Elsevier

Thin Solid Films

Volume 534, 1 May 2013, Pages 107-110
Thin Solid Films

Optimization of nitrogen plasma source parameters by measurements of emitted light intensity for growth of GaN by molecular beam epitaxy

https://doi.org/10.1016/j.tsf.2013.02.013Get rights and content

Abstract

A comprehensive analysis of operating parameters of Addon RF nitrogen plasma source was made in order to determine how a ratio of different active nitrogen species depends on operating parameters of the source such as supplied power and nitrogen flow. We show that output signal of the optical sensor that measures intensity of the light emitted by the plasma is a direct measure of the amount of active nitrogen available for growth. Results of optical emission spectroscopy and measurements of growth kinetics show that nitrogen excited metastable molecules are the species mainly contributing to the growth of GaN under Ga-rich conditions. A procedure is presented allowing to find an optimal conditions of the plasma cell for high-quality GaN growth. Under these conditions the nitrogen flux contains maximum amount of excited metastable molecules and minimal amount of ionic and atomic nitrogen species to minimize GaN lattice damage, even at high growth rates.

Highlights

► Operating parameters of Addon radio-frequency nitrogen plasma source studied ► Their influence on molecular beam epitaxy (MBE) growth of GaN analyzed ► MBE growth rate of GaN well correlates with output of the plasma emission sensor. ► Optical emission spectroscopy measurements of the nitrogen plasma made ► Nitrogen excited molecules mainly contribute to plasma-assisted MBE growth of GaN

Introduction

It is well established already that by plasma-assisted molecular beam epitaxy (PAMBE) the highest quality GaN layers are grown under Ga-rich conditions when nitrogen flux controls the growth kinetics [1], [2]. The most widely used are the radio frequency (RF) nitrogen sources, in which inductively coupled RF power ignites and sustains a plasma discharge in the source cavity. RF nitrogen plasma sources usually produce a complex mixture of nitrogen species relative concentrations of which depend on construction of the source and its operating parameters. Theoretical studies of reaction of different nitrogen species with GaN suggest that excited molecular nitrogen may be preferable for GaN growth due to a specific reaction mechanism when an excess energy of reacting molecule is carried away as a kinetic energy of one of two impinging atoms. In opposite, relatively high reaction energy (3.87 eV) of atomic nitrogen, superior to the bonding energy of GaN (2.2 eV), may be transferred directly to the growing film eventually causing damage of the crystal lattice [3]. Experimental studies suggest that neutral and ionic atomic nitrogen promote enhanced decomposition of GaN above 700 °C, while metastable molecular nitrogen leads to a stable growth of GaN layers with improved electrical quality [4], [5]. Also Arehart et al. reported a direct correlation between density of certain point defects with percentage of ionic and atomic nitrogen species emitted from the source, suggesting plasma damage of GaN layer [6]. Other sources indicate a role of high energy ions and boron contamination as additional sources of defects [7], [8]. The problem becomes especially important when the power supplied to the plasma source increases in order to increase GaN growth rate.

All these examples show that for high growth efficiency and good quality of GaN films precise control of concentrations of various nitrogen species produced by RF plasma source is crucial. However, relative contributions of them strongly depend on the source used, so direct comparison of results obtained with the sources manufactured by different producers is difficult. Ptak et al. compared efficiency of GaN growth with the use of two RF sources, Oxford CARS-25 and EPI Unibulb [4], [5]. They observed that the Oxford source produced primarily atomic nitrogen flux that was found relatively inefficient for growing GaN. The EPI source, producing significantly less atomic N and more excited nitrogen molecules, gave much higher growth rate and better quality layers. This finding allowed them to conclude that the metastable molecular nitrogen component of the total nitrogen flux governed growth kinetics of GaN. Iliopoulos et al. studied the influence of operating parameters of the Oxford HD25 RF nitrogen source on the PAMBE growth rate of GaN [9]. Their results indicate prime role of molecular nitrogen in the growth kinetics with a minor contribution of nitrogen atoms. Also a higher solubility of metastable molecular nitrogen in liquid gallium has been predicted, thus making these species the most available for GaN growth in Ga-rich mode [10].

In this work we carried a comprehensive analysis of operating parameters of Addon nitrogen RF plasma source to study their influence on composition of nitrogen flux produced and on kinetics of PAMBE growth of GaN. A procedure is presented allowing to find an optimal plasma conditions under which a maximum amount of excited metastable molecules and a minimal amount of atomic nitrogen are produced for every given growth rate, thus leading to high-quality GaN growth. An optical sensor was used to measure intensity of the light emitted by the source. Our results show that output signal of the sensor is well correlated with the growth rate of GaN under Ga-rich conditions. Thus, the sensor signal is a direct measure of the amount of active nitrogen available for growth. To get insight into the nature of species present in the nitrogen plasma optical emission spectroscopy measurements have been made. Our results suggest, in agreement with previous reports, that nitrogen excited metastable molecules are the species mainly contributing to the PAMBE growth of GaN. We show that by varying operating parameters of the source the growth rate of GaN can be changed in a wide range. For each growth rate a range of supplied power and nitrogen flow values is found which corresponds to maximal amount of excited molecular nitrogen in the mixture of different species produced by the cell. In that way the plasma source parameters are optimized for growth of GaN with low defect densities and at high growth rates.

Section snippets

Experimental procedure

An Addon RFN-50-63 plasma source was used to produce active nitrogen flux. The source was installed in a Riber Compact 21 MBE system equipped with conventional Knudsen cells supplying group-III atoms. An optical Si sensor delivered with the cell was coupled to a viewport of the RF source to measure intensity of the light emitted by the nitrogen plasma in the wavelength range of 750 to 850 nm. As will be shown below, its output voltage U is a direct measure of amount of nitrogen active species

Results and discussion

It is well known that nitrogen plasma intensity, and so the optical sensor output voltage U, depends both on the RF power coupled to the source cavity and the nitrogen gas flow. Fig. 1 shows results of experiment in which nitrogen flow was increased stepwise while for each step the RF power was adjusted to get a fixed value of sensor output voltage U. For each U value this procedure was repeated as long as the plasma stayed in the high brightness mode. As seen, the range of source operating

Summary

In conclusion, we performed a comprehensive analysis of operating parameters of Addon nitrogen plasma source and studied their influence of PAMBE growth of GaN. Our results show that output signal of the optical sensor that measures intensity of the light emitted by the plasma is a direct measure of the amount of active nitrogen available for growth. Thus, the sensor is fast and efficient tool for controlling the growth rate of GaN under Ga-rich conditions. The growth rate of GaN was found to

Acknowledgments

The authors are grateful to P. Nowakowski and L. Dmowski for their contribution to the experiments. This work was partially supported by the European Union within European Regional Development Fund, through grant Innovative Economy (POIG.01.01.02-00-008/08 NanoBiom). One of us (MS) thanks for support from European Social Fund through Human Capital Programme and local authorities (Samorząd Województwa Mazowieckiego – “Potencjał naukowy wsparciem dla gospodarki Mazowsza – stypendia dla

References (13)

  • N. Newman

    J. Cryst. Growth

    (1997)
  • E. Iliopoulos et al.

    J. Cryst. Growth

    (2005)
  • B. Heying et al.

    Appl. Phys. Lett.

    (2000)
  • C.R. Elsass et al.

    Appl. Phys. Lett.

    (2000)
  • A.J. Ptak et al.

    Appl. Phys. Lett.

    (1999)
  • T.H. Myers et al.

    J. Vac. Sci. Technol. B

    (1999)
There are more references available in the full text version of this article.

Cited by (24)

  • Strain control in graphene on GaN nanowires: Towards pseudomagnetic field engineering

    2022, Carbon
    Citation Excerpt :

    GaN NWs were grown on in-situ nitridated Si(111) substrate in a Riber Compact 21 system with an elemental source of Ga and Al. A radio frequency Addon nitrogen plasma cell, controlled by an optical sensor of plasma light emission was used as the source of active nitrogen species [46]. The growth temperature and time were varied to prepare NWs with different: interdistances and number of NWs on the surface.

  • Effects of nitrogen flow rate on the morphology and composition of AlGaN nanowires grown by plasma-assisted molecular beam epitaxy

    2019, Journal of Crystal Growth
    Citation Excerpt :

    The optical signal became stronger with an increase in the nitrogen flow rate. Generally, the amount of active nitrogen species, which are directly associated with the nitrogen flux, is proportional to the light intensity of the nitrogen plasma [30–34]. Therefore, the optical signal indicates that the nitrogen flux increased logarithmically with a nitrogen flow rate from 0.3 to 2.0 sccm.

  • Contactless electroreflectance studies of the Fermi level position at the air/GaN interface: Bistable nature of the Ga-polar surface

    2017, Applied Surface Science
    Citation Excerpt :

    The growth conditions were controlled using reflection high-energy electron diffraction (RHEED), laser reflectometry and optical pyrometry to maintain two-monolayer gallium surface coverage at the growth temperature of 720 °C. The nitrogen flux which determines the GaN growth rate under the Ga-rich growth conditions was controlled with the use of an optical sensor attached to the source as described elsewhere [33]. The concentration of magnesium in GaN:Mg layers was determined by secondary ion mass spectroscopy (SIMS).

View all citing articles on Scopus
View full text