Projects

Getting every watt off the array

The team was budgeting array power from cell datasheets, which ignores everything that happens between a bare cell and an encapsulated module on a moving vehicle.

• Characterise Maxeon Gen III cells by outdoor four-wire IV sweep on a Keithley 2461, and analyse 3×6 modules for mismatch loss to inform cell binning and array layout.
• Lead infrared and electroluminescence inspection to find hotspots, cracks and interconnect faults; piloted a new encapsulation method for durability and water-ingress resistance.
• Build Python tools modelling bypass-diode configurations and temperature-driven efficiency loss, to inform thermal and cooling design for the next-generation array.

Saving energy in Stanford's residential buildings

• Audited 320 common spaces across 45 buildings and roughly 2,900 lighting elements to map occupancy-sensor coverage. Only 36.9% of spaces had them.
• Built an ROI model from the audit data projecting 66,421 kWh and $11,941 in annual savings against $65,650 upfront, a payback of about 5.5 years.
• Evaluated vendors and specified hardware for a window-sensor HVAC pilot.

Cleanroom fixtures

Sample geometry was quietly a variable in our deposition results. A masking split during sputtering gave 8/8 working devices with a modified mask against 4/8 without, which was enough to justify designing proper fixtures. Substrate holders and evaporation masks are in use; the latest sputter mask is designed but not yet fabricated Not yet made.

Organic transistors and OLEDs

• Fabricated bottom-gate PDBT-co-TT organic thin-film transistors with a PFBT self-assembled monolayer on the gold contacts, confirmed by goniometry (36.7° to 64.3° contact angle). Extracted threshold voltage and saturation mobility from transfer curves and measured 13% higher mobility, consistent with reduced interfacial trap density.
• Built F8BT OLEDs with PEDOT:PSS and PFN-Br injection layers and a 100 nm aluminium cathode evaporated at SNF. Used matched electroluminescence and photoluminescence peaks to attribute an anomalously high turn-on voltage to spin-coating defects rather than interlayer energetics.

Remote sensing of wetland change

Geospatial analysis in Google Earth Engine, ArcGIS and R, with fieldwork in Louisiana and sample processing back on campus. Wrote the sample-prep SOP for cryomilling and digestion, and was second author on an AGU abstract. My first research position, and where I learned to handle data before I learned to make devices.

Composite pressure vessels for cryo-compressed hydrogen

Project work on COPV design considerations for a hydrogen storage startup.