Over the past decade, HIT-type solar cells with back electrodes have significantly improved the performance of commercially available photovoltaic panels; nevertheless, there is still scope for even higher performance cells based on crystalline silicon. A high-efficiency silicon cell made under laboratory conditions, the Passivated Emitter Rear Locally diffused cell (PERL), is shown in Fig. 1. This cell designed in the late 1980s had efficiencies as high as \( 23\% \), a huge improvement in silicon cell efficiency of \( 17\% \) – the highest value just 7 years earlier. Since then, further refinement of PERL-type cells has brought their efficiency to \( 25\% \).
The PERL-type cell shares many features with the cell whose electrodes are placed at the back. Similarly, there is complete shielding in the passivating oxide layer and small surface contacts. However, the PERL type cell is a more robust design, more tolerant to environmental conditions, in view of the poor surface passivation and poor performance of classic cells.
Major improvements made in the PERL-type cell in recent years include a significant increase in the oxide passivation of the top surface, which allows the direct application of a two-layer anti-reflection coating [1]. An "annealing" sequence for this oxide and localized top contact points are used to increase the open circuit voltage and improve back surface passivation, and metallization resistance is chosen to improve the fill factor.
To maximize cell performance, as much light from useful wavelengths as possible should be absorbed by the cell. To achieve this result, modern cell designs, such as the PERL cell, incorporate several optical solutions.
In this case, the optical losses in the front of the cell are reduced by implementing a textured inverted pyramid structure ( Fig. 1 ), coated with an anti-reflective layer, which allows the absorption of the reflected light a second time, reducing the transmission losses. The electrodes at the front were made to have the smallest possible surface area, which increases the intensity of light entering the photovoltaic cell.
The inverted pyramids along the top surface are primarily for optical purposes. Most of the radiation quanta will hit one of the side walls of the pyramids, which causes the rays to be reflected, thus increasing the optical path of the light. The reflected light ray gives the light, at least, a second chance to return to the cell and be absorbed. Some quanta near the bottom of the pyramids sometimes have more chances.

Pyramids are coated with a layer of oxide thick enough to act as an anti-reflective coating. In more recent designs, the oxide layer is thin and a two-layer anti-reflection coating is used [1].
The quantum absorbed by the cell moves diagonally through the cell towards the back surface of the cell. In this way, the radiation quantum has a longer path over which it can be absorbed by the cell. The unabsorbed light reaching the back surface is reflected by a highly efficient reflector formed by combining a back oxide layer covered with a layer of aluminium [2]. The reflectance from this combination depends on the angle of incidence and the thickness of the oxide layer, but typically exceeds \( 95\% \) for incidence angles near \( 0_{}^{o} \) (normal). The reflection coefficient decreases below \( 90\% \) when the incident angle approaches the angle of total internal reflection at the silicon/oxide interface ( \( 24.7_{}^{o} \)) and increases again to near \( 100\% \) when this angle is exceeded.
On the back surface of the photovoltaic cell, spot electrodes are used in conjunction with thermal oxide passivation layers to reduce unwanted recombination at the surface in the uncontacted region. Silicon heavily doped with boron (p+) acts as a local back surface to limit minority electron recombination.
Light reflected from the back moves toward the top surface. Some quanta of radiation reach the surface and can leave the photovoltaic cell without producing an exciton. Others undergo total internal reflection. This causes about half of the light radiation directed at the front surface from the inside to be reflected back into the cell toward the back electrode.
The number of light quanta leaving the photovoltaic cell after the first reflection depends on the geometry of the pyramids. The loss of light energy can be reduced by destroying some of the symmetries used, for example by using tilted inverted pyramids or by using the "tiler's pattern" method. The latter approach is currently used in PERL cell designs.
The combination of inverted pyramids and a back reflector creates a very efficient way to increase light absorption by increasing the path length of a light ray in a photovoltaic cell. The effective light ray path length enhancement factors are measured [2]. The increase in absorption occurs primarily in the infrared region.
The external responsivity (that is, the response of the photovoltaic panel (in amperes) to 1 watt of incident light) of PERL-type cells reaches higher values at longer wavelengths than conventional silicon cells, with values of 0.75 \( \frac{A}{W} \) measured at 1.02 \( \mu m \). The energy conversion efficiency for some wavelengths is even greater than \( 45\% \) [3] compared to classic cells.

Current PERL-type cells lose approximately \( 5\% \) of the incoming light due to absorption loss or reflection associated with the metal electrodes, combined with reflection from the non-metalized top surface of the cell. The optical losses that occur are also due to reflection and absorption by the top metal electrodes of the cell. One can minimize this reflection by sizing the electrodes, height, width, and shape. One can also try to redirect light rays into the cell by bypassing the electrodes [4]. Thus, there is some room for small to moderate increases in efficiency by further improving the optical properties of these cells.

Advanced cell designs have been used in spacecraft and terrestrial high-value applications such as solar car racing [5]. Very expensive multi-step photolithographic processes were used there. Such photovoltaic cells are too expensive for wide applications. Nevertheless, recent advances in cell construction using laser technologies make it possible to produce PERL-type cells at low cost. As mentioned, the best cell efficiency is \( 25\% \), and the best photovoltaic panel efficiency is \( 22.9\% \) [6].

It appears that crystalline silicon wafer technology will be the dominant photovoltaic technology for at least the next decade. This is indicated by the investments made in photovoltaic manufacturing plants and the reduction in PV panel prices. Bifacial Cell Technology (PERL) photovoltaic panels have also recently emerged.

In summary, photovoltaic cells made with PERL technology have many advantages:

• higher cell efficiency achieved by passivation of n-type c-Si material,
• only one temperature step in the deposition technology, both phosphor and boron, that minimizes the loading on silicon wafers,
• short production time of PERL cells,
• different cell production technologies, single-sided or double-sided cells,
• higher, up to \( 30\% \), energy efficiency of double-sided cell.