Breaking the photoswitch speed limit

The study addresses the need for dramatically faster photoisomerization in solid-state materials used in optoelectronics, artificial muscles, data encryption, and catalysis. Conventional approaches that tune solvent polarity and viscosity typically yield only modest (≈ one order of magnitude) improvements and solid-state performance is often limited by close packing and intermolecular interactions that hinder large structural rearrangements and, for some switches (e.g., spiropyrans), formation of zwitterionic intermediates. The authors hypothesize that engineering a solvent-free, confined yet spacious and rigid environment around photoswitches can both remove solvent stabilization of zwitterionic states and accommodate the required conformational changes, thereby enabling ultrafast photoisomerization in the solid state. They aim to demonstrate this using spiropyran, hydrazone, and diarylethene derivatives embedded in porous metal-organic frameworks (MOFs), comparing rates in solution, solvent-filled MOFs, and evacuated (solvent-free) MOFs, and to explore generality and mechanisms via spectroscopy and TD-DFT modeling.

• Photoisomerization rates in solution can be modulated by solvent polarity/viscosity but usually within one order of magnitude. Solid-state switching is typically slower due to close packing, π–π stacking, and H-bonding which impede large structural changes and zwitterion formation (notably for spiropyrans).
• Spiropyrans switch between neutral spiropyran (SP) and zwitterionic merocyanine (MC) via excited-state C–O bond cleavage, involving large conformational change and increased dipole; this hinders solid-state switching.
• Diarylethenes undergo a 6π electrocyclic ring-closure between two neutral isomers with minimal structural rearrangement and often switch efficiently in the solid state.
• Hydrazones isomerize (E↔Z) via rotation/inversion about C=N and can involve significant spatial reorganization; solid-state performance can be constrained by packing and intermolecular forces.
• Prior confined-environment approaches and surface grafting provided only slight rate enhancements; achieving solution-like rates in bulk solids has been a field benchmark. Reported solution kinetics for similar systems are typically in the 0.0006–0.2 s⁻¹ range under comparable conditions, aligning with literature on these classes of switches.

• Systems studied: three classes of photochromic compounds (spiropyrans 1–3; hydrazones 4–6, 7–9; diarylethenes 11–13). Compounds chosen based on reported photophysical data to enable direct comparison and control for instrument/conditions.
• Environments compared: solution; bulk solid; MOF-confined with pores filled by synthesis solvent (e.g., DMF); and evacuated (solvent-free) MOFs.
• Scaffold selection criteria: (i) framework integrity under chemical modification and UV/vis irradiation; (ii) voids (≈12–16 Å) to accommodate structural changes (≈90° for spiropyran, ≈180° for hydrazone, length changes for diarylethene); (iii) metal nodes enabling coordination of photoswitch linkers to prevent leaching.
• Integration strategies: (a) De novo pillar incorporation between 2D layers constructed from Zn paddlewheel nodes and H4DBTD linkers to form Zn2(photoswitch)(DBTD); (b) Post-synthetic coordination of carboxylate-functionalized photoswitches to unsaturated Zr sites (defects) in UiO-67 (Zr6O4(OH)4(BPDC)6). Examples: Zn2(3)(DBTD) (spiropyran) via de novo; UiO-67+2 (spiropyran), UiO-67+5 and UiO-67+6 (hydrazones), UiO-67+11 (diarylethene) via post-synthetic installation.
• Characterization: PXRD to confirm scaffold integrity; 1H NMR of digested MOFs and TGA to quantify loading and defect levels; no leaching detected by supernatant spectroscopy after solvent exposure.
• Kinetic measurements: UV–vis absorbance for solutions; diffuse reflectance for solids/MOFs using integrating sphere setups. Samples irradiated with 365 nm LED, followed by visible attenuation (400–900 nm) to monitor MC→SP or E↔Z relaxations. Thickness controlled to be within excitation penetration depth. Kinetics fitted with first-order monoexponential functions; measurements performed in triplicate.
• Solvent effects: compared MOF samples with pores containing synthesis solvent (e.g., DMF) vs evacuated (solvent-free) to assess polarity/viscosity and electrostatic stabilization effects.
• Theoretical modeling: TD-DFT to construct excited-state PES for spiropyrans 2 and 3, monitoring r1 (C2–O1) and r2 (C1–O1) as reaction coordinates. Compared non-integrated (solution-like) vs coordinatively integrated (anchoring groups fixed) to assess changes in excited-state barriers.
• Dual-switch platform: Co-integration of spiropyran 2 and hydrazone 5 into UiO-67 (UiO-67+2+5); distribution probed by epifluorescence microscopy and SEM-EDX after Cu2+ binding to MC of 2 to infer statistical distribution.
• Additional methods: SEM-EDX conditions detailed; instrument models for NMR, FTIR, PXRD, UV–vis, diffuse reflectance; TGA to estimate defects and thermal stability.

• Solution-like rates preserved in solvent-filled MOFs:
  • Spiropyrans: UiO-67+2 in DMF k = 0.061 s⁻¹ vs 2 in DMF 0.036 s⁻¹; Zn2(3)(DBTD) in DMF 0.054 s⁻¹ vs 3 in DMF 0.08 s⁻¹.
  • Hydrazones and diarylethenes also maintained comparable rates to solution when confined with solvent.
• Record solid-state acceleration for spiropyrans under solvent-free confinement:
  • UiO-67+2 evacuated (solvent-free) k = 31.2 s⁻¹ vs 2 in DMF 0.036 s⁻¹ (~867×) and vs UiO-67+2 in DMF 0.061 s⁻¹ (~512×).
  • Zn2(3)(DBTD) evacuated k = 53 s⁻¹ vs 3 in DMF 0.08 s⁻¹ (~662×); vs Zn2(3)(DBTD) in DMF 0.054 s⁻¹ (~981×).
  • This constitutes ~1000× enhancement in the solid state, surpassing previously conceived “speed limits” for spiropyrans.
• Minor effects for neutral-switch classes under evacuation:
  • Hydrazones: UiO-67+5 DMF 0.005 s⁻¹ → solvent-free 0.013 s⁻¹ (~2.6×); UiO-67+6 DMF 0.029 s⁻¹ → solvent-free 0.026 s⁻¹ (no significant increase).
  • Diarylethenes: Minimal variations between solution and MOF; Zn2(12)(DBTD) in DMF 0.037 s⁻¹; solid-state 12 at 0.36 s⁻¹; evacuation of Zn2(12)(DBTD) not possible due to degradation.
• Dual-switch MOF (UiO-67+2+5):
  • As-synthesized (solvent in pores): k(2) = 0.084 s⁻¹ (λmax 576 nm), k(5) = 0.0004 s⁻¹ (λmax 363 nm).
  • After evacuation: k(2) ≈ 55 s⁻¹ (~1000× increase), k(5) ≈ 0.00097 s⁻¹ (~2.5× increase).
  • Enables complementary modulation of absorption across 300–650 nm with a dynamic rate range spanning ~10⁵ in response times (10⁻1 to 10¹ s⁻¹).
• Mechanistic insight via TD-DFT: Excited-state barrier for 2 changes minimally upon coordination (ΔEmax 26.66 kcal/mol non-integrated vs 26.57 kcal/mol integrated), indicating the rate enhancement arises primarily from engineered pore environments (reduced electrostatic stabilization of zwitterionic MC, unhindered structural rearrangement) rather than intrinsic barrier lowering by rigidity.
• Practical considerations: No leaching of coordinated photoswitches detected; framework integrity maintained under irradiation for studied systems (confirmed by PXRD).

Embedding spiropyrans in rigid, solvent-free MOF pores removes stabilizing interactions of polar solvents with the zwitterionic merocyanine state and provides sufficient free volume to accommodate large conformational changes during SP↔MC interconversion. This dual effect—electrostatic de-stabilization of MC in the absence of solvent and reduced steric/viscous hindrance—drives order-of-magnitude acceleration in solid-state photoisomerization, far exceeding solution performance. In contrast, neutral-switch classes (hydrazone, diarylethene) show minimal rate enhancement upon evacuation, consistent with the hypothesis that solvent stabilization of a zwitterionic intermediate is the key lever for spiropyrans. TD-DFT indicates the excited-state barrier is not substantially altered by mere coordination/rigidity, underscoring the importance of pore-environment engineering. The preservation of solution-like behavior in solvent-filled MOFs demonstrates that confinement per se does not impede kinetics, and that rates can be finely tuned by toggling pore solvation. Co-integration of distinct switches enables simultaneous, complementary control over spectral regions and kinetic timescales within a single solid material, suggesting utility for multilevel anticounterfeiting, information encryption, and programmable release in photopharmacology. Overall, the results validate the central hypothesis and establish confined, solvent-free MOF environments as a general strategy to surpass conventional solid-state switching limitations.

The work demonstrates a conceptually distinct route to break the perceived speed limit of photochromic switching in solids by engineering solvent-free, rigid, and spacious MOF environments. Spiropyran derivatives achieve ~1000-fold rate enhancement in the solid state relative to solution, while maintaining reversibility and stability. Neutral hydrazone and diarylethene switches retain near-solution behavior, with only modest increases upon evacuation, highlighting the specificity of the mechanism to zwitterion-forming switches. A dual-switch MOF (spiropyran + hydrazone) affords complementary optical modulation across 300–650 nm and a very broad dynamic range of rates (approx. 10⁻1 to 10¹ s⁻¹), enabling programmable response times spanning five orders of magnitude. Future research directions include: expanding to other zwitterion-forming switches; rational pore-environment design (polarity, functionality) to tailor rates and selectivity; integrating into device architectures for ultrafast optoelectronics, real-time sensing, and controlled photorelease; and ensuring scaffold stability under evacuation for broader classes of frameworks.

• Framework stability under evacuation can limit solvent-free studies: Zn2(12)(DBTD) degraded upon evacuation, precluding solvent-free kinetic measurements for that diarylethene system.
• The dramatic rate enhancement is most pronounced for zwitterion-forming spiropyrans; hydrazone and diarylethene classes show only slight changes, which may limit generalizability across all photoswitches.
• Kinetic analysis relies on first-order monoexponential fits and specific irradiation/probe conditions; different light intensities, wavelengths, or sample geometries could influence extracted rate constants.
• Photoswitch loading levels (e.g., low coverage in UiO-67) and defect distributions may affect local environments; while leaching was not detected and distributions appeared statistical, heterogeneities at the nanoscale cannot be fully excluded.