The relentless pursuit of higher efficiency and lower cost in photovoltaic technology drives continuous innovation at every layer of a solar cell. In my research on copper indium gallium selenide (CIGS) thin film solar panels, a particular focus has been on the transparent front electrode. This component serves the critical dual function of admitting light into the absorbing layer while efficiently collecting the generated electrical current. The common choice, aluminum-doped zinc oxide (AZO), presents a fundamental trade-off: thicker films offer better conductivity but poorer transparency, and vice-versa. In this article, I will detail the conception, development, and successful implementation of a novel patterned multilayer front contact architecture that decouples this trade-off, leading to a measurable boost in the performance of CIGS thin film solar panels.
The heart of the challenge lies in the properties of the transparent conductive oxide (TCO). For CIGS thin film solar panels, AZO is favored due to its good transparency, non-toxicity, and relatively low cost. Its sheet resistance ($R_{sh}$) and optical transmittance ($T$) are primarily governed by its thickness ($d$) and the doping level. A simple model for the conductivity is given by:
$$ \sigma = n e \mu $$
where $\sigma$ is the conductivity, $n$ is the carrier concentration (enhanced by Al doping), $e$ is the electron charge, and $\mu$ is the carrier mobility. While increasing $d$ lowers $R_{sh}$ ($R_{sh} = 1 / (\sigma d)$), it also increases optical absorption in the AZO layer itself, reducing the photon flux that reaches the CIGS absorber. This is quantified by the parasitic absorption loss. The conventional solution in our lab was a 300 nm thick AZO layer, which offered a balance but was not optimal.
Researchers have explored AZO/metal/AZO (AMA) structures, where an ultra-thin metal layer (often silver, Ag) is sandwiched between two AZO layers. The metal layer drastically improves lateral conductivity. However, the continuous metal film, despite being thin, introduces significant optical absorption across the entire active area of the cell. This absorption loss often negates the electrical benefit for a thin film solar panel, where maximizing light in-coupling is paramount. The optical transmittance of such a multilayer stack can be approximated by considering interference effects, but the strong absorption coefficient ($\alpha_{Ag}$) of the metal is the dominant limiting factor:
$$ T_{total} \approx T_{AZO(top)} \cdot e^{-\alpha_{Ag} \cdot d_{Ag}} \cdot T_{AZO(bottom)} $$
This led me to a key insight: the metal’s electrical function is only crucial in specific locations—directly beneath the current-collecting metal grid lines that are deposited on top of the final cell. The grid lines, typically made of Ni/Ag, are necessary to collect current from the large area TCO but themselves block light. Therefore, placing a continuous conductive Ag layer across the entire cell sacrifices optical performance unnecessarily.
I proposed a novel architecture: an AZO/Patterned-Ag/AZO front contact. In this design, the intermediate Ag layer is not continuous. Instead, it is deposited through a shadow mask, forming discrete lines that have the identical pattern, width, and alignment as the final Ni/Ag grid lines. This patterned Ag layer sits directly underneath the grid. The rest of the cell area comprises only the two AZO layers. The conceptual advantage is clear:
- Electrical: The patterned Ag lines provide a low-resistance path for current flow directly beneath the main grid, effectively reducing the series resistance ($R_s$) contribution from the TCO in the most critical regions.
- Optical: Since the Ag pattern is perfectly shadowed by the opaque top grid, it introduces zero additional optical loss to the cell. The areas between grid lines, where light must enter, are free of the absorbing metal film.
This architecture cleverly separates the optical and electrical pathways in the front contact of the thin film solar panel.
To validate this concept, I fabricated and characterized several front contact schemes on glass substrates. The structures and their key properties are summarized in the table below. All AZO layers were deposited by RF magnetron sputtering under optimized conditions (base pressure: $1.2 \times 10^{-5}$ Pa, working pressure: 0.1 Pa). The Ag layers were deposited by e-beam evaporation. The “patterned” Ag was defined using a mechanical mask.
| Front Contact Structure | Layer Thickness (nm) | Average Weighted Transmittance (400-1100 nm) (%) | Sheet Resistance, Rsh (Ω/□) | Optical-Electrical Figure of Merit* |
|---|---|---|---|---|
| AZO (Reference) | 300 | 77.5 | 16.2 | ~1870 |
| AZO (Thin) | 180 | 79.9 | 63.1 | ~1010 |
| AZO/Ag/AZO (Continuous) | 50/8/50 | 62.9 | 6.0 | ~660 |
| AZO/Patterned-Ag/AZO (Novel) | 50/8/50 | 79.7 | 38.2 | ~1660 |
*A common figure of merit for TCOs is $\Phi = T^{10} / R_{sh}$, used here for qualitative comparison.
The data powerfully supports the hypothesis. The novel AZO/Patterned-Ag/AZO structure achieves a transmittance virtually identical to the thin, high-transparency 180 nm AZO film (~79.7% vs. 79.9%). Crucially, it does this while cutting the sheet resistance nearly in half (38.2 Ω/□ vs. 63.1 Ω/□). Compared to the standard 300 nm AZO reference, our novel structure offers superior transparency and a manageable increase in $R_{sh}$. In stark contrast, the continuous AZO/Ag/AZO stack, while having the lowest $R_{sh}$, suffers from a catastrophic drop in transmittance (>15% absolute), which is detrimental for a thin film solar panel.
The ultimate test was the integration into complete CIGS thin film solar panels (substrate size: 0.5 cm²). The device stack was: soda-lime glass / Mo (700 nm) / CIGS (2 µm, co-evaporated) / CdS (CBD) / i-ZnO (50 nm) / Front Contact / Ni(50nm)/Ag(1000nm) Grid. The front contact was the variable. Precise alignment ensured the patterned Ag lines sat directly under the final Ni/Ag grid. The current-voltage (J-V) characteristics under standard AM1.5G illumination were measured, and the key parameters are extracted in the following table.
| Front Contact Type | Open-Circuit Voltage, Voc (mV) | Short-Circuit Current Density, Jsc (mA/cm²) | Fill Factor, FF (%) | Series Resistance, Rs (Ω·cm²)* | Conversion Efficiency, η (%) |
|---|---|---|---|---|---|
| AZO (300 nm) | 647 ± 3 | 31.35 ± 0.05 | 68.67 ± 0.23 | ~2.7 | 13.83 ± 0.03 |
| AZO (180 nm) | 640 ± 6 | 32.54 ± 0.22 | 66.69 ± 0.14 | ~7.7 | 13.75 ± 0.09 |
| AZO/Ag/AZO (Continuous) | 631 ± 3 | 20.20 ± 0.42 | 66.87 ± 0.51 | ~2.1 | 8.52 ± 0.16 |
| AZO/Patterned-Ag/AZO (Novel) | 647 ± 3 | 32.63 ± 0.66 | 67.70 ± 0.41 | ~4.6 | 14.14 ± 0.39 (Champion: 14.53) |
*Estimated from J-V curve fitting.
The performance trends correlate directly with the optical and electrical data. The cell with the continuous AZO/Ag/AZO contact shows a severely depressed $J_{sc}$, consistent with its low transmittance, leading to very poor efficiency. The cell with the thin 180 nm AZO contact shows a higher $J_{sc}$ than the 300 nm reference due to better transparency, but its high $R_s$ (from high $R_{sh}$) degrades the FF, resulting in similar overall efficiency.
The cell with our novel AZO/Patterned-Ag/AZO front contact delivers the best of both worlds:
- It matches the high $J_{sc}$ of the thin AZO cell (and surpasses the reference), thanks to its excellent optical transmittance.
- It maintains a $V_{oc}$ on par with the reference cell.
- It achieves a FF significantly higher than the thin AZO cell because its lower $R_{sh}$ translates to a lower $R_s$.
The synergy of these factors results in a clear efficiency gain. The champion cell achieved an efficiency of 14.53%, representing an absolute increase of 0.7% (a relative improvement of approximately 5%) over the standard 300 nm AZO baseline. This enhancement can be primarily attributed to the increased photocurrent, governed by the relationship:
$$ J_{sc} = q \int_{\lambda} \phi(\lambda) \cdot EQE(\lambda) \, d\lambda $$
where $q$ is the electron charge, $\phi(\lambda)$ is the photon flux, and $EQE$ is the external quantum efficiency. The higher transmittance of our novel front contact directly boosts the $EQE$ across the spectrum, increasing the integral and thus $J_{sc}$.
In conclusion, the AZO/Patterned-Ag/AZO front contact architecture successfully breaks the traditional transparency-conductivity trade-off in TCO design for CIGS thin film solar panels. By strategically placing conductive material only where it is electrically necessary and optically inconsequential, we achieved a net gain in device performance. This design principle—minimizing parasitic optical losses while optimizing electrical collection—is a powerful guiding concept for advancing not only CIGS technology but potentially other thin-film photovoltaic and optoelectronic devices where transparent electrodes are employed. Future work will focus on optimizing the pattern geometry (line width, spacing), exploring other metal candidates, and scaling the patterning process for large-area module manufacturing to fully unlock the potential of this approach for commercial thin film solar panels.