It is no surprise that the birth of the semiconductor chip (IC) industry has boosted technology and manufacturing to a large extent in the photovoltaic (PV) industry. Both of them use single crystal silicon wafers (to a large extent for now in the PV industry) and fabrication techniques and tools that bear a lot of similarity. In spite of these commonalities, they have diverged away from each other as they matured.
Increase in the computing speed and reducing the size of the IC components (the transistor) has been the driver for performance (processing speed) and reduced cost. This is based on the well known Moore’s Law which is defined as the doubling of transistors per unit area every eighteen months. This law is central towards defining standards and building the roadmap for the IC industry and has certainly proven to be a tremendous success in the modern computing age. Translating Moore’s law to the technology level, it is the reduced gate length that increases the performance and reduces the cost (due to smaller transistor size).
The PV industry has certainly matured with some highs and lows over the last few decades. With the concern on energy crisis and climate issues, renewable energy has taken the front seat on innovation and capital spending, of which solar, primarily PV is a key energy source to address this concern. Similar to processing speed, efficiency is the key driver for any energy industry. This in turn along with an increased economy of scale drives cost down. For the PV industry, there are several technology approaches to raise the efficiency bar. However, the spectrum is so wide that it appears unclear what the roadmap needs to be for the industry to reach the high efficiency and low cost target. Since PV and IC industries are similar at a fundamental level, even though they have become divergent lately, taking a few pages out of the IC industry learning experience such as the Moore’s law is certainly a perceivable thought for the PV industry.
A recent article [1] described using the volume of material in the PV cell as the parameter to translate Moore’s law at the technology level similar to the surface area used for transistors in the IC industry. The analogy between the two is the increased power per unit volume for the PV cell versus the increased number of transistors per unit area for the IC. For a given volume, area is seldom changed (in order to maintain a large collection area), thereby making thickness as the knob to decrease the material volume. Reducing the material cell thickness as simple as it may sound, is one of the key challenges in the PV industry. As in reducing the gate length which comes with resolving other issues due to scaling such as yield and transistor parametrics, so does reducing the thickness of the PV cell, which entails resolving recombination issues, a key parameter that needs to be extremely low for efficiency improvement along with light absorption issues that is dependent on band-gap/doping profiles. Another example that is an emerging market, even though might not be a direct comparison to conventional PV but deserves mention is the concentrated PV technology where the volume of material used is extremely small. Here, the technology needs orders of magnitude higher input radiation from the sun as opposed to conventional PV cells.
The bright side towards drawing this parallel using Moore’s law between the two industries is that this would help define a path to more focused innovations, capital spending and set energy policies, regulations and standards in the PV industry. Contingent upon the success of this law would set a conceptual framework for predicting the speed of innovation and could arguably be also extended to other sources of the energy industry.
[1] Future Photovoltaics, May 2010
Nagarajan Sridhar, Technologist, Solar Lab
