Self-Association of Acetic Acid in Dilute Deuterated Chloroform. Wide

Nov 2, 2010 - Rebecca C. Quardokus , Natalie A. Wasio , Ryan D. Brown , John A. Christie , Kenneth W. Henderson , Ryan P. Forrest , Craig S. Lent , St...
0 downloads 0 Views 1MB Size
Download PDF
12168

J. Phys. Chem. A 2010, 114, 12168–12175

Self-Association of Acetic Acid in Dilute Deuterated Chloroform. Wide-Range Spectral Reconstructions and Analysis using FTIR Spectroscopy, BTEM, and DFT Martin Tjahjono,* Shuying Cheng, Chuanzhao Li, and Marc Garland Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (A*STAR), 1 Pesek Road, Jurong Island, Singapore 627833, Singapore ReceiVed: July 20, 2010; ReVised Manuscript ReceiVed: October 6, 2010

The binary solution of acetic acid in CDCl3 was studied at room pressure on the interval T ) 293-313 K with a series of acetic acid concentrations up to 0.16 M. In-situ Fourier transform infrared (FTIR) spectroscopy measurements on the interval of 400-3800 cm-1 were utilized as the analytical method to monitor the spectral changes due to self-association of acetic acid. The band-target entropy minimization (BTEM) algorithm was employed to reconstruct the underlying pure component spectra. Analysis successfully provided two major spectral estimates of acetic acid, namely, the monomer (primarily in the form of monomer-CDCl3 complex) and the centrosymmetric cyclic dimer. In addition, analysis provided one minor spectral estimate containing signals from both noncyclic dimers and higher aggregates. Also, spectral estimates were obtained for phosgene and water which were present at trace levels even though considerable precaution was taken to conduct the experiments under anhydrous and anaerobic conditions. Density functional theory (DFT) calculation was performed to assign the acetic acid structures corresponding to the BTEM spectral estimates. Since the structure of dilute acetic acid has been the subject of numerous studies, the present investigation helps to resolve some issues concerning the speciation of acetic acid at low concentrations in low polarity solvents. In particular, the present study provides for the first time, wide-range spectral reconstructions of the species present. 1. Introduction Molecular self-association occurs in numerous intermolecular hydrogen-bonding situations involving O-H · · · O, N-H · · · N, N-H · · · O, and so forth. Such self-association, particularly in dilute solutions, typically results in the formation of small aggregates. Small aggregates of water,1 ammonia,2 methanol,3 and so forth have been the subject of spectroscopic as well as first-principal studies. Studies of molecular self-association have been of fundamental interest in numerous physical,4 chemistry, and biology5 as well as astrophysical contexts.6 The chemistry of carboxylic acids, and in particular the selfassociation of acetic acid, has also been the subject of numerous spectroscopic and theoretical investigations.7-33 This interest is due in part to the relatively simple structure of acetic acid as well as to the presence of both O · · · H-O and relatively weaker O · · · H-C hydrogen bonding. Infrared spectroscopic measurements of acetic acid have been performed in argon and nitrogen matrixes,9-11,21,29,31 gas phase,7,8,15,28 liquid phase,8,22 and solution.12-14,18,20,23,33 These measurements have revealed the presence of several acetic acid species, such as cis- and transmonomers,29 cyclic dimer, open (linear) dimers,21,28,31 and higher aggregates.13,16 The matrix isolation spectroscopy investigations have been particularly effective in varying the ratios of monomer to dimer (because of changes in pulse duration etc.),21 and this has aided the assignment of infrared bands. Studies in the liquid phase are highly dependent on the polarity of the solvent, which affects the degree of self-association and the distribution of species present.18,26 In contrast to gas-phase studies, considerable line broadening takes place in the condensed phase, and this certainly complicates spectral interpretation especially when there is considerable spectral overlap between species. * To whom correspondence should be addressed. Tel.: (65) 6796-3960; fax: (65) 6316-6185; e-mail: [email protected].

The above examples suggest that experimental studies of selfassociation suffer from difficulties with spectral interpretation. In part, this is due to the inability to isolate any of the species present and is partly due to ever-present spectral overlap. Conventional band fitting deconvolution techniques are typically utilized to overcome these problems, but this is often restricted to rather narrow spectral windows.12,16-18,20,22,23 Alternatively, a spectral reconstruction technique could be applied to overcome the problems of multicomponent spectral overlap and the restriction to narrow spectral windows. The resulting pure component spectral estimates can provide a much clearer assessment of the bands present as well their relative intensities. In the present study, a series of solutions of acetic acid in CDCl3 were prepared and Fourier transform infrared (FTIR) spectroscopic measurements were performed at room pressure on the interval T ) 293-313 K. The collected spectra were analyzed with the band-target entropy minimization (BTEM) algorithm34 to obtain the underlying pure component spectra of the nonisolable acetic acid species. The wide-range spectral reconstructions clearly showed the simultaneous signatures corresponding to the various types of C-O, CdO, O-H, and C-D vibrations. The identities of these species were then confirmed by comparing with density functional theory (DFT) predictions. The use of a deuterated solvent considerably simplified the analysis since nearly all C-H vibrations originated from the acetic acid species. Since wide-range spectral reconstructions are rarely reported, the present study shows the utility of such signal processing in order to better understand the speciation occurring in self-associating systems. 2. Experimental Section Materials. The solvent deuterated chloroform (Cambridge Isotope Laboratories, Inc., 99.8%) was dried over activated

10.1021/jp106720v  2010 American Chemical Society Published on Web 11/02/2010

Self-Association of Acetic Acid molecular sieves type 4 Å (Aldrich) and was kept under argon and was stored in the dark. The argon was purified prior to use by passage through a column containing 100 g of reduced BTScatalyst (Fluka AG Buchs, Switzerland) and 100 g of 4 Å molecular sieves to adsorb trace oxygen and water, respectively. The solute glacial acetic acid (Merck, pro analysi, 99.99%) from freshly opened bottles was used without further purification. To avoid any further exposure to moisture, the acetic acid bottle was stored in a drybox under dry nitrogen. Equipment. The experimental setup consisted of a 25 mL glass jacketed reactor equipped with a magnetic stirrer, a Teflon membrane pump (Cole-Parmer), and a flow-through infrared cell. The temperature of the reactor was kept isothermal using a temperature bath circulator (Polyscience 9105 with temperature stability ( 0.05 K). The solution was pumped in a closed-loop from the reactor through the pump to the infrared flow-through cell and back to the reactor. Argon was used to maintain a total system pressure of 1.013 × 105 Pa. Mid-infrared spectra were collected using a mid-infrared VERTEX 70 FT-IR Bruker spectrometer equipped with a DTGS detector. The KBr single crystal windows used (Korth Monokristalle, Kiel, Germany) had dimensions of 20 mm diameter by 2 mm thickness. Two sets of Viton gaskets provided sealing, and Teflon spacers (thickness of 0.1 mm) were used between the windows. A purified air system (Specken-Drumag, Germany) was used to purge the spectrometer. Experimental Aspects. Three semibatch experiments were carried out with solutions of acetic acid in CDCl3 measured at T ) 283.15, 293.15, and 313.15 K. In a typical experiment, background infrared spectra of the empty chamber were first recorded. Then, ca. 24 mL dry CDCl3 was transferred to the reactor under argon. The stirrer was turned on and liquid was circulated throughout the entire system. The infrared spectrum of the solvent was subsequently recorded. Next, acetic acid was continuously injected (with infusion rates between 0.5 and 1.5 µL/min) using a gastight syringe and a programmable syringe pump (Harvard Apparatus, model PHD 4400) to vary the acetic acid concentrations, and at the same time, the infrared spectra were automatically collected every 2 min. The maximum concentration of acetic acid used in this study was ca. 0.16 M. The absorbance spectra of the liquid in the cell (with path length ) 0.1 mm) were measured in the range of 400-3800 cm-1 with a spectral resolution of 2 cm-1 and 10 scans coadded. 3. Computational Section Spectral Data Analysis. The raw infrared spectra measured at different temperatures were first consolidated into a single absorbance data matrix. These combined data were separately analyzed in two spectral regions with wavenumber ranges of 400-640 cm-1 and 940-3800 cm-1. For each region, singular value decomposition (SVD) and thereafter band-target entropy minimization (BTEM) analysis34 were performed to identify and to reconstruct the underlying pure component spectra of the acetic acid species as well as other constituents. The signal percentage was calculated to quantify the signal contributions from each individual constituent. Finally, relative concentrations were evaluated from the absorbance data and the BTEM pure component spectra estimates following the Lambert-BeerBougert-Law equation. DFT Calculations. The DFT calculations were performed with Gaussian 03.35 Several acetic acid speciessincluding monomer and monomer-CDCl3 complex, dimers, trimer, and tetramerswere modeled, and calculations were carried out to obtain the fully optimized geometry and their corresponding

J. Phys. Chem. A, Vol. 114, No. 46, 2010 12169

Figure 1. Infrared spectra of acetic acid in CDCl3 measured at a temperature of 293.15 K and at several concentrations. Inset: the enlarged infrared spectra in the wavenumber range of 400-640 cm-1.

predicted infrared spectra. All theoretical calculations were carried out at room temperature (298.15 K) and by considering chloroform as a solvent. IEFPCM method was used to account for the solvent model system. B3LYP functional and several basis sets were first used for the acetic acid monomer and the hydrogen-bonded cyclic dimer to achieve good correspondence between the BTEM spectral estimates and the DFT predicted spectra. The finally chosen level of theory (i.e., B3LYP/6311++g(d,p)) was then applied for monomer-CDCl3 complex and other dimers, trimer, and tetramer to better understand the possible structure of the species observed in this study, including species “X” in Figure 2 (vide infra). 4. Results and Discussion Typical raw infrared spectra of binary solution of acetic acid in CDCl3 measured at several different concentrations are shown in Figure 1. Numerous spectral changes primarily due to acetic acid species are clearly identifiable. Spectral changes in the region 640-940 cm-1 are not observed because of the strong absorption of the solvent CDCl3. Accordingly, the spectral analyses were carried out in two separate regions, namely, in the wavenumber ranges of 400-640 cm-1 and 940-3800 cm-1. SVD and BTEM analysis were employed in each region to identify and reconstruct the underlying pure component spectra of acetic acid species and other species observed in the solution. Spectral Analysis in the Range of 940-3800 cm-1. The infrared spectra measured at T ) 283.15, 293.15, and 313.15 K were collated. Singular value decomposition (SVD) followed by BTEM analysis were performed on this combined data set in the wavenumber range of 940-3800 cm-1. Some distinct carbonyl vibrations with maxima at 1713, 1724, and 1757 cm-1 were observed from the right singular vectors (obtained from singular value decomposition), and these bands were subsequently used as the band-targets in the BTEM analysis. These band-targets result in three different spectral estimates of acetic acid species as shown in Figure 2. The first two spectral estimates in Figure 2 are rather consistent with acetic acid monomer and acetic acid cyclic dimer. The bands shown by these two spectral estimates are similar to the bands of acetic acid monomer (trans-conformer)7,9,31

12170

J. Phys. Chem. A, Vol. 114, No. 46, 2010

Tjahjono et al. TABLE 2: Comparison of the Experimental Infrared Spectrum (in cm-1) of Acetic Acid Cyclic Dimer Obtained by BTEM and the Theoretical DFT Prediction at the B3LYP/6-311++g(d,p) Level of Theory experimental DFT

BTEM analysisa

vaporb

Arc

assignmentd

475.7 632.4 904.6 946.6 1028.9 1068.6 1310.5 1389.3 1445.9 1462.3 1468.8 1726.8

480 627 n/a n/a 1016 1052 1293 1359

623 896 942 1014 1066 1290 1430

485.1 636.2 900.4 979.1 1019.7 1052.7 1309.3 1364.3 1420.6 1433.8

1737

1720.1

δ(C-C-O) δ(O)C-O) ν(C-C) τ(O-H) Fs(CH3) Fas(CH3) ν(C-O) δs(CH3) δs(CH3) δs(CH3) δas(CH3) νas(CdO)

1415 1712

(1308) (1364) (1421) (1721)

a

From experimental IR measurement performed in CDCl3. Reference 15. c Reference 31 and ref 10 are given in parentheses. Notes: ν ) stretching, δ ) bending, F ) rocking, τ ) twisting, s ) symmetric, as ) asymmetric.

b d

Figure 2. The reconstructed pure component infrared spectra of three acetic acid species in CDCl3 obtained using BTEM in the range of 940-3800 cm-1. 10,15,31

and acetic acid centrosymmetric cyclic dimer, which were previously studied in the gas phase, N2, and Ar matrixes. Both experimental and theoretical DFT vibrational bands as well as band assignments of the monomer and the cyclic dimer obtained from this study are provided in Tables 1 and 2, respectively. The vibrational bands previously studied in the gas phase, N2, and Ar matrixes are also provided for comparison. The possibility of solute-solvent interaction between acetic acid monomer and solvent CDCl3 in the present study needs to be addressed. Indeed, a previous study by Nyquist et al. showed that an intermolecular hydrogen-bonded complex can be formed between the carbonyl oxygen atom of acetic acid monomer and the H atom of CHCl3.36 This solute-solvent interaction can be inferred from the substantial red-shift observed for the CdO vibration of acetic acid monomer in chloroform as compared to the CdO vibration of acetic acid monomer in the gas-phase or matrix studies. In the present study, a solute-solvent interaction is confirmed by the blue-shift of C-D stretch of CDCl3. In Figure 2, a BTEM

estimate of the monomer shows a small band at ca. 2258 cm-1, and this band is blue-shifted ca. 5 cm-1 from the C-D stretch of the pure solvent CDCl3 (at 2253 cm-1). The blue-shift of the C-D vibration observed in this study is analogous to the blueshift observed for C-H in cyclic ketone-CHCl3 complexes.37 The theoretical DFT vibrational bands of the acetic acid monomer-CDCl3 complex is reported in Table 1. Further discussion regarding the monomer complex and its structure is provided in a later section. In addition to the monomer and cyclic dimer species, the present spectroscopic study also revealed the presence of another acetic acid species, noted as species X in Figure 2. In previous studies, it has been shown that a linear (or open) dimer and higher aggregates could also be formed during the selfassociation of acetic acid both in the gas phase and in the condensed phase.10,13,16,18,22,28,31 The spectral estimate of species X obtained in this study shows some major vibrational bands which have similar band characteristics to those previously reported for the linear dimer acetic acid species.10,28,31 This includes some vibrational bands of C-O at 1237 cm-1, CdO

TABLE 1: Comparison of the Experimental Infrared Spectrum (in cm-1) of Acetic Acid Monomer Obtained by BTEM and the Theoretical DFT Predictions at the B3LYP/6-311++g(d,p) Level of Theory DFT monomer-CDCl3 complex

BTEMa

vaporb

433.7 541.4 588.4 644.6 866.9 1003.9 1064.6 1196.2

437 567 596 634 873 1008 1067 1219

435 543 586 n/a n/a 996 1050 1185

534 581 642 847 989 1048 1182

1326.6 1403.4 1463.8 1468.7 1768.9

1347 1408 1463

1281 1384 1443

1264 1382 1430

1766 2336 3689

1757 2258 3520

1788

3445.9 a

e

experimental

monomer

3583 b

Arc

N2d

assignmente

(428) 534.1 (535) 580.4 (581) 638.2 (639) 849.4 985.3 (987) 1047.2 (1044) 1179.4, 1150.1 (1181, 1152) 1324.4, 1259.1 (1259) 1379.5 (1380) 1430.8, 1438.8 (1434, 1439) 1779.2 (1779)

432 548 588 659 857 992 1049 1185

δ(C-C-O) τ(O-H) δ(O)C-O) τ(C-CdO) ν(C-C) Fs(CH3) Fas(CH3) ν(C-O)

1284 1389 1442,1437

δ(C-O-H) δs(CH3) δs(CH3) δas(CH3)

1778

3564.0

3547

ν(CdO) ν(C-D) ν(O-H)

c

(3566)

From experimental IR measurement performed in CDCl3. Reference 7. Reference 31 and ref 9 are given in parentheses. d Reference 9. Notes: ν ) stretching, δ ) bending, F ) rocking, τ ) twisting, s ) symmetric, as ) asymmetric.

Self-Association of Acetic Acid

J. Phys. Chem. A, Vol. 114, No. 46, 2010 12171 TABLE 3: Percentage of Signal Arising from Each Component in the Measured Spectra

a

component

signal contribution (%)

CDCl3 acetic acid, cyclic dimer acetic acid, monomera acetic acid, species X water phosgene moisture CO2 total

81.65 16.33 1.47 0.04 0.15 0.04 0.09 0.06 99.83

Primarily as monomer-CDCl3 complex.

Figure 3. BTEM pure component spectra estimates in the range of 940-3800 cm-1 of other components observed in this study. Artifacts arising because of the strong absorption of the solvent CDCl3 at wavenumber 2253 cm-1 are indicated by the dotted line.

at 1724 cm-1, and broad O-H bands at ca. 3300 cm-1. However, the spectral estimate obtained in this study is somewhat more complex since it consists of three bands (instead of two) in the carbonyl stretching region. It is highly probable that the spectral estimate X does not belong to a pure linear dimer. Rather, it could be the combination between the linear dimers and the higher aggregates. Support from the DFT calculations is provided in a later section. In addition to these three spectral estimates for acetic acid, BTEM analysis also identified some other pure component spectra belonging to the solvent CDCl3, traces of phosgene and water, and vapor-phase moisture and carbon dioxide from the background (IR chamber). The pure spectra estimates of these components are shown in Figure 3. Although BTEM analysis identified the presence of phosgene and water in the solutions, the signal-to-noise ratios of these spectral estimates are relatively low. This indicates that these species are present at very low concentrations (indeed, a calibration experiment indicates that the maximum concentration of water is less than 3 × 10-4 mole fraction). Phosgene as an impurity can be formed because of the oxidation of CDCl3 in the presence of oxygen and light. To confirm the presence of phosgene, separate infrared measurements were conducted on dried CDCl3 (treated with molecular sieves under argon) and on untreated CDCl3. The spectrum of untreated CDCl3 had pronounced bands at ca. 1807 cm-1 and at 1601 cm-1 corresponding to phosgene and water, respectively. This result supports the use of treated CDCl3 in order to minimize the presence of the phosgene and water impurities in solution. The BTEM spectral estimates provided in Figures 2 and 3 were used to assess the signal contribution from each component. Accordingly, the percentage of signal arising from each component was used to distinguish the major and minor species. The calculated percentage of signal arising from each component is provided in Table 3. The results clearly show that there are only two major solute species, namely, the monomer-CDCl3 complex and the cyclic dimer. The species X identified by BTEM is shown to be a minor component.

Figure 4. Relative concentration profiles for two major solutes, namely, the cyclic dimer (red) and the monomer-CDCl3 complex (blue), from three experimental runs measured at T ) 293.15, 303.15, and 313.15 K.

Figure 4 shows the calculated relative concentrations of two major acetic acid species in the solution. The spectra number shown on the x-axis directly correlates with the infusion time. The nonsmooth concentration profiles are due to changes made to the infusion rates (see Experimental Section for details). The relative concentrations of the remaining solutes are not shown since they were identified as minor constituents in Table 3. The relative concentrations of the moisture and CO2 are not shown in this figure since they are not associated with the solution. Spectral Analysis in the Range of 400-640 cm-1. Similarly, singular value decomposition followed by BTEM analysis was performed on the combined absorbance spectra in the wavenumber range 400-640 cm-1 to identify and reconstruct the pure component spectral estimates of the acetic acid species. Relatively weak bands were observed in this lower region (see inset of Figure 1). Two distinct bands having maxima at 543 cm-1 and 627 cm-1 were observed from the right singular vectors (obtained from singular value decomposition), and they were subsequently used as the band-targets in the BTEM analysis. The respective two spectral estimates obtained from BTEM are shown in Figure 5a. The two spectral estimates of acetic acid species shown in Figure 5a have similarities to the spectra of acetic acid monomer and the acetic acid cyclic dimer which were previously reported for measurements in the gas phase and Ar or N2 matrixes in Tables 1 and 2. The spectral estimate for the minor acetic acid species, namely, species X (shown in Figure 2), could not be

12172

J. Phys. Chem. A, Vol. 114, No. 46, 2010

Tjahjono et al.

Figure 5. The reconstructed pure component infrared spectra of (a) acetic acid species and (b) other species obtained using BTEM in the range of 400-640 cm-1.

Figure 6. The comparison of pure component spectra of acetic acid monomer and monomer-CDCl3 complex obtained from experiment and from DFT calculations in the infrared region of (a) 400-640 cm-1 and (b) 940-3800 cm-1.

successfully obtained using BTEM in this region. This is probably due to its particularly low concentration. In addition to the two primary spectral estimates for the acetic acid species, namely, the monomer-CDCl3 complex and the cyclic dimer, BTEM could successfully reconstruct the pure component spectra of the solvent CDCl3 as well as the moisture (in the IR chamber). The results are shown in Figure 5b. Other components such as carbon dioxide as well as trace water and phosgene, which were previously identified in the range of 940-3800 cm-1, could not be identified using BTEM in the low wavenumber region. This may be due to the lack of vibrational signals in this region or a combination of weak signal and low concentrations. DFT Spectral Predictions. B3LYP functional and several basis sets were used in the initial DFT calculations. These calculations were performed to achieve good correspondence between BTEM spectral estimates and the DFT predicted spectra for the acetic acid monomer and the centro-symmetric cyclic dimer (see Supporting Information Figures S1 and S2, respectively). The finally chosen basis set (i.e., 6-311++g(d,p)) was found to reliably predict the spectra for the acetic acid monomer and the centro-symmetric cyclic dimer. Accordingly, this basis set was used to more confidently assign the vibrations present in these species. In the final step, the theoretical calculations

were further used to better understand the spectrum of the monomer-CDCl3 complex as well as the minor species X. Monomer Acetic Acid and Monomer-CDCl3 Complex. B3LYP functional and several basis sets were used to obtain a good and reliable prediction for the infrared spectrum of the acetic acid monomer species. The results show that the spectrum predicted by B3LYP/6-311++g(d,p) level of theory is quite similar to the experimental spectra. The DFT predicted spectra and the experimental spectra are shown in Figure 6a and b for both the regions 400-640 cm-1 and 940-3800 cm-1, respectively. The optimized geometry of the corresponding monomer acetic acid species is given in Figure 7a. The detailed results from the other DFT calculations for the monomer can be found in the Supporting Information (Figure S1). In addition to the acetic acid monomer, the infrared spectrum of the monomer-CDCl3 complex was also predicted. The DFT predicted spectrum is shown in Figure 6a and b for both the regions 400-640 cm-1 and 940-3800 cm-1, respectively. The optimized geometry of the corresponding monomer-CDCl3 complex is shown in Figure 7b. As can be seen, the predicted infrared spectrum of monomer-CDCl3 complex is more consistent with the BTEM estimate than that of monomer alone. In particular, the relative intensities in the region of 400-640 cm-1 are quite similar, and the blue-shifted C-D stretch can

Self-Association of Acetic Acid

J. Phys. Chem. A, Vol. 114, No. 46, 2010 12173

Figure 7. The optimized geometries of the major solutes (a) monomer, (b) monomer-CDCl3 complex, and (c) cyclic dimer.

be seen in the range of 940-3800 cm-1. Other geometries of monomer-CDCl3 complexes were also considered, for example, hydrogen bonding through the hydroxyl oxygen atom. However, these additional geometries were inconsistent with the BTEM spectral estimate. The predicted spectra of the other geometries are provided in Figure S4 (available in the Supporting Information). Acetic Acid Cyclic Dimer. B3LYP functional and several basis sets were also used in order to obtain a good and reliable prediction for the acetic acid cyclic dimer. Similar to the monomer case, the spectrum predicted by B3LYP/6-311++g(d,p) level of theory is quite similar to the experimental spectra. The comparison between the predicted spectrum and the experimental spectrum is shown in Figure 8a and b for the wavenumber regions of 400-640 cm-1 and 940-3800 cm-1, respectively. The corresponding optimized geometry for the centro-symmetric cyclic dimer structure is shown in Figure 7c. Most of the DFT predicted spectral bands, with the notable exception of those corresponding to the broad OH band, match with the experimental spectra. The broadening of the OH vibrations in this condensed phase spectrum is similar to the broadening effect previously observed in gas-phase or argon matrix studies.11,21,31,32 The broadening effect of the OH band for the cyclic dimer has been addressed in greater detail elsewhere.32,38 The fine structure of the BTEM estimate in the OH region is also seen in the experimental gas-phase spectrum of the cyclic dimer. The detailed results from the other DFT calculations for the cyclic dimer can be found in the Supporting Information (Figure S2). Acetic Acid Species X. Table 3 indicates that the spectrum X contributes only 0.04% of the total signal and represents ca. 0.2% of the acetic acid signal. For very weak signals, various artifacts can arise during spectral reconstructions. In particular, collinearities (superpositions) from the signals of various components can occur. Since the B3LYP/6-311++g(d,p) level of theory satisfactorily predicts the spectra for both the monomer and the cyclic dimer, this level of theory was used to predict the infrared spectra of two noncyclic (or open) dimers, trimer, and tetramer (structures shown in Figure 9). Figure 10 shows the predicted spectra of the linear dimer and the side-on dimer which are considered to be the two most stable noncyclic dimers in gas-phase studies28 as well as higher aggregates, namely, trimer and tetramer.22 These predicted spectra are subsequently compared to the experimental spectrum for species X obtained from BTEM in the wavenumber range of 940-3800 cm-1. The BTEM estimate of species X shows quite broad and complex signals in the C-O region (ca. 1200-1250 cm-1), the methylene region (ca. 1250-1500 cm-1), and the carbonyl region (ca. 1650-1800 cm-1). The complexity of the signals in these regions definitely suggests that more than one species is contributing to the spectral estimate. The most obvious example is the CdO region which has at least three strong vibrations, and these seem to arise from dimers as well as higher aggregate contributions. Signals from higher aggregates, particularly tetramer, appear to be present in the spectral estimate

Figure 8. The comparison pure component spectra of acetic acid cyclic dimer obtained from experiment and from DFT calculation in the infrared region of (a) 400-640 cm-1 and (b) 940-3800 cm-1. Inset: comparison of infrared spectra in the wavenumber range 940-2000 cm-1.

X as shown by the very distinctive O-H vibration at ca. 3670 cm-1. A quite significant portion of the signals in spectral estimate X appears to arise from the linear dimer. Indeed, as shown in Table 4, some of the bands correspond quite well with previous matrix and gas-phase studies.

12174

J. Phys. Chem. A, Vol. 114, No. 46, 2010

Tjahjono et al.

Figure 9. The optimized geometries of the minor acetic acid solutes (a) linear dimer, (b) side-on dimer, (c) trimer, and (d) tetramer.

TABLE 4: Comparison of the Experimental Infrared Vibrational Bands (in cm-1) of Acetic Acid Species X Obtained by BTEM and the Linear (Open) Dimer Reported in Literature BTEM analysisa 1005 1050 1223 (sh) 1237 1297 1324 1361,1368 1390 1422 1699 1724 1754b 3200 (broad) 3455 3672c

ref 31

ref 28

994.6 1056.8 1246.8

1237 1319

1370.1 1390.8 1441.2 1738.5 1760.7 3251.8, (3250)d 3417.0, (3416)d

1769 3300

a From experimental IR measurement performed in CDCl3. Tentatively assigned to side-on or higher aggregates. c Tentatively assigned to higher aggregates (i.e., tetramer). d Reference 10 given in parentheses. b

Figure 10. Comparison of the pure component spectra of acetic acid species X obtained from experiment as well as dimers, trimer, and tetramer obtained from DFT calculations in the infrared region of 940-3800 cm-1.

Other species could also be contributing to the signals present in spectral estimate X. For example, the presence of the sideon dimer in solution has also been considered from a theoretical viewpoint in a previous study.23,25,27,28 The DFT predicted spectrum of this side-on dimer shows two different carbonyl band characteristics at wavenumbers 1726 cm-1 and 1759 cm-1, and these vibrations are also similar to vibrations seen in the spectral estimate of X. Finally, it is certainly possible that dimers and higher aggregates might interact with the solvent CDCl3, as in the case of monomer, to form hydrogen-bonded complexes. However, in the present study, the intensities of any C-D vibrations in

the BTEM spectral estimates of the cyclic dimer and species X are quite weak. This suggests that there may be minimal hydrogen bonding of solvent with these aggregates. This may be because the carbonyl oxygen is already coordinated via selfassociation of acetic acid. Therefore, in this study, further consideration of the wide variety of the possible hydrogenbonded structures with the solvent was not undertaken. 5. Conclusions This study demonstrates for the first time wide-range infrared spectral reconstructions of the acetic acid solutes present in dilute solution. The band-target entropy minimization (BTEM) analysis used for spectral reconstruction successfully provided two major spectral estimates of acetic acid, namely, the monomer (primarily in the form of monomer-CDCl3 complex) and the cyclic dimer, as well as one minor spectral estimate containing signals from both noncyclic dimers and higher aggregates. The DFT calculations performed in this study were useful in the assignment of the structures corresponding to BTEM spectral estimates

Self-Association of Acetic Acid of acetic acid species. The present investigation using widerange spectral reconstructions helps to provide information concerning the speciation of acetic acid at low concentrations in low polarity solvents. This type of approach should be useful in the study of other systems where self-associating or association/hydrogen bonding occurs between dissimilar species. Acknowledgment. This work was supported by Agency for Science, Technology, and Research (A*STAR), Singapore. The authors would like to thank one of the reviewers who raised issues concerning interactions between acetic acid and the solvent CDCl3. Supporting Information Available: Comparison of BTEM spectral estimates of the monomer and cyclic dimer to several DFT predicted spectra are provided in Figures S1 and S2, respectively. Figure S3 shows various possible geometries of the acetic acid-CDCl3 complexes, and Figure S4 shows their corresponding DFT infrared spectra. This information is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Bentwood, R. M.; Barnes, A. J.; Orville-Thomas, W. J. J. Mol. Spectrosc. 1980, 84, 391–404. Huisken, F.; Kaloudis, M.; Kulcke, A. J. Chem. Phys. 1996, 104, 17–25. Kim, K.; Jordan, K. D. J. Phys. Chem. 1994, 98, 10089–10094. Estrin, D. A.; Paglieri, L.; Corongiu, G.; Clementi, E. J. Phys. Chem. 1996, 100, 8701–8711. Buck, U.; Huisken, F. Chem. ReV. 2000, 100, 3863–3890. (2) Slipchenko, M. N.; Kuyanov, K. E.; Sartakov, B. G.; Vilesov, A. F. J. Chem. Phys. 2006, 124, 24110. Kulkarni, S. A.; Pathak, R. K. Chem. Phys. Lett. 2001, 336, 278–283. Beu, T. A.; Buck, U. J. Chem. Phys. 2001, 114, 7853–7858. (3) Dixon, J. R.; George, W. O.; Hossain, Md. F.; Lewis, R.; Price, J. M. J. Chem. Soc., Faraday Trans. 1997, 93, 3611–3618. Hagemeister, F. C.; Gruenloh, C. J.; Zwier, T. S. J. Phys. Chem. A 1998, 102, 82–94. Buck, U.; Huisken, F. Chem. ReV. 2000, 100, 3863–3890. (4) Pimentel, G. C.; McClellan, A. L. The Hydrogen Bond; Freeman: San Fransisco, CA, 1960. Scheiner, S. Hydrogen Bonding: A Theoretical PerspectiVe; Oxford University Press: New York, 1997. Xantheas, S. S. Recent Theoretical and Experimental AdVances in Hydrogen Bonded Clusters; Kluwer Academic: Boston, MA, 2000. (5) Ohya, Y. Hydrogen Bonding. In Supramolecular Design for Biological Applications; Yui, N., Ed.; CRC Press: Boca Raton, FL, 2002; Chapter 3. Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural Chemistry and Biology; Oxford University Press Inc.: New York, 1999. (6) Vigasin, A. A.; Slanina, Z. Molecular Complexes in Earth’s Planetary, Cometary and Interstellar Atmosphere; World Scientific: River Edge, NJ, 1998. (7) Haurie, M.; Novak, A. J. Chim. Phys. 1965, 62, 137. (8) Bellamy, L. J.; Lake, R. F.; Pace, R. J. Spectrochim. Acta 1963, 19, 443–449. (9) Berney, C. V.; Redington, R. L.; Lin, K. C. J. Chem. Phys. 1970, 53, 1713–1721. (10) Grenie, Y.; Cornut, J.-C.; Lassegues, J.-C. J. Chem. Phys. 1971, 55, 5844–5846. (11) Redington, R. L.; Lin, K. C. J. Chem. Phys. 1971, 54, 4111–4119. (12) Bulmer, J. T.; Shurvell, H. F. J. Phys. Chem. 1973, 77, 256–262.

J. Phys. Chem. A, Vol. 114, No. 46, 2010 12175 (13) Goldman, M. A.; Emerson, M. T. J. Phys. Chem. 1973, 77, 2295– 2299. (14) Dervil, E.; Odiot, S. J. Mol. Struct. 1979, 51, 107–115. (15) Marechal, Y. J. Chem. Phys. 1987, 87, 6344–6353. (16) Ng, J. B.; Shurvell, H. F. J. Phys. Chem. 1987, 91, 496–500. (17) Borschel, E. M.; Buback, M. Z. Naturforsch. 1988, 43a, 207–214. (18) Fujii, Y.; Yamada, H.; Mizuta, M. J. Phys. Chem. 1988, 92, 6768– 6772. (19) Turi, L.; Dannenberg, J. J. J. Phys. Chem. 1993, 97, 12197–12204. (20) Malijevska´, I.; Pola´sˇek, M. Collect. Czech. Chem. Commun. 1995, 60, 1094–1100. (21) Halupka, M.; Sander, W. Spectrochim. Acta, Part A 1998, 54, 495– 500. (22) Nakabayashi, T.; Kosugi, K.; Nishi, N. J. Phys. Chem. A 1999, 103, 8595–8603. (23) Nishi, N.; Nakabayashi, T.; Kosugi, K. J. Phys. Chem. A 1999, 103, 10851–10858. (24) Burneau, A.; Ge´nin, F.; Quile`s, F. Phys. Chem. Chem. Phys. 2000, 2, 5020–5029. (25) Nakabayashi, T.; Sato, H.; Hirata, F.; Nishi, N. J. Phys. Chem. A 2001, 105, 245–250. (26) Nakabayashi, T.; Nishi, N. J. Phys. Chem. A 2002, 106, 3491– 3500. (27) Chocholousˇova´, J.; Vacek, J.; Hobza, P. J. Phys. Chem. A 2003, 107, 3086–3092. (28) Emmeluth, C.; Suhm, M. A. Phys. Chem. Chem. Phys. 2003, 5, 3094–3099. (29) Mac¸oˆas, E. M.; Khriachtchev, L.; Fausto, R.; Ra¨sa¨nen, M. J. Phys. Chem. A 2004, 108, 3380–3389. (30) Lewandowski, H.; Koglin, E.; Meier, R. J. Vib. Spectrosc. 2005, 39, 15–22. (31) Sander, W.; Gantenberg, M. Spectrochim. Acta, Part A 2005, 62, 902–909. (32) Emmeluth, C.; Suhm, M. A.; Luckhaus, D. J. Chem. Phys. 2003, 118, 2242–2255. (33) Tjahjono, M.; Allian, A. D.; Garland, M. J. Phys. Chem. B 2008, 112, 6448–6459. (34) Widjaja, E.; Li, C. Z.; Garland, M. Organometallics 2002, 21, 1991– 1997. Chew, W.; Widjaja, E.; Garland, M. Organometallics 2002, 21, 1982– 1990. Widjaja, E.; Li, C. Z.; Chew, W.; Garland, M. Anal. Chem. 2003, 75, 4499–4507. (35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision C.2; Gaussian, Inc.: Wallingford, CT, 2004. (36) Nyquist, R. A.; Clark, T. D.; Streck, R. Vib. Spectrosc. 1994, 7, 275–286. (37) Mukhopadhyay, A.; Mukherjee, M.; Pandey, P.; Samanta, A. K.; Bandyopadhyay, B.; Chakraborty, T. J. Phys. Chem. A 2009, 113, 3078– 3087. (38) Dreyer, J. J. Chem. Phys. 2005, 122, 184306.

JP106720V