Effect of gas flow rates on PECVD-deposited nanocrystalline silicon thin film and solar cell properties

https://doi.org/10.1016/j.solmat.2007.09.013Get rights and content

Abstract

Nanocrystalline silicon films have been deposited at a plasma excitation frequency of 54.24 MHz by varying the flow rates of SiH4+H2 mixture in the reaction chamber. It has been found that with increase in gas flow rate from 100 to 300 sccm the defect density, microstructural defect fraction and the crystalline volume fraction in the film decrease. Films deposited at optimum total gas flow rate of 200 sccm with comparable crystalline volume fraction have shown better structural and optoelectronic properties compared to the films deposited at 100 sccm total gas flow rate for application in solar cell. Solar cells have been fabricated using these layers as absorber layers and the maximum cell efficiency obtained is 6.2% (AM1.5, 28 °C) at 200 sccm total gas flow rate. It has been found that material prepared using higher total gas flow rate of 200 sccm together with higher hydrogen dilution is better suited for solar cell application.

Introduction

Microcrystalline and nanocrystalline hydrogenated silicon thin films have attracted considerable attention for their promising applications in high efficiency and stable solar cells. As absorption coefficients of these silicon layers are low in the visible region of solar radiation, the thickness of the absorber layer should be more than 1 μm. So the deposition rate should be high to make solar cell commercially viable. For this purpose high power, pressure and very high plasma excitation frequency have been used. Plasma excitation frequency higher than conventional 13.56 MHz for deposition of nanocrystalline Si:H thin film reduces the maximum energy of ions impinging on the substrate while applying high excitation frequency and power to obtain an effective gas dissociation and a high deposition rate. Deposition rate also can be improved by supplying film-forming precursors at higher rate, i.e. by increasing the total gas flow rate. This also improves the film property by reducing the affect of powder formation on film growing surface [1], [2]. Mai et al. [1], [3] studied the effect of total gas (Tfl) flow and plasma power density mainly on the solar cell properties and found that higher Tfl have favourable effect on the solar cell performance. They also discussed the variation of crystalline volume fraction with Tfl and correlate them with abstraction reaction of atomic hydrogen. They suggested that hydrogen dilution should be increased at higher Tfl to reduce the effect of abstraction reaction. But they have not done a detailed analysis of the material prepared with different total gas flow rate. Niikura et al. [4] studied the effect of total gas flow rate using a special type of cathode and at a very high power pressure regime of 4 W/cm2 power density and 9.3 Torr chamber pressure. They found crystallinity decreases with increase in total gas flow rate and their sample become amorphous at very high total gas flow rate. So in the present article we have varied the total gas flow rate from 100 to 300 sccm and also varied other parameters to keep the film growth in microcrystalline or nanocrystalline region. We have studied the properties of films with similar crystallinity but deposited with different total gas flow rate, have examined their role as an absorber layer of solar cell and have correlated these studies. Higher total gas flow rate reduces gas phase polymerisations. But for a microcrystalline growth at Tfl more than 300 sccm, hydrogen dilution greater than 98% is required which in turn reduces the deposition rate drastically.

Section snippets

Experimental

Undoped hydrogenated nanocrystalline silicon thin films have been prepared decomposing mixture of silane (SiH4) and hydrogen gas (H2) in a multichamber PECVD system. The multichamber has a load lock and separate chambers for the deposition of different layers of pin solar cell. The plasma excitation frequency (fex) and substrate temperature (Ts) were kept at 54.24 MHz and 180 °C, respectively. Total gas flow rate (Tfl) was varied from 100 to 300 sccm. The hydrogen dilution [Y=(H2/(SiH4+H2))×100]

Deposition rate

Fig. 1 shows variation of deposition rate with total gas flow rate for samples deposited under different deposition conditions. Deposition rate increases as Tfl increases from 100 to 300 sccm under all conditions but the rates of increase are different. This increase in deposition rate is due to higher supply of film forming precursors at higher Tfl. Maximum deposition rate of 7.8 Å/s obtained at Tfl of 300 sccm. But the increase in deposition rate is not proportional to Tfl due to lower gas

Discussions

The crystalline volume fraction of the films deposited at higher total gas flow rate is lower while deposited under same power, pressure condition. The decrease in crystallinity at higher total gas flow rate is mainly due to abstraction reaction which reduces the atomic density of H atoms in the growth zone and crystalline volume fraction of the film. So with increase in Tfl the hydrogen dilution should be increased. From Table 1 it can be noted that for Pw=0.5 W/cm2, Tfl=100 sccm the value of Xc

Conclusions

The effect of total gas flow rate on nanocrystalline silicon thin films have been investigated in this paper. Abstraction reaction of atomic hydrogen is more prominent when total gas flow rate is higher. Deposition rate increases as Tfl increases, but rate of increase in deposition rate with Tfl saturates for Tfl above 200 sccm for lower power deposited films. Both microstructural defect fraction and crystalline volume fraction decrease as total gas flow rate increases. Higher total gas flow

Acknowledgments

The steady state photo carrier grating (SSPG) measurements have been done by Dr. C. Longeaud of LGEP, France. TEM studies were done under Nano Science and Technical Initiative Programme, DST, Government of India. The work has been carried out under a project funded by Ministry of Non-Conventional Energy Sources, Government of India. One of the authors (SM) gratefully acknowledges the financial support of the Council of Scientific and Industrial Research (CSIR), Government of India.

References (11)

  • C. Niikura et al.

    J. Non-Cryst. Solids

    (2004)
  • A. Shah et al.

    Sol. Energy Mater. Sol. Cells

    (2003)
  • J.K. Rath

    Sol. Energy Mater. Sol. Cells

    (2003)
  • A.H. Mahan

    Sol. Energy Mater. Sol. Cells

    (2003)
  • Y. Mai et al.

    J. Appl. Phys.

    (2005)
There are more references available in the full text version of this article.

Cited by (0)

View full text