Solar energy storage at an atomically defined organic-oxide hybrid interface

Molecular photoswitches provide an extremely simple solution for solar energy conversion and storage. To convert stored energy to electricity, however, the photoswitch has to be coupled to a semiconducting electrode. In this work, we report on the assembly of an operational solar-energy-storing organic-oxide hybrid interface, which consists of a tailor-made molecular photoswitch and an atomically-defined semiconducting oxide film. The synthesized norbornadiene derivative 2-cyano-3-(4-carboxyphenyl)norbornadiene (CNBD) was anchored to a well-ordered Co3O4(111) surface by physical vapor deposition in ultrahigh vacuum. Using a photochemical infrared reflection absorption spectroscopy experiment, we demonstrate that the anchored CNBD monolayer remains operational, i.e., can be photo-converted to its energy-rich counterpart 2-cyano-3-(4-carboxyphenyl)quadricyclane (CQC). We show that the activation barrier for energy release remains unaffected by the anchoring reaction and the anchored photoswitch can be charged and discharged with high reversibility. Our atomically-defined solar-energy-storing model interface enables detailed studies of energy conversion processes at organic/oxide hybrid interfaces.

3 measurements were performed in a 1-cm path length cuvette scanning the wavelength from 700 to 290 nm on either a Cary 50 Bio or a Cary 100 UV-vis spectrophotometer, coupled with Peltier temperature control. Photoswitching for bulk conversion to CQC was performed using a Vilber Lourmet TLC lamp at 610 μw/cm 3 with a wavelength of 365 nm. Photoswitching at wavelength 310 nm was performed using Thorlabs M310L3 LED lamps. The thermal back reaction was performed by heating the sample (cuvette) by a Peltier unit in the UV-vis spectrophotometer.
Quantum yields were measured by a published procedure in a high concentration regime (absorption above 2 at 300 nm) using potassium ferrioxalate and tris-phenanthroline iron (II) complex as a chemical actinometer. 11 The cuvette was irradiated perpendicularly in a fixed setup with stirring ensuring no movement during the experiment, using a Thorlabs LED lamp M300L4 with an attached collimator. HRMS spectra were acquired by atmospheric pressure chemical ionisation (APCI) using an Agilent 1260 Infinity instrument fitted with an Agilent 6120 quadrupole. Elemental analyses were performed at London Metropolitan University.
Alkyne 1 was made according to a previously reported protocol. 12 To stirring ice cooled DMF (100 mL, 1.29 mol) POCl3 (15 mL, 160 mmol) was slowly added under a nitrogen atmosphere.
The ice bath was removed and the vessel allowed to stir for 15 min. The vessel was re-immersed in ice, 4'-iodoacetophenone (12.51 g, 50.8 mmol) was added, and the flask was heated to 50 °C for 3 h. The cooled solution was poured into 20% aqueous NaOAc (300 mL) and then allowed to cool overnight. The solid was collected by suction filtration and washed with water (3 x 100 mL). Subsequently, the solid was dissolved in CHCl3 (200 mL), and I2 (12.00 g, 47.2 mmol) and 28% aqueous NH3 (100 mL) were added to this stirring solution. After stirring for 3h at RT, saturated aqueous NaS2O3 (200 mL) was added and the phases were separated. The organic phase was dried over Na2SO4, filtered and the solvent was removed in vacuo. The residue was taken up in THF (200 mL), and aqueous NaOH (2.39 g, 59.8 mmol, in 10 mL H2O) was added to this stirring solution. After 4 h, saturated NaHCO3 (200 mL) was added, the mixture extracted with Et2O (2 x 200 mL) and the combined organics washed with saturated brine (100 mL). The organic extracts were dried over Na2SO4, filtered and the solvent removed. The residue was purified by flash column chromatography (gradient elution of CH2Cl2/petroleum spirit 1:4 to CH2Cl2/petroleum spirit 3:7) to obtain 1 (8.40 g, 65%) as a light orange solid. Rf = 0.62   Single crystal X-ray crystal structure. A single crystal was mounted in paratone-N oil on a plastic loop. X-ray diffraction data were collected at 150(2) K on an Oxford X-calibur single crystal diffractometer using Mo K radiation. 13 The data set was corrected for absorption using a multi-scan method, and structure solved by direct methods using SHELXS-2014 and refined by full-matrix least squares on F2 by SHELXL-2014, 14,15 interfaced through the program X-Seed. 16 All non-hydrogen atoms were refined anisotropically and hydrogen atoms were included as invariants at geometrically estimated positions. The X-ray experimental data and refinement parameters for CNBD are given below. Perspective views of the single X-ray crystal structure of CNBD are shown below highlighting the NBD framework and the planarity of the chromophore. Crystals were grown from CH2Cl2/nheptane and data collected as described above.

Supplementary Figure 25:
Single crystal X-ray crystal structure of CNBD, perspective 1.

Supplementary Figure 26:
Single crystal X-ray crystal structure of CNBD, perspective 2.

Supplementary Figure 27:
Single crystal X-ray crystal structure of CNBD, perspective 3.

Supplementary Discussion
Calculation of quantum efficiency and conversion probability. The (CN) region of the IRAS data were fitted using the KolXPD software package. The density of CNBD molecules in one monolayer was assumed to be identical to the number of binding sites at the surface Co 2+ ions, which is 3.6 nm -2 . 17 Other thicknesses were estimated from the deposition time relative to the monolayer value. The external quantum efficiency QE was calculated as where n(CNBDconsumed) is the density of CNBD converted, and nphotons is the number of incident UV photons per area as calculated from the photon flux density and the irradiation time. The relative conversion probability per molecule P was calculated as = ∆ (CNBD) ∆ photons × (CNBD) (2) where ∆ (CNBD) are the CNBD molecules converted per irradiation step and ∆ photons are the number of incident photons per irradiation step.
Effect of the molecular orientation on the conversion probability. We assume that the CNBD molecules are randomly oriented in the multilayer and so are the transition dipole moments over the sphere as illustrated below. The incident light is directed along the z axis and the not polarized.
Supplementary Figure 28: Orientation of transition dipole moments within volume integral. 20 The fraction of transition dipoles ( . . /2) which have a polar angle between and 2 ⁄ is: The active component par of the transition dipole moment ( ) which is parallel to the xy plane is calculated as The corresponding absorption probability (or photoconversion probability) is where is an experimental constant.
Using the above result, the photoconversion probability can be expressed as a fraction of the total number of absorbers as: Consequently, we would expect the transition probability to decrease to 75% of its original value at a conversion of = 0.5 and to 19% of its original value at a conversion of = 0.9. The experimental values are much larger, showing that the molecular orientation is not the dominating factor for coverage dependence of the photoconversion probability. Also the inhomogeneity of the light intensity over the irradiated area is by far too small to account for the very large variations in photoconversion probability, which exceed two orders of magnitude between = 0 and = 0.9.
Effective electric field at the CNBD/Co3O4(111) interface. The CNBD/Co3O4 interface is irradiated by UV light at normal incidence. Due to the low thickness of the Co3O4 film (8 nm), we consider the light reflection at the underlying Ir single crystal. From the Fresnel formula, the reflection coefficient at normal incidence is = = 1 − 2 + 2 1 + 2 − 2 (7) where 1 is the refractive index of the external medium, i.e. vacuum or CNBD film, and 2 and 2 are the real and the complex (absorption) part of the refractive index of the Ir crystal. From the above formula, the phase shift between the incident and the reflected wave is calculated as tan = 2 2 1 1 2 − 2 2 − 2 2 (8) and the amplitude of the reflected wave is calculated as | | = | |√ ( 1 − 2 + 2 ) 2 ( 1 + 2 − 2 ) 2 (9) Using the refractive index of Ir at 360 nm ( 2 = 1.9 and 2 = 3.3) we calculate a phase shift of 154° at the vacuum interface ( 1 = 1), an amplitude of the reflected wave of | | = 0.78| | and a resulting field at the interface of | + | interface = 0.30| |.
For an organic film with 1 = 1.5, we calculate | | = 0.70| |, so that the maximum field strength above the surface will be | + | max = 1.70| |. As the UV absorption will be proportional to 2 , the maximum ratio between absorption in the thick film and monolayer film will be ( | + | max | + | interface ) 2 = ( 1.70 0.30 ) 2 ≈ 32 (10) The ratio compares well with the ratio between the initial reaction probabilities in the thick multilayer film and in the monolayer.