Electron Spin Relaxation of C60 Monoanion in Liquid Solution

Article pubs.acs.org/JPCA

Electron Spin Relaxation of C60 Monoanion in Liquid Solution: Applicability of Kivelson−Orbach Mechanism Krishnendu Kundu,† Daniel R. Kattnig,‡ Boryana Mladenova,§ Günter Grampp,*,§ and Ranjan Das*,† †

Tata Institute of Fundamental Research, Homi Bhabha Road, Mumbai 400005, India Physical and Theoretical Chemistry Laboratory, University of Oxford, South Parks Road, Oxford OX1 3QZ, United Kingdom § Institute of Physical and Theoretical Chemistry, Graz University of Technology, Stremayrgasse 9/Z2, A-8010 Graz, Austria ‡

S Supporting Information *

ABSTRACT: We report the results of our investigation on the electron spin relaxation mechanism of the monoanion of C60 fullerene in liquid solution. The solvent chosen was carbon disulfide, which is rather uncommon in EPR spectroscopy but proved very useful here because of its liquid state over a wide temperature range. The conditions for exclusive formation of the monoanion of C60 in CS2 were first determined using electrochemical measurements. Using these results, only the monoanion of C60 was prepared by chemical reduction using Hg2I2/Hg as the reducing agent. The EPR line width was measured over a wide temperature range of 120−290 K. The line widths show weak dependence on temperature, changing by a factor of only about 2, over this temperature range. We show that the observed temperature dependence does not obey the Kivelson−Orbach mechanism of electron spin relaxation in liquids, applicable for radicals with low-lying, thermally accessible excited electronic states. The observed temperature dependence can be empirically fitted to an Arrhenius type of exponential function, from which an activation energy of 74 ± 3 cm−1 is obtained. From the qualitative similarities in the characteristics of the spin relaxation rates of C60 monoanion radical and the cyclohexane type of cation radicals reported in the literature, we propose that a pseudorotation-induced electron spin relaxation process could be operating in the C60 monoanion radical in liquid solution. The low activation energy of 74 cm−1 observed here is consistent with the pseudorotation barrier of C60 monoanion, estimated from reported Jahn−Teller energy levels.

■

INTRODUCTION C60 fullerene, possibly the most symmetric nonlinear molecule known, has 3-fold degenerate LUMOs. When its anion radical is formed, the unpaired electron occupies one of its degenerate LUMOs. Such degenerate electronic ground states usually have nonzero orbital angular momentum. Modulation of the resulting spin−orbit interaction, due to molecular motions in liquids or lattice vibrations in solids, can cause very efficient electron spin relaxation of such radicals. An early instance of efficient spin relaxation process was found by Townsend and Weissman1 in the EPR spectra of the anion radical of symmetric benzene in liquid solution. They reported its EPR lines to be broader and more difficult to saturate than those of less symmetric radicals such as anthracene anion. This observation shows that both spin−lattice relaxation (T1) and spin−spin relaxation (T2) times of benzene anion radical are much shorter than those of less symmetric radicals. The unpaired electron of C6H6 anion radical occupies one of the doubly degenerate e2u MOs, a consequence of the high D6h symmetry of C6H6. Townsend and Weissman1 reported EPR lines of similar characteristics for the anion radicals of less symmetric triphenylene and coronene. In spite of their lower symmetry, these two molecules also have doubly degenerate LUMOs. In fact, the existence of at least C3v symmetry in a © 2015 American Chemical Society

molecule is sufficient to have degenerate molecular orbitals. C60 has a much higher symmetry of Ih. Triplet states of most organic molecules have large zero-field splitting, whose modulation in the liquid state causes efficient electron spin relaxation, making their EPR spectra too broad to be observed. In contrast, because of its very high symmetry, the magnitude of the zero-field splitting of triplet C60 is very small. As a result, its EPR spectrum has a very narrow line of only 0.14 G width in degassed methylcyclohexane, which can be easily observed in the liquid state.2 The degeneracy of the HOMO of C60 is 5fold, and that of LUMO is 3-fold. The spin relaxation process of the C60 anion radical is therefore expected to be even more efficient than that of the C6H6 anion radical. Qualitatively this has been observed: the EPR spectra of C60− show that its lines are much broader than that of C6H 6− under similar experimental conditions. In this work, we investigated the possible electron spin relaxation mechanism of C60− radical in liquid solution. Since the observation of Townsend and Weissman,1 many models have been proposed to explain the enhanced electron Received: December 18, 2014 Revised: March 10, 2015 Published: March 19, 2015 3200

DOI: 10.1021/jp5126409 J. Phys. Chem. A 2015, 119, 3200−3208

Article

The Journal of Physical Chemistry A

exclusively form. To that end, we chose carbon disulfide as the solvent and prepared C60− in it by chemical reduction of C60. The spin relaxation rates, estimated from the EPR line widths, showed only very weak temperature dependence, which was not consistent with the mathematical expression for the temperature dependence given by the KO relaxation mechanism. Unlike the KO expression, the weak temperature dependence of the line width was found to be Arrhenius type. By analogy with the electron spin relaxation process observed in cyclohexane type of radicals,20 we proposed that a pseudorotation-induced relaxation mechanism could be operating in the C60 anion as well.

spin relaxation processes associated with the broad solution EPR lines of organic radicals, such as the C6H6 anion radical, with degenerate electronic ground states.3−7 Of these models, that of Kivelson6,7 gives an explicit expression for the temperature dependence of T1, which can be verified experimentally. This model, called the electric field fluctuation model, is adapted from the Orbach mechanism8 of phononinduced electron spin relaxation of systems with low-lying electronic states in the solid state. Highly symmetric radicals with degenerate ground state are expected to undergo Jahn−Teller distortion, which lowers their symmetry. While the ground state then becomes nondegenerate, low-lying thermally accessible electronic excited states may exist if the distortion is not large. In the electric field fluctuation model of Kivelson,6,7 the radical experiences timedependent electric field fluctuations because of collisions with solvent molecules and its own molecular vibrations. Such fluctuations can cause an electric-field-induced transition to a thermally accessible excited state as well as a de-excitation from the excited to the ground state. These transitions, combined with the spin−orbit interaction, can cause the electron spins to flip and thus become a spin relaxation process. We call this Kivelson−Orbach (KO) relaxation mechanism.9 Even though this mechanism was proposed more than 5 decades ago, there has been no quantitative experimental test of it until recently. We have recently shown how the KO mechanism can be used to rationalize the temperature dependence of the spin−lattice relaxation (SLR) times of the anion radicals of benzene and several of its derivatives in liquid solutions.9 Our results show good correlation between the observed electron spin−lattice relaxation rates and the energy gap of the two lowest electronic statesthe Jahn−Teller splitting energyof these anion radicals. With this successful application of the KO mechanism to the benzene anion radical and its derivatives, it was natural to expect that the same mechanism should explain the relaxation process of the C60− radical in solution as well. We undertook the present work to investigate this. Since the discovery of C60, the anions of C60 have been studied intensively because many interesting physical properties, such as ferromagnetism,10,11 superconductivity,12 etc., have been observed in these systems. This is also the reason why spin relaxation studies of various C60 anions have been carried out in frozen solution13−15 or in the form of a solid salt.15,16 As there was no specific aim to investigate the electron spin relaxation behavior only in the liquid state, these studies often covered a broad temperature range, with some EPR measurements extending from room temperature, where the sample was in the liquid state, all the way down to liquid helium temperature.13,14 The general observation in these studies is a strong temperature dependence of the EPR line width in the frozen state, whereas it is very weakly dependent in liquid solutions. In many such studies, the authors interpreted their EPR data of C60− of liquid and frozen states together, without making any distinction.14,18 Many authors have also tried to explain their solid state C60 anion data using the Orbach mechanism.16,17,19 Eaton’s group14 and Kadish’s group18 have reported line widths of C60− in various solvents in the liquid state. However, no specific relaxation mechanism was addressed in these studies. As our aim was to investigate the applicability of the KO mechanism to the electron spin relaxation of the monoanion of C60, we needed a solvent that remained liquid over a wide temperature range and in which the C60 monoanion could

■

EXPERIMENTAL SECTION Chemicals. C60 was purchased from Sigma with a stated purity of 98% and was used for EPR sample preparation without further purification. Carbon disulfide (99% purity) was purchased from Merck; it was dried by refluxing with CaH2, distilled over 4-Å molecular sieves, and stored under argon. Dry pyridine (water content 3V2/√5. From their studies on the vibronic spectra of C60−,35 they reported the value of the vibrational energy ℏω = 687 cm−1 and V1 = 1.84. Using these values, we calculated the energies ED3d and ED2h over a wide range of allowed values of V2 and V3 (−0.85 < V2 < 0.85, and 0.9 > V3 > 3V2/√5). The calculated values showed that the energy difference between ED3d and ED2h, the pseudorotation barrier, could be anything from 0 to ca. 550 cm−1. In contrast, quantum chemical calculations of Ramanantoanina et al., based on multidimensional density functional theory and “without fitting of parameters to experimental data”,37 indicate that all the three distortions D5d, D3d, and D2h represent energy minima. Their energies are very similar, differing by at most 15 cm−1. Thus, the magnitude of the pseudorotation barrier also should be around 15 cm−1. The above estimates indicated a very small pseudorotation barrier in C60−. Our observation of 74 cm−1 as the activation energy in the temperature dependence of the EPR line widths and assigning the process to pseudorotation is consistent with that.

■

SUMMARY AND CONCLUSION We adopted a previously reported chemical reduction method to synthesize the monoanion of C60. For the temperature dependence study of line widths in liquid solution over a broad range, CS2 was chosen as the solvent. The C60− in this solution remains in the liquid state down to 130 K. The EPR line width of C60− in CS2 solvents was measured in the temperature range 290−130 K. The observed temperature dependence of line 3206

DOI: 10.1021/jp5126409 J. Phys. Chem. A 2015, 119, 3200−3208

Article

The Journal of Physical Chemistry A

Application of Kivelson−Orbach Mechanism of Electron Spin Relaxation. Mol. Phys. 2014, 112, 1577−1588. (10) Hebard, A. F.; Rosseinsky, M. J.; Haddon, R. C.; Murphy, D. W.; Glarum, S. H.; Palstra, T. T. M.; Ramirez, A. P.; Kortan, A. R. Superconductivity at 18 K in Potassium-Doped C60. Nature 1991, 350, 600−601. (11) Stephens, P. W.; Mihaly, L.; Lee, P. L.; Whetten, R. L.; Huang, S. M.; Kaner, R.; Deiderich, F.; Holczer, K. Structure of Single-phase Superconducting K3C60. Nature 1991, 351, 632−624. (12) Wang, H. H.; Kini, A. M.; Savall, B. M.; Carlson, K. D.; Williams, J. M.; Lykke, K. R.; Wurz, P.; Parker, D. H.; Pellin, M. J.; Gruen, D. M.; et al. First Easily Reproduced Solution-phase Synthesis and Confirmation of Superconductivity in the Fullerene KxC60 (Tc = 18.0 ± 0.1 K). Inorg. Chem. 1991, 30, 2838−2839. (13) Eaton, S. S.; Kee, A.; Konda, R.; Eaton, G. R.; Trulove, P. C.; Carlin, R. T. Comparison of Electron Paramagnetic Resonance Line Shapes and Electron Spin Relaxation Rates for C60− and C603− in 4:1 Toluene:Acetonitrile and Dimethyl Sulfoxide. J. Phys. Chem. 1996, 100, 6910−6919. (14) Schell-Sorokin, A. J.; Mehran, F.; Eaton, G. R.; Eaton, S. S.; Viehbeck, A.; O’Toole, T. R.; Brown, C. A. Electron Spin Relaxation Times of C60− in Solution. Chem. Phys. Lett. 1992, 195, 225−232. (15) Dubois, D.; Jones, M. T.; Kadish, K. M. Electroreduction of Buckminsterfullerene, C60, in Aprotic Solvents: Electron Spin Resonance Characterization of Singly, Doubly, and Triply Reduced C60 in Frozen Solutions. J. Am. Chem. Soc. 1992, 114, 6446−6451. (16) Völkel, G.; Pöppl, A.; Simon, J.; Hoentsch, J. Evidence of the Jahn-Teller Splitting of C60− in C60-tetraphenylphosphoniumchloride from an Electron-Spin-Relaxation Study. Phys. Rev. B 1995, 52, 10188−10193. (17) Gotschy, B.; Völkel, G. The Universality of the Electron Spin Relaxation in C60 Mono Radical Anion Salts: EPR Studies of the Model System (P(C6H5)4]2C60X. Appl. Magn. Reson. 1996, 11, 229− 238. (18) Rataiczak, R. D.; Koh, W.; Subramanian, R.; Jones, M. T.; Kadish, K. M. Electron Spin Resonance Characterization of the Anion Radicals of Singly, Doubly and Triply Reduced C60 in Liquid Solutions. Synth. Met. 1993, 56, 3137−3141. (19) Fukuzumii, S.; Mori, H.; Suenobu, T.; Imahori, H.; Gao, X.; Kadish, K. M. Effects of Lowering Symmetry on the ESR Spectra of Radical Anions of Fullerene Derivatives and the Reduction Potentials. J. Phys. Chem. A 2000, 104, 10688−10694. (20) Borovkov, V. I.; Beregovaya, I. V.; Shchegoleva, L. N.; Bagryanskii, V. A.; Molin, Yu. N. Pseudorotation as a Possible Origin of Fast Paramagnetic Relaxation in Radical Ions with a Degenerate or Quasi-degenerate Ground State. Dokl. Phys. Chem., Part 2 2009, 426, 108−112. (21) Hirsch, A. The Chemistry of the Fullerenes; Thieme Medical Publishers, Inc.: New York, 1994; pp 44−47. (22) McVitt, J. T.; Ching, S.; Sullivan, M.; Murry, R. W. Fluid Electrolyte Solutions for Electrochemistry at Near Liquid Nitrogen Temperatures. J. Am. Chem. Soc. 1989, 111, 4528−4529. (23) Stinchcombe, J.; Pénicaud, A.; Bhyrappa, P.; Boyd, P. W. D.; Reed, C. A. Buckminsterfulleride(1−) Salts: Synthesis, EPR, and the Jahn-Teller Distortion of C60−. J. Am. Chem. Soc. 1993, 115, 5212− 5217. (24) Boulas, P.; Subramanian, R.; Kutner, W.; Jones, M. T.; Kadish, K. M. Facile Preparation of the C60 Monoanion in Aprotic Solvents. J. Electrochem. Soc. 1993, 140, L130−L132. (25) Dubois, D.; Moninot, G.; Kutner, W.; Jones, M. T.; Kadish, K. Electroreduction of Buckminsterfullerene, C60, in Aprotic Solvents. Solvent, Supporting Electrolyte, and Temperature Effects. J. Phys. Chem. 1992, 96, 7137−7145. (26) Bennati, M.; Grampp, A.; Bäuerle, P.; Dinse, K. P.; Schweitzer, P.; Baumgarten, M. EPR Study of Fullerene Radicals Generated in Photosensitized TiO2 Suspensions. J. Phys. Chem. 1995, 99, 8782− 8789. (27) Paul, P.; Kim, K. C.; Sun, D.; Boyd, P. D. W.; Reed, C. A. Artifacts in the Electron Paramagnetic Resonance Spectra of C60

width shows that the Kivelson−Orbach mechanism of electronspin relaxation does not operate in this radical in the temperature range studied here. The temperature dependence of line width fits to an empirical function, ΔBpp = C exp (−ΔE/ kBT), with ΔE = 74 ± 3 cm−1. This temperature dependence of SLR rate is similar to the very fast SLR rates recently reported for cyclohexane type of cation radicals. From the similarities of their relaxation behavior, we propose that a pseudorotation induced relaxation process could be operative in C60− in solution. Our work should stimulate further investigation into the relaxation mechanism of the C60 monoanion in liquid solution to explore the role of solvents in influencing pseudorotation and build theoretical models to calculate the pseudorotation barrier of C60−.

■

ASSOCIATED CONTENT

S Supporting Information *

Complete citation of the ref 12 containing the names of all the authors. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (G.G.). *E-mail: [email protected] (R.D.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS R.D. sincerely thanks Janette L. Dunn of University of Nottingham, U.K., for useful discussions on the pseudorotation of fullerenes and help in estimating its pseudoraotation barrier. This work was partly funded by an India-Austria (DST/ BMWF) Joint Research Project, for which R.D. thanks the Department of Science and Technology (DST), Govt. of India (project grant no. INT/AUA/BMWF/P-05/2011), and G.G. thanks the Austrian Exchange Service (OeAD) for financial support within the program WTZ-IN-03-2011.

■

REFERENCES

(1) Townsend, M. G.; Weissman, S. I. Possible Symptom of the JahnTeller Effect in the Negative Ions of Coronene and Triphenylene. J. Chem. Phys. 1960, 32, 309−310. (2) Closs, G. L.; Gautam, P.; Zhang, D.; Paul, J. K.; Hill, S. A.; Wasserman, E. Steady-State and Time-Resolved Direct Detection EPR Spectra of Fullerene Triplets in Liquid Solution and Glassy Matrices. Evidence for a Dynamic Jahn-Teller Effect in Triplet C60. J. Phys. Chem. 1992, 96, 5228−5231. (3) McConnell, H. M.; McLachlan, A. D. Nuclear Hyperfine Interactions in Orbitally Degenerate States of Aromatic Ions. J. Chem. Phys. 1961, 34, 1−12. (4) McConnell, H. M. Spin-Orbit Coupling in Orbitally Degenerate States of Aromatic Ions. J. Chem. Phys. 1961, 34, 13−16. (5) Freed, J. H. ESR Hyperfine Linewidths from Benzene Anion Distortions. J. Chem. Phys. 1965, 43, 1427−1428. (6) Kivelson, D. Theory of ESR Linewidths for Benzene Negative Ion Radicals. J. Chem. Phys. 1966, 45, 751−752. (7) Kivelson, D. Electric-Field Fluctuations and Spin Relaxation in Liquids. J. Chem. Phys. 1966, 45, 1324−1332. (8) Orbach, R. On the Theory of Spin-Lattice Relaxation in Paramagnetic Salts. Proc. Phys. Soc. London, A 1961, 77, 821−826. (9) Kundu, K.; Das, R. Saturation Recovery Electron Spin−lattice Relaxation Studies on Benzene Anion Radical and its Derivatives: 3207

DOI: 10.1021/jp5126409 J. Phys. Chem. A 2015, 119, 3200−3208

Article

The Journal of Physical Chemistry A Fullerene Ions: Inevitable C120O Impurity. J. Am. Chem. Soc. 2002, 124, 4394−4401. (28) Al’tshuler, S. A.; Valiev, K. A. On the Theory of Longitudinal Relaxation of Paramagnetic Salt Solution. Sov. Phys. JETP 1959, 8, 661−668. (29) de Coulon, V.; Martins, J. L.; Reuse, F. Electronic Structure of Neutral and Charged C60 Clusters. Phys. Rev. B 1992, 45, 13671− 13676. (30) Borovkov, V. I.; Beregovaya, I. V.; Shchegoleva, L. N.; Potashov, P. A.; Bagryansky, V. A.; Molin, Yu. N. Radical Ions with Nearly Degenerate Ground State: Correlation between the Rate of SpinLattice Relaxation and the Structure of Adiabatic Potential Energy Surface. J. Chem. Phys. 2012, 137, 104305-1−104305-14. (31) Hands, I. D.; Dunn, J. L. Analysis of the T1u ⊗ hg Jahn-Teller System as a Model for C60 Molecules. Phys. Rev. B 1995, 52, 5996− 6005. (32) Hands, I. D.; Dunn, J. L.; Bates, C. A. Quantitative Assessment of the Vibronic Coupling in C60− via Doping Effects in Raman Scattering Experiments. Phys. Rev. B 2001, 63, 245414-1−245414-12. (33) Hands, I. D.; Dunn, J. L.; Bates, C. A. Theoretical Calculation of Pseudorotation Rates Applicable to the C60− Ion. Phys. Rev. B 2006, 73, 235425-1−235425-11. (34) Hands, I. D.; Dunn, J. L.; Bates, C. A. Determination of Pseudorotation Rates in E ⊗ e Jahn-Teller Systems. Phys. Rev. B 2006, 73, 014303-1−014303-9. (35) Hands, I. D.; Dunn, J. L.; Bates; Michael, C. A.; Hope, J.; Meech, S. R.; Andrews, D. L. Vibronic Interactions in the Visible and Near-Infrared Spectra of C60− Anions. Phys. Rev. B 2008, 77, 1154451−115445-5. (36) Hands, I. D.; Dunn, J. L.; Bates, C. A. Visualization of Static Jahn-Teller Effects in the Fullerene Anion C60−. Phys. Rev. B 2010, 82, 155425-1−155425-15. (37) Ramanantoanina, H.; Gruden-Pavlovic, M.; Zlatar, M.; Daul, C. Density Functional Theory Study of the Multimode Jahn-Teller Problem in the Fullerene Anion. Int. J. Quantum Chem. 2013, 113, 802−807.

3208

DOI: 10.1021/jp5126409 J. Phys. Chem. A 2015, 119, 3200−3208