ARTICLE pubs.acs.org/JPCA
On the Origin of the Solid-State Thermochromism and Thermal Fatigue of Polycyclic Overcrowded Enes Pance Naumov,*,†,§ Nobuo Ishizawa,‡ Jun Wang,‡ Ljupco Pejov,§ Martin Pumera,|| and Sang Cheol Lee^ †
)
Department of Material and Life Science, Graduate School of Engineering, Osaka University, 21 Yamada-oka, Suita, Osaka, Japan ‡ Ceramics Research Laboratory, Nagoya Institute of Technology, Tajimi, Gifu, Japan § Institute of Chemistry, Faculty of Science, SS Cyril and Methodius University, Arhimedova 5, MK1001 Skopje, Macedonia Division of Chemistry and Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, 21 Nanyang Link, 637371 Singapore ^ School of Advanced Materials and Systems Engineering, Kumoh National Institute of Technology, Gumi 730701, Korea
bS Supporting Information ABSTRACT: Aimed at unraveling the relative contribution of the folding, twisting and bending in the mechanism of the solid-state thermochromism of overcrowded polycyclic aromatic enes (PAEs), the structures of two typical heteromeric and homomeric representatives, 2-(thioxanthen-9-ylidene)indane-1,3-dione (1) and 9,90 -bi-9(10H)-anthracenylidene-10,100 -dione (bianthrone, 2), were studied by temperature-resolved single crystal X-ray diffraction (120530 K) and solid-state UVvisible spectroscopy. Aside from negligibly small unfolding of the tricyclic moiety of 1, this first direct diffraction study of the high-temperature structures of solid PAEs did not unravel any significant and detectable changes in the time- and space-averaged intramolecular structures, thus, showing that the PAE-type thermochromism is not due to phase transitions or to major and permanent molecular distortions of a large portion of the material that would be caused by folding, twisting and/or bending. Instead, the experimental observations and theoretical modeling indicated that the color change is probably due to a dynamic process, where the absorption spectrum changes as a result of enhanced thermal oscillations of the two halves of the molecules around the central bridge. In addition to the reversible coloration, we also observed irreversible processes of thermal fatigue that afford stable chemical products that absorb in the visible region. We showed that the stable products are conductive and they act electrocatalytically toward oxidation of several biomarkers.
1. INTRODUCTION From the first report of the thermochromism of bianthrone (2, Figure 1) solution by Meyer in 1909,1,2 to later observations using crystalline overcrowded polycyclic aromatic enes (PAEs), there has been continual interest in this phenomenon by organic chemists. The renewed interest in overcrowded PAEs stems not only from the fundamental importance of their conformational energy landscapes affected by the tendency to avoid closer-thanvan-der-Waals interatomic contacts in the fjord regions,3,4 but also from the related optical and charge-transfer properties as well as their application as photoactive switchable units in inclusion compounds or multiphase mixtures.5,6 Notably, PAEs are convenient models for chiro-optical switching and small molecular motors.710 Despite the extensive theoretical treatment given to the in vacuo models,1115 a scientific explanation for the interplay among the twisting, folding, and bending (Figure 1) in the evolution of the high-temperature (HT) thermochromic forms of crystalline overcrowded PAEs, supposedly identical with the respective piezochromic and low-temperature (LT) photochromic species, has remained unresolved for a century.16,17 From r 2011 American Chemical Society
Figure 1. Chemical structures, with atom labeling, of the PAEs 1 and 2. Right: schematic representation of the folding, twisting and bending of PAEs, exemplified with portions of 1 and exaggerated for clarity.
as many as 12 proposed mechanisms,18 based on experimental evidence, the initially postulated radical pathways19 have been Received: April 30, 2011 Revised: June 19, 2011 Published: June 22, 2011 8563
dx.doi.org/10.1021/jp2040339 | J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A eventually discredited,20,21 and the thermochromic effect was identified as intramolecular phenomenon. An early attempt to elucidate the structure of the thermochromic form of the dixanthylene based on room-temperature (RT) diffraction data on two polymorphs was inconclusive.22 Recently, a convincing crystallographic evidence was provided18 of two distinctly colored polymorphs (yellow and purple) of a bistricyclic PAE, which was in accordance with the earlier indications23 that the thermochromism of the PAEs is due to switching between folded and twisted potential minima. Although such a large conformational change can explain the color difference between different polymorphs, and it also appears to be a viable mechanism for the thermochromism in solution, it appears less probable in pure crystal where it would enforce prohibitively large perturbation on the environment for a process which is known to proceed typically in a single-crystal-tosingle-crystal manner. Aside from the problem of overcrowding in relation to the photophysics and photo/thermochromism, which has been addressed in detail by Agranat et al.11 by thorough theoretical study of the conformational spaces of several systems related to the PAEs studied here, the existing data on treatments of solid-state thermochromic systems at high level of theoretical sophistication are rather scarce. Dorogan and Minkin24 utilized isolated cluster calculations to explain the thermal isomerization of perimidinespirocyclohexadienones in the solid state. Lukes and Breza25 reported the relation between the thermochromism and the thermal motion of aromatic rings in bithiophene and 3-butylbithiophene based on a combination of molecular dynamics (MD) simulations with the AM1 Hamiltonian and subsequent calculation of the optical spectra for series of snapshots from the MD trajectory. The relation between the thermochromism and the internal rotation in polythiophenes was elaborated by Elmaci and Yurtsever26 by employing computation of the absorption spectra by TD-DFT methods at various angles of the intramolecular torsion. In the attempt to explain the thermochromism in poly(3-alkylthiophene), lattice energy and Monte Carlo simulations have been carried out by Corish et al.,27 particularly with respect to the conformational energy profiles of the butyl side chains. Recently, periodic DFT calculations were utilized by Filhol et al.28 to study the coloration properties of polydiacetylenes. Of all experimental methods that have been employed to study the thermochromism of PAEs, the most direct information about the origin of this phenomenon could be obtained by X-ray diffraction (XRD) analysis of their structures at HT. Generally, the HT XRD technique has been rarely applied in the studies of organic thermochromism in the past.29 We report herein the crystal structures of the indanedione 1 (2-(thioxanthen-9-ylidene)indane-1,3-dione) and bianthrone 2 (Figure 1) determined at temperatures up to their melting points with a genuine setup for in situ single-crystal HT XRD. The results provided a direct insight into the origin of the thermochromism of PAE-type compounds. Surprisingly, the analysis showed absence of major XRD-detectable conformational changes of these molecules. The excited-state calculations substantiated the results, and confirmed that the solid-state thermochromism is effected by modification of the excited-state potential energy surface instigated by strong thermal oscillations. Moreover, the HT electronic spectra indicated that the thermochromism of crystalline PAEs is not an entirely reversible phenomenon, and that more than one molecular mechanism is necessary to account completely for all observations related to the color change of different PAEs.
ARTICLE
2. EXPERIMENTAL AND THEORETICAL DETAILS 2.1. X-ray Diffraction. The crystal data and structure refinement details are listed in Tables S1 and S2 (Supporting Information). All temperature values were corrected linearly for differences between the measured and the actual temperature at the sample position, and the latter were determined independently by direct measurement using an appropriate thermocouple (the linear temperature dependence between the measured and the real values was confirmed for the whole temperature interval used in this study). In the LT single-crystal XRD experiments, the crystals were glued to the tip of a glass rod and cooled to 120 K with an open-flow cooling system (Japan Thermal Co.) using double sheath of nitrogen gas recycled from air. The diffraction data were collected in ω-oscillation mode (0.3 image width) with Bruker APEX three-circle diffractometer equipped with CCD detector, using graphite-monochromatized MoKR X-ray radiation (0.71073 Å) obtained from sealed-tube X-ray source. The data were collected with the software SMART30 and then merged and integrated with SAINT.30 The data were further processed by XPREP and corrected for absorption effects with SADABS.30 The structures were solved with direct methods31 and refined on F02,32 assigning anisotropic displacement parameters to all non-hydrogen atoms. In the experiments performed at RT and at temperatures higher than the RT, the reflections (0.5 image width, 35 s exposure for each frame) were collected with APEX diffractometer operating with the APEX2-W2K/NT software (Bruker AXS), equipped with CCD detector, by using MoKR X-ray radiation (0.71073 Å) obtained from a rotating anode source.33 To control the temperature of the sample, a specially designed double-gas heating device (Japan Thermal Co.) with precisely controllable temperature of the nitrogen gas was employed.34 After the data were collected at 296 K, the temperature of the crystal was slowly increased (10 K/min) and additional data sets were collected at various points of constant temperature. The accumulation of the data at each temperature took about 24 h; in certain cases, the data collection time had to be reduced to avoid significant sample decomposition. Due to the difficulties related to the tendency for sublimation in the hot gas stream (in the case of 1), in order to prevent direct exposure of the sample, for the HT measurements the single crystalline samples were supported by L0.1 mm soda glass rods and enclosed in L0.3 mm silica glass capillaries. Several sets of experiments, employing various heating regimes (heating/cooling speed and time/temperature of annealing), were performed in order to study in detail the thermal behavior of the samples, which depended critically on the conditions, and to check the reproducibility. The use of long-term exposure (hours) to very HT inevitably affected the quality of the data, which occasionally necessitated use of fresh samples for the experiments performed at different temperatures; the reported values represent the best out of several sets of temperature-controlled experiments. For better comparison of the refined data, samples of similar size were always used in such cases. The rate of temperature change was 10 K/min in most cases. In the case of compound 2, it was found that rapid heating (100 K/min) causes melting, while slow heating results in maintenance of the solid phase (in addition to the thermochromic change, as explained in the main text). The integrated and scaled data were empirically corrected for absorption effects with SADABS.35 All data sets were corrected for absorption effects, and except for the data of 1 at 120 and 296 K, all data were also 8564
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A
ARTICLE
Table 1. Selected Intramolecular Parameters in the Molecular Structures of 1 and 2 in Single Crystalline State χ(C90 )d ()
χ(C9)c ()
d(C9C90 ) (Å)
θa ()
ωb ()
120
42.59(7)
11.72(66)
3.27(35)
14.72(36)
1.3675(29)
4.69
298
43.84(12)
10.99(1.11)
3.15(58)
14.08(61)
1.3732(43)
5.16
T (K)
R1 (%)
2-(thioxanthen-9-ylidene)indane-1,3-dione (1)
400
44.93(14)
10.76(1.22)
2.37(63)
14.99(69)
1.3612(46)
5.34
454
45.26(17)
10.54(1.36)
2.12(69)
15.31(77)
1.3702(51)
5.75
465
45.16(16)
10.38(1.26)
3.94(64)
14.69(72)
1.3587(47)
5.71
487
45.48(19)
10.75(1.55)
3.23(79)
14.47(87)
1.3634(58)
5.64
296
44.02(12)
11.05(1.12)
2.09(58)
15.96(62)
1.3682(43)
5.18
+0.25(1.23)
0.0041(87)
e
change 367
+2.89(19)
0.97(2.21)
+0.07(1.14)
0
0
9,9 -bi-9(10H)-anthracenylidene-10,10 -dione (bianthrone, 2) 120
39.54(4)
0
4.13(27)
+4.13(27)
1.3580(25)
3.85
296
39.84(6)
0
3.65(39)
+3.65(39)
1.3529(34)
4.52
432
39.99(7)
0
3.44(42)
+3.44(42)
1.3543(36)
5.16
508
40.20(8)
0
3.96(47)
+3.96(47)
1.3571(41)
3.53
530
40.30(10)
0
4.33(64)
+4.33(64)
1.3552(51)
4.92
+0.20(91)
0.0028(76)
e
change 410
+0.76(14)
0
0.20(91)
θ, folding angle: the dihedral between the best planes through the two terminal benzene rings. b ω, ethylene twist: ω = 1/2[τ(C8AC9C90 C8A0 ) + τ(C9AC9C90 C9A0 )]. c Pyramidalization at C9,1618,54 χ(C9) = [τ(C9AC9C90 C8A) mod 360] 180. d Pyramidalization at C90 ,1618,54 χ(C90 ) = [τ(C9A0 C9C90 C8A0 ) mod 360] 180. e Difference between the values at the highest and the lowest temperatures. The standard deviations were calculated as sums of the respective deviations. The values exceeding the 2σ threshold are bolded. a
corrected for extinction. The structures were solved by using direct methods36 and refined on Fo2 with SHELXL.32 Except for two atoms (C8A and C40 ), which were treated with isotropic displacement parameters because of their very large motions (these atoms are not included in the calculated deformation parameters and, therefore, they do not contribute to the reported structure changes) in the structure of 1 determined at 465 K, all nonhydrogen atoms were assigned anisotropic parameters and freely refined, and all H-atom positions were treated within the riding model (the bonds to the respective non-H atoms are corrected for temperature effects) and corrected for the thermal effects. The anisotropic tensors of the strongly disordered atoms of 2 at 465 K were also restrained to be approximately isotropic. The absolute value of the pyramidalization angles of molecule 1 observed in the present study correspond well with the values 3 and 15 reported by Stezowski, Agranat et al.23 The pyramidalization angles of 2 (Table 1) are close to the values reported in ref 37, which were cited in refs 12, 38 and 39, +3.6 and 3.6. In the structure of (E)-2,6,20 ,60 -tetra-t-butylbianthrone,40 the respective values are also similar, but somewhat smaller, +2.93 and 2.93. 2.2. Spectroscopic Characterization. The UVvisible spectra were recorded in reflectivity mode from either pure samples or samples diluted with NaCl and sandwiched between two glass plates with ColorEye 3100 UV/vis reflectance spectrophotometer (Gretag Macbeth, U.S.A.), equipped with a customdesigned temperature controlling accessory. 2.3. Electrochemical and Microscopic Characterization. All voltammetric experiments were performed on a μAutolab Type III electrochemical analyzer (Eco Chemie) connected to a personal computer and controlled by General Purpose Electrochemical Systems version 4.9 software (Eco Chemie). Electrochemical
experiments were performed in a 5 mL voltammetric cell at room temperature using a three-electrode configuration. A platinum electrode (Autolab) served as an auxiliary electrode, while an Ag/AgCl electrode (CH instruments) served as a reference electrode. All electrochemical potentials in this paper are stated versus a standard Ag/AgCl reference electrode. A scanning electron microscope (field emission type, JEOL, Japan) was used to study the morphology of the samples. A field emission transmission electron microscope (JEOL, Japan) working at 100 kV was employed to obtain images. All chemicals were from Sigma-Aldrich. 2.4. Theoretical Details. To get a deeper insight into the physical basis of the thermochromism in the presently studied systems, we employed two theoretical approaches based on the time-dependent density functional theory (TD-DFT). In the first approach, the electronic transitions of compounds 1 and 2 were calculated at the fixed experimental geometries determined by X-ray crystallographic methods at 120 and 487 K (for 1) and 120 and 530 K (for 2). In the second approach, the molecules of 1 (at 120 and 487 K) and 2 (at 120 and 530 K) were subjected to BornOppenheimer molecular dynamics (BOMD) simulations, followed by TDDFT calculations on selected snapshots from the equilibrated BOMD trajectories. The BOMD simulations were performed at AM1 level of theory using the Verlet integrator. The temperature was kept constant with the velocity scaling method. The longest simulations were carried out for time periods larger than 3.5 ps. All TD-DFT calculations were carried out within the adiabatic approximation.4143 Two combinations of exchange and correlation functionals were used for these calculations: (i) Becke’s three-parameter adiabatic connection exchange functional (B3)44 in combination with the LeeYangParr (LYP)45 correlation functional (B3LYP), and (ii) PBE1 exchange 8565
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A
ARTICLE
Figure 2. Temperature effects on the reflectance UVvisible spectra (powder samples dispersed in NaCl) and the color of single crystals of 1 (a) and 2 (b) induced by heating (black curves, v) and cooling (red curves, V; rate: 2 K/min for 1 and 5 K/min for 2). Note the sublimation of the microcrystalline debris from 1 in the inset in the left panel.
functional combined with the PBE correlation one (PBE1PBE).46 While the B3LYP combination of functionals has exhibited a remarkable performance for a variety of purposes, the PBE1PBE combination is characterized with certain improvements over the other DFT functionals. It was constructed with the main aim of improving the well-known deficiency in their long-range behavior, as it provides an accurate description of the linear response of the uniform electron gas, correct behavior under uniform scaling, as well as a smoother potential.46 Both computational methodologies employed are hybrid HF-DFT (that is, they include an admixture of HF exchange energy), in contrast to the “pure” DFT ones. It has been reported in the literature that adiabatic hybrid functionals usually provide accurate low-lying excitation energies of molecular systems.47,48 To test the basis set convergence effects, two basis sets were used for orbital expansions (6-31+G(d) and 6-311+G(d)), solving the KohnSham (KS) SCF equations iteratively for each particular purpose of this study, along with the “ultrafine” (99, 590) grid for numerical integration (99 radial and 590 angular integration points). All calculations were carried out with the Gaussian 03 series of codes.49 The TD-DFT calculations were carried out within the linear response formalism, in which the electronic subsystem, initially residing in its ground stationary state (DFT ground state is here taken as a reference state), is subject to a time-dependent perturbation that modifies its external potential v.42 The linear response TD-DFT accounts for both the exchange and Coulomb electronhole correlation and it also includes the electron configuration relaxation upon excitation induced by external perturbation. Vertical excitation energies were calculated as poles of the density response functions (instead of simple differences between energies of the KS orbitals). Oscillator strengths corresponding to the electronic transitions were computed by the relation: 2me fi f j ¼ 2 ðεj εi Þjμij j2 3p where me is the electron mass, p = h/2π, ε denotes the (KS) orbital eigenvalue, while |μij| denotes the transition dipole moment between the states i and j: Z ψi ð B r Þμ^ψj ð B r ÞdrB jμij j ¼
3. RESULTS AND DISCUSSION 3.1. Spectroscopic Analysis of the PAEs. Compounds 123,50
and 251 were selected for this study as typical overcrowded PAEs
with diametrically different molecular structures that exhibit different thermochromism in single crystalline state: whereas 1 is a heteromeric, folded-twisted syn-pyramidalized ene with short CH 3 3 3 O contacts (∼2.39 and 2.73 Å), which turns red at HT,23 2 is a homomeric, folded, antipyramidalized bistricyclic structure with short CH 3 3 3 HC contacts (∼2.93 Å), occasionally also referred to as “HH bonds”,52 which turns green at HT.53 The selection was also based on the relevance of 2 as the first and one of the most studied PAEs, and the existence of the acetic acid (AcOH) solvate of 1, 1 3 AcOH23 in addition to the unsolvated crystal, which is valuable to corroborate the conclusions by considering the structurecolor relationship between 1 and 1 3 AcOH. Crystals of 1 and 2, obtained by vacuum sublimation and slow evaporation, respectively, were analyzed with temperature-resolved optical microscopy, UVvisible spectroscopy, and XRD. Heated single crystals of 1 underwent seemingly reversible, gradual color change from orange (LT) to red (HT). This color change was due to increased absorption of the low-energy peak and the visible absorption edge (Figure 2a). However, after cooling the sample, the spectrum did not return to the original state, indicating that the color change was not entirely reversible. Although the absorption tail on the low-energy side (550650 nm) of the HT form was retained, because of the similar colors of the HT and RT forms (orange-red and red, respectively; see the inset in Figure 2a), such a minor residual permanent change in coloration could not be distinguished visually. In the case of the canary yellow crystals (LT) of 2, the color change was also gradual, but much more drastic: while rapid heating (100 K min1) resulted in melting, by slow heating (10 K min1) to 508 K, the crystallinity was preserved. By increasing the temperature, the color of the yellow single crystals initially changed to effluorescent green (see the inset in Figure 2b), followed by dark green, and finally to black, although their diffraction ability was retained. The original yellow color of the green crystals, and even of the black crystals, which have not been annealed at HT, was recovered by cooling, although the resulting yellow crystals maintained a green shade. Such a switching between two distinct colors of 2 (yellow and nearly black) presents a very unusual example of a complete closure of the visible optical absorption gap in an organic compound. The reversibility of this process strongly depended on the state of the sample, the heating regime and the annealing time (Figure 3a). Annealing at 508 K for 2 h or heating to a higher temperature (578 K) without annealing resulted in permanent change; neither the yellow color nor the diffraction ability of the crystals could be recovered by cooling, as illustrated by the temperature profile of the absorbance 8566
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A at 650 nm (Figure 3b). Unlike the original crystals, the resulting black solid was strongly absorbing, opaque and amorphous, although it retained the habit and integrity of the original crystal. Surprisingly, the product exhibited an extraordinary mechanical hardness and thermal stability: the blocks remained compact in the hot gas stream and did not exhibit any morphological changes at least up to 873 K. 3.2. Temperature-Resolved X-ray Diffraction. Single crystals of 1 and 2 were analyzed by temperature-resolved X-ray diffraction at seven and five temperatures, respectively, from the ranges 120487 K for 1 and 120530 K for 2, to the highest temperatures at which the single crystals remained stable (the refinement details are deposited as SI). At HT, the overall crystal symmetry of the crystals was preserved, but the thermal motions of the atoms in both molecules were strongly enhanced. The oscillations were particularly strong at the terminal regions of the molecules: the tricyclic portion of 1 (especially, the sulfur atom and the atoms C2C7), with the two terminal rings oscillating out of their best planes, and at the phenylene ring of the indane1,3-dione half, which oscillated within its plane (Figure 4). The overall thermal motion at HT was reminiscent of twisting of the larger (2-thioxanthene) half and wagging of the smaller (indane1,3-dione) half of the molecule around the central double bond.
Figure 3. (a) Heating and annealing (inset) effects on the reflectance spectrum of pure 2. (b) Temperature profile of the value of the KubelkaMunk function at 650 nm for pure crystals of 2 subjected to different thermal regimes, and micrographs of the crystals before (canary yellow) and after (black, with metallic luster) irreversible thermal treatment.
ARTICLE
Consequently, the thermal parameters of the atoms at the central double bond and their immediate neighbors remained more isotropic than the other atoms. The motion of the terminal rings in the two symmetry-related halves of 2 can be qualitatively described as a combination of wagging of both halves and folding at the tricyclic moiety. The temperature-resolved structure analysis (for the definitions of distortion parameters1618,54 see Table 1) showed that, except for slight unfolding of 1 (Δθ = +2.89(19); the unfolding of 2 is negligible), all (space-averaged) parameters in both molecules remain constant within the whole temperature range. As will be demonstrated by the theoretical calculations below, the slight unfolding of 1, which is the only value that surmounts the 2σ threshold significantly, is too small to account for the drastic color change. Significant pyramidalization (Δχ)54 at the bridge could not be detected; if at all present, it is within the experimental uncertainty, and significant twisting (Δω) could not be observed either. As shown by the values listed in Table 1, in both compounds, the length of the central bridging bond C9C90 , as well as the length of the carbonyl group C11O1 in 2, remained practically constant, and we could not determine significant changes in the intermolecular interactions. 3.3. Theoretical Analysis of the Thermochromism. The theoretical analysis of the static and dynamic aspects related to the electronic structure of the studied compounds was undertaken in order to get a deeper insight into the physical basis of the overcrowded PAE-type thermochromism. The results from the TD-DFT calculations at fixed experimental geometries are summarized in Tables S3S6 (SI). First of all, it can be seen that the excitation energies as well as the oscillator strengths, are well-converged with the basis set size. Moreover, except for the order of the third and the fourth excited states, the two hybrid functionals afford essentially identical results with both basis sets. The predicted electronic ground states of both 1 and 2 are singlet, and the lowest transitions in both cases were identified as the first excited singlet state. As can be seen from Tables S3S6 in the SI, the temperature-induced changes in the averaged solidstate molecular geometries determined by the X-ray crystallographic techniques are expected to result in only subtle changes in the optical absorption properties. If one considers the snapshots from the BOMD trajectory, on the other hand (typical
Figure 4. ORTEP-style plots (drawn at 30% probability level) of the temperature-resolved structures of 1 (top) and 2 (bottom). For 1, the structures determined at 465 K (heating) and at 296 K (after cooling) are not shown (the data are deposited as SI). 8567
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A
ARTICLE
Figure 5. Computed lowest-lying electronic transitions of 1 and 2 for typical snapshots from the BOMD trajectories (left), and selected geometries overlapped at the bridge with the respective experimental structures (right).
results for 1 and 2 are presented in Tables S7S9), the differences become significantly larger, indicating a dynamic character of the thermochromic effect in the studied compounds. This is even better illustrated in Figure 5, where the simulated δ-like spectral patterns for 1 and 2 for typical snapshots from the BOMD trajectories at the two different temperatures (left), and selected geometries overlapped at the bridge with the respective experimental structures (right) are shown. Therefore, the current theoretical results inevitably show that the electronic transition energies of both molecules are practically insensitive (in both the quantitative and qualitative senses) to the small thermally induced changes in the average molecular geometries obtained from the XRD data, the differences being only 23 nm. However, inclusion of the dynamic effects by calculations on the snapshots from the BOMD trajectories strongly shifted the transition energies of the thermally induced, momentarily perturbed geometries, occasionally causing reordering of the higherenergy electronic states (Figure 5). Thus, based on these results, the thermochromic behavior of the PAEs can be safely identified as a thermally induced dynamic effect. 3.4. Analysis of the Stable Black Products. It is well-known that heat treatment of various organic low- and high-molecular weight substances under an inert atmosphere results in carbonbased conductive materials at temperatures as low as 500 C.5558 Such carbon materials are widely used as electrodes in sensing and energy storage applications. We were intrigued by the thermal stability of the black product with metallic luster obtained during the slow heating of 2 in the HT XRD experiments, and, in order to characterize it more closely, we synthesized larger amount of the material by subjecting bulk amount of
Figure 6. (a, b) Scanning electron microscopy (SEM) image showing morphology and size distribution of carbonized microparticles from thermally treated 2. (c, d) Transmission electron microscopy (TEM) image showing detailed morphology of the microparticles.
2 to heat treatment in an air atmosphere. The resulting black powder material was carefully characterized by electron microscopy and electrochemistry in order to understand its morphology and its performance for electrochemical applications. Scanning electron microscopy (SEM) revealed that carbonization produced particles with dimensions from tens up to 100 μm (Figure 6a). Closer SEM observation does not reveal any porous structure of the materials observable under SEM resolution 8568
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A
ARTICLE
oxidation of dopamine originates at +204 mV with a maximum at +744 mV and a peak current of 19.2 μA at bare SPE while at the modified SPE it originates at +142 mV with a maximum at +692 mV and a peak current of 64.1 μA. The oxidation of uric acid originates at +318 mV with a maximum at +563 mV and a peak current of 7.5 μA at bare SPE while at the modified SPE it originates at +214 mV with a maximum at +476 mV and a peak current of 33.5 μA. The low-potential oxidation of ascorbic acid at the modified electrode is a result of the electrocatalytic effect of the black product while the enhanced currents are caused by the large surface area of the modified electrode. Our experiments clearly show that the product is conductive and that it acts electrocatalytically toward the oxidation of these biomarkers.
Figure 7. Cyclic voltammograms of 5 mM (A) ascorbic acid, (B) dopamine, and (C) uric acid at bare SPE (black line) and modified (red line) electrodes. Conditions: Scan rate = 100 mV s1. Phosphate buffer (50 mM) at pH 7.4.
4. CONCLUSIONS In conclusion, by employing HT XRD analysis with an original experimental setup, we provided direct experimental evidence that, at least in single crystalline state, the solid-state thermochromic change of the PAEs 1 and 2 is not a result of switching between different conformational minima. Instead, the change of the excited-state potential surface appears to be a dynamic effect and was identified as a result of the large molecular distortions due to increased thermal atomic oscillations. We also observed irreversible thermal fatigue processes leading to stable chemical products that absorb in the visible region, which might be similar to the photochemical reactions of 2.31 These products do not appear to be the main contributors to the thermochromism, because they remain stable even after cooling of the sample, which retains its thermochromic properties. The results also demonstrate that the in situ HT XRD technique is a useful method for the analysis of organic thermochromism because it can provide a complementary evidence to the conclusions based on polymorphism. ’ ASSOCIATED CONTENT
(Figure 6b). We turned to transmission electron microscopy (TEM) to ascertain any fine surface structure of the materials (Figure 6c,d). TEM revealed that the particles do not exhibit any observable meso- or macropores at the resolution of the electron microscope used. As mentioned above, the carbonization procedure is similar to that which is used for the preparation of conductive and electroactive pyrolytic films from organic materials.5558 Subsequently, we examined the electrochemical properties of the black particles. We investigated the electrochemical response of modified screen-printed electrodes (SPE) toward oxidation of major biomarkers, such as dopamine, ascorbic acid, and uric acid. Determination of these three biomarkers is of major significance in biomedical studies. From the electrochemical point of view, the dopamine is an example of a compound with heterogeneous electron transfer that strongly depends on the electrocatalysis of the electrode surface. Ascorbic acid and uric acid are typical interferences of dopamine in body fluids; their electrochemistry also depends on the electrode properties.59 Figure 7 shows voltammetric profiles of bare SPE and the modified SPE electrodes. The oxidation of ascorbic acid (Figure 7a) at a bare SPE originates at +160 mV and reaches a maximum at +420 mV with a peak current of 12.6 μA. At the modified electrode, the oxidation of ascorbic acid originates at about +30 mV and reaches a maximum at +420 mV with a peak current of 76 μA. The
bS
Supporting Information. The crystallographic data were deposited within the Cambridge Structural Database under the numbers CCDC 763082763093. Crystallographic data in CIF and tabular (Tables S1 and S2) formats, additional experimental details, micrographs of crystals (Figures S1S5), 1H NMR spectra (Figure S6), and complete ref 49. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*Tel.: +81-(0)6-6879-4574. Fax: +81-(0)6-6879-7806. E-mail:
[email protected];
[email protected].
’ ACKNOWLEDGMENT This work was partially supported by the Grants-in-Aid for Scientific Research No. 18206071 from the JSPS, the Japanese Science and Technology Agency, JST (FRBforGYR program), and partially also by a Global COE program, “the Global Education and Research Center for Bio-Environmental Chemistry” from the Ministry of Education, Culture, Sports, Science and Technology, Japan. We thank Hwee Ling Poh for performing part of the measurements. 8569
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
The Journal of Physical Chemistry A
’ REFERENCES (1) Meyer, H. Ber. Dtsch. Chem. Ges. 1909, 42, 143–145. (2) Meyer, H. Monatsh. Chem. 1909, 30, 165–177. (3) Cheng, P.-T.; Nyburg, S. C. Acta Crystallogr. 1973, B29, 1358–1359. (4) Coppens, P. Acta Crystallogr. 1973, B29, 1359–1360. (5) Huck, N. P. M.; Jager, W. F.; de Lange, B.; Feringa, B. L. Science 1996, 273, 1686–1688. (6) Feringa, B. L.; Huck, N. P. M.; van Doren, H. A. J. Am. Chem. Soc. 1995, 117, 9929–9930. (7) Koumura, N.; Zijlstra, R. W. J.; van Delden, R. A.; Harada, N.; Feringa, B. L. Nature 1999, 401, 152–155. (8) Koumura, N.; Geertsema, E. M.; Meetsma, A.; Feringa, B. L. J. Am. Chem. Soc. 2000, 122, 12005–12006. (9) Feringa, B. L. Acc. Chem. Res. 2001, 34, 504–513. (10) Koumura, N.; Geertsema, E. M.; van Gelder, M. B.; Meetsma, A.; Feringa, B. L. J. Am. Chem. Soc. 2002, 124, 5037–5051. (11) Pogodin, S.; Rae, I. D.; Agranat, I. Eur. J. Org. Chem. 2006, 5059–5068. (12) Levy, A.; Pogodin, S.; Cohen, S.; Agranat, I. Eur. J. Org. Chem. 2007, 5198–5211. (13) Pogodin, S.; Suissa, M. R.; Levy, A.; Cohen, S.; Agranat, I. Eur. J. Org. Chem. 2008, 2887–2894. (14) Biedermann, P. U.; Stezowski, J. J.; Agranat, I. Eur. J. Org. Chem. 2001, 15–34. (15) Agranat, I.; Tapuhi, Y. J. Am. Chem. Soc. 1979, 101, 665–671. (16) Shoham, G.; Cohen, S.; Suissa, R. M.; Agranat, I. Stereochemistry of strained, overcrowded ethylenes. In Molecular Structure: Chemical Reactivity and Biological Activity, IUCr Crystallographic Symposia 2; Stezowski, J. J., Huang, J.-L., Shao, M.-C., Eds.; Oxford University Press: Oxford, 1988; pp 290312. (17) Biedermann, P. U.; Stezowski, J. J.; , Agranat, I. Overcrowded Polycyclic Aromatic Enes. In Advances in Theoretically Interesting Molecules; Thummel, R. P., Ed.; JAI Press: Stanford, CN, 1998; Vol. 4, pp 245322. (18) Biedermann, P. U.; Stezowski, J. J.; Agranat, I. Chem.Eur. J. 2006, 12, 3345–3354. (19) Nielsen, W. G.; Fraenkel, G. K. J. Chem. Phys. 1953, 21, 1619–1619. (20) Theilacker, W.; Kort€um, G.; Elliehausen, H.; Wilski, H. Z. Naturforsch. 1954, B9, 167–168. (21) Theilacker, W.; Kort€um, G.; Elliehausen, H. Chem. Ber. 1956, 89, 1578–1592. (22) Mills, J. F. D.; Nyburg, S. C. J. Chem. Soc. 1963, 308–321. (23) Stezowski, J. J.; Biedermann, P. U.; Hildenbrand, T.; Dorsch, J. A.; Eckhardt, C. J.; Agranat, I. J. Chem. Soc., Chem. Commun. 1993, 213–215. (24) Dorogan, I. V.; Minkin, V. I. Mol. Cryst. Liq. Cryst. 2005, 431, 123–127. (25) Lukes, V.; Breza, M. J. Mol. Struct. (THEOCHEM) 2007, 820, 35–39. (26) Elmaci, N.; Yurtsever, E. J. Phys. Chem. A 2002, 106, 11981–11986. (27) Corish, J.; Feeley, D. E.; Morton-Blake, D. A.; Beniere, F.; Marchetti, M. J. Phys. Chem. B 1997, 101, 10075–10085. (28) Filhol, J-S.; Deschamps, J.; Dutremez, S. G.; Boury, B.; Barisien, T.; Legrand, L.; Schott, M. J. Am. Chem. Soc. 2009, 131, 6976–6988. (29) Naumov, P.; Lee, S. C.; Ishizawa, N.; Jeong, Y. G.; Chung, I. H.; Fukuzumi, S. J. Phys. Chem. A 2009, 113, 11354–11366. (30) SMART-NT, ver. 5.624; SAINT-Plus-NT, ver. 6.02; and SADABS; Bruker AXS, Inc.: Madison, Wisconsin, U.S.A., 1999. (31) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; Burla, M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27, 435–436. (32) Sheldrick, G. M. E. SHELXL-97, Structure refinement program; University of G€ottingen: G€ottingen, Germany, 1997. (33) APEX2-W2K/NT, v1.0; Bruker Analytical X-ray Systems Inc.: Madison, WI, 2006. (34) Ishizawa, N.; Kondo, S.; Hibino, H.; Igarashi, S.; Nakamura, M.; Saho, R. Annual Report of the Ceramics Research Laboratory 2006, Nagoya
ARTICLE
Institute of Technology: Nagoya, Japan, 2007; 6, 1216. For an example of a successful application of the high-temperature diffraction to study phase transitions in inorganic materials, see: Ishizawa, N.; Tateishi, K.; Kondo, S.; Suwa, T. Inorg. Chem. 2008, 47, 558–566. (35) Sheldrick, G. M. SADABS, University of G€ottingen: G€ottingen, Germany, 1996. (36) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112–122. (37) Agpar, P. A.; Wasserman, E. 1978, unpublished results (Allied Chemical). (38) Shoham, G.; Cohen, S.; Suissa, R. M.; Agranat, I. Stereochemistry of strained, overcrowded ethylenes. In Molecular Structure: Chemical Reactivity and Biological Activity, IUCr Crystallographic Symposia 2; Stezowski, J. J., Huang, J.-L., Shao, M.-C., Eds.; Oxford University Press: Oxford, 1988; pp 290312. (39) Biedermann, P. U.; Stezowski, J. J.; Agranat, I. Overcrowded Polycyclic Aromatic Enes. In Advances in Theoretically Interesting Molecules; Thummel, R. P., Ed.; JAI Press: Stanford, CN, 1998; Vol. 4, pp 245322. (40) (a) M€uller, U.; Enkelmann, V.; Adam, M.; M€ullen, K. Chem. Ber. 1993, 126, 1217–1225. (b) Becker, R. S.; Earhart, C. E. J. Am. Chem. Soc. 1970, 92, 5049–5057. (41) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109, 8218–8224. (42) Casida, M. E.; Jamorski, C.; Casida, K. C.; Salahub, D. R. J. Chem. Phys. 1998, 108, 4439–4449. (43) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454–464. (44) Becke, A. D. Phys. Rev. 1988, A38, 3098–3100. (45) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. 1988, B37, 785–789. (46) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865–3868. (47) Peng, B.; Yang, S.; Li, L.; Cheng, F.; Chen, J. J. Chem. Phys. 2010, 132, 34305–19. (48) Tao, J.; Tretiak, S.; Zhu, J. J. Phys. Chem. 2008, B112, 13701–13710. (49) Frisch, M. J. et al. , Gaussian 03, Revision E.01; Gaussian, Inc.: Wallingford, CT, 2004 (the complete reference is deposited as Supporting Information). (50) Sch€onberg, A.; Mustafa, A.; Sobhy, M. E. E.-D. J. Am. Chem. Soc. 1953, 75, 3377–3378. (51) Ault, A.; Kopet, R.; Serianz, A. J. Chem. Educ. 1971, 48, 410–411. (52) Wolstenholme, D. J.; Cameron, T. S. J. Phys. Chem. 2006, 110, 8970–8978. (53) Harnik, E.; Schmidt, G. M. J. J. Chem. Soc. 1954, 3295–3302. (54) Radhakrishnan, T. P.; Agranat, I. Struct. Chem. 1991, 2, 107–115. (55) Murali, S.; Dreyer, D. R.; Valle-Vigon, P.; Stoller, M. D.; Zhu, Y.; Morales, C.; Fuertes, A. B.; Bielawski, C. W.; Ruoff, R. S. Phys. Chem. Chem. Phys. 2011, 13, 2652–2655. (56) Oberlin, A. Carbon 1984, 22, 521–541. (57) Lewis, I. C. Carbon 1982, 20, 519–529. (58) Harris, P. J. F. Philos. Mag. 2004, 84, 3159–3167. (59) McCreery, R. L. Chem. Rev. 2008, 108, 2646–2687.
8570
dx.doi.org/10.1021/jp2040339 |J. Phys. Chem. A 2011, 115, 8563–8570
