Reduction of threading dislocation density in Al0.12Ga0.88N epilayers by a multistep technique

T. Lang*, M. A. Odnoblyudov, V. E. Bougrov, S. Suihkonen, O. Svensk, P. T. Törmä, M. Sopanen, H. Lipsanen

*Corresponding author for this work

    Research output: Contribution to journalArticleScientificpeer-review

    4 Citations (Scopus)


    Although suitable for the reduction of the threading dislocation density in GaN layers the widely used two-step MOCVD method does not work as efficiently for AlGaN. This is due to slow surface diffusion of the Al species. In the present paper, the previously reported in situ multistep method for MOCVD growth of high-quality GaN films is adopted for the growth of Al0.12Ga0.88N films on c-plane sapphire. The developed method for AlGaN growth is virtually GaN free in the sense that no continuous film of GaN is needed near the substrate interface. Crack-free layers of Al0.12Ga0.88N with a thickness of about 2 μm are grown by the method. A sparse distribution of 3D GaN nucleation islands and stimulation of threading dislocation reactions enable a reduction of the threading dislocation density down to 5×108 cm-2 in the Al0.12Ga0.88N films. The threading dislocation density is evaluated by etch-pit density measurements. High-resolution X-ray diffraction and transmission electron microscopy are used to study the crystallinity of the Al0.12Ga0.88N layers. Reflectometry is utilized to analyze film growth in situ. The surface morphology of GaN nucleation layers and Al0.12Ga0.88N epilayers is characterized by atomic force microscopy.

    Original languageEnglish
    Pages (from-to)276-280
    Number of pages5
    JournalJournal of Crystal Growth
    Publication statusPublished - Jan 2007
    MoE publication typeA1 Journal article-refereed


    • A1. Nucleation
    • A1. Threading dislocations
    • A3. MOCVD
    • B1. AlGaN

    Fingerprint Dive into the research topics of 'Reduction of threading dislocation density in Al0.12Ga0.88N epilayers by a multistep technique'. Together they form a unique fingerprint.

    Cite this