Toward Improving Triplet Energy Transfer from Tetracene to Silicon Using a Covalently Bound Tetracene Seed Layer

Silicon solar cells are operating close to the theoretical maximum efficiency limit. To increase their efficiency beyond this limit, it is necessary to decrease energy losses occurring for high-energy photons. A sensitizing layer of singlet-fission material can in principle double the current generated by high-energy photons, and significantly reduce energy losses from high-energy photons within the solar cell. Here, we construct a model of such a solar cell, using Si(111) surfaces and tetracene. To increase the energy transfer between the two layers, a series of tetracene derivatives was synthesized, and the molecules were covalently attached onto the silicon surface as a seed layer. Using X-ray diffraction, a shift in crystal structure and ordering of the tetracene close to the seed layer can be observed. Unfortunately, the effect on the energy transfer was limited, showing a need for further investigations into the effect of the seed layer.

Tetracene deposition 30 nm or 100 nm tetracene layers were deposited on silicon surfaces using an evaporation chamber by Angstrom Engineering Inc, at a base pressure below 7*10 -7 mbar. Tetracene was purchased from Sigma-aldrich (99.99% purity) and used as is. The deposition was 1Å/s in all cases.
Magnetic-field-dependent photocurrent measurements were performed using a home-built setup. The magnetic field is applied by an electromagnet, made up by two Helmholtz coils and calibrated using a Hall effect sensor. The magnetic field is applied by sending a current of up to 5 Amperes through the magnet, resulting in a magnetic field of up to 0.35 T. The field is oriented parallel to the sample surface. The excitation source is a 520 nm diode laser, installed in a Thorlabs temperature controlled laser housing. The cw laser power is around 10 mW with a laser spot size of approx. 1 mm. The photocurrent is measured with a Keithly 2636A Source measure unit.

Synthesis of tetracene
Tetracene was prepared according to the procedure described in Kulkarni et al. (2018). 1 In short, 2.50 g (9.68 mmol, 1 equiv) of 5,12-naphthacenequinone was dissolved in 220 ml each of methanol and THF.
While stirring, 1.65 g (43.62 mmol, 4.5 equiv) of NaBH 4 was slowly added, and the reaction mixture was allowed to stir for another hour at room temperature. Then, the mixture was neutralized with (glacial) acetic acid, and transferred to an extraction funnel. Here, 150 ml of both brine and DCM are added, and the organic layer is separated. The organic layer is washed with 2x 150 ml of brine, and the combined water layers are back-extracted with 150 ml of DCM. The organic layers are combined, dried over magnesium sulfate, and concentrated on a rotavapor to yield 2.34 g (8.93 mmol) of 5,12-dihydrotetracene-5,12-diol as a yellow solid.

5-bromotetracene
In a three-necked flask equipped with a stir bar, 1.00 g (4.38 mmol, 1 equiv) of tetracene was suspended in 250 ml of dry DCM, and the suspension was sonicated for 30 minutes. Meanwhile, an addition funnel was charged with 0.86 g (4.83 mmol, 1.1 equiv) of n-bromosuccinimide (NBS), 70 ml of dry DCM, and 15 ml of dry DMF. This solution was flushed with argon for 10 minutes, after which the addition funnel was installed onto the three-necked flask under argon flow. A condenser was added, and the tetracene suspension was heated to 50 °C under argon. Then, the NBS solution was added dropwise, over a period of 2h. Next, the reaction mixture was allowed to stir overnight at 50 °C, in the dark, under argon. The mixture was cooled down to room temperature, and the DCM was removed on a rotavapor. 30 ml of methanol was added to the resulting slurry, and the mixture was cooled down to 0 °C. 5 ml of water was added dropwise, and the red solids were collected by vacuum filtration. The solids were washed with cold water and methanol to yield 1.25 g (4.08 mmol, 93%) of 5-bromotetracene as a red solid.

Synthesis of 2-ethynyltetracene (2)
For the synthesis of 2-ethynyltetracene, 2-bromotetracene was first prepared according to a procedure from Zhao et al. (2020). 3 Then, an adapted procedure from the one described above for 5-ethynyltetracene was followed.

2-bromotetracene
The fresh 8-bromotetracene-5,12-dione was dissolved in 140 ml of isopropanol, and 2.18 g (57.63 mmol, 14 equiv) of NaBH 4 was added. A condenser was installed onto the flask, and the reaction mixture was allowed to reflux at 95 °C for 24h in the dark. After cooling down, the mixture was placed on an ice bath, and 140 ml of a 2M aqueous HCl solution was added dropwise. The mixture was heated to 105 °C, and allowed to reflux for 3h. After cooling, the solids were collected using vacuum filtration, and washed with water, methanol, acetone, and hexane, to yield 0.97 g (3.14 mmol, 75%) of 2-bromotetracene as an orange solid.   1H). This compound is unstable, and could therefore not be further characterized. Full characterization was therefore performed on 2-((trimethylsilyl)ethynyl)tetracene. 4-fluoro-α,α,α,α-tetrabromoxylene was prepared according to the procedure described by Zhao et al.  15 (s, 1H). This compound is unstable, and could therefore not be characterized further. Full characterization was therefore performed on 5-((trimethylsilyl)ethynyl)-8-fluorotetracene.

Surface functionalization and backfilling
Double side polished silicon (111) surfaces of 1x1 cm were cleaned before etching and functionalisation by sonicating them for 10 minutes each in isopropanol, hexane, acetone, and DCM. After the final sonication step, the surfaces were blown dry with a stream of nitrogen, and submitted to a 10-minute plasma treatment in a plasma oven, using oxygen gas to generate the plasma. Then, the surfaces were transported into an oxygen-free glovebox (O 2 = 0%). Here, they were etched with a solution of 40% NH 4 F in water for 15 minutes, under constant bubbling with nitrogen gas to remove bubbles formed during etching. After etching, the surfaces were washed extensively with deoxygenated water, and blown dry with a stream of nitrogen. The thus prepared surfaces were immediately used for functionalisation by fully immersing them in either an argon-purged solution of tetracene derivative in dry toluene, or neat 1-pentyne for the reference surface. The immersed surfaces were heated to 90 °C, and allowed to react overnight. The next day, the surfaces were removed from the solution, and sonicated for 10 minutes in dry DCM to remove any physisorbed 1-pentyne or tetracene derivative. The surfaces functionalised with tetracene derivative were then submerged in neat 1-pentyne for backfilling. Again, the surfaces were allowed to react at 90 °C overnight, and cleaned by sonicating in dry DCM.