Interplay between electrical transport properties of GeMn thin films and Ge substrates

We present evidence that electrical transport studies of epitaxial $p$-type GeMn thin films fabricated on high-resistivity Ge substrates are severely influenced by parallel conduction through the substrate, related to the large intrinsic conductivity of Ge due to its small band gap. Anomalous Hall measurements and large magnetoresistance effects are completely understood by taking a dominating substrate contribution as well as the measurement geometry into account. It is shown that substrate conduction persists also for well-conducting, degenerate, $p$-type thin films, giving rise to an effective two-layer conduction scheme. Using $n$-type Ge substrates, parallel conduction through the substrate can be reduced for the $p$-type epilayers, as a consequence of the emerging $\mathit{pn}$-interface junction. GeMn thin films fabricated on these substrates exhibit a negligible magnetoresistance effect. Our study underlines the importance of a thorough characterization and understanding of possible substrate contributions for electrical transport studies of GeMn thin films.

Figure 1

(a) Scaled schematic of samples with Hall bar mesa. The principle setup for measuring the longitudinal ($V_{\mathit{xx}}$) and Hall ($V_{\mathit{xy}}$) voltage as well as total current ($I$) is also indicated. (b) Geometry of samples where the van der Pauw method has been employed.

Figure 2

Sample resistance versus temperature for the GeMn sample (green solid) and the Ge substrate reference (blue dashed). Inset: Same data as function of inverse temperature. A straight line (grey dash-dotted) corresponding to an activation energy of 10.9 meV can be fitted to the extrinsic freeze-out.

Figure 3

(a) MR for various temperatures for the GeMn sample. (b) MR for the bare high resistivity Ge substrate. (c) Tangent of the Hall angle for (left) the GeMn sample and (right) the Ge substrate. The color code is the same for all panels.

Figure 4

Sample resistance as a function of temperature for the Ge:B sample (red solid) with a doping concentration of $5\times {10}^{19}$ cm${}^{-3}$. The reference measurement of the substrate (blue dashed) is also depicted. The inset shows a close-up for small temperatures, indicating the metallic conductance of the Ge:B sample, when parallel conduction through the substrate ceases.

Figure 5

(a) Sample resistance versus magnetic field of the Ge:B sample. (b) Hall resistance of the Ge:B sample. The dip of $R$ at zero field and the peak of $R_{\mathit{xy}}$ at low fields mark the onset of parallel conduction through the substrate at $9\mathrm{\hspace{0.5em}}\mathrm{K}$. The color code is the same in both panels.

Figure 6

Sheet (left) and Hall resistance (right) versus magnetic field for different temperatures of the Ge:B sample affected by parallel conduction. Shown is the measurement, the computation according to Eqs. (2), and the individual contributions of the substrate and metallic epilayer.

Figure 7

MR for various temperatures for the GeMn sample (open symbols) and the bare Ge substrate (filled symbols).

Figure 8

Sample resistance versus temperature for the GeMn sample (green dashed) and the Ge:B sample (red solid) grown on the n-type Ge substrate (blue dotted). The GeMn sample is affected by parallel conduction above $250\mathrm{\hspace{0.5em}}\mathrm{K}$ as the $\mathit{pn}$ barrier becomes inefficient. The Ge:B sample shows metallic conduction up to RT.

Figure 9

MR for various temperatures for the GeMn sample grown on the $n$-type Ge substrate.