W6C

-74.28

-57.47

-11.06 pTS−.H+W12A

-148.03

-114.86

-22.33

pTS−.H+W12B

-148.10

-114.68

-22.02

pTS−.H+W7A

-86.60

-67.41

-13.82

-14-


Page 15 of 35 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


Figure 5: Cluster binding energy as a function of the number of water molecules (blue line/symbols). Note the larger stabilization gain from n =1 to n = 2 than any other step, shown dramatically in the binding energy per water plot (right axis, red line/symbols).

Table I and Figure 5 show the free energy of formation of the lowest energy clusters for each n as a function of n. There is a large gap of 3.5 kcal·mol-1 between the monohydrate and dihydrate binding energy; subsequent water molecule additions have a consistent binding energy of about 1.7 kcal·mol-1. The result is a slope change or “kink” at n = 2. Enhanced binding per water for n = 2 is highlighted by the per molecule binding energy, shown in Figure 5. Stabilization at 2 implies that the “preferred” number of H2O’s per acid molecule is 2. Thus, the dihydrate is expected to be over-represented in a system containing populations of various (pTSA)(H2O)n units. This is similar to the situation that causes n=21 to be a “magic number” among H+(H2O)n clusters.20,66-68 Details of the source of this unusual binding energy is under investigation. Notice (Table 1) that binding energies are much smaller at 253 K than at 0 K. The entropic contribution to water cluster free energy is huge near ambient temperature. This is a known phenomenon. Similar results were obtained for solute-free (H2O)n clusters, 2  n  10, by Shields et al.69 In Shields’ Table 8, Ee, E0, and GT (T = 10, 200, 298) notations correspond to E0, -15-



E0K, and GT (T = 253) here. For (H2O)n clusters, Shields et al found that binding energies become much smaller at 200 K than at 0 K and become positive at 298 K, implying that small water clusters are ultimately unstable at 298 K and evaporate over time to become H2O monomers. Here, G253 values are negative so these solute-water clusters are predicted to be stable at 253 K; we likewise find that the entropy term greatly affects both the absolute and the relative binding energies (i.e. relative ordering of isomers for a fixed n) of pTSA-water clusters.

Identification of the 2835 and 3642 cm-1 resonances What is the source of the 2835 cm1 signal? Armed with over 150 (pTSA)(H2O)n cluster optimizations and frequency calculations, we reasoned that either the mode of interest was present somewhere among the OH stretch signals already computed, or it derived from an OH stretch whose local bonding environment is very similar to one computed. (The exact frequency of an OH stretch is highly sensitive to its donor types (da, daa, etc.) and (if H-bonded) of its acceptor.70-73) Any mode that fell in the range 283515 cm1 is called a “match” and modes in the range 2835 10 cm1 are labeled “strong matches:” 28 matches were found. Twelve of these are the OH stretch of the sulfonic acid H donating to water in pTSA.W2A: the dihydrate. These are all strong matches, ranging from 2830 cm-1 to 2845 cm-1. This stretch is seen repeatedly because the database includes separate optimizations for tweaks of the cluster pTSA.W2A in which the toluene is rotated or the direction of the free H’s is flipped. The remaining 16 matches occur in various clusters with n  5. In every case the mode has a 3-fold symmetric stretch of an H3O+ embedded in an Eigen ion. As noted the Eigen generates a very broad signal even under the best experimental conditions, so Eigen ion modes can be ruled out. We therefore tentatively assign the 2835 cm-1 mode to the S-O-H… O of the dihydrate. -16-


Page 16 of 35

Page 17 of 35 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


As noted above, the dihydrate is a priori likely to be overrepresented, and hence is expected to provide an unusually strong signal, so dovetails nicely with this interpretation. We remark that the other O-H stretches of the dihydrate are either low-intensity free O-H’s or water-water H-bonds which are too broad at 253 K to detect separately. The full predicted stretch spectrum of the dihydrate is provided in the Supplementary Information. The same procedure is applied to the 3642 cm-1 signal; 37 matches are obtained. One match is the free O-H stretch of dry pTSA, at 3645 cm-1. This assignment would require the implausible scenario that a significant fraction of the pTSA, which comes as the mono-hydrate because it is so hygroscopic, lost all its water to the CCl4 solvent. The largest group of matches, 14 in all, arise from bonds from a dda water to one O of the SO3. All these bonds are strained (large H-O—O angle), coupled with other stretches, or both. The 3642 cm-1 position is on the blue end of the range for dda…O-S stretches and only occurs if the H…O distance is unusually long, or if the frequency is altered by strong coupling, or both. Such bonds are among the weakest to start with and intermittently break and re-form at 20 C, making them poor candidates for generating a narrow signal. Similarly, 8 matches are strained and long dda-daa bonds. The dda-daa is the weakest category of water-water bonds so it is implausible that the weak blue end of this category could generate a narrow signal while other water-water bonds only contribute to the broad signal. Seven more matches are likewise dismissed as weak, strained, and coupled instances of dda…ddaa or ddaa…O-S. The remaining six matches are water donating to the benzene ring. All are strong matches, their range being 3641 cm-1 to 3651 cm-1. It has already been noted that stretches for water…benzene are known to be around 3660 cm-1. Electron donating groups on the benzene ring, such as the sulfonate substituent, are expected to redden the stretch by some 20 cm-1.74 One -17-



consequence of this assignment is that no water…benzene bonding can occur until n is reasonably large. Attempts to optimize a smaller cluster with a H2O…benzene bond resulted in the water being ripped away to make more favorable bonds as part of the droplet at the sulfonyl head group. Water…benzene bonds do not occur until the sulfonyl is “satisfied” and water molecules begin to “spill over” toward the benzene as in Figure 4h. We could not determine at what point this occurs but calculations with larger clusters suggest n will be somewhere in range of 16 to 20 before “spilling” occurs for a global minimum or near-minimum cluster. The 3642 cm-1 signal is therefore taken as a marker for “large cluster” without a particular size or geometry being specified. Our “matching” method relies on gas phase computations so one can ask whether solvent interactions could significantly alter frequencies. Dangling H’s that are exposed to the CCl4 become red-shifted by 35 cm-1 to 50 cm1 according to both calculation and the known examples of the monomer’s symmetric and asymmetric stretches. However, all dangling H’s in all clusters resonate at least 90 cm1 bluer than 3642 cm-1, so this cannot be the explanation.

AIMD Results Gas phase NVE AIMD runs for the dihydrate pTSA.W2A and the trihydrates pTSA.W3A and pTSA.W3B were conducted. Data were collected for 12.2 psec. (pTSA.W2A and pTSA.W3B) or for 20.3 psec. (pTSA.W3A) following a 1 psec. equilibration interval. For the dihydrate, temperature based on kinetic energy was 268  26 K. Referring to atoms as labeled in Figure 4(a), the O1—O4 H-bond remained intact throughout; the O4—O5 H-bond was present during 98.1% of the snapshots. We followed Kumar’s geometric criteria for OD-H*- -OA to be deemed H-bonded, namely d(H*, OA) < 2.5 Å, d(OD,OA) < 3.4, and =angle(H*-OD-OA ) < cut = 30.75 By this criterion, the O5—O2 bond broke and re-formed several times. Starting at 2.84 psec. it remained -18-


Page 18 of 35

Page 19 of 35 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


Figure 6: Evolution of the torsional angles – measuring internal rotations – during the AIMD run. The methyl group (blue) spins freely. In contrast, there is no net rotation about the S-O bond.

severed for 55 fsec. and when it re-formed, O5 had become attached to O3 via H5 in lieu of bonding via H4 to O2, and that pattern persisted for the remainder of the run. Overall an H-bond was present from O5 to either O2 or O3 during 93.2% of snapshots (98.1% if cut=40). Flipping of the dangling H’s occurred often.

Figure 6 shows the instantaneous torsional angles; these measure internal rotations. The methyl group spun freely, completing nearly 10 net revolutions during the run. There was no net rotation around the S—O bond, which oscillated across a mean position of about 90. The transition state for rotation around the S—O has Gts253 = 3.1 kcal·mol-1 and occurs at orientations

-19-



of 6 and 174 for the S-O1-C1-C2 dihedral angle. In the optimized geometry, the dihedral angle is 81. The 20.36 psec. trihydrate run starting with pTSA.W3A likewise revealed a fairly stable bonding pattern. Temperature was 241 ± 29 K. Referring to Figure 4(b), the O1-H1- -O4 bond was intact for the entire run and O4-H2- -O5 was present for 99.9% of snapshots (100.0% if cut=35). There was an H-bond from O5 to O6 in 98.0% of snapshots and this bond flipped once from using H4 to using H5 mid-way through the simulation. The bond from O6 to O2 flipped from using H6 to using H7 at 17.3 psec., and “walked” to an O6-H6- -O3 bond at 19.5 psec. and “walked” back to O6-H7- -O2 at 20.3 psec. just before run completion. During the interval 2.3 to 3.6 psec., an additional O5-H5- -O3 H-bond formed, turning the topology into pTSA.W3C, and likewise the system became topologically pTSA.W3C during 19.9 to 20.1 psec. when an O5-H4- -O2 bond occurred. The total of 1.5 psec. with the additional bond, versus 18.8 psec. without it, yields a ratio of 12.5 favoring pTSA.W3A over pTSA.W3C. This is consistent with the ratio predicted by Table I. The G253 difference between pTSA.W3A and pTSA.W3C of 1.26 kcal/mol equals 2.5 kT and predicts an occurrence ratio of 12.2. Presumably other trihydrate topologies including pTSA.W3B would also occur in a long enough run but the barriers to interchange are higher and were not overcome in this limited simulation. A separate trihydrate run was performed starting at the pTSA.W3B topology (Figure 4(c)). In this trihydrate run, temperature was 238 ± 23 K. The pTSA.W3B topology demonstrated less bond stability than either the dihydrate or the previous pTSA.W3A trihydrate. Referring to Figure 4(c), the O1-H1- -O4 bond was intact for all but 6 fsec. when  crested at 31.2, and H1 “almost” transferred once when d(O1,H1) surged briefly to 1.17 Å. The O5 side of the cluster remained -20-


Page 20 of 35

Page 21 of 35 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


relatively stable: an O4-H2- -O5 bond (resp. an O5—O2 bond via either H4 or H5) was present in 88.9% (resp. 84.9%) of snapshots; this climbed to 98.1% (resp. 93.3%) when cut was raised to 40, suggesting similar behavior to the dihydrate. However, the O6 water roamed much freer. The corresponding figures on the O6 side were just 69.2% for the O4-H3- - O6 bond (resp. 42.4% for any O6—O3) and, for cut=40, 76.0% (resp. 47.0%). The configuration topology matched the high-energy isomer pTSA.W3D 43.2% of the time, including the entire final 3.5 psec. when the O6 water appeared to be “flailing” on the end of the O4—O6 bond and remained far removed from the sulfonate. By contrast it matched the low-energy pTSA-W3B, Figure 4(c), just 14.6% of the time. As with the dihydrate, the methyl group spun freely but there was only slow oscillation around the S—O bond. In this run the methyl completed 7 revolutions in the first 4 psec., then switched direction and turned 10 more revolutions before the end of the run. This is consistent with a classical calculation of the moment of inertia of the methyl group around the C—C bond, which yields 1.31 THz as the frequency when rotational kinetic energy is ½kT. It is expected that bond rearrangements happen more slowly in solvent than in gas phase, nonetheless this picture provides insight relevant to the matrix. The AIMD runs support the idea that the dihydrate is stable. While the trihydrate in the pTSA.W3A state (with about 8% as pTSA.W3C) also appears stable, if bond directions change to convert it to pTSA.W3B, the picture changes. Then the trihydrate holds one or the other of its flanking waters (i.e. O5 and O6) so loosely that they could be readily prone to “evaporation” into the matrix monomer population at this temperature. Conversely a dihydrate or other cluster that encounters a monomer may have the monomer join it for a while. The overall picture is a dynamic equilibrium in which waters are exchanged among (pTSA)(H2O)n “islands” via the monomer population. Calculations are under way to see how the dynamics may differ once n ≥ 4 and the cluster dissociates. A preliminary look -21-



Page 22 of 35

at the case of n=7 finds frequent interchange among the top four geometries (e.g. Figures 4(e), 4(f)) but no proton transfers that affect the embedded H3O+.

D and H position preferences in D2O When the water cluster contains both 1H and D and they are free to distribute themselves into the various available positions for H, a “positional isotope effect” can be observed. For example, at 80 K the central H of the Zundel cation is 2.3 times more likely to be H than D, compared with prediction based on random assignment.13 When the HOD monomer binds to benzene, the preferred orientation has O-D making the H-bond.

76-78

A previous study explored

these “positional isotope effects” computationally for a large database of (H2O)n clusters and found that a J-shaped curve accurately predicted the zero-point energy of replacing a single position with D, as a function of the O-H* distance in the optimized structure.79 The same study demonstrated mild cooperativity effects when replacing two H’s by D’s on the same or on adjacent water molecules. Cooperativity effects are small, so a reasonable approximation is that the effect of multiple substitutions is additive. Some experiments reported herein involved addition of D2O followed by looking at O-1H stretch resonances. In these, a few of the exchangeable H’s are 1H that is present in the original monohydrate (3 1H’s per acid molecule) and a tiny amount comes from 1H2O that is present in “dry” CCl4, but the remainder (10, 14, or 18 hydrogen atoms per acid molecule) is D. Clusters therefore consist of D2O with a small number of H substitutions. Are positional isotope effects affecting the results, and should they be considered a factor in the interpretation? To answer this, for both for pTS.H+W11X ( Figure 4(h)) as a representative of a cluster having an H-bond to the benzene ring and for the dihydrate, we followed the example of reference79 by computing the free

-22-


Page 23 of 35 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


energy difference between the cluster with all D and the cluster when each position is replaced by H. The results are listed in Table II: ZPE(DH) is the effect of single D-to-H substitution at the indicated hydrogen, while G253(DH) is the calculated free energy difference at 20 C. With just one 1H in the cluster, assuming a Boltzmann distribution at 20 C, the last column gives the probability of each position having the 1H. Hydrogen labels refer to Figure 4(a,h). With no positional isotope effect these probabilities would be 1/23 = 0.043 and 1/5 = 0.2 for the two-water clusters. Results closely follow the theory outlined in reference 79; positions in short H-bonds associated with the Eigen ion have the smallest substitution energy, followed by dangling H’s and lastly those in H2O—H2O H-bonds. In Figure 4(h) the H3O+ is marked “+” and its three hydrogens are H5, H6, H7. The benzene-bonding H is H23 and in the cluster overall it is one of the more favored positions to be 1H. However, at 20 C the effect is small and it is only about 10% more likely to be 1H than pure chance would predict. For the dihydrate, the sulfonic acid H is 30% more likely to be 1H than the other bonding H’s (H2 and H4). This may be moot however because we suspect that many pTSA molecules never dissociate in the low-water environment, and those that end up as dihydrate may keep the protium atom that they start with. For pTSA.W2A, a Boltzmann distribution may determine 1H and D at H2 through H5 but not at H1.

-23-



Page 24 of 35

Table II. For two clusters, calculated ZPE change and free energy change (at 20C) from switching a single position from D to H, along with the Boltzmann probability for that position to be H when there is just one H in the cluster. ZPE

G253

(DH)

(DH) 

prob

pTS.H+ W11X

ZPE (DH)

G253 (DH)





prob

pTS.H+ 



W11X

H1

2.152

2.205

0.042

H16

2.032

2.154

0.046

H2

2.036

2.155

0.046

H17

2.156

2.213

0.041

H3

2.179

2.238

0.039

H18

2.183

2.238

0.039

H4

2.158

2.200

0.042

H19

2.168

2.235

0.039

H5

1.978

2.009

0.062

H20

2.147

2.219

0.041

H6

2.120

2.155

0.046

H21

2.157

2.222

0.040

H7

2.038

2.069

0.055

H22

1.955

2.130

0.048

H8

2.176

2.235

0.039

H23

2.007

2.142

0.047

H9

2.173

2.231

0.040

pTSA.W2A

H10

2.031

2.152

0.046

H1

2.043

2.078

0.234

H11

2.163

2.214

0.041

H2

2.152

2.211

0.179

H12

2.148

2.220

0.040

H3

2.017

2.153

0.201

H13

2.145

2.217

0.041

H4

2.145

2.214

0.179

H14

2.176

2.226

0.040

H5

1.986

2.139

0.207

H15

2.180

2.238

0.039

-24-


Page 25 of 35 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


Discussion The three systems presented above – pTSA-(H2O)n, pTSA-H2O-(D2O)n-1, and pTS-Na+(H2O)m – along with theoretical calculations provide strong evidence supporting two structures: one with two hydrating water molecules localized on the nonionized acid giving rise to the resonance at 2835 cm-1 and the other a large water cluster d-OH donation to the  system giving rise to the resonance at 3642 cm-1. Growth of both these with increasing water suggests a source consisting of relatively dry acid. Since the lowest water content is many times saturation, there must be many water-water bonds. The above two are singled out as follows. The most intense water resonances are expected to be those associated with the Eigen or Zundel ions, however these are known to be very broad. Similarly, the many non-hydronium-associated water-water bonds are not observed because they are also broad under ambient thermal conditions. Among the waternonwater interactions, the dihydrate (S-)O-H — OH2 stretch is overrepresented, high-intensity, and at just the right frequency to match the resonance observed at 2835 cm-1. Likewise the waterπ signal is experimentally known80 both to be at the right frequency (3642 cm-1) and to be narrower than water-water bands. Hence, this section focuses on these two resonances (2835 cm-1 and 3642 cm-1) plus the smaller shoulder at 3660 cm-1. It concludes with a three-structure model: the dihydrate, the multi-water single-acid cluster π donation, and a relatively dry acid source that is consistent with observations. 2835 cm-1 The resonance at 2835 cm-1 has several attributes that support assignment to an acid resonance. Most significantly, it is not present in the salt solution even with many times more water than is added to the acid. Absence in the salt solution eliminates assigning the resonance to any C-H resonance associated with the methyl group, the benzenoid ring, or any solvent impuritywater stretch. Three additional observations narrow potential assignments: the frequency is -25-



unaffected by added water, the frequency does not shift with substitution of D2O for H2O to dissolve the acid, and the intensity increases with added water of either isotopic variant. Thus, the responsible structure must be stable. Our AIMD simulations of the dihydrate structure (Figure 4a) supports that this structure is stable at -20 °C. Interestingly, both the acid and salts63 support an IR-observable two-water structure. In the acid system, having two water molecules permits completion of a ring, stabilizing the dihydrate. Indeed, calculation of the binding energy of pTSA-(H2O)n isomers at -20°C for n = 1-12 (Table I and Figure 5) indicates that on a per water basis, the dihydrate structure is preferred over other cluster sizes from n = 1-12. Thus, there is strong evidence that the dihydrate exists in solution. Importantly, the resonance at 2835 cm-1 is not only present, but shows enhanced intensity gain upon using D2O to dissolve the acid monohydrate relative to the gain using normal water. This suggests that the 2835 cm-1 resonance is site specific in both normal water and D2O. Lack of H/D exchange in the deuterated system does not preclude ionization (i.e. transfer of the H to its acceptor O), at various points in the process, so long as the H never departs from the S-O-HO H-bond in which it starts. 3642 cm-1 and 3660 cm-1 shoulder Like the 2835 cm-1 resonance, the resonance at 3642 cm-1 has several characteristics that aid association with a vibrational mode. The resonance is not detected in the pTS-Na+ salt system, it is somewhat more intense when the acid monohydrate is dissolved in normal water than when dissolved in heavy water, and it is accompanied by a shoulder at 3660 cm-1 that is more prominent for heavy water than normal water. (To facilitate discussion, the region from 3600 cm-1 to 3800 cm-1 will be referred to as the “free OH” region.) The free OH region resonances are accompanied by a broad hydrogen-bonded feature from 3000 cm-1 to 3550 -26-


Page 26 of 35

Page 27 of 35 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


cm-1. The broad resonance is both more intense and more structured for normal water than for heavy water and is nearly non-detectable for the salt. Thus, the 3642 cm-1 feature must be associated with hydrogen-bonded water that clusters due to and is polarized by the acid. The most intense among the many d-OH resonances is the d-OH donating into the aromatic  cloud; consistent with the π cloud polarizing the O-H bond thus enhancing the oscillator strength. There is ample precedent for this assignment.80-88 Significantly, Pribble and Zwier80 report a 4 cm1

wide benzene-water stretch at 3642 cm-1 for n = 5: the same frequency reported in this work.

Alternate assignment to a sulfonate donating OH is eliminated since the observed intensity grows with n. Formation of this cluster is reversible as indicated by its disappearance on warming and reappearance on recooling, consistent with the relative weakness of the OH-π bond. The small shoulder at 3660 cm-1 is assigned to the d-OH of HOD water. There is precedent for this assignment; a 3660 cm-1 resonance was observed previously in the CCl4 matrix for mixed isotope water.89 The heavy water system shows a larger 3660 cm-1 peak due to both d-OH of clusters and d-OH of HOD monomers. Consistent with this assignment, there is a d-OD resonance at 2689 cm-1 that grows significantly with added D2O. Regarding the broad hydrogen-bonded resonance extending from 3000 cm-1 to 3550 cm-1, we note that this broad feature is nearly non-detectable in the salt system but is noticeable in the acid. We believe that this is due to stronger polarization of the water by the acid.

Tying 3642 cm-1 and 2835 cm-1 Assignment of the 2835 cm-1 resonance to the dihydrate and the 3642 cm-1 resonance to the OH-π peak of a stabilized, larger water cluster suggests that both simultaneously exist in solution. -27-



Consider a model consisting of hydrated many-acid clusters breaking into smaller units as water is added. The size and nature of the units is an interesting aspect of the CCl4 matrix: injection of neat water supports only monomers. The matrix accommodates much more water with the acid. Emergence of dihydrates (2:1 water:acid) from the 6:1 water:aid ratio requires some units with larger water:acid ratios. This would be another example of asymmetrical fission of a water-organic droplet as argued by Donaldson, Tuck, and Vaida90,91 based on thermodynamics. The exceptional stability of the dihydrate likely drives the dihydrates to separate from a multi-acid multi-water cluster. Alternately, units of many sizes may initially separate, but subsequent exchange via the monomer population, as suggested by the AIMD run of the trihydrate pTSA.W3B, leads to an equilibrium distribution among cluster sizes in which dihydrates dominate for the reasons described previously. The stable dihydrate signature at 2835 cm-1 suggests that the acid H stays with the sulfonate; it does not exchange with D. As water is added to the pool, additional dihydrates detach increasing the 2835 cm-1 resonance. D2O is slightly more efficient at this detachment, hence the larger 2835 cm-1 resonance enhancement with D2O addition. Larger water-acid clusters also detach or result from dihydrates that collect water. Ultimately, these many-water-acid structures feature a d-OH that dangles over the π system. Polarization due to donation boosts the oscillator strength. Slight bias for a d-OH rather than a d-OD in the mixed isotope system coupled with enhanced cluster splitting with D2O results in a nearly isotope independent growth of the 3642 cm-1 resonance. Further experiments will test this three-structure model.

-28-


Page 28 of 35

Page 29 of 35 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


Summary Experimental and computational data provides new insights into the clustering of water and p-toluenesulfonic acid in a CCl4 matrix. The principal experimental findings are narrow peaks at 2835 cm-1 and 3642 cm-1; these increase with either H2O or D2O addition. With the help of computational studies, these are assigned respectively to the O-H stretch of S-O-H in (pTSA)(H2O)2 and to O-H stretches of water donating to the benzene ring. The model is that the matrix contains a spread of cluster sizes in which the dihydrate is a major species if not the dominant one. Because the dihydrate contains acid:water at just a 2:1 molar ratio, clusters with a larger water:acid ratios must also be present. These large clusters contain water-benzene bonding and are tentatively assigned to 3642 cm-1 signal. Dominance of the dihydrate is buttressed by static ab initio and AIMD studies that suggest it is akin to a “magic number” in terms its relative stability. All clusters are believed to be dynamical in that they can evaporate or condense water molecules via the medium of a small but important monomer population. This study underscores the complexity of a three-species system like water-acid-CCl4.

Acknowledgements TK and MJS would like to thank Tufts University Summer Scholars for supporting this work through a summer grant. We also thank Dr. Patrick Bisson for his technical expertise.

-29-



References 1. Kassner, J. L.; Hagen, D. E., "Clustering of Water on Hydrated Protons in a Supersonic Free Jet Expansion - Comment," J. Chem. Phys. 1976, 64, 1860-1861. 10.1063/1.432330 2. Kebarle, P.; Haynes, R. N.; Collins, J. G., "Competitive Solvation of Hydrogen Ion by Water and Methanol Molecules Studied in Gas Phase," J. Am. Chem. Soc. 1967, 89, 5753-5757. 10.1021/Ja00999a003 3. Kebarle, P.; Searles, S. K.; Zolla, A.; Scarboro.J; Arshadi, M., "Solvation of Hydrogen Ion by Water Molecules in Gas Phase. Heats and Entropies of Solvation of Individual Reactions - H+(H2O)n-1 + H2O → H+(H2O)n," J. Am. Chem. Soc. 1967, 89, 6393-6399. 10.1021/Ja01001a001 4. Schwarz, H. A., "Gas-Phase Infrared-Spectra of Oxonium Hydrate Ions from 2 to 5 μ," J. Chem. Phys. 1977, 67, 5525-5534. 10.1063/1.434748 5. Stace, A. J.; Moore, C., "Solvation of Hydrogen-Ions in Mixed Water Alcohol Ion Clusters," J. Am. Chem. Soc. 1983, 105, 1814-1819. 10.1021/Ja00345a022 6. Wang, Q.; Hammes-Schiffer, S., "Hybrid quantum/classical path integral approach for simulation of hydrogen transfer reactions in enzymes," J. Chem. Phys. 2006, 125. Artn 184102 10.1063/1.2362823 7. Pfaendtner, J.; Yu, X. R.; Broadbelt, L. J., "Quantum chemical investigation of low-temperature intramolecular hydrogen transfer reactions of hydrocarbons," J. Phys. Chem. A 2006, 110, 10863-10871. 10.1021/jp061649e 8. Sakaguchi, N.; Hirano, S.; Matsuda, A.; Shuto, S., "Radical reactions with 2-bromobenzylidene group, a protecting/radical-translocating group for the 1,6-radical hydrogen transfer reaction," Org. Lett. 2006, 8, 3291-3294. 10.1021/ol061162o 9. Haji-Akbari, Amir; Debenedetti, Pablo G. , "Perspective: Surface Freezing in Water: A Nexus of Experiments and Simulations," J. Chem. Phys. 2017, 147, 060901:1-11. 10.1063/1.4985879 10. Zhang, Zhengcai; Guo, Guang-Jun, "The Effects of Ice on Methane Hydrate Nucleation: A Microcanonical Molecular Dynamics Study," Phys. Chem. Chem. Phys. 2017, 19, 19496-19505. 10.1039/C7CP03649C 11. Sloan, E. D.; Koh, C. A., Clathrate Hydrates of Natural Gases Third Edition ed. (CRC Press, 2008) Vol. 119 pgs. 701. 12. Biswas, R.; Carpenter, W.; Fournier, J. A.; Voth, G. A.; Tokmakoff, A., "IR spectral assignments for the hydrated excess proton in liquid water," J. Chem. Phys. 2017, 146. 10.1063/1.4980121 13. Devlin, J. P.; Severson, M. W.; Mohamed, F.; Sadlej, J.; Buch, V.; Parrinello, M., "Experimental and computational study of isotopic effects within the Zundel ion," Chem. Phys. Lett. 2005, 408, 439-444. 10.1016/j.cplett.2005.04.087 14. Kaledin, M.; Wood, C. A., "Ab Initio Studies of Structural and Vibrational Properties of Protonated Water Cluster H7O3+ and Its Deuterium Isotopologues: An Application of Driven Molecular Dynamics," J. Chem. Theory Comput. 2010, 6, 2525-2535. 10.1021/ct100122s 15. Kulig, W.; Agmon, N., "Both Zundel and Eigen Isomers Contribute to the IR Spectrum of the GasPhase H9O4+ Cluster," J. Phys. Chem. B 2014, 118, 278-286. 10.1021/jp410446d 16. Wolke, C. T.; Fournier, J. A.; Dzugan, L. C.; Fagiani, M. R.; Odbadrakh, T. T.; Knorke, H.; Jordan, K. D.; McCoy, A. B.; Asmis, K. R.; Johnson, M. A., "Spectroscopic snapshots of the proton-transfer mechanism in water," Science 2016, 354, 1131-1135. 10.1126/science.aaf8425 17. Fournier, J. A.; Wolke, C. T.; Johnson, C. J.; Johnson, M. A.; Heine, N.; Gewinner, S.; Schollkopf, W.; Esser, T. K.; Fagiani, M. R.; Knorke, H.; Asmis, K. R., "Site-specific vibrational spectral signatures of water molecules in the magic H3O+(H2O)20 and Cs+(H2O)20 clusters," Proc. Natl. Acad. Sci. USA 2014, 111, 18132-18137. 10.1073/pnas.1420734111

-30-


Page 30 of 35

Page 31 of 35 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


18. Fournier, J. A.; Wolke, C. T.; Johnson, M. A.; Odbadrakh, T. T.; Jordan, K. D.; Kathmann, S. M.; Xantheas, S. S., "Snapshots of Proton Accommodation at a Microscopic Water Surface: Understanding the Vibrational Spectral Signatures of the Charge Defect in Cryogenically Cooled H+(H2O)n=2-28 Clusters," J. Phys. Chem. A 2015, 119, 9425-9440. 10.1021/acs.jpca.5b04355 19. Headrick, J. M.; Diken, E. G.; Walters, R. S.; Hammer, N. I.; Christie, R. A.; Cui, J.; Myshakin, E. M.; Duncan, M. A.; Johnson, M. A.; Jordan, K. D., "Spectral signatures of hydrated proton vibrations in water clusters," Science 2005, 308, 1765-1769. 10.1126/science.1113094 20. Shin, J. W.; Hammer, N. I.; Diken, E. G.; Johnson, M. A.; Walters, R. S.; Jaeger, T. D.; Duncan, M. A.; Christie, R. A.; Jordan, K. D., "Infrared signature of structures associated with the H+(H2O)(n) (n=6 to 27) clusters," Science 2004, 304, 1137-1140. 10.1126/science.1096466 21. Pogorelov, V.; Doroshenko; G.Pitsevich; V.Balevicius; V.Sablinskas; B.Krivenko; L.G.M.Pettersson, "From Clusters to Condensed Phase – Ft IR Studies of Water," J. Mol. Liq. 2017, 235, 7-10. 10.1016/j.molliq.2016.12.037 22. Simon, A.; Noble, J. A.; Rouaut, G.; Moudens, A.; Aupetit, C.; Iftner, C.; Mascetti, J., "Formation of Coronene:Water Complexes: Ft-IR Study in Argon Matrices and Theoretical Characterisation," Phys. Chem. Chem. Phys. 2017, 19, 8516-8529. 10.1039/C6CP08559H 23. Takano, Miwako; Takahashi, Toko; Iijima, Kazuki; Yamada, Yasuhiro, "The Acetylacetone-Water Complex in a Low-Temperature Solid Argon Matrix," J. Mol. Spec. 2017, 333, 27-35. 10.1016/j.jms.2017.02.001 24. Searcy, J. Q.; Fenn, J. B., "Clustering of water on hydrated protons in a supersonic free jet expansion," J. Chem. Phys. 1974, 61, 5282-5288, Article. 10.1063/1.1681876 25. Vaden, T. D.; Lisy, J. M.; Carnegie, P. D.; Pillai, E. D.; Duncan, M. A., "Infrared spectroscopy of the Li+(H2O)Ar complex: the role of internal energy and its dependence on ion preparation," Phys. Chem. Chem. Phys. 2006, 8, 3078-3082. 10.1039/b605442k 26. Walters, Richard S.; Pillai, E. Dinesh; Duncan, Michael A., "Solvation Dynamics in Ni+(H20)n Clusters Probed with Infrared Spectroscopy," J. Am. Chem. Soc. 2005, 127, 16599 - 16610. 27. Craig, S. M.; Menges, F. S.; Duong, C. H.; Denton, J. K.; Madison, L. R.; Mccoy, A. B.; Johnson, M. A., "Hidden role of intermolecular proton transfer in the anomalously diffuse vibrational spectrum of a trapped hydronium ion," Proc. Natl. Acad. Sci. USA 2017, 114, E4706-E4713. 10.1073/pnas.1705089114 28. Kuo, M.; Kamelamela, N.; Shultz, M. J., "Rotational structure of water in a hydrophobic environment: Carbon tetrachloride," J. Phys. Chem. A 2008, 112, 1214-1218. 10.1021/jp7097284 29. Shultz, M. J.; Vu, T. H., "Hydrogen Bonding between Water and Tetrahydrofuran Relevant to Clathrate Formation," J. Phys. Chem. B 2015, 119, 9167-9172. 10.1021/jp509343x 30. Shultz, M. J.; Vu, T. H.; Meyer, B.; Bisson, P., "Water: A Responsive Small Molecule," Accounts Chem. Res. 2012, 45, 15-22. 10.1021/ar200064z 31. Vu, T. H.; Kalin, S. D.; Shultz, M. J., "Spectroscopic Identification of Water-Propane Interaction: Implications for Clathrate Nucleation," J. Phys. Chem. A 2010, 114, 6356-6360. 10.1021/jp10178z 32. Vu, T. H.; Shultz, M. J., "Competitive Binding of Methanol and Propane for Water Via MatrixIsolation Spectroscopy: Implications for Inhibition of Clathrate Nucleation," J. Phys. Chem. A 2011, 115, 998-1002. 10.1021/jp109872m 33. Ionic equilibrium Tables. Organic acids Tables. Data., International Union of Pure and Applied Chemistry. Commission on Equilibrium, Ionisation constants of organic acids in aqueous solutionIUPAC chemical data series Serjeant, E. P. and Dempsey, B. editor (Pergamon Press, Oxford ; New York, 1979) pgs. xi, 989 p. 0080223397. 34. Bandy, A. R.; Ianni, J. C., "Study of the hydrates of H2SO4 using density functional theory," J. Phys. Chem. A 1998, 102, 6533-6539. 10.1021/Jp980270s

-31-



35. Givan, A.; Larsen, L. A.; Loewenschuss, A.; Nielsen, C. J., "Matrix isolation mid- and far-infrared spectra of sulfuric acid and deuterated sulfuric acid vapors," J. Mol. Struct. 1999, 509, 35-47. 10.1016/S0022-2860(99)00209-4 36. Hintze, P. E.; Feierabend, K. J.; Havey, D. K.; Vaida, V., "High-resolution spectroscopy of H2SO4, HDSO4, and D2SO4 vapor in the region 1200-10,000 cm-1," Spectrochim. Acta A 2005, 61, 559-566. 10.1016/j.saa.2004.05.006 37. Lane, J. R.; Kjaergaard, H. G.; Plath, K. L.; Vaida, V., "Overtone spectroscopy of sulfonic acid derivatives," J. Phys. Chem. A 2007, 111, 5434-5440. 10.1021/jp0688005 38. Gnanendran, N.; Amin, R., "The effect of hydrotropes on gas hydrate formation," J. Petrol. Sci. Eng. 2003, 40, 37-46. 10.1016/S0920-4105(03)00082-2 39. Gnanendran, N.; Amin, R., "Equilibrium hydrate formation conditions for hydrotrope-waternatural gas systems," Fluid Phase Equilibr. 2004, 221, 175-187. 10.1016/j.fluid.2004.04.013 40. Pierola, I. F.; Agzenai, Y., "Ion Pairing and Anion-Driven Aggregation of an Ionic Liquid in Aqueous Salt Solutions," J. Phys. Chem. B 2012, 116, 3973-3981. 10.1021/jp212202b 41. Chemistry, Turnkey Computational, "PQS ab initio version 4.0 released," 2011. 42. Becke, A. D., "Density-Functional Thermochemistry .3. The Role of Exact Exchange," J. Chem. Phys. 1993, 98, 5648-5652. 10.1063/1.464913 43. Lee, C. T.; Yang, W. T.; Parr, R. G., "Development of the Colle-Salvetti Correlation-Energy Formula into a Functional of the Electron-Density," Phys. Rev. B 1988, 37, 785-789. 10.1103/PhysRevB.37.785 44. Anick, D. J., "Comparison of Hydrated Hydroperoxide Anion (HOO-)(H2O)n Clusters with Alkaline Hydrogen Peroxide (HOOH)(OH-)(H2O)n-1 Clusters, n=1-8, 20: An ab Initio Study," J. Phys. Chem. A 2011, 115, 6327-6338. 10.1021/jp110558y 45. Bryantsev, V. S.; Diallo, M. S.; van Duin, A. C. T.; Goddard, W. A., "Evaluation of B3LYP, X3LYP, and M06-Class Density Functionals for Predicting the Binding Energies of Neutral, Protonated, and Deprotonated Water Clusters," J. Chem. Theory Comput. 2009, 5, 1016-1026. 10.1021/ct800549f 46. Ganesan, A.; Dreyer, J.; Wang, F.; Akola, J.; Larrucea, J., "Density functional study of Cu2+phenylalanine complex under micro-solvation environment," J. Mol. Graph. Model. 2013, 45, 180-191. 10.1016/j.jmgm.2013.08.015 47. Gao, B.; Wyttenbach, T.; Bowers, M. T., "Protonated Arginine and Protonated Lysine: Hydration and Its Effect on the Stability of Salt-Bridge Structures," J. Phys. Chem. B 2009, 113, 9995-10000. 10.1021/jp903307h 48. Lambrecht, D. S.; Clark, G. N. I.; Head-Gordon, T.; Head-Gordon, M., "Exploring the Rich Energy Landscape of Sulfate-Water Clusters SO42- (H2O)(n=3-7): An Electronic Structure Approach," J. Phys. Chem. A 2011, 115, 11438-11454. 10.1021/jp206064n 49. Lee, H. M.; Tarkeshwar, P.; Kim, K. S., "Structures, energetics, and spectra of hydrated hydroxide anion clusters," J. Chem. Phys. 2004, 121, 4657-4664. 10.1063/1.1779566 50. Momany, F. A.; Appell, M.; Willett, J. L.; Bosma, W. B., "B3LYP/6-311++G** geometryoptimization study of pentahydrates of alpha- and beta-d-glucopyranose," Carbohyd Res. 2005, 340, 16381655. 10.1016/j.carres.2005.04.020 51. Ramondo, F.; Bencivenni, L.; Caminiti, R.; Pieretti, A.; Gontrani, L., "Dimerisation of urea in water solution: a quantum mechanical investigation," Phys. Chem. Chem. Phys. 2007, 9, 2206-2215. 10.1063/b617837e 52. Senthilkumar, L.; Umadevi, P.; Nithya, K. N. S.; Kolandaivel, P., "Density functional theory investigation of cocaine water complexes," J. Mol. Model. 2013, 19, 3411-3425. 10.1007/s00894-0131866-0

-32-


Page 32 of 35

Page 33 of 35 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


53. Smith, A.; Vincent, M. A.; Hillier, I. H., "Mechanism of acid dissociation in water clusters: Electronic structure studies of (H2O)nHX (n = 4, 7; X = OH, F, HS, HSO3, OOSO2H, OOH·SO2)," J. Phys. Chem. A 1999, 103, 1132-1139. 10.1021/Jp984216n 54. National Institute of Standards and Technology, Precomputed vibrational scaling factors, 2016. 55. Anick, D. J., "Polyhedral water clusters, II: correlations of connectivity parameters with electronic energy and hydrogen bond lengths," J. Mol. Struc.Theochem. 2002, 587, 97-110. 10.1016/S01661280(02)00100-8 56. Anick, D. J., "Topology-energy relationships and lowest energy configurations for pentagonal dodecahedral, (H2O)20X clusters, X=empty, H2O, NH3, H3O+: The importance of O-topology," J. Chem. Phys. 2010, 132, 164311:1-12. 10.1063/1.3397812 57. Chen, H. N.; Xu, J. Q.; Voth, G. A., "Unusual Hydrophobic Interactions in Acidic Aqueous Solutions," J. Phys. Chem. B 2009, 113, 7291-7297. 10.1021/jp9025909 58. Kuo, J. L.; Ciobanu, C. V.; Ojamae, L.; Shavitt, I.; Singer, S. J., "Short H-bonds and spontaneous self-dissociation in (H2O)20: Effects of H-bond topology," J. Chem. Phys. 2003, 118, 3583-3588. 10.1063/1.1538240 59. Vaden, T. D.; Forinash, B.; Lisy, J. M., "Rotational structure in the asymmetric OH stretch of + Cs (H2O)Ar," J. Chem. Phys. 2002, 117, 4628-4631. 10.1063/1.1503310 60. Vaden, T. D.; Weinheimer, C. J.; Lisy, J. M., "Evaporatively cooled M+(H2O)Ar cluster ions: Infrared spectroscopy and internal energy simulations," J. Chem. Phys. 2004, 121, 3102-3107. 10.1063/1.1774157 61. Weinheimer, C. J.; Lisy, J. M., "Gas-phase cluster ion vibrational spectroscopy of Na+(CH3OH)(27)," J. Phys. Chem. 1996, 100, 15305-15308. 10.1021/Jp9621787 62. Weinheimer, C. J.; Lisy, J. M., "Vibrational predissociation spectroscopy of Cs+(H2O)(1-5)," J. Chem. Phys. 1996, 105, 2938-2941. 10.1063/1.472160 63. Bisson, P.; Xiao, H.; Kuo, M.; Kamelamela, N.; Shultz, M. J., "Ions and Hydrogen Bonding in a Hydrophobic Environment: CCl4," J. Phys. Chem. A 2010, 114, 4051-4057. 10.1021/jp9106712 64. Max, J. J.; Chapados, C., "Isotope effects in liquid water by infrared spectroscopy," J. Chem. Phys. 2002, 116, 4626-4642. 10.1063/1.1448286 65. Zhou, X. Y.; Wang, C. L.; Wu, F. M.; Feng, M.; Li, J. Y.; Lu, H. J.; Zhou, R. H., "The ice-like water monolayer near the wall makes inner water shells diffuse faster inside a charged nanotube," J. Chem. Phys. 2013, 138, 204710:1-6. 10.1063/1.4807383 66. James, T.; Wales, D. J., "Protonated water clusters described by an empirical valence bond potential," J. Chem. Phys. 2005, 122, 134306:1-11. 10.1063/1.1869987 67. Wei, S.; Shi, Z.; Castleman, A. W., "Mixed Cluster Ions as a Structure Probe - ExperimentalEvidence for Clathrate Structure of (H2O)20H+ and (H2O)21H+," J. Chem. Phys. 1991, 94, 3268-3270. 10.1063/1.459796 68. Yang, X. L.; Castleman, A. W., "Large Protonated Water Clusters H+(H2O)n (1≤ n>60) - the Production and Reactivity of Clathrate-Like Structures under Thermal Conditions," J. Am. Chem. Soc. 1989, 111, 6845-6846. 10.1021/Ja00199a056 69. Shields, Robert M.; Temelso, Berhane; Archer, Kaye A.; Morrell, Thomas E.; Shields, George C., "Accurate Predictions of Water Cluster Formation, (H2O)n=2−10," J. Phys. Chem. A 2010, 114, 11725-11737. 10.1021/jp104865w 70. Anick, D. J., "O-H stretch modes of dodecahedral water clusters: A statistical ab initio study," J. Phys. Chem. A 2006, 110, 5135-5143. 10.1021/jp055632s 71. Buck, U.; Ettischer, I.; Melzer, M.; Buch, V.; Sadlej, J., "Structure and spectra of three-dimensional (H2O)n clusters, n = 8, 9, 10," Phys. Rev. Lett. 1998, 80, 2578-2581. 10.1103/PhysRevLett.80.2578

-33-



72. Sadlej, J., "Theoretical study of structure and spectra of cage clusters (H2O)n, n=11,12," Chem. Phys. Lett. 2001, 333, 485-492. 10.1016/S0009-2614(00)01397-X 73. Sadlej, J.; Buch, V.; Kazimirski, J. K.; Buck, U., "Theoretical study of structure and spectra of cage clusters (H2O)n, n=7-10," J. Phys. Chem. A 1999, 103, 4933-4947. 10.1021/Jp990546b 74. Barth, H. D.; Buchhold, K.; Djafari, S.; Reimann, B.; Lommatzsch, U.; Brutschy, B., "Hydrogen bonding in (substituted benzene)·(water)n clusters with n

Matrix Isolation Spectroscopy: Aqueous p-Toluenesulfonic Acid

Recommend Documents