Working principle of carrier selective poly-Si/c-Si junctions: Is tunnelling the whole story?

https://doi.org/10.1016/j.solmat.2016.05.045Get rights and content

Highlights

  • Plausible arguments for additional current transport mechanisms in POLO junctions besides tunneling.

  • Presentation of an alternative picture based on local current flow.

  • Excellent agreement between model and experimental data for reasonable input parameters.

  • Deduction of an optimization strategy for p+ poly-Si/c-Si junctions.

  • Record J0 values for n+ POLO junctions with wet chemically (thermally) grown interfacial oxide – 1.5 (0.7) fA/cm2.

  • Record J0 values for p+ POLO junctions with wet chemically grown interfacial oxides – 8 fA/cm2.

Abstract

We present arguments that additional effects besides laterally homogenous tunnelling might occur in carrier-selective poly-Si/c-Si junctions: (i) the symmetrical electrical behaviour of n+ and p+ poly-Si/c-Si junctions, (ii) direct observation of structural modifications of the interfacial oxide upon thermal treatment by transmission electron microscopy, even for poly-Si/c-Si junctions with good passivation quality, and (iii) the achievement of low junction resistances even for interfacial oxide thicknesses >2 nm after thermal treatment. We present an alternative picture, essentially based on a localized current flow through the interfacial oxide, mediated either by local reduction of the oxide layer thickness or by pinholes. In consequence, the local current flow implies transport limitations for both minority and majority carriers in the c-Si absorber, and thus a correlation between recombination current and series resistance. Thus, a poly-Si/c-Si junction can also be explained within the framework of a classical pn junction picture for a passivated, locally contacted emitter, e.g. by the model of Fischer. Both electron selective contacts (n+ poly-Si) and hole selective contacts (p+ poly-Si) can be described consistently when using reasonable input parameters. Especially for p+ poly-Si/c-Si junctions, our model could guideline further improvement.

Introduction

Carrier selective junctions based on silicon rich, mostly polycrystalline (poly-) layers deposited on an interfacial oxide grown on a crystalline (c-) silicon absorber are currently attracting significant research interest. Excellent junction quality with recombination current densities J0 down to 0.7 fA/cm2 [1], as well as efficiencies up to 25.1% [2] have been demonstrated with this approach.

Regarding the physical working principle of poly-Si/c-Si junctions, different pictures exist [3], [4]. For interfacial oxide thicknesses of ~2 nm or less, it is safe to assume that tunnelling through the oxide takes place. Nevertheless, we present in this work strong support that, at least for our samples with initial interfacial oxide thicknesses >2 nm, additional effects are relevant, which could in particular explain the excellent performance of our p+ poly-Si/c-Si junctions.

Section snippets

Arguments for additional relevant effects

Our first argument for additional current transport mechanisms besides lateral homogeneous tunnelling is the experimental observation of low junction resistances ρc and low recombination current densities J0 for both n+ poly-Si/c-Si and p+ poly-Si/c-Si junctions. This might not be consistent with the tunnelling model, as explained in the following:

As indicated in Fig. 1, the poly-Si is inherently highly defective, i.e., a high density of energy states within the band gap of the poly-Si could

Alternative picture

The alternative picture presented in the following is a generalized version of our previous model according to Ref. [4]. A comparison of both approaches, as well as a summary of differences and similarities is given in the appendix.

As known from secondary ion mass spectrometry (SIMS) and electrochemical capacitance voltage (ECV) measurements [7], [16], dopants diffuse from the poly-Si into the c-Si during the junction formation process, especially for boron doped poly-Si layers (Fig. 4(a)).

Conclusions

We present arguments that laterally homogeneous tunnelling through an unaltered interfacial oxide is possibly not the only current mechanism present in poly-Si/c-Si junctions. First, excellent p+ poly-Si/c-Si junction properties (low recombination current, low junction resistance) have been achieved despite a “wrong tunnelling barrier height ratio” for electrons and holes, which seems inconsistent to the expectations of the tunnelling mode. Second, structural modifications of the interfacial

Acknowledgments

We would like to thank Susanne Mau, Bianca Gehring and Renate Winter for their help with sample processing, and Andrea Lissel for her help with the TEM sample preparation. We furthermore would like to thank Nils-Peter Harder and Bianca Lim for valuable discussions at his/her ISFH times. We furthermore would like to thank Uwe Höhne and Jan-Dirk Kähler from Centrotherm for the LPCVD (low-pressure chemical vapor deposition) poly-Si depositions.

This work is mainly performed in the framework of the

References (31)

  • U. Römer et al.

    Ion Implantation for Poly-Si Passivated Back-Junction Back-Contacted Solar Cells

    IEEE J. Photovolt.

    (2015)
  • J. Y. Gan, R.M. Swanson, Polysilicon emitters for silicon concentrator solar cells, in: Proceedings of the 21st IEEE...
  • Sentaurus Device User Guide, Version A-2007.12, pp. 455–456, table 79,...
  • I.R.C. Post et al.

    Polysilicon emitters for bipolar transistors: a review and re-evaluation of theory and experiment

    IEEE Trans. Electron. Dev.

    (1992)
  • F. Feldmann, M. Bivour, C. Reichel, A. Moldovan, H. Steinkemper, A. Richter, J. Benick, M. Hermle, S.W. Glunz, Advanced...
  • Cited by (188)

    • Photon management in silicon photovoltaic cells: A critical review

      2024, Solar Energy Materials and Solar Cells
    View all citing articles on Scopus
    View full text