Design of high-mobility p -type GaN via the piezomobility tensor

Jie-Cheng Chen, Joshua Leveillee, Chris G. Van de Walle, Feliciano Giustino · 2025

Read Design of high-mobility p -type GaN via the piezomobility tensor by Jie-Cheng Chen, Joshua Leveillee, Chris G. Van de Walle, Feliciano Giustino (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-06723.txt.

Introduction

Chen, Leveillee, Van de Walle, and Giustino hunt for strain recipes that make holes move faster in gallium nitride, a semiconductor whose weak hole flow blocks whole device families. Prior strain work was piecemeal. Their 2025 study introduces a piezomobility tensor, a compact map from every strain pattern to the resulting hole speed, computed from first-principles physics with refined energy corrections.

Three strain patterns stand out: two simple squeezes and one shear. A two percent squeeze along one axis promises room-temperature hole mobility up to 164 units, with shear close behind at 148. Both far exceed the material's usual sluggish pace. The authors offer the tensor method as a general strain design tool for semiconductors.

Catalog record

CREATORJie-Cheng Chen, Joshua Leveillee, Chris G. Van de Walle, Feliciano Giustino
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-06723

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Design of high-mobility p -type GaN via the piezomobility tensor Jie-Cheng Chen, Joshua Leveillee, Chris G. Van de Walle, Feliciano Giustino Abstract Gallium nitride (GaN) is a wide-bandgap semiconductor of significant interest for applications in solid-state lighting, power electronics, and radio-frequency amplifiers. An important limitation of this semiconductor is its low intrinsic hole mobility, which hinders the development of p -channel devices and the large-scale integration of GaN CMOS in next-generation electronics. Prior research has explored the use of strain to improve the hole mobility of GaN, but a systematic analysis of all possible strain conditions and their impact on the mobility is lacking. In this study, we introduce a piezomobility tensor notation to characterize the relationship between applied strain and hole mobility in GaN. To map the strain-dependence of the hole mobility, we solve the ab initio Boltzmann transport equation, accounting for electron-phonon scattering and GW quasiparticle energy corrections. We show that there exist three optimal strain configurations, two uniaxial strains and one shear strain, that can lead to significant mobility enhancement. In particular, we predict room-temperature hole mobility of up to 164 cm 2 /Vs for 2% uniaxial compression and 148 cm 2 /Vs for 2% shear strain. Our methodology provides a general framework for investigating strain effects on the transport properties of semiconductors from first principles. I Introduction Wurtzite gallium nitride (GaN) has played a fundamental role in the development of solid-state lighting, and is currently used in power electronics and wireless communications. 1 ; 2 ; 3 ; 4 ; 5 ; 6 ; 7 ; 8 ; 9 Owing to its wide bandgap of 3.4 eV, GaN offers lower leakage current and higher critical field as compared to conventional silicon, thus enabling reduced energy dissipation, faster switching, and higher operating temperatures. 3 ; 2 GaN exhibits high electron mobility, in excess of 1000 cm 2 /Vs at room temperature for nominally-undoped single crystal samples, 10 ; 11 ; 12 ; 13 which is comparable to the electron mobility in silicon. 2 However, the room-temperature mobility of holes in GaN is only 30 cm 2 /Vs. 2 ; 14 ; 15 ; 16 ; 17 ; 18 ; 19 ; 20 ; 21 ; 22 This imbalance hinders the development of complementary metal-oxide-semiconductors (CMOS) field-effect transistors and their large-scale integration in integrated circuits. One potential avenue to improve the hole mobility of GaN is to reduce the hole effective masses via strain engineering. 23 ; 24 ; 25 ; 26 ; 27 ; 28 The rationale for this approach is twofold: (i) the dominant hole-phonon scattering mechanism in GaN is acoustic deformation potential (ADP) scattering, 27 which is sensitive to the electronic density of states and hence the band effective masses; (ii) the topmost valence bands are heavy holes with an effective mass m^{*}_{\rm hh}=1.94\,m_{e} along the c -axis direction 29 ( m_{e} is the free electron mass). Therefore, reducing the hole effective mass would simultaneously increase the numerator and decrease the denominator in Drude’s formula \mu=e\tau/m^{*} , where \mu , e , \tau are the mobility, electron charge, and relaxation time, respectively. Several theoretical studies investigated potential strategies to improve the hole mobility of GaN via strain engineering. In Refs. 27 ; 29 , using the ab initio Boltzmann transport equation ( ai BTE), the authors found that biaxial tensile strain in the basal plane or uniaxial compressive strain along the c -axis leads to the split-off ( sh ) holes being lifted above the heavy-holes ( hh ), with a corresponding enhancement of the mobility by 230% for 2% strain. In Ref. 30 , the authors employed a semiempirical {\bf k}\!\cdot\!{\bf p} Hamiltonian to show that a uniaxial compression is the best approach for optimizing in-plane hole mobility in the GaN/AlN 2D hole gas. In Ref. 28 , using the ai BTE, the authors predicted that lattice-matching GaN to ZnGeN 2 or MgSiN 2 substrates would induce an increase in the hole mobility by 50% and 260%, respectively. Additional efforts in this direction have been reported in Refs. 31 ; 32 ; 33 ; 34 ; 35 ; 36 ; 37 . All these studies support the notion that strain engineering offers a promising avenue to increase the mobility of p -type GaN. However, a broad and systematic investigation of possible strain configurations has not been performed yet. To fill this gap, we here introduce the piezomobility tensor as a conceptual tool for examining the effect of strain on carrier transport in a systematic manner. We evaluate this tensor by performing first-principles calculations of the phonon-limited carrier mobility in GaN via the ai BTE, and we include spin-orbit coupling (SOC) and GW quasiparticle corrections for high-accuracy results. We find three optimal strain configurations, two corresponding to uniaxial strain and one to shear strain, that induce a significant enhancement of the theoretical hole mobility. Specifically, we obtain a room-temperature hole mobility of up to 164 cm 2 /Vs for a configuration corresponding to single-component 2% compressive \epsilon_{3} strain (uniaxial, c -axis; in Voigt notation, see Fig. 1 for crystal orientation). In specifying this uniaxial compression, we keep all other strain components equal to zero. Experimentally, such single-component strain is not achievable, but as discussed in Sec. IV.5 we find that it constitutes a very good approximation to what can be obtained in realistic conditions. The second optimal configuration corresponds to \epsilon_{1} (uniaxial, x -axis), resulting in 159 cm 2 /Vs; and the third to \epsilon_{4} ( yz shear) strain, resulting in 148 cm 2 /Vs. The first of these strain configurations was already discussed in the literature, 27 ; 29 while the other two configurations are new. To the best of our knowledge, the use of shear strain has not been discussed before. We also rationalize our findings in terms of the band effective mass, and find that the conductivity effective mass constitutes a reliable approximate descriptor of the change of hole mobility under strain, both single- and multi-component. Figure 1: Top view and side view of the crystal structure of wurtzite GaN, with Ga in green and N in gray. The primitive lattice vectors and the convention for the Cartesian reference frame employed throughout this manuscript are shown. The manuscript is organized as follows. In Sec. II we introduce the notion of piezomobility tensor and state the properties of this tensor for wurtzite GaN. Section III summarizes the computational setup employed in this work, as well as the key components of the ai BTE methodology. In Sec. IV we describe our results. In particular, Sec. IV.1 provides a concise overview of our findings for single-component strain, and Sec. IV.2 analyzes the best case scenarios identified here in greater detail. In Sec. IV.3 we investigate the importance of effects beyond the linear order in the strain, and in Sec. IV.4 we show how the hole mobility enhancement in GaN primarily arises from the reduction of the hole effective mass. Section IV.5 considers a few realistic strain scenarios where multiple strain components are simultaneously nonvanishing. In Sec. V , we discuss our results in the light of recent experimental work and we draw our conclusions. II Piezomobility tensor To map out the relation between strain and

Read the complete text