Process
1. Substrate Preparation
- Ultrasonically cleaned ITO-coated glass in IPA, followed by DI water.
- Masked the substrate edges with Kapton tape to preserve an exposed ITO contact.
2. Cu₂O Electrodeposition
- Electrolyte: 0.25 M CuSO₄ with 3 M lactic acid, adjusted to pH 10 with NaOH and held at ~50 °C.
- Deposition: 1.0 V from a DC power supply for 40 min in a two-electrode cell: the ITO glass as the cathode (negative lead) and a graphite counter electrode as the anode (positive lead).
- Rinsed in DI water and dried.
3. Thermal Oxidation
- Removed the masking and annealed in a box furnace at 425 °C for 1 h, converting the Cu₂O to CuO.
- Allowed the samples to cool to room temperature in the furnace before removal.
4. ZnO Reactive Sputtering
- Re-masked the contact edges and sputtered from a Zn target in pure O₂ for ~5 min.
5. Aluminum Top Contact
- With the masking in place, sputtered from an Al target in Ar for ~8 min.
- Performed a final anneal at 425 °C for 1 h, oxidizing the thin Al layer.
Microscopy Results
AFM of Cu₂O After Deposition
AFM after Cu₂O deposition showed pronounced surface texturing. Some peaks appeared incomplete, likely due to inspecting the sample mid-deposition.
SEM Results
SEM was taken after each fabrication step to follow how the surface changed through the stack.
Cu₂O after electrodeposition
The electrodeposited Cu₂O formed a dense, continuous film of faceted, pyramid-shaped crystallites about 1–2 µm across, with no bare substrate visible across the 41.5 µm field of view. Sharply faceted grains like these are typical of Cu₂O deposited from an alkaline (pH 10) lactate bath, and the full coverage indicates uniform current density across the substrate.
Feature formed during annealing
Annealing converted the Cu₂O to CuO, and needle-like nanowires appeared across the surface. They were scattered over the film (left), grew densely out of a crack (center), and lined the film edge (right). Nanowire growth like this is common when copper oxides are thermally oxidized in air at around 400 °C, and is generally attributed to stress in the growing oxide. The clustering at cracks and edges fits that explanation, since those are points of high stress.
ZnO after sputtering
The sputtered ZnO formed a fine-grained coating over the CuO. In the right panel the faceted grains beneath are still visible, which shows that the thin ZnO follows the texture of the layer under it rather than smoothing it out. The left and middle panels also show a crack and loose surface debris. Defects like these can become shunt paths through the junction, which matches the low shunt resistance seen in the I-V measurements.
Thin aluminum oxide layer
The thin aluminum layer, oxidized during the final anneal, did not form a continuous film. The underlying grains are still visible through it, and bright rounded nodules, from about 0.2 µm to over 1 µm across, sit on the surface either alone or in clusters. This suggests the aluminum gathered into droplet-like nodules rather than wetting the surface. The resulting incomplete coverage is a likely cause of the poor top contact.
Failed deposition: island growth
In a failed deposition, the Cu₂O grew as separate islands instead of a continuous film. At low magnification (right), the islands are grouped in streaks with bare substrate between them. At 2 kx (center), they are rounded and beginning to merge. At 10 kx (left), each island turns out to be made of many small faceted crystals. A loose electrical contact during deposition is the likely cause: it made the current density uneven, so growth concentrated at a few spots instead of covering the substrate evenly.
IV Characterization
Dark IV confirmed diode behavior. Under illumination, the curves suggested significant recombination losses consistent with the surface defects and non-uniform coverage visible in SEM. Shunt resistance was low, indicating leakage paths likely from the incomplete aluminum coverage and porous CuO layer.
Challenges & Lessons
- Handling samples between furnace steps introduced contamination; cleaning between each step (ultrasonic IPA + DI water) would reduce surface defects
- Loose contact during copper deposition on one sample caused non-uniform current density, visible as discontinuous regions and spherical particles in SEM
- Without introducing O₂ gas flow during oxidation, the growth rate diverged from simulation predictions
Future Work
- Tighter process control at each step, particularly cleaning between depositions
- Introduce proper dry O₂ flow during thermal oxidation steps
- Investigate conformal ZnO deposition to reduce interface defects
