Structure and Tunneling Splitting Spectra of Methyl Groups of

Jul 28, 2014 - and TMP BRA complexes, the inelastic neutron backscattering spectra were ... No tunneling splitting is observed in the case of the TMP ...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCA

Structure and Tunneling Splitting Spectra of Methyl Groups of Tetramethylpyrazine in Complexes with Chloranilic and Bromanilic Acids A. Piecha-Bisiorek,*,† G. Bator,† W. Sawka-Dobrowolska,† L. Sobczyk,† M. Rok,† W. Medycki,‡ and G. J. Schneider§ †

Faculty of Chemistry, University of Wrocław, Joliot-Curie 14, 50-383 Wrocław, Poland Institute of Molecular Physics, Polish Academy of Sciences, Smoluchowskiego 17, 60-179 Poznań, Poland § Jülich GmbH, Jülich Centre for Neutron Science, Outstation at FRM II, Lichtenbergstrasse 1, 85747 Garching, Germany ‡

S Supporting Information *

ABSTRACT: The crystal and molecular structure of the 2,3,5,6-tetramethylpyrazine (TMP) complex with 2,5-dibromo-3,6-dihydroxy-p-quinone (bromanilic acid, BRA) has been studied and the results are compared with TMP CLA (2,5-dichloro-3,6-dihydroxy-pquinone (chloranilic acid, CLA) complex. The X-ray structure of TMP BRA complex indicates the formation of dimeric units, in which two BRA− anions are connected by two O−H···O (2.646(2) Å) hydrogen bonds, whereas the cations and anions are joined together by strong N+−H···O− (2.657(2) Å) hydrogen bonds. The results are analyzed in terms of both the methyl group surroundings and the C−H···O and N+−H···O− (or N···H−O) bridge formations. Both effects, the strength of the N+−H···O− hydrogen bonds and steric hindrance for the rotations, are responsible for the CH3 group dynamics. For the TMP CLA and TMP BRA complexes, the inelastic neutron backscattering spectra were also investigated. In the case of TMP CLA, four tunneling signals have been observed in the energy range ±30 μeV, which indicates four inequivalent methyl groups in the crystal structure at the lowest temperature. No tunneling splitting is observed in the case of the TMP BRA complex, most probably due to the overlapping with the elastic peak. The tunneling results are consistent with the 1H NMR spin−lattice relaxation time investigations in a wide temperature range, which also point to the CH3 group tunneling effect in the case of TMP CLA.

1. INTRODUCTION

groups that is reflected in the dynamics of CH3 groups (see Figure 1). The review of problems connected with complex formation of TMP with various organic acids (proton donors) and particularly those with hydrogen bonds has been reported in ref 5. The objects of our interest were the complexes with picric acid,4 squaric acid,3,7 CLA,6 and hydrogen iodine, HI3.8 In addition to the classical physicochemical methods (e.g., nuclear magnetic resonance (1H NMR) as well as theoretical analysis) in the studies on dynamics, the methods of neutron scattering were used. Particularly, the tunneling spectra were collected, which yield important information about a rotation of methyl groups. Such studies require low temperatures, and in the case of good separation one can conclude about the differentiation of methyl groups in the crystal structure, which can be deduced from the complex crystal structure as well. The most important thing, however, is a determination of the parameters of the rotational potential. This will be described later. The objects of our particular interest in the present paper are the crystalline complexes of TMP with BRA and CLA. The

A forecast of the strength of hydrogen bonds in the molecular complexes is important in the designing of the novel organic materials exhibiting polar or semiconducting properties. In such materials, besides the formation of relatively strong hydrogen bonds, we expect either the formation of infinite chains of molecules in the crystal lattice or the stacking alignment of molecules, which favors the desired electrical or optical properties. The 2,5-dichloro-3,6-dihydroxy-p-quinone (chloranilic acid, CLA) or 2,5-dibromo-3,6-dihydroxy-p-quinone (bromanilic acid, BRA) acids as a relatively strong proton donors, are able to form extended supramolecular systems due to two equivalent OH groups.1,2 The strength of the hydrogen bond, however, depends not only on the proton donor property of the acid molecule but also on the proton affinity of the base molecule. 2,3,5,6-Tetramethylpyrazine (TMP) belongs to interesting compounds from the viewpoint of the methyl group dynamics. The molecule is symmetric: the nitrogen atoms are located in the plane vis-à-vis, and all four methyl groups possess identical positions (in an isolated molecule). The complex formation via the engagement of the lone electron pairs as the proton acceptors and electron donors substantially changes the symmetry of molecules and environment of methyl © 2014 American Chemical Society

Received: June 12, 2014 Revised: July 27, 2014 Published: July 28, 2014 7159

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

Article

Figure 1. Examples of molecular complexes with single or double protonated TMP.

complex with CLA was already studied by us, both its X-ray diffraction structure6 and the tunneling splitting, which was presented in details by Prager et al. in refs 9 and 10. In these papers, we reported the crystal structures of the TMP CLA complex at 293, 100, and 14 K. The TMP CLA complex crystallizes (at 293 and 100 K) in the monoclinic P21/c space group, and both TMP and CLA molecules are located on the inversion center. The lattice consists of the infinite chains along the a-axis with strong O−H···N hydrogen bonds (2.708(1) Å), while the TMP and CLA rings are oriented to each other at an angle equal to 84.4(1)°. At 82 K, a phase transition takes place, leading to a doubling of the unit cell. Below the phase transition, the two CLA molecules lie on the inversion centers in space group P21/n, and the TMP molecule lies in a general position. An alternation of the O−H···N hydrogen bond’s lengths linking to the TMP and CLA molecules along the c-axis is observed (see Table S1 and Figure S1 in the Supporting Information). The tunneling splitting in TMP CLA was measured with the backscattering spectrometer BSS of Forschungszentrum Jülich. These results showed that at least three tunnelling peaks might be distinguished.10 One of them was assigned to the two types of methyl groups in the crystal structure (it was the most intensive peak), and the other two were assigned to the two remaining nonequivalent groups. The intensities of two latter peaks should be the same, as it results from theory. However, the observed intensities were different. The BSS experiments were technically challenging at the time they were undertaken; the focus was on qualitative features and on the position of inelastic lines, not their form or intensity. In retrospect, we cannot assess whether the discrepancy between observed and expected tunnelling intensities is an artifact of the spectrometer or a genuine anomaly of our sample. These measurements required a revision, and the third generation instrument

SPHERES at FRM II, Garching (Germany) with its high resolution seems ideally suited to verify the tunnelling splitting observed in Jülich. Thus, the first aim of the present study was the verification of the published tunneling splitting in TMA CLA by means of SPHERES. On the basis of the crystal structures available in the Cambridge Structural Database (version 5.32, November 2013) it turns out that only few organic compounds, with BRA, have been studied so far (for example, with phenazine, pyrazine or quinoxine).11−14 The complex of TMP with BRA has not been studied so far. Its crystal structure was not known, although one could expect that it should be analogous to the structure of TMP CLA with some slight differences of structure and tunneling splitting spectra. The most basic aim of our present paper is to gain a knowledge about the environment effect on the tunnelling splitting and particularly about the interactions of C−H bonds with oxygen atoms.

2. EXPERIMENTAL SECTION Synthesis. The title compound was prepared by mixing equimolar amounts of TMP (Sigma-Aldrich ≥ 98%) and BRA (TCI > 98%) in acetone−acetonitrile solutions. After a few days, the dark violet single crystals were obtained by a slow evaporation from the solution. The purity of the compound was confirmed by an elemental analysis, which gave the following mass percentages: C: 38.73% (theor. 38.74%), N: 6.46% (theor. 6.45), H: 3.21% (theor. 3.25). Thermal and Spectroscopic Measurements. Simultaneous thermogravimetric analysis (TGA) and differential thermal analysis (DTA) were performed on a Setaram SETSYS 16/18 instrument in the temperature range 300−750 K (see Figure S2). 7160

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

Article

NMR. The NMR measurements were made using an ELLAB TEL-Atomic PS 15 (operating at 25 MHz in temperatures from room temperature down to liquid nitrogen) and Tecmag Scout (operating at 24.8 MHz in temperatures below liquid nitrogen) spectrometers. Spin−lattice relaxation times, T1, were measured using a saturation sequence of π/2 pulses followed by a variable time interval τ and a reading π/2 pulse. The magnetization was found to recover exponentially within experimental error at all temperatures. The temperature down to liquid nitrogen of the sample was controlled by a UNIPAN 650 temperature controller operating on Pt 100 sensor providing long time temperature stability better than 1 K. Below temperatures of liquid nitrogen, the helium-cooled Leybold cryostat was applied. The sample of powdered TMP BRA and TMP CLA were evacuated at room temperature and then sealed under vacuum in a glass ampule. All measurements were made on heating the sample. The errors in the measurements of T1 were estimated to be about 5%. 3. Results and Discussion. 3.1. Crystal and Molecular Structures. The TMP BRA complex crystallizes at 100 K in the P1̅ space group with two formula units per unit cell. The asymmetric part of the unit cell, shown in Figure 2, consists of

Fourier Transform Infrared (FT-IR). FT-IR spectra were recorded in the range of 400−4000 cm−1 on a Bruker 113v FTIR spectrophotometer using KBr pellets. X-ray. The X-ray diffraction studies in a case of the TMP BRA complex at 100 K were performed by using a KUMA KM4 CCD κ-axis four circle diffractometer equipped with an Oxford Cryosystem cooler. The graphite monochromated MoKα radiation was applied. The correction for the Lorentz polarization as well as for absorption was taken into account. The structure was solved by direct methods with SHELXS-9715 and refined by the full-matrix least-squares methods by using the SHELXL-97 program.16 The non-hydrogen atoms were refined with anisotropic thermal parameters. All the H atoms were found from the different synthesis. The H atoms bonded to N(1) and O(2) were refined isotropically. The atoms of the methyl groups were allowed to ride, with isotropic values of 1.5 Ueq(C). The crystal data and structure refinements are presented in Table 1. Full crystallographic data are deposited at the Cambridge Crystallographic Data Center (CCDC No 990054). Table 1. Crystal Data and Structure Refinement for TMP BRA at 100 K TMP BRA formula formula weight T (K) crystal system space group a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) V (Å3) Z crystal size (mm) θ range (deg) Dcalc(g cm−3) index ranges

μ(Mo Kα) (mm−1) absorption correction Tmin, Tmax no. of reflections collected no. of independent reflections goodness-of-fit on F2 final R1, wR2 indices [F > 4σ F)] final R1, wR2 indices (all data) Δρmax, min (e Å−3)

C14H14Br2N2O4 434.09 100(2) triclinic P1̅ 8.5101(6) 8.8367(9) 11.4836(13) 80.979(9) 72.450(8) 67.242(9) 758.49(13) 2 0.20 × 0.15 × 0.15 2.84 to 27.5 1.901 −9 ≤ h ≤ 11 −11 ≤ k ≤ 11 −14 ≤ l ≤ 14 5.363 analytical 0.414, 0.500 6642 3433 (Rint = 0.018) 0.969 0.0223, 0.0567 0.0351 0.0528 0.427/ −0.347

Figure 2. Independent part of the unit cell with the atom numbering of the TMP BRA complex. Dashed lines indicate N+−H···O− and C− H···O hydrogen bonds.

the BRA and TMP molecules. Selected interatomic distances and angles are listed in Table 2. In the TMP BRA complex, the Table 2. Selected Bond Lengths (Å) and Angles (°) TMP BRA −

BRA Br(1)−C(11) Br(2)−C(14) O(2)−C(12) O(3)−C(13) O(5)−C(15) O(6)−C(16) TMPH+ N(1)−C(2) N(1)−C(6) N(2)−C(3) N(2)−C(5) TMPH+ C(2)−N(1)−C(6) C(3)−N(2)−C(5)

INS. High-resolution neutron spectra were measured on the backscattering spectrometer SPHERES17 of the Jülich Centre for Neutron Science at Forschungsneutronenquelle Heinz Maier-Leibnitz FRM II (Technische Universität München (TUM), Garching) at temperatures between 2 and 50 K in the energy range ±30 μeV. Software SLAW and FRIDA were used for data treatment and fitting the curves, respectively. The theoretical description of the fitting procedure is presented in the Supporting Information. 7161

1.879(2) 1.888(2) 1.327(2) 1.246(3) 1.250(2) 1.215(3) 1.355(3) 1.343(3) 1.334(3) 1.345(3) 123.4(2) 119.7(2)

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

Article

Figure 3. Dimeric unit of the TMP BRA complex in the crystal lattice. In the crystal structure, the pairs of BRA anions, related by a center of symmetry, are linked through OH···O hydrogen bonds to form the dimer. Each BRA− anion in the dimeric unit is linked to the TMPH+ cation through N+−H···O− hydrogen bond. Dashed lines denote hydrogen bonds.

Figure 4. Part of molecular structure along the a axis. Dashed lines indicate hydrogen bonds in the (bc) plane.

There is no hydrogen-bonding interaction involving the nitrogen atom (N2) at the 4-position of the pyrazine ring, and, as a consequence, no infinite (D···A)∞ linkage was observed, in contrast to the structures of TMP CLA described in the Introduction. In addition to the main O−H···O and N+− H···O− hydrogen bonds, the two auxiliary C(21)−H(23)··· O(5) and C(61)−H(61)···O(6) hydrogen bonds are also formed within the complex moiety (see Figures 3 and 4). Furthermore, the molecules are linked by the C(61)−H(62)··· Br(1)d, C(51)−H(53)···O(2)c (Figure 4) and also C(31)− H(32)···O(2)b and C(61)−H(63)···O(5)e interactions and form a three-dimensional (3D) hydrogen-bonded network (Table 3). In the crystal lattice of the complex, the Br(1)··· Br(2) (x, y − 1, z) contact is also observed (3.744(2) Å), which is shorter than the sum of the van der Waals radii (3.90 Å). 3.2. Infrared Spectra. The comparison of the infrared spectra for TMP CLA and for TMP BRA (Figure 5a,b) confirms univocally the structural data: (i) the asymmetric O−

H atom from the acidic part (hydroxyl group) is shifted to the base species and forms the ionic units of the BRA− and TMPH+ type. It should be underlined that this conclusion is also confirmed by a detailed analysis of the BRA and TMP geometry (see Table 2), especially with regard to the selected bond length (BRA) and angles (TMP). The proton transfer from the acid to base molecule is supported by the shortening of the C−O bonds in the bromoanilic acid molecule (from 1.327(2) to 1.250(2) Å). Moreover, the protonation of the N−H group in TMP widens the CN−C angles (from 119.7(2)◦ to 123.4(2)◦). The molecule of the TMP cation is almost planar and the angle between the planes of the TMP cation and BRA anion is equal to 9.4(1)◦. The structure of TMP BRA complex indicates the formation of dimeric units (see Figure 3), in which the two BRA− anions are connected by two O−H···O (2.646(2) Å) hydrogen bonds, whereas the cations and anions are joined together by strong N+−H···O− (2.657(2) Å) hydrogen bonds. 7162

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

Article

presented by Press18 and in the paper devoted to the molecular complex of Phen PA (1:2).19 Some tunneling results of the TMP CLA complex between 2 and 50 K in the energy range ±30 μeV are shown in Figure 6a.

Table 3. Hydrogen Bonds and Short Contacts for TMP BRA D−H···A (Ǻ )

D−H (Å)

H···A (Å)

D···A (Å)

< D−H···A (°)

N(1)−H(1)···O(5) O(2)−H(2)···O(3)a O(2)−H(2)···O(3) C(21)−H(23)···O(5) C(31)−H(32)···O(2)b C(51)−H(53)···O(2)c C(61)−H(61)···O(6) C(61)−H(62)···Br(1)d C(61)−H(63)···O(5)e

0.96(3) 0.82(2) 0.82(2) 0.96 0.96 0.96 0.96 0.96 0.96

1.72(3) 1.97(2) 2.19(2) 2.39 2.62 2.77 2.31 2.98 2.76

2.657(2) 2.646(2) 2.656(2) 3.190(3) 3.320(3) 3.230(3) 3.257(3) 3.905(2) 3.332(3)

164(3) 140(3) 116(3) 141 130 110 168 162 119

Symmetry code: a1 − x, 1 − y, 2 − z; b1 − x, 1 − y, 1 − z; cx, y, −1 + z; d1 − x, −y, 1 − z; e−x, 1 − y, 1 − z.

Figure 6. Tunneling results for (a) TMP CLA complex at several temperatures and (b) TMP BRA complex at 3 K in the energy range ±30 μeV. The fitting procedure has been performed in an analogous way as it was described in ref 19 (see also Supporting Information).

In the present studies, no tunnelling was found for the complex of TMP BRA (Figure 6b). The most probable interpretation of this fact is related to a difference in the crystal structure of TMP BRA and TMP CLA. The relatively short C−H···O contacts determine the dynamics of the CH3 groups of TMP BRA and most probably cause the larger rotational barrier in comparison to TMP CLA. It should be emphasized, however, that the comparison of the CH3 group surroundings for both compounds is difficult. We have only the structure of TMP BRA at 100 K and any phase transition below this temperature is unlikely (triclinic symmetry at 100 K). The structure of TMP CLA at 100 K is different than at 14 K, therefore it is difficult to compare one of these structures with TMP BRA. The temperature evolution of the tunneling spectra in the energy range ±5 μeV and numbers of bands are more specifically depicted in Figure 7a, while panel b illustrates the ln Δ Ei = ln[ℏwi (T = 0) − ℏwi] plot versus 1000/T. The tunneling frequencies for TMP CLA at 2 K were as follows: (1) ±2.2 μeV (1, 1′), (2) ±3.7 μeV (2, 2′), (3) ±22 μeV (3, 3′), and (4) ±31 μeV (4, 4′) (see Figures 6a and 7a). The activation energy, E01, for Peak 2 is equal to 14.5 meV (more details in the Supporting Information). It should be noted that it is extremely difficult to carry out the analysis of line widths since the peaks overlap one another. However, taking into account the comparable intensities of two of them (at ±2.2 and ±3.7 μeV), we can conclude that they correspond to the two nonequivalent methyl groups in the TMP CLA complex while the two other peaks observed at energies ±22 and ±31 μeV have been assigned to the other nonequivalent CH3 groups. 3.4. NMR Relaxation. The temperature dependencies of the 1H NMR spin−lattice relaxation time (T1) for TMP BRA (red) and TMP CLA (red open) are shown in Figure 8. The behavior of the proton relaxation times with temperature is distinctly different for the two crystals investigated, TMP CLA and TMP BRA. The temperature dependence of the spin−

Figure 5. Infrared spectra of the powdered TMP BRA and TMP CLA complexes in KBr pellets at 300 K. Absorbance in panel a was multiplied by 3. The broad band at ca. 3430 cm−1 corresponds to the water molecules existing in the KBr powder used for the pellet preparation.

H···N hydrogen bonds in TMP CLA6 and (ii) the proton transfer in the case of TMP BRA. The essential difference in the infrared spectra are observed in the whole wavenumber range. The most interesting behavior is related to the stretching vibrations of the O−H···N hydrogen bonds (TMP CLA complex) as well as to the N+−H···O− hydrogen bond (TMP BRA complex with the proton transfer). In the latter case, the dimers of the BRA anionic forms are created. Therefore, we deal with two types of the strong hydrogen bonds. The N+− H···O− bonds in the TMP BRA complex are markedly weaker than the O−H···N ones in TMP CLA. This means that the NH+ bonds are significantly stronger than the O−H ones, and this is why they are characterized by the intensive relatively narrow band at 3185 cm−1. The other band below 3000 cm−1 exhibits complex structure, which consists of the C−O and C O stretching and bending vibration overtones. This band corresponds to the markedly broader band assigned to the O− H···N hydrogen bonds in the TMP CLA complex. 3.3. INS Spectra. High-resolution inelastic neutron backscattering was applied to find scattering function, S(Q,ω), where hQ is momentum transfer and ℏωenergy transfer for the molecular complexes of TMP CLA and TMP BRA in the temperature range between 2 and 50 K. The single-particle model used for description of the CH3 group rotation has been 7163

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

Article

Figure 7. (a) INS spectra of the TMP CLA complex at several temperatures for selected peaks (1 and 2). (b) Arrhenius plot ln[E(0 K) − E(T)] versus 1000/T for the position of Peak 2.

The practical constant value of spin−lattice relaxation time below 50 K suggests a nonclassical mechanism of relaxation and tunneling jumps as the dominant mechanism of the spin− lattice relaxation at low temperatures. This observation is qualitatively consistent with conclusions drawn from the inelastic neutron backscattering experiment. Further analysis of the 1H NMR results is required using different quantum mechanical models, which is planned in the future.

4. CONCLUSIONS In this paper, the results of our studies show the essential role of tunnelling scattering measurements in the analysis of methyl group dynamics. Simultaneously it has been shown that the SPHERES instrument allowed us to obtain valuable results. The most important result is related to a comparison of the tunnel splitting spectrum of the TMP CLA complex with the temperature dependence of the spin−lattice relaxation times from the 1H NMR experiment obtained in the same temperature region. It seems that results of the tunneling splitting yield convincing information about the low-temperature structure and symmetry of methyl groups. The X-ray diffraction studies obtained at 100 K of TMP CLA proved two sets of different methyl groups in the crystal structure. We have observed, however, four tunneling peaks at the lowest temperature, which may be assigned to four methyl groups different in the crystal structure. It is consistent with the suggestion that the TMP CLA complex undergoes a structural phase transition at 82 K leading to the doubling of the unit cells.10 The TMP molecule lies no more in a special position, and the four methyl groups differentiate. Taking into account the observations mentioned above, a discussion of the results obtained by us using SPHERES and by Prager et al.10 in Jülich seems to be worthwhile. According to the latter one only three energy transitions were found, at ±29, ± 21, and ±3.3 μeV. Moreover, the line at ±21 μeV is rather insensitive to the temperature, while the other two transitions follow the normal shift toward the elastic line while broadening. Our measurements disclosed four well-defined peaks at about

Figure 8. Temperature dependence of the spin−lattice relaxation time, T1, for TMP CLA and TMP BRA.

lattice relaxation time for TMP BRA has a form of one wide, perhaps doubled, minimum with T1min = 37 ms at 87 K. Below 28 K, the spin−lattice relaxation time is almost constant at a value of about 3.4 s. We cannot exclude that the quadrupolar effects considerably influence the spin−lattice relaxation at low temperature. The T1 temperature dependence reveals different slopes on each side of the minimum. The slopes allow us to determine the activation energy values to be equal to 8.2 kJ/ mol (1.97 kcal/mol =85.5 meV) and 3.3 kJ/mol (0.8 kcal/mol =35 meV) for the left and right slope, respectively. In turn, the spin−lattice relaxation time dependence for TMP CLA exhibits two minima at 33 and 15 K with T1min equals to 361 and 266 ms, respectively. Solely the left slope of the T1 temperature dependence up to 50 K is similar to the corresponding slope for TMP BRA described above, but the value of activation energy determined is, however, lower and equals to 2.5 kJ/mol (0.6 kcal/mol = 26 meV). On cooling to 6.6 K, a plateau of T1 is observed with rather short values of T1, below 1 s. The above activation energy values correspond to the classical jumps over the barrier. 7164

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

Article

±2.2, ±3.7, ±22 and ±31 μeV. As it was mentioned, we assigned them to four independent methyl groups in the crystal structure of TMP CLA at 2 K. The outermost tunneling peaks show quite different behavior with temperature: the peak at ±22 μeV shifts very unusually toward higher energies, while the peak at ±31 μeV shifts toward the elastic line. Moreover, both analyzed peaks disappear just above 16 K. The line at ±2.2 μeV is rather insensitive to the temperature, whereas the other one, at about ±3.7 μeV shifts toward the elastic line and broadens. One should also consider the discrepancy in the activation energy values estimated by us and by Prager et al.,10 which represent the methyl librational energy. In our measurements the activation energy value, E01, for Peak 2 (±3.7 μeV; see Figure 7) is equal to 14.5 meV and is almost 1.6 times larger than that presented by Prager et al., however, for the peak at 29 μeV (3 K). In our case, it was impossible to estimate the activation energy for the peak at 31 μeV (2 K) since it disappears below 20 K. Nevertheless, the value of E01 estimated by us for Peak 2 is still larger than that found for similar compounds.6,19,20 The lack of the tunneling splitting observation in the TMP BRA complex is due most probably to the stronger interaction of the C−H methyl group with oxygen atoms, and the size of splitting is too small. The splitting signals most probably undergo an overlapping with the central elastic peak. This result is in agreement with the X-ray diffraction studies, which show that the C−H···O bridges in the TMP BRA complex are shorter than in the case of TMP CLA. For comparison, Tables 4 and 5 present the values of activation

bonds (of the length of 2.668(2) and 2.714(2) Å). In the case of TMP BRA, we deal with a completely different situation. Two BRA− ions form dimeric units with two O−H···O hydrogen bonds. On both sites of the (BRA−)2 dimer, the strong N+−H···O− hydrogen bonds are formed with proton transfer to the TMP molecules (there is no proton transfer in TMP CLA). In this way, the assemblies of four moleculestwo BRA− and two TMPH+are created. There are not infinite chains, which could be similar to those observed in the TMP CLA complex. Moreover, the angles between the planes of TMP and acid molecules in these two cases are completely different. They are equal to ca. 90◦ and 10◦ in TMP CLA and TMP BRA, respectively. As it was mentioned above, the C− H···O contacts in TMP BRA are shorter than those in a case of TMP CLA. This different molecular alignment causes different surroundings for the methyl groups and consequently different rotational potential. This gives as a result the four tunnelling peaks in TMP CLA, and no tunnelling is observed for TMP BRA. There are two methods with regards to the solid state that are sensitive to the CH3 group tunneling effects at low temperature: (i) the inelastic backscattering neutron scattering and (ii) 1H NMR (temperature dependence of the spin−lattice relaxation time, T1). Both require the presence of the hydrogen atoms. The results observed by us both in the INS and 1H NMR experiment with regards to the tunneling of the methyl groups at low temperature are qualitatively consistent. Obviously they require quantitative elaboration, which is planned in the future in a separate paper. Nevertheless, we can state that the methods are adequate, and the results shed more light on the problem of CH3 tunneling at low temperature.

Table 4. Activation Energies for Pure TMP Amine and Complex with Tetracyanobenzene Obtained from QENS Measurements compounds TMP (tetramethylpyrazine) TMP TCNB (1:1) (tetracyanobenzene)

activation energy

type of spectra

ref.

51.8 meV 22.8 meV

QENS QENS

20 20

■

ASSOCIATED CONTENT

S Supporting Information *

Details of the thermal (DTA and DTG) and X-ray diffraction measurements (for TMP CLA complex at T = 14 K) as well as theoretical description of the INS fitting procedure. This material is available free of charge via the Internet at http:// pubs.acs.org.

energies for the classical rotation of the CH3 groups obtained from qusielastic neutron scattering (QENS) and the tunnelling activation energies for the molecular complexes containing TMP, respectively. The differences observed both in the infrared and INS spectra for TMP CLA and TMP BRA may be attributed to the different crystal structures of the molecular complexes under study. The differences consist in the alignment of molecules in the lattice. In the crystal structure of TMP CLA, the chains of molecules are formed in such a way that the TMP and CLA molecules lie alternatively. At 14 K, the infinite chains are formed along the a-axis via two types of the O−H···N hydrogen

■

AUTHOR INFORMATION

Corresponding Author

*Address: University of Wrocław, Faculty of Chemistry, F. Joliot-Curie 14, 50-383 Wrocław, Poland. E-mail: anna.piecha@ chem.uni.wroc.pl. Phone: +48 71 375 72 88. Fax: +48 71 328 23 48. Notes

The authors declare no competing financial interest.

Table 5. Comparison of the Hydrogen Bonds Geometry with Tunneling Frequencies and Tunneling Activation Energies for Known Supramolecular Complexes of TMP hydrogen bond lengths (Å) complexes TMP TMP TMP TMP

CLA (1:1) (chloranilic acid) H2SQ (1:1) (squaric acid) PIC (1:2) (picric acid) BRC (bromanilic acid)

D−H···A

D···A(Å)

type of HB

tunneling frequencies

activation energy

type of spectra

ref.

O−H···N O···H−N O···H−N O···H−N

2.692 2.619 2.016 2.657

HB(a,b) PT(a) PT(a,b) PT(a)

3.7 μeV 1.55 μeV 4.20 μeV 3.16 μeV 4.24 μeV −

14.5 meV 15.0 meV 13.2 meV 6.9 meV 6.3 meV −

INS INS INS −

3 10 4 present work

PT - the proton transfer; HB - the hydrogen bond without proton transfer; a,b - the number of protonated nitrogen atoms in TMP (see also Figure 1). 7165

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166

The Journal of Physical Chemistry A

■

Article

(19) Bator, G.; Sobczyk, L.; Sawka-Dobrowolska, W.; Wuttke, J.; Pawlukojć, A.; Grech, E.; Nowicka-Scheibe, J. Structural, Spectroscopic and Theoretical Studies on 3,4,7,8- Tetramethyl-1,10- Phenantroline Complex with Picric Acid. Chem. Phys. 2013, 410C, 55−65. (20) Krawczyk, J.; Nowina-Konopka, M.; Janik, J. A.; Steinsvol, O.; Bator, G.; Pawlukojć, A.; Grech, E.; Nowicka-Scheibe, J.; Sobczyk, L. Quasi-Elastic Neutron Scattering (QENS) Studies on the 1:1 Tetramethylpyrazine-1,2,4,5-tetracyanobenzene Complex. Collect. Czech. Chem. Commun. 2009, 74, 73−84.

ACKNOWLEDGMENTS We acknowledge the support of the Ministry of Science and Higher Education, Poland, under the International Cofinanced Project in 2013.

■

REFERENCES

(1) Klinman, J. P.; Mu, D. Quinoenzymes in Biology. Annu. Rev. Biochem. 1994, 63, 299−344. (2) Badruz Zaman, Md.; Tomura, M.; Yamashita, Y. Crystal Engineering Using Anilic Acids and Dipyridyl Compounds through a New Supramolecular Synthon. J. Org. Chem. 2001, 66, 5987−5995. (3) Prager, M.; Wischnewski, A.; Bator, G.; Grech, E.; Pawlukojć, A.; Sobczyk, L. INS Spectroscopic Study of the 1:1 Tetramethylpyrazine (TMP) Squaric Acid (H2SQ) Complex. Chem. Phys. 2007, 334, 148− 153. (4) Sawka-Dobrowolska, W.; Bator, G.; Sobczyk, L.; Grech, E.; Nowicka-Scheibe, J.; Pawlukojć, A.; Wuttke, J. The (2:1) Complex of Picric Acid with Tetramethylpyrazine: The Structure, IR Spectra and Tunnel Splitting of Methyl Groups. J. Mol. Struct. 2010, 975, 298−302. (5) Bator, G.; Sobczyk, L. Strong Hydrogen Bonds in Selected Complexes of Organic Acid with Tetramethylpyrazine. Wiad. Chem. 2011, 65, 839−869. (6) Sawka-Dobrowolska, W.; Bator, G.; Sobczyk, L.; Grech, E.; Nowicka-Scheibe, J.; Pawlukojć, A. Structure and Vibrational Spectra of 1:1 Chloranilic acid (CLA)−Tetramethylpyrazine (TMP) Complex. Struct. Chem. 2005, 16, 281−286. (7) Nowicka-Scheibe, J.; Grech, E.; Sawka-Dobrowolska, W.; Bator, G.; Sobczyk, L. Structure and Vibrational Spectra of the 1:1 Squaric Acid (H2SQ)−Tetramethylpyrazine (TMP) Adduct. Polym. J. Chem. 2007, 81, 643−652. (8) Sobczyk, L.; Bator, G.; Sawka-Dobrowolska, W.; NowickaScheibe, J.; Grech, E.; Pawukojć, A. Assembly of Protonated Tetramethylpyrazine (TMP) in Triiodide. Vibrational Spectra and DFT Simulations. Polym. J. Chem. 2009, 83, 957−965. (9) Prager, M.; Pawlukojć, A.; Sobczyk, L.; Grech, E.; Grimm, H. Inelastic Neutron Scattering Study of Tetramethylpyrazine in the Complex with Chloranilic Acid. J. Phys.: Condens. Matter 2005, 17, 5725−5739. (10) Prager, M.; Pietraszko, A.; Sobczyk, L.; Pawlukojć, A.; Grech, E.; Seydeln, T.; Wischnewski, A.; Zamponi, M. X-ray Diffraction and Inelastic Neutron Scattering Study of 1:1 Tetramethylpyrazine Chloranilic Acid Complex: Temperature, Isotope and Pressure Effects. J. Chem. Phys. 2006, 125, 194525−1−194525−11. (11) Tomura, M.; Kumai, R.; Takura, Y. J. Proton-Displacive Ferroelectricity in Neutral Cocrystals of Anilic Acids with Phenazine. J. Mater. Chem. 2009, 19, 4421−4434. (12) Horiuchi, H.; Kumai, R.; Takunaga, Y.; Tokura, Y. Proton Dynamics and Room-Temperature Ferroelectricity in Anilate Salts with a Proton Sponge. J. Am. Chem. Soc. 2008, 130, 13382−13391. (13) Zaman, M. B.; Udachin, K. A.; Ripmeester, J. A. Supramolecular Grid and Layer Architectures. Hydrogen Bonds and Halogen− Halogen Interactions Influenced by Bromo-, Chloro-, and CyanoSubstituted Anilic Acids. Cryst. Growth. Des. 2004, 4, 585−589. (14) Tomura, M.; Yamashita, Y. One-Dimensional Supramolecular Tapes in the Co-Crystals of 2,5-Dibromo-3,6-dihydroxy-1,4-benzoquinone (Bromanilic Acid) with Heterocyclic Compounds Containing a Pyrazine Ring Unit. CrystEngComm 2000, 2, 92−95. (15) Sheldrick, G. M. SHELXS-97, Program For Solution of Crystal Structure; University of Götingen, Germany, 1997. (16) Sheldrick, G. M. SHELXL-97, Program For Structure Refinement of Crystal Structure; University of Götingen, Germany, 1997. (17) Wuttke, J.; Zamponi, M. Simulation-Guided Optimization of Small-Angle Analyzer Geometry in the Neutron Backscattering Spectrometer SPHERES. Rev. Sci. Instrum. 2013, 84, 115108−1− 115108−6. (18) Press, W. Single-Particle Rotations in Molecular Crystals; Springer Tracts in Modern Physics; Springer-Verlag: Berlin, 1981; Vol 92. 7166

dx.doi.org/10.1021/jp5058429 | J. Phys. Chem. A 2014, 118, 7159−7166