Direct observations of slow photo-induced structural and physical changes

doi:10.21203/rs.3.rs-4811531/v1

Download PDF

Article

Direct observations of slow photo-induced structural and physical changes

https://doi.org/10.21203/rs.3.rs-4811531/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Photoexcitation and the subsequent physical or chemical processes are often too rapid to capture using direct, real-time observation methods1-6. Current approaches rely mainly on time-resolved spectra derived from short-pulsed lasers with appropriate wavelengths and pulse widths for the target phenomena7-10, supplemented by some experimental techniques combined with theoretical simulations11-14. In contrast, if a material exhibits sufficiently slow changes after photoexcitation, its structures and physical properties can be directly observed and compared with theoretical and spectroscopic results. This article explores the magnetic and structural properties of a new material that displays slow relaxation from UV-excited states. The study demonstrates that both valence and inner-shell electrons are cooperatively photoexcited, transitioning to a new state influenced by spin, charge, and orbital degrees of freedom. Unlike traditional theoretical descriptions, which assume time-independent atomic orbitals as the basis for molecular orbitals, this research reveals that atomic orbitals are time-dependent during these processes. These findings offer a foundational understanding of the relaxation processes from photoexcited states and provide insights into general chemical reactions.

Physical sciences/Physics/Chemical physics

Physical sciences/Materials science/Condensed-matter physics/Magnetic properties and materials

Scientists have long aspired to observe the detailed processes of dynamic changes triggered by photoexcitation in real time^15-18. Such processes often occur in mixtures that lack well-defined atomic positions. This lack of structure makes real-time observation of detailed changes in structures and physical properties even more challenging, even when some processes are slow. Previously, exceptional materials based on charge-transfer complexes containing photoreactive species were found to exhibit unusually slow or progressive changes upon UV irradiation in single crystals^19,20. However, their photoexcitation and relaxation processes remained too rapid to observe real-time structural changes and the evolution of physical properties using standard experimental methods. Additionally, UV irradiation caused serious disorder or amorphization in the crystals, preventing detailed structural analyses. In this work, we synthesized a new charge-transfer salt, MV[Au(dmit)₂]₂ (MV salt; Figs. 1a and b and Extended Data Figs. 1a and b), which undergoes sufficiently slow changes in the molecular structure after UV irradiation (~240–450 nm, 5 min) on single crystals, enabling detailed structural analyses during the change. Based on single-crystal X-ray structural analyses, the change mainly involved Au atoms at the centre of the Au(dmit)₂ molecules and lasted for months (~2,500 h, Extended Data Fig. 2) with intermittent structural changes. Before UV irradiation, all Au(dmit)₂ molecules in the MV salt were almost completely planar, as is typical for Au(III) complexes²¹. Approximately 200 h after the cessation of brief (~5–10 min) UV irradiation (t = 0 h) of the single crystals, 2%–3% of the Au atoms began to deviate from the molecular plane as disordered atoms (Fig. 1c and Extended Data Figs. 1–2). The non-planar coordinates of the Au(III) complexes were unusual, suggesting exceptional situations such as transition states during chemical reactions. The distances between the deviated Au atoms and the molecular plane, d (Å), in non-planar Au2 and Au3 alternately increased and decreased at t ³ 200 h. Specifically, the Au atoms fluctuated between three sites following UV irradiation; thus, the structure also fluctuated among the three structures: planar Au1, non-planar Au2, and non-planar Au3. Structural changes occurred after the cessation of UV irradiation, excluding heating effects or decomposition by UV irradiation as the cause of the structural changes. However, single-crystal X-ray structural analyses indicated that 97%–98% of the Au atoms adopted the original planar coordinates (Extended Data Figs. 1–2). The X-ray oscillation photographs used in the structural analyses changed depending on the elapsed time t, indicating that the sample fluctuated between highly and poorly crystalline states (Fig. 1d and Extended Data Figs. 3–4). Because single-crystal X-ray structural analyses reveal only long-range ordered atomic arrangements, and because UV excitation may seriously disturb the atomic arrangement and electron distribution in the single crystals, we cross-checked the UV-induced structural change using a different measurement method requiring no long-range order in the samples. We examined the real-time X-ray absorption fine structure (XAFS)²² of the Au L_III-edge using single crystals of the MV salt (Extended Data Figs. 5a–f) and the same UV light source used in the single-crystal X-ray structural analyses. The XAFS spectra demonstrated that most Au atoms (>95%) in the sample had similar coordinate structures and that the Au local structure was retained during and until 270 min after the cessation of irradiation.

The structure of the MV salt progressively changed, while retaining the appearance of a single crystal. Because such disorder effects are particularly evident in the early stages of relaxation, deviated Au atoms were hardly observed at t £ 200 h using single-crystal X-ray structural analyses. Further analysis of the X-ray oscillation photographs revealed that most crystal planes exhibited long-lasting lattice distortions with fluctuations in the single crystal of the MV salt after UV irradiation (Extended Data Figs. 6a and b). Based on single-crystal X-ray structural analyses, we clarified that UV excitation caused the Au atoms to deviate from the molecular plane and intermittently changed sites, thereby forming planar Au1 and non-planar Au2 and Au3 configurations. The MV salt gradually transitioned from a dark state to a different (meta)stable state.

Upon UV-induced distortion of the molecular structure of Au(dmit)₂ from planar Au1 to non-planar Au2 or Au3, the Au–C distances decreased when the Au atoms deviated (i.e., Au1 > Au2 and Au3), reaching values comparable to the van der Waals distance (Au–C = 3.36 Å)²³ between adjacent molecules (Fig. 2a and Supplementary Table 1), suggesting that UV irradiation induced the intermolecular interactions. Intermolecular interactions are the starting points of chemical reactions^24,25, phase transitions^26,27, and various other phenomena^28,29. While these interactions stabilize the total (Gibbs) energy of the MV salt, the distortion of the Au coordinate destabilizes the delocalization energy of the Au(dmit)₂ molecules. The disorder at the Au site decreases the Gibbs energy by increasing entropy; however, it is unfavourable for intermolecular interactions between pairs of facing molecules, such as MV–Au(dmit)₂ and Au(dmit)₂–Au(dmit)₂. To quantify the net effects of these opposing factors on stability, we calculated the intermolecular interactions and thermodynamic properties of the observed structures using density functional theory (DFT)³⁰ (Extended Data Fig. 7). The calculated molecular orbitals (MOs) show that intermolecular interactions after UV irradiation due to extensive delocalization of electrons over neighbouring molecules. This delocalization is more prominent around the non-planar coordinated Au(dmit)₂ (Au2(dmit)₂ or Au3(dmit)₂) molecules compared to that of the planar Au(dmit)₂ (Au1(dmit)₂) molecules (Fig. 2b). This UV excitation induces delocalization. Similarly, the Mulliken charges³¹ on the Au atoms (Fig. 2c) and the Gibbs energy (Fig. 2d) consistently demonstrate that the Au(dmit)₂ molecules exchange energy and charge between neighbouring molecules. However, the thermodynamic quantities of each structure (Au1(dmit)₂, Au2(dmit)₂, and Au3(dmit)₂) repeatedly increase and decrease with t (Extended Data Fig. 7). Meanwhile, they retain their average values for each structure until t ~ 2,000 h. The same calculation also shows the importance of the Au2 and Au3 non-planar structures in the charge and spin distributions when the Au(dmit)₂ molecules possess unpaired electrons (Fig. 2e). In summary, UV-irradiation-induced intermolecular interactions cause continuous fluctuations in the energies, charges, and structures of the MV salt, driven by charge, spin, and orbital degrees of freedom.

X-ray photoelectron spectroscopy (XPS) is well known as a powerful tool for surface analyses³² and is useful for selective analyses of the net UV-affected region in bulk materials, as UV penetrates solids much deeper (a few micrometres) than the photoelectron escape depth (a few Angstroms)³³. The first-principles calculations (Figs. 2b and e) indicate that the Au and S atoms dominate the electron and spin densities in the MV salt. Therefore, we focused on the XPS of the Au and S atomic orbitals, which were continuously monitored before (dark), during (t = 0), and after UV irradiation (t > 0) (Figs. 3a–f, Extended Data Figs. 8a–d, and Supplementary Figs. 5a–e). Upon UV irradiation, all observed XPS peaks (Figs. 3a–c and e and Extended Data Figs. 8a–d) significantly shift to shallower binding energies (BEs). Such peak shifts are generally understood as changes in the oxidation states of the atoms³⁴. In this work, photoexcitation causes optical transitions of electrons, which correspond to transient changes in the oxidation states of the atoms. In other words, the observed XPS spectral changes indicate that the energy of the MV salt is enhanced by the photon energy. The Au 5s (Fig. 3b) and S 2p (Extended Data Figs. 8c and d) spectra recover their original (dark) BE values by ~3,000 h, whereas the remaining atomic orbitals change to shallower BE values quite different from those during UV irradiation or under dark conditions. These differences in relaxation behaviour are correlated with the degrees of freedom, viz., the Au 4f and 4d orbitals exhibit more complicated and larger responses than the s and p orbitals of both the Au and S atoms. Furthermore, the Au 4f_5/2 and Au 4f_7/2 peaks broaden after UV irradiation (Fig. 3a and Supplementary Fig. 5a) and then gradually become sharper. This finding indicates that UV irradiation produces various photoexcited states close to each other in energy and that they gradually converge into fewer states during the relaxation processes. The peak separation between the Au 4f_5/2 and Au 4f_7/2 peaks, viz., the spin–orbit coupling in the Au 4f orbitals, increases with time, which indicates that the orbitals change in shape (see Supplementary Notes 1–3 in the Supplementary Information)³⁵. We observed the electrons and MOs in transition to a different steady state after the cessation of UV irradiation. Furthermore, the 4f_5/2–4f_7/2 peak intensity ratio becomes increasingly imbalanced over time (Fig. 3d), indicating that spin polarization occurs around the Au 4f orbitals. A similar trend is observed for the Au 4d orbitals (Fig. 3f). The MV salt is magnetized spontaneously and transiently.

To observe the occurrence and evolution of spin polarization after UV irradiation, we repeated the measurements of the temperature dependence of the magnetic susceptibility of the MV salt before (dark) and after UV irradiation (Figs. 4a–e). Under the dark conditions, the temperature (T)-independent small susceptibility (~3 ´ 10^–⁴, Fig. 4a) indicates that the salt is effectively diamagnetic except for a negligible number of antiferromagnetically interacting spins manifested at T £ 10 K. This finding is consistent with the fact that practically no unpaired electrons exist in the MV salt in the ground state, except for a small number of unpaired electrons produced by weak MV–Au(dmit)₂ CT interactions, as observed for a related compound²⁰. However, after UV irradiation, paramagnetic behaviour is observed with large hysteresis between the zero-field cooling (ZFC) and field-cooling (FC) processes (Figs. 4b–e). Note that the susceptibility c is negative, in contrast to the standard paramagnetism (c > 0)³⁶. This magnetic susceptibility can be explained if the unpaired electrons occupy significantly high energy levels with large orbital angular momenta. Such a situation is possible in the excited states caused by the Au 4f–Au nf and Au 4d–Au nd transitions (4 < n). This result is consistent with the above discussion of XPS, indicating that we observed how the spins polarized. Additionally, alternating increases and decreases in the hysteresis occur between FC and ZFC (Figs. 4b–e). The observed period (~30 h) depends on the time interval of the measurements and should be independent of the intrinsic oscillation periods of the spin system. However, this characteristic still demonstrates that the spin system fluctuates significantly during the relaxation processes after the cessation of UV irradiation.

This work supports the analysis of physical and chemical processes in time-resolved spectroscopy^37,38. Different (meta)stable states are separated by an energy barrier, preventing spontaneous transitions between states. Consistently, in this work, some of the atomic orbitals spontaneously destabilized themselves (Fig. 2d) with alternating stabilization and destabilization after the cessation of photoexcitation, whereas the total (Gibbs) energy of the system decreased during these processes. Because the chemical species in the system do not simultaneously behave in the same manner, one transiently observes the averaged structure, properties, and energy of different species as a snapshot of fluctuation. The XPS spectra also indicate that the core electrons cooperate to achieve a new stable state by exchanging their energies with each other and with their surroundings using their spin, charge, and orbital degrees of freedom. The observed structural and property changes should include the general aspects of chemical and physical changes. For example, in the material design and discussion of the mechanisms of electro-³⁹ and photoluminescence⁴⁰ and solar cells⁴¹, a few characteristic MOs around the valence bands/orbitals of key compounds are usually paid the most attention. The gradual and persistent changes in the closed-shell core electrons observed in this work contrast sharply with the rule of thumb that valence electrons dominate chemistry. All atomic orbitals cooperate to rearrange their chemical bonds for relaxation in response to large perturbations, such as photoexcitation and chemical reactions.

During slow structural changes after UV excitation, the valence and inner-shell electrons cooperate to polarize their spins and charges by exchanging their energies with each other and with their surroundings using the available degrees of freedom. This behaviour can be macroscopically observed as fluctuations in structure, properties, and energy. Based on the orbital approximation, by gradually and continuously adjusting the energy levels and spin states, the atomic orbitals at deeper levels possessing larger degrees of freedom can play dominant roles in stabilizing both the transient and final states during molecular orbital reconstruction. Some atomic orbitals had energy levels higher than their initial states after spontaneous changes. These findings provide an experimental basis for the currently proposed or assumed mechanisms of chemical reactions, photoexcitation, and other non-equilibrium phenomena to provide a deeper understanding of material dynamics.

All methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information, details of author contributions, competing interests, and statements of data and code availability are available at https://doi.org/.

Pettine, J., et al. Light-driven nanoscale vector current. Nature 626, 984–989 (2024) (DOI 10.1038/s41586-024-07037-4)
Nabben, D., Kuttruff, J., Stolz, L., Ryabov, A. & Baum, P. Attosecond electron microscopy of sub-cycle optical dynamics. Nature 619, 63–67 (2023) (DOI 10.1038/s41586-023-06074-9)
Gruhl, T., et al. Ultrafast structural changes direct the first molecular events of vision. Nature 615, 939–944 (2023) (DOI 10.1038/s41586-023-05863-6)
Wu, L., et al. Optically induced charge-transfer in donor-acceptor-substituted p- and m-C₂B₁₀H₁₂ carboranes. Nat. Commun. 15, 3005 (2024) (DOI 10.1038/s41467-024-47384-4)
Soorkia, S., Jouvet, C. & Grégoire, G. UV photoinduced dynamics of conformer-resolved aromatic peptides. Chem. Rev. 120, 3296–3327 (2020) (DOI 10.1021/acs.chemrev.9b00316)
Pattengale, B., Ostresh, S., Schmuttenmaer, C. A. & Neu, J. Interrogating light-initiated dynamics in metal–organic frameworks with time-resolved spectroscopy. Chem. Rev. 122, 132–166 (2022) (DOI 10.1021/acs.chemrev.1c00528)
Gong, Y., et al. Boosting exciton mobility approaching Mott-Ioffe-Regel limit in Ruddlesden-Popper perovskites by anchoring the organic cation. Nat. Commun. 15, 1893 (2024) (DOI 10.1038/s41467-024-45740-y)
Shrestha, S., et al. Long carrier diffusion length in two-dimensional lead halide perovskite single crystals. Chem 8, 1107–1120 (2022) (DOI 10.1016/j.chempr.2022.01.008)
Wagner, K., et al. Nonclassical exciton diffusion in monolayer WSe₂. Phys. Rev. Lett. 127, 076801 (2021) (DOI 10.1103/PhysRevLett.127.076801)
Quan, L. N., et al. Vibrational relaxation dynamics in layered perovskite quantum wells. Proc. Natl Acad. Sci. USA. 118, e2104425118 (2021) (DOI 10.1073/pnas.2104425118)
Kim, H. S., Lee, S. H., Yoo, S. & Adachi, C. Understanding of complex spin up-conversion processes in charge-transfer-type organic molecules. Nat. Commun. 15, 2267 (2024) (DOI 10.1038/s41467-024-46406-5)
Wang, Z., et al. Role of interfacial donor-acceptor percolation in efficient and stable all-polymer solar cells. Nat. Commun. 15, 1212 (2024) (DOI 10.1038/s41467-024-45455-0)
Dong, H.-H., et al. Theoretical and experimental investigations of Al³⁺ ion-suppressed phase-separation structures in rare-earth-doped high-phosphorus silica glasses. Phys. Chem. Chem. Phys. 26, 3869–3879 (2024) (DOI 10.1039/d3cp04758j)
Wen, X. & Jia, B. New insight into carrier transport in 2D layered perovskites. Chem 8, 904–906 (2022) (DOI 10.1016/j.chempr.2022.03.014)
Otsuka, H., et al. Transient chemical and structural changes in graphene oxide during ripening. Nat. Commun. 15, 1708 (2024) (DOI 10.1038/s41467-024-46083-4)
Larsson, A., et al. Dynamics of early-stage oxide formation on a Ni-Cr-Mo alloy. npj Mater. Degrad. 8, 39 (2024) (DOI 10.1038/s41529-024-00463-9)
Puntel, D., et al. Out-of-equilibrium charge redistribution in a copper-oxide based superconductor by time-resolved X-ray photoelectron spectroscopy. Sci. Rep. 14, 8775 (2024) (DOI 10.1038/s41598-024-56440-4)
Zhang, X., et al. Enhancing photocatalytic H₂O₂ production with Au cocatalysts through electronic structure modification. Nat. Commun. 15, 3212 (2024) (DOI 10.1038/s41467-024-47624-7)
Naito, T. Development of control method for conduction and magnetism in molecular crystals. Bull. Chem. Soc. Jpn 90, 89–136 (2017) (DOI 10.1246/bcsj.20160295)
Naito, T., et al. A molecular crystal with an unprecedentedly long-lived photoexcited state. Dalton Trans. 48, 12858–12866 (2019) (DOI 10.1039/c9dt02377a)
Bertrand, B., Williams, M. R. M. & Bochmann, M. Gold(III) Complexes for Antitumor Applications: an Overview. Chemistry. Chem. 24, 11840–11851 (2018) (DOI 10.1002/chem.201800981)
Kirichkov, M. V., et al. Application of X-ray based modern instrumental techniques to determine the heavy metals in soils, minerals and organic media. Chemosphere 349, 140782 (2024) (DOI 10.1016/j.chemosphere.2023.140782)
Bondi, A. Van der Waals volumes and radii. J. Phys. Chem. 68, 441–451 (1964) (DOI 10.1021/j100785a001)
Brändén, G. & Neutze, R. Advances and challenges in time-resolved macromolecular crystallography. Science 373, eaba0954 (2021) (DOI 10.1126/science.aba0954)
Baiz, C. R., et al. Vibrational spectroscopic map, vibrational spectroscopy, and intermolecular interaction. Chem. Rev. 120, 7152–7218 (2020) (DOI 10.1021/acs.chemrev.9b00813)
Hervé, M., et al. Ultrafast and persistent photoinduced phase transition at room temperature monitored by streaming powder diffraction. Nat. Commun. 15, 267 (2024) (DOI 10.1038/s41467-023-44440-3)
Wang, Y., et al. Theoretical understanding of photon spectroscopies in correlated materials in and out of equilibrium. Nat. Rev. Mater. 3, 312–323 (2018) (DOI 10.1038/s41578-018-0046-3)
Chen, X., Zhang, X., Xiao, X., Wang, Z. & Zhao, J. Recent developments in understanding charge transfer in molecular electron donor-acceptor systems. Angew. Chem. Int. Ed. 62, e202216010 (2023) (DOI 10.1002/anie.202216010)
Yersin, H., et al. Intersystem crossing, phosphorescence, and spin-orbit coupling. Two contrasting Cu(I)-TADF dimers investigated by milli- to micro-second phosphorescence, femto-second fluorescence, and theoretical calculations. Coord. Chem. Rev. 478, 214975 (2023) (DOI 10.1016/j.ccr.2022.214975)
Frisch, M. J., et al. Gaussian16, revision. C. 01. Gaussian Inc. Wallingford CT (2019)
Zhao, J., Zhu, Z.-W., Zhao, D.-X. & Yang, Z.-Z. Atomic charges in molecules defined by molecular real space partition into atomic subspaces. Phys. Chem. Chem. Phys. 25, 9020–9030 (2023) (DOI 10.1039/d2cp05428k)
Greczynski, G. & Hultman, L. A step-by-step guide to perform x-ray photoelectron spectroscopy. J. Appl. Phys. 132, 011101 (2022) (DOI 10.1063/5.0086359)
Zhang, N. & Xiong, Y. Dynamic characterization of artificial photosynthesis through in situ X-ray photoelectron spectroscopy. Curr. Opin. Green Sustain. Chem. 41, 100796 (2023) (DOI 10.1016/j.cogsc.2023.100796)
Bagus, P. S., Nelin, C. J. & Brundle, C. R. Chemical significance of X-ray photoelectron spectroscopy binding energy shifts: A perspective. J. Vac. Sci. Technol. A 41, 068501 (2023) (DOI 10.1116/6.0003081)
Wada, S., Tsutsumi, T., Saita, K. & Taketsugu, T. Ab initio molecular dynamics study of intersystem crossing dynamics for MH₂ (M = Si, Ge, Sn, Pb) on spin-pure and spin-mixed potential energy surfaces. J. Comp.45, 552–562 (2024) (DOI 10.1002/jcc.27271)
Kittel, C. Introduction to Solid State Physics. 8th ed. 11 297–317 (Fremont, CA: John Wiley & Sons, 2005)
Dürr, R. N., et al. T. From NiMoO₄ to g-NiOOH: Detecting the active catalyst phase by time resolved in situ and operando Raman spectroscopy. ACS Nano 15, 13504–13515 (2021) (DOI 10.1021/acsnano.1c04126)
Dürr, R. N., et al. Correction to “From NiMoO₄ to g-NiOOH: detecting the active catalyst phase by time resolved in situ and operando Raman spectroscopy.” ACS Nano 15, 20693 (2021) (DOI 10.1021/acsnano.1c10145)
Kumar, K. Charge transporting and thermally activated delayed fluorescence materials for OLED applications. Phys. Chem. Chem. Phys. 26, 3711–3754 (2024) (DOI 10.1039/d3cp03214k)
Hernández-Castillo, D., Eder, I. & González, L. Guidelines to calculate non-radiative deactivation mechanisms of ruthenium tris(bipyridine) derivatives. Coord. Chem. Rev. 510, 215819 (2024) (DOI 10.1016/j.ccr.2024.215819)
Kim, D., Yun, T., An, S. & Lee, C.-L. How to improve the structural stabilities of halide perovskite quantum dots: Review of various strategies to enhance the structural stabilities of halide perovskite quantum dots. Nano Converg. 11, 4 (2024) (DOI 10.1186/s40580-024-00412-x)

Synthesis procedures

The starting materials, MV·I₂⁴² and TBA[Au(dmit)₂]⁴³, were prepared following the literature. Acetonitrile (FUJIFILM Wako Pure Chemical Corporation, super-dehydrated for organic synthesis) was procured and used without further purification. Single crystals of MV[Au(dmit)₂]₂ were synthesized using a slow metathesis approach called the ‘diffusion method’. MV·I₂ (1,1'-dimethyl-4,4'-bipyridinium diiodide; 1.32 mg, 3.0 mmol) and TBA[Au(dmit)₂] (tetrabutylammonium bis(1,3-dithiole-2-thione-4,5-dithiolato)aurate(III); 5.41 mg, 6.5 mmol) were precipitated in CH₃CN (90 mL) in separate compartments of a three-compartment H-tube with fine porous glass frits (No. 3) in a N₂ atmosphere under dark conditions at 296 K. The two reactants were allowed to dissolve slowly in the solvent, diffuse, and react. After ~2–3 weeks, black diamond-shaped platelets ~0.5-1 mm in size (b-MV[Au(dmit)₂]₂) were obtained, along with an a polymorph of black hexagonal platelets ~0.5-1 mm in size (a-MV[Au(dmit)₂]₂). The a-type salt did not exhibit similar physical and structural properties after UV irradiation. Therefore, this paper only refers to the b-type salt without specifying the type. The UV–vis absorption spectra of MV·I₂, TBA[Au(dmit)₂], and MV[Au(dmit)₂]₂ are shown in Supplementary Fig. 1. The UV–vis–NIR spectra in the solution and solid state (powder) were recorded using a JASCO V-630 in transmittance mode at 295 K. MV·I₂ and TBA[Au(dmit)₂] were dissolved in CH₃CN (1.92 ´ 10^–⁵ M) and CH₃OH (2.06 ´ 10^–⁵ M), respectively, and the baseline (the absorption by the quartz cell and solvent) was independently recorded and subtracted from each spectrum. The measurement conditions were as follows: data response = fast, scan rate = 200 nm min^–¹, and data interval = 2 nm. The single crystals of MV·I₂ and MV[Au(dmit)₂]₂ were separately well ground with Al₂O₃ (white fused alumina, 3000 mesh, Fujimi Incorporated) in an agate mortar, sandwiched between optical quartz plates, and subjected to spectrum measurement. The background (the quartz plates and Al₂O₃ powder) was independently recorded and subtracted from the sample spectra. TBA[Au(dmit)₂] was finely powdered in an agate mortar without Al₂O₃, sandwiched between the quartz plates, and subjected to spectrum measurement under identical measurement conditions. The background (the quartz plates) was independently recorded and subtracted from the sample spectra.

UV irradiation

Single-crystal X-ray structural analyses were performed under different UV irradiation conditions (light source, bandpass filter, mirror module, light intensity, and irradiation duration). The photoresponses did not qualitatively change when the irradiation wavelengths were in the range of ~240–500 nm. This observation is consistent with the solid-state absorption spectra of MV[Au(dmit)₂]₂ and related materials (SupplementaryFig. 1), which show main absorption peaks in this region.

Two UV light sources were used in this study: a SUPERCURE-203S (200 W, Hg/Xe lamp, SAN-EI ELECTRIC, 220–1,100 nm) for the XAFS and XPS analyses and MAX-350 (300 W, Xe lamp, ASAHI SPECTRA, 250–1,050 nm) for the single-crystal X-ray structural analyses. For X-ray structural analysis, which requires intense light to deeply penetrate single crystals, the more powerful MAX-350 light source was chosen. Both light sources were equipped with input power adjusters, multimode optical quartz fibres (core diameter: 5 mm, length: 1 m), bandpass filters, and mirror modules to select the output wavelengths and adjust the light intensities. The distance between the single-crystal sample and the optical fibre end was ~13 mm for both XAFS and the single-crystal X-ray structural analyses. The light intensity at the sample position was estimated using a Si-diode power metre (NOVA, OPHIR); the specifications of the light sources were provided by the suppliers. The wavelengths used were 250–450 nm for XAFS and 300–600 nm for the single-crystal X-ray structural analyses. The intensities were maintained constant through different measurement runs at ~1 W cm^–² for both XAFS and single-crystal X-ray structural analyses. The sample temperature during the UV irradiation was not corrected. Other details of the UV irradiation have been described in our previous paper.^20,44

Single-crystal X-ray structural analyses

Prior to performing X-ray structural analysis and physical property/spectroscopic measurements (electrical resistivity, magnetic susceptibility, XPS, and XAFS) on the new sample, the crystallinity was examined using X-ray oscillation photographs obtained using a VariMax RAPID/a at the Advanced Research Support Centre (ADRES), Ehime University. High-quality single crystals (accuracy ³90%) were first analysed at 298 K in the dark to confirm their initial structures before UV irradiation. Data were collected at ADRES using a Rigaku VariMax Saturn CCD724/a detector equipped with a Mo rotor anode (l = 0.71073 Å). For each single crystal, X-ray oscillation photographs (1,440—2,880 frames) were collected, a process that generally requires 2–4 h per crystal.

The structural analysis temperature was carefully chosen by considering the relaxation rates from the UV-excited states and the thermal motion of atoms, both of which can depend on temperature. Lower temperatures are advantageous for observing crystal structures due to reduced thermal motion.⁴⁵ However, relaxation at extremely low temperatures, such as 77 K, was expected to be too slow to monitor until complete relaxation, based on ESR experiments conducted on a related material.²⁰ Considering the possible temperature increase under UV irradiation, the initial temperature for data collection was tentatively set to 128 K. For data collection at 128 K, the samples were cooled to 128 K at a rate of -1 K min^–¹ under dark conditions to avoid unnecessary disorder, inhomogeneity, or possible supercooling. After 10 hours, corresponding to the initial rapid relaxation period estimated from the behaviour of related materials, the samples were warmed to 298 K at a rate of +1 K min^-¹.²⁰

For structural analysis during relaxation, a single crystal was first exposed to UV radiation for 5 min. The crystal was then maintained at 128 K for 10 h following the cessation of UV irradiation. After 10 h, the crystal was stored at 298 K, except during the data collection at 128 K. The temperature variation rates were 1 K min^-¹ between 128 K and 298 K. All single-crystal X-ray structural analyses were performed at 128 K, and all structures were based on the same single crystal. The structure during UV irradiation could be completely solved with (~0.2%) or without non-planar-coordinated Au atoms (structures Au2 and Au3) with equal reliability. Therefore, the structure under UV irradiation is not discussed in detail.

Regardless of whether the crystal was exposed to UV irradiation, 18 frames of oscillation photographs were taken at both the beginning and end of each data collection run to detect any artefacts such as (partial) decomposition, microcracks, and other types of deterioration of the crystal caused by temperature variation or UV irradiation. The collected data were first processed using CrysAlisPro ver.41_64.93a (Rigaku) and then analysed using Olex 2.⁴⁶ The details of data collection, data processing, and structural analyses, which mainly consisted of standard procedures, are summarized in crystal information format (cif) files deposited to The Cambridge Crystallographic Data Centre (SupplementaryTable 2).

XAFS

The X-ray absorption spectra were obtained using beamline 15A1 at the Photon Factory of the High Energy Accelerator Research Organization. X-rays from the 2.5 GeV storage ring were monochromatized using a Si(111) double crystal. The spectra were recorded at ambient temperature and pressure. The samples were single crystals of MV[Au(dmit)₂]₂, freshly prepared, stored in the dark, and examined by single-crystal X-ray structural analyses to confirm they were high-quality single crystals without non-planar coordinated Au atoms. Controlling the thickness of the single crystals was not possible, leading to variations between crystals that significantly affected the signal-to-background ratio in both XAFS detection modes (fluorescence and transmittance). The spectra were expected to depend on the beam positions on the crystal surface because of possible inhomogeneity during and after UV irradiation. Accordingly, micro-XAFS was performed using a beam size of 17 mm (horizontal)×32 mm (vertical) mm². The X-ray beam (~12 keV) was incident upon the most developed faces (the ab planes) of the diamond-shaped crystals at an incident angle of 45°, to detect both the fluorescence and transmitted X-rays simultaneously using a custom-built Lytle detector and an ion chamber, respectively. The spectra of three single crystals in each salt (~0.4×0.3×d mm³ in dimension: d = 0.008–0.07 mm) were recorded to examine sample dependence and reproducibility. Every time a single crystal was set in the holder, a 2D map of the fluorescence/transmission intensity was collected by raster scanning the sample with a 20 mm step. The Au L_III absorption edge energy was calibrated using Au foil. While aligning the polarization angles of the incident X-rays along and perpendicular to the molecular planes of Au(dmit)₂ could detect the deviated Au atoms most sensitively, no such polarization angle existed due to the different stacking directions of Au(dmit)₂ columns in the unit cell. Preliminary spectra obtained with the polarization angles of the // and ^ b-axis in the ab plane exhibited only minor differences; hence, polarization dependence was not examined further. Initially, the ‘dark’ spectra were obtained as described. Next, the same sample was continuously exposed to UV light, during which the spectra were collected. The distance between the sample and the end of the light guide was 10 mm. To confirm the UV-irradiation time and sample dependencies of XAFS, we recorded spectra using different crystals under the same conditions, except for the UV-irradiation time (5 min). The obtained spectra quantitatively agreed with each other. Immediately after the cessation of UV irradiation (t = 0 min), the spectra were repeatedly collected under the same conditions. Each spectrum was the average of 5–10 recordings, each taking 15–30 min. The spectra did not change after being repeatedly recorded at the same position on the crystal within this time. However, the XAFS spectra change irreversibly after prolonged (³45 min) synchrotron X-ray irradiation at ~12 keV. We collected several XAFS spectra at different positions on the same sample to confirm reproducibility and homogeneity. The samples used in the XAFS analysis were checked afterwards by single-crystal X-ray structural analyses to confirm that they retained their original structure and crystal quality without decomposition or deterioration.

XPS

XPS spectra were recorded using the single crystals covering a conducting C tape (~1.5 ´ 1.5 mm², Nisshin EM) and a spectrometer JEOL JPS 9200 (Mg Ka (1,253.6 eV, 12 kV, 40 mA)). The base pressure in the main chamber was ~6.6 ´ 10^–⁷ and ~7.8 ´ 10^–⁶ Pa with and without X-ray irradiation, respectively. The charge-up effects on the BE were corrected using the C 1s peak of the C tape (graphite). An electrostatic objective lens was used, whereas a magnetic objective lens was not. The field of view was 3.0 mm. We examined the Au 4d_3/2, Au 4d_5/2, Au 4f_5/2, Au 4f_7/2, Au 5s, S 2s, S 2p_1/2, and S 2p_3/2 spectral regions at ~298 K. First, the stabilities of the samples under continuous X-ray irradiation were examined. The spectra remain unchanged after several months of continuous X-ray irradiation. The best measurement conditions were determined to be in the dark (Extended Data Table 1). The S 2p1/2 and S 2p 3/2 peaks could not be resolved as separate peaks because the energy difference between them is smaller than the spectrometer's resolution (~0.9 eV), resulting in the observation of a single merged peak. The sample was then removed from the spectrometer and exposed to UV irradiation in air for 5 min using the Hg/Xe lamp described above (200 W, the range of 250—450 nm was selected using a bandpass filter and an Al mirror). The distance between the end of the optical fibre and the sample was ~1 cm. For comparison, we performed XPS on an independently prepared sample of the same material under dark conditions, and we then exposed the sample (in the main chamber) to UV light from the same light source through the quartz window of the main chamber, during which we conducted XPS under UV irradiation (~1 h). The distance between the sample and the end of the optical fibre was ~30 cm. The XPS spectra after cessation of UV irradiation were consistent between the two UV irradiation conditions, demonstrating that the results were qualitatively independent of the UV irradiation conditions. Because Ar sputtering irreversibly changes the spectra with poor reproducibility, the sample surface was not etched prior to collecting the XPS spectra presented in this paper. Recording the series of spectra required ~1 h. The obtained spectra were distinguished by the elapsed time t (h) after the cessation of UV irradiation, which was defined as the time at which we began collecting every series of XPS spectra. The spectrum at t = 0 (h) was obtained during UV irradiation. The XPS spectra were continuously recorded until t = 300 h, after which spectra were collected every 24 h until t = 3,000 h. The obtained spectra were analysed by curve fitting using SpecSurf ver. 1.9.2.12 (2012, JEOL Ltd.) to obtain the BEs and intensities. The intensity is the peak area derived from the photoelectron counts (cps) multiplied times the acquisition time (s) and integrated along the BE (eV) across the peak widths.

Physical property measurements

Electrical resistivity was measured at 296 K using crystals, a custom-built cryostat, and the standard two-probe method with a constant direct current. Magnetic susceptibility was assessed using polycrystalline samples with a magnetic property measurement system MPMS-XL7 (Quantum Design). Detailed methods for these measurements can be found in our previous publication.^20,43,47,48

Theoretical calculations

DFT calculations (RwB97XD/LanL2DZ(5d, 7f)) were performed using Gaussian 16³¹ and GaussView 6.1.1.⁴⁹ To obtain the energies of the three structural types of Au(dmit)₂ anions in the solid state, X-ray molecular structures were utilized without structural optimization. The electronic energy and thermodynamic quantities were calculated for a simplified system containing only the Au(dmit)₂ anion and MV cation. To examine model dependence, we performed similar calculations using systems with one molecule (MV or Au(dmit)₂) and two molecules (MV–Au(dmit)₂ or Au(dmit)₂–Au(dmit)₂). The results indicated that the molecular systems depicted in Supplementary Figs. 4a–c were more appropriate than the remaining models.

Data availability

Source data (crystallographic information format (cif) files) for Figs. 1b–d, 2a, and 2c as well as Extended Data Figs. 1–4, and 6 were deposited at the Cambridge Crystallographic Data Centre, with the registry numbers summarized in Supplementary Table 2. The data were obtained free of charge from The Cambridge Crystallographic Data Centre at www.ccdc.cam.ac.uk/data_request/cif. The source data for Figs. 2b–e, 3a–f, and 4a–e as well as Extended Data Figs. 5a–f, 7a–d, and 8a–d are provided with this paper in the Supplementary Information.

42. Naito, T., et al. Simultaneous control of carriers and localized spins with light in organic materials. Adv. Mater. 24, 6153–6157 (2012) (DOI 10.1002/adma.201203153)

43. Steimecke, G., Sieler, H.-J., Kirmse, R. & Hoyer, E. 1.3-Dithiol-2-thion-4.5-dithiolat aus Schwefelkohlenstoff und alkalimetall. Phosphorus Sulfur Relat. Elem. 7, 49–55 (1979) (DOI 10.1080/03086647808069922)

44. Naito, T., et al. UV-vis-Induced vitrification of a molecular crystal. Adv. Funct. Mater. 17, 1663–1670 (2007) (DOI 10.1002/adfm.200600583)

45. Wright, J. D. Molecular motion in crystals’ in Molecular Crystals. 2^nd ed, Chap. 5 74–96 (Wiltshire: Cambridge University Press, 2024), GB

46. Dolomanov, O. V., Bourhis, L. J., Gildea, R. J., Howard, J. A. K. & Puschmann, H. OLEX2: a complete structure solution, refinement and analysis program. J. Appl. Crystallogr. 42, 339–341 (2009) (DOI 10.1107/S0021889808042726)

47. Naito, T., et al. Molecular photoconductor with simultaneously photocontrollable localized spins. J. Am. Chem. Soc. 134, 18656–18666 (2012) (DOI 10.1021/ja306260b)

48. Naito, T., Inabe, T., Niimi, H. & Asakura, K. Light-induced transformation of molecular materials into devices. Adv. Mater.16, 1786–1790 (2004) (DOI 10.1002/adma.200400308)

49. Dennington, R., Keith, T. A. & Millam, J. M. GaussView. version 6.1.1 (Shawnee Mission KS: Semichem Inc., 2016)

Acknowledgements The authors are grateful to T. Tsuneda at Hokkaido University for valuable discussions regarding the DFT calculations. The authors would like to thank S. Mori and R. Konishi at ADRES, Ehime University, for their single-crystal X-ray structural analyses, experimental advice, assistance with troubleshooting of the hardware and software, and productive discussions. The authors are grateful to K. Konishi for assistance and valuable advice in using the SQUID. They are also grateful to K. Asakura at the Institute for Catalysis, Hokkaido University, for the valuable assistance with the XAFS experiments. The XAFS experiments were performed with the approval of the Photon Factory Program Advisory Committee (Proposal No. 2022G104). This work was partially funded by a Grant-in-Aid for Challenging Exploratory Research (18K19061) from the JSPS, the Iketani Science and Technology Foundation (ISTF; 0331005-A), the Research Grant Program of the Futaba Foundation, the Casio Science Promotion Foundation, an Ehime University Grant for Project for the Promotion of Industry/University Cooperation, and the Canon Foundation (Science and Technology that Achieve a Good Future).

Author contributions T.N. conceived and supervised the project, designed almost all experiments, confirmed the experimental results, checked the analyses, and wrote the paper. Under the instruction of T.N., Y.M. synthesized the compound, including its single crystals; S.S. performed preliminary studies on the single-crystal X-ray structural analyses and theoretical calculations; M.N. and Y.N. performed the single-crystal X-ray structural analysis and the first-principles calculations; K.M. recorded the XPS spectra; and K.M. and M.N. measured the magnetic susceptibility. Y.T., Y.N., and T.N. collaborated to obtain the XAFS spectra. All the authors discussed the results and commented on the manuscript.

Competing interests The authors declare no competing interests.

Additional information

Supplementary Information The online version contains supplementary material available at https://doi.org/......

Materials & Correspondence

Correspondence and requests for materials should be addressed to Toshio Naito ([email protected]).

Reprints and permissions information is available at http://www.nature.com/reprints.

There is NO Competing Interest.

Download PDF

Version 1

posted

You are reading this latest preprint version

Direct observations of slow photo-induced structural and physical changes

Status:

Version 1

Abstract

Figures

UV-excitation-responsive materials

UV-induced intermolecular interactions

Real-time atomic-level observation

Real-time observation of electrons

Discussion

Online content

References

Methods

Declarations

Additional Declarations

Supplementary Files

Status:

Version 1