Manufacturing

Through wet chemical etching, wafers are cleaned from residual contaminants and the saw damage is removed. Texturing is the most common technology used in the reduction of optical losses in monocrystalline silicon solar cells. This process increases the collected photons and thus, improves cell efficiency. Alkaline texturing process creates square based pyramids randomly distributed on the surface of the wafer so that the sunlight gets reflected rather than trapped. Wafers run through a cascading rinsing process called Etching. It starts with a KOH rinse followed by kOH and Additives. The wafers are rinsed with DI (deionized water), which is the purest form of water, with the mobility of electrons.

To establish a separation of the photo-generated charge carriers, the solar cell needs a p/n-junction. For p-type based solar cells, this is realized by a n-type emitter. The dopant is Phosphorus (P), coming from a liquid (POCl3) source. The dopant is diffused into the silicon by a combined oxidation/heat treatment in the POCl3 diffusion system.

For further improvement of efficiency and to reduce cost, Laser Doping Selective Emitter (LDSE) has been introduced. With this technology, we can go up to sheet resistance of 180-200 ohm/sq . A laser is used to burn a pattern onto the wafer in the area where fingers will be printed later in the manufacturing process. Typical sheet resistance in the lasered area is about 70-90 ohm/sq after the process. The PSG on the wafer contains a lot of phosphorous and acts as a diffusion source for the process. The laser ablates the PSG, melts the silicon locally at the surface (temperature >1400 °C) and causes a fast in-diffusion of the phosphorous atoms.

Mono PERC technology requires a well-polished surface on the rear side (pyramid removed) for better compatibility with TMA (Trimethyl Aluminium) during the rear side PECVD process. During the diffusion process, the n-layer is also formed on edges and results in shorting between p-layer and n-layer of the cell. In case shorting is not removed, unwanted current will flow resulting in shunted cells. This process isolates the junction from the edges by etching HF and HNO3 chemical. In the same bath, additives are used to polish the rear surface for better cell efficiency. The emitter layer on the rear side of the wafer, generated during the diffusion process, is unilaterally isolated from the front side of the wafer in order to prevent malfunction of the solar cell. The layer thickness depends on the diffusion parameters and is between 20-50 nm. Further, the PSG needs to be removed before silicon nitride deposition because it lacks appropriate refractive index and has a poor surface passivation. The etching of PSG is a very stable process as the PSG itself is etched very fast in diluted hydrofluoric (HF) acid. The bare silicon has a very slow etch speed so even if the wafers are still in the HF solution after PSG is removed, there is only a minor risk of etching back the emitter.

A simplified diffusion furnace with wafers loaded back to back is used to form a silicon oxide layer on the front of the wafer. The thin oxide on the front of the wafer serves two purposes:
1. Silicon oxide passivates n-type surfaces very well (better than SiN)
2. The oxide layer warrants PID stability of the cell
The process also further heals the laser damage and further creates a slightly deeper diffusion of phosphorous. To avoid optical losses, the layer thickness is kept below 15 nm.

In this process, a thin layer of Al2O3 (AlOx) is deposited on the rear p-type surface (PECVD) and then capped with a layer of SiNx (antireflection coating). Generally, direct plasma tube type PECVD is used for rear side AlOx/SiNx deposition. The AlOx layer is about 8-12 nm thick and makes an almost perfect passivation to the p-type wafer. The capping layer i.e. SiNx layer is about 85-100 nm thick and caps the AlOx as well as delivers the required thickness for good optical reflection. After AlOx/SiNx deposition, a laser is used for the local opening on the rear dielectric layers.

To further enhance the efficiency of the solar cell, it is a good approach to deposit an anti-reflection coating on the cell’s n-side, which is exposed to the sun. This coating is made up of silicon nitride and traps the light by making constructive interference pattern of light inside the solar cell and destructive interference outside the solar cell. Since the coating increases the light absorption property of solar cells and reduces reflection loss, it directly increases the short circuit current and hence, increases efficiency. For use as antireflection coating of silicon solar cells, a silicon nitride layer of minimum 75 nm (on flat surfaces) at a refractive index of 2.0 to 2.1 is recommended.

A critical step in the manufacturing process of PERC solar cells is Laser Contact Opening (LCO), where laser ablation is used to perforate a thin passivation layer onto the rear side of the solar cell. This process reduces electrical losses in the cell, resulting in approximately 1% (absolute) higher conversion efficiency.

To complete a solar cell, electrical contacts have to be applied to the front side and the back side of the processed wafer. This is carried out by deposition of metal paste via screen-printing. The wafer is placed on a printing-table where a screen is lowered onto the wafer and a spreading knife presses the printing paste through the screen onto the wafer. The back contact is made of Silver soldering pads. A direct soldering of aluminium is not possible due to immediate formation of Aluminium oxide, which prevents wetting by solder. After Silver soldering pads are printed, the back side is printed with Aluminium grid shaped lines to create a back surface field. This field pushes the electron towards the n-side which are created by low energy photons and hence, increases efficiency. The front contact is made of Silver and is grid-shaped for maximum light transmission.

The front contact must penetrate the antireflective coating (ARC) and contact the emitter near the highly doped surface in order to get low contact resistance. On the rear side at the aluminium silicon interface a melt is produced and during cooling a highly aluminium doped layer is crystallized. During the firing process, hydrogen stored in the silicon surface and in the silicon nitride antireflective coating must be driven into the bulk as fast as possible. However, the process must be stopped before the hydrogen effuses again from the bulk. Based on these process demands, the firing process must be done by a temperature pulse of a few seconds with very fast heating and cooling ramps.

Light-induced degradation (LID) can cause significant losses in output power of crystalline silicon solar cells and modules. The main reason for LID in monocrystalline Cz-Si solar cells is the formation of a recombination active Boron-Oxygen (BO) complex in the silicon bulk. After the firing process, the cell is passed through a regeneration tool where very high intensity illumination at elevated temperature (~100-300 °C) is maintained. This makes the hydrogen atoms from the SiN layer mobile and they diffuse into the silicon wafer and passivate the Boron-Oxygen complex. The process intends to reduce Light Induced Degradation (LID) in the finished mono-crystalline PERC solar cell. Typically, LID is in the range of 3-10% without this process. With this process, it reduces to <1.5%.

In the last step, the cells are tested and sorted according to their electrical, EL and optical quality. In the cell tester unit all the electrical parameters of the solar cells (efficiency, short circuit current, open circuit voltage, fill factor, shunt resistance, series resistance etc.) are measured and the cells are sorted into different bins according to their efficiency and optical quality.

This process creates hills and valleys on the surface of the wafer so that the sunlight gets reflected rather than trapped. Wafers run through a cascading rinsing process called ETCHING. It starts with a KOH rinse followed by HF,HNO3 and Additives. The wafers are rinsed with DI (deionized water) which is the purest form of water, with the mobility of electrons.

To establish a separation of the photo-generated charge carriers, the solar cell needs a p/n-junction. For p-type based
solar cells, this is realized by an n-type emitter. The dopant is
Phosphorus (P), coming from a liquid (POCl3) source. The
dopant is diffused into the silicon by a combined oxidation /
heat treatment in the POCl3 diffusion system.

Edge Isolation – The diffusion process also results in N-layer being formed on the edges which results into shorting
between P and N region from the edge of solar cells. If
shorting is not removed, the unwanted current will flow which will result in shunted cells. The Edge Isolation process isolates the junction from the edges by doing etching to some micron using HF and HNO3 chemical.
PSG Removal – During Emitter diffusion, an oxide is built on
the wafer surface made up of high concentrations of
Phosphorous, thus being called phosphorous silicate glass
(PSG). The layer thickness depends on the diffusion
parameters and may vary between 20 and 50 nm. The PSG
needs to be removed before silicon nitride deposition,
because it lacks appropriate refractive index and has a poor
surface passivation. Thus PSG Removal is done.
The etching of the PSG is a very stable process, as the PSG
itself is etched very fast in diluted hydrofluoric (HF) acid, the bare silicon has a very slow etch speed. So even if the wafers are still in the HF solution after PSG is removed there is only a minor risk of etching back the emitter. Thus, the only parameters that influence the PSG removal itself are the HF concentration and the etch time.
Anti PID unit – SiO2 layer produced in the process of
ozonation. The silicon oxide layer is produced between the
silicon substrate and silicon nitride. Due to strong oxidation
power of ozone, when the silicon substrate is oxidized with
ozone, the bottom surface of the silicon layer of the silicon
oxide layer can be generated quickly, that is adding an
additional SiO2 dielectric layer between emitter and Anti
reflecting coating by inline ozone generator .The SiO2 films
grown by ozone, though only having a thickness of 1–2 nm,
still shows a good stability in PID. This thin layer of SiO2 is
confirmed through water drop test as it’s hydrophilic in nature. Thus, SiO2 dielectric layer ensures excellent PID.

Plasma Enhanced Chemical Vapour Deposition
In this step, an anti-reflective coating is put on the wafer to
ensure minimum reflection of the sunlight.

Once the wafer gets blue color, after PECVD, they go through printing cycle. Electrical contacts are applied to the front and
back side of processed wafer, which renders the functionality
of solar cell. This is carried out by deposition of metal paste
via screen-printing.

The front contact must penetrate the antireflective coating
(ARC) and contact the emitter near the highly doped surface
in order to get low contact resistance. On the rear side at the
aluminum silicon interface a melt is produced and during
cooling down a highly aluminum doped layer is crystallized.

In the final step the cells are tested and sorted according to
their electrical, EL and optical quality. In cell tester unit all the electrical parameters of the solar cells like efficiency, short circuit current, open circuit voltage, fill factor, shunt resistance, series resistance etc are measured and then the cells are sorted in to the different bins according to their efficiency, EL and optical quality into 48 different classes.