---
title: Geometry-Induced Anisotropic Magnetoconductance (GAMC)
url: https://www.emergentmind.com/topics/geometry-induced-anisotropic-magnetoconductance-gamc
type: topic
---

# Geometry-Induced Anisotropic Magnetoconductance (GAMC)

Searching arXiv for recent and foundational papers on geometry-induced anisotropic magnetoconductance and closely related magnetotransport mechanisms.
Geometry-induced anisotropic magnetoconductance (GAMC) denotes a class of magnetotransport phenomena in which the dependence of conductance or resistance on magnetic-field direction, current direction, or magnetization orientation is controlled by geometry rather than by conventional spin-dependent scattering alone. In the literature surveyed here, the relevant “geometry” may be a corrugated quantum-well surface, an open curved nanostructure, a finite-thickness topological surface state, a symmetry-lowered tunnel barrier, a finite scattering region that supports Fabry–Pérot-like interference, a momentum-space mode-overlap constraint, or the quantum metric of Bloch states. The common feature is that geometry converts field orientation into a change in effective magnetic texture, available channels, interference conditions, or nonlinear response coefficients, thereby producing anisotropic magnetoconductance or magnetoresistance [1706.06919] [1408.1775].

## 1. Conceptual scope and relation to adjacent magnetotransport effects

The GAMC label is best understood as an umbrella for several transport responses that are reported under different names in different subfields. Closely related papers use the language of anisotropic magnetoresistance (AMR), ballistic anisotropic magnetoresistance (BAMR), tunneling anisotropic magnetoresistance (TAMR), planar Hall effect (PHE), magnetochiral anisotropy (MCA), and electric magnetochiral anisotropy (eMChA). In these works, the anisotropy is traced not to a single universal microscopic mechanism but to geometry-dependent orbital coupling, geometry-controlled mode matching, finite wavefunction extent, symmetry lowering at interfaces, or quantum geometry in momentum space [1611.05236] [1201.1439] [2103.05901] [2109.05188] [2502.09141].

| Mechanism in the literature | Representative systems | Transport signature |
|---|---|---|
| Surface corrugation or finite thickness | SiGe/Si/SiGe quantum wells; Bi\(_2\)Se\(_3\) thin films | OOP/IP anisotropic magnetoconductance or magnetoresistance |
| Curved or open real-space geometry | Rolled-up nanotubes; carbon nanoscrolls; bent nanoribbons | Field-direction-dependent channel count, positive magnetoconductance, BAMR/GAMC |
| Symmetry-lowered transmission geometry | Dirac–Weyl junctions; single-Mn tunnel diode; Rashba weak link | AMR-like conductance, TAMR-like transmission, AC-phase-sensitive anisotropy |
| Ballistic trajectory selection | Sinai billiards; rough graphene nanoribbons | Geometry-dependent transmission/reflection and magnetoconductance dips |
| Quantum-confined topological or quantum-geometric transport | QAH films; TI nanowires; tellurene | Giant AMR, MCA, eMChA, second-harmonic nonlinear response |

A recurrent distinction in the cited papers is between conventional AMR and geometry-induced anisotropy. Conventional AMR is associated with spin-dependent scattering and spin-orbit interaction in ferromagnets, whereas several GAMC examples are explicitly described as orbital, geometric, or mode-overlap effects, or as finite-thickness interference effects that would not exist in an idealized strictly two-dimensional or symmetry-restored limit [1611.05236] [1408.1775] [1706.06919].

## 2. Real-space geometry as an effective magnetic-field transformer

A canonical GAMC realization is the ultra-high mobility SiGe/Si/SiGe quantum well, where orientation-dependent in-plane-field magnetoresistance was traced to intrinsic ridges on the quantum-well surface [1706.06919]. Atomic-force microscopy revealed a cross-hatched ridge pattern aligned along [110] and [-110], with period \(\sim 10~\mu\text{m}\) and depth \(\sim 50~\text{nm}\). Because the surface is corrugated rather than flat, an in-plane magnetic field \(B_\parallel\) acquires a spatially varying local perpendicular component \(B_\perp\). The 2D electron gas therefore experiences a geometry-induced orbital field in addition to the usual spin-polarization response. The anisotropy depends on the orientation of \(I\) and \(B_\parallel\) relative to the ridges: for \(I \perp B_\parallel\), neighboring \(B_\perp\)-stripes act like narrow Hall bars and opposite Hall currents largely cancel in the longitudinal measurement; for \(I \parallel B_\parallel\), the stripes are perpendicular to the current and the sample behaves more like a Corbino-like geometry, enhancing the resistance. The effect is strongest when current is parallel or perpendicular to the ridges, and it becomes essentially the same for \(I \parallel B_\parallel\) and \(I \perp B_\parallel\) when the current is along [100], about \(45^\circ\) to the ridge directions. The paper proposed the mapping
\[
R(I\parallel B_\parallel)=R(I\perp B_\parallel)\left(1+\frac{\mu^2B_\parallel^2}{L}\int\limits_0^L\frac{B_\perp^2}{B_\parallel^2}dr\right),
\]
with \(\mu=1/n_\text{s}eR_\square(I\perp B_\parallel)\), and reported agreement between the recalculated and measured \(I\parallel B_\parallel\) curves [1706.06919].

Rolled-up semiconductor nanoarchitectures provide a second archetype. In compact rolled-up nanotubes, the open curved geometry and hard-wall boundaries generate a magnetic spectrum containing quasi-1D Landau-like states, snake states, and edge/skipping states, and rotating the field changes how many subbands cross the Fermi level [1408.1775]. The anisotropy is quantified as
\[
\mathrm{BAMR}(\theta)=\frac{G(\theta)-G(\pi/2)}{G(\pi/2)},
\]
with \(\theta\) measured relative to the edge axis. A major result is the approximate scaling \(\mathrm{BAMR}\propto 1/w\) with winding number \(w\), reflecting the growing weight of bulk-like magnetic subbands relative to edge states as the number of windings increases. The paper emphasized that this anisotropy is orbital and geometric, not spin-orbit-scattering-based, and stated that it can persist up to room temperature for realistic GaAs rolled-up structures [1408.1775].

A closely related but more recent bent-ribbon construction makes the geometric field conversion explicit. For a bent 2DEG nanoribbon, the left and right halves see effective fields
\[
B_{\rm eff,L} = -B_0 \cos(\theta+\phi), \qquad B_{\rm eff,R} = B_0 \cos(\theta-\phi),
\]
where \(B_0\) is the external field, \(\theta\) its orientation, and \(\phi\) the bending angle [2508.15385]. For \(\phi=\pi/2\), one field orientation produces a uniform effective field, while the orthogonal orientation produces an effective magnetic dipole with opposite signs on the two halves. This selectively creates bulk cyclotron/Landau states, interface snake states, and edge skipping states, and the ballistic conductance \(G=N\,2e^2/h\) therefore becomes strongly orientation dependent. The reported GAMC ratio
\[
{\rm GAMC}(\theta)=\frac{G(\theta)-G(\pi/2)}{G(\pi/2)}
\]
reaches about \(100\%\) in a single-bend geometry, about \(200\%\) in a twice-bent geometry, and about \(300\%\) for a once-bent ribbon with \(\phi'=2\arctan(2)\). The effect is reported to remain large at \(300\) K, with approximately \(88\%\), \(131\%\), and \(190\%\) for three example geometries, and to be observable even at weak fields, particularly in graphene [2508.15385].

Carbon nanoscrolls realize an allied curved-geometry mechanism, but with positive rather than suppressive magnetoconductance [2408.03518]. In a two-winding nanoscroll subject to an axial field, the field changes tangential quantization and creates additional zero-energy doublet modes at a half-integer flux quantum in the actual open-ended geometry. Because the zero-temperature conductance is \(G(E,0)=\frac{2e^2}{h}N_s\), the extra zero modes increase the mode count near charge neutrality, yielding a positive magnetoconductance coefficient
\[
\text{PMC}=\frac{G(\phi)-G(0)}{G(0)}.
\]
For a two-winding nanoscroll at about \(10\) T, the reported low-density conductance is tripled, corresponding to a PMC of about \(200\%\), and the effect persists under disorder or mild inter-turn misalignment [2408.03518].

## 3. Geometry-controlled mode matching, symmetry lowering, and transmission anisotropy

In Dirac–Weyl magnetic junctions, GAMC appears as a momentum-space overlap effect rather than a real-space orbital-field conversion [1611.05236]. The device consists of doped Dirac semimetal leads and a ferromagnetic Weyl semimetal barrier of thickness \(L\), with magnetization confined to the \(x\)-\(z\) plane. Exchange shifts the Weyl Fermi surface in momentum space by \(k_0\sin\theta\) in the projected \(k_y\)-\(k_z\) plane, so the overlap between lead modes and barrier modes depends on the angle \(\theta\) between magnetization and current. The total conductance is computed from the Landauer formula,
\[
G(\theta)=2\frac{e^2}{h}\frac{S}{(2\pi)^2}\int_D T(k_y,k_z)\,dk_y\,dk_z,
\]
and for small shifts the paper derived the AMR-like form
\[
G(\theta)\simeq G(0)-\Delta G\sin^2\theta.
\]
The conductance is independent of \(k_0\) at \(\theta=m\pi\), minimal at \(\theta=(m+1/2)\pi\), and periodic as \(G(\theta+\pi)=G(\theta)\). The paper stressed that this mechanism is completely different from conventional AMR, because it arises from geometric mode matching in projected \(k\)-space, not from spin-dependent scattering [1611.05236].

A localized, wavefunction-shape version of the same principle appears in the GaMnAs single-impurity tunnel diode [1201.1439]. There, the tunnel current is controlled by the spatial shape of a spin-orbit-coupled Mn acceptor wavefunction near an AlGaAs barrier. The transmission proxy is
\[
T \propto \int_{z>0} \left| \psi_h(\mathbf r) \right|^2 \, d^3r.
\]
Rotating the Mn \(5/2\) spin changes the orbital character and spatial extent of the acceptor through \(p\)-\(d\) exchange and strong spin-orbit coupling, while the barrier and reduced \(C_{2v}\) symmetry convert that change into anisotropic leakage. For a Mn atom \(1\) monolayer from the barrier, the bound-state eigenenergy changes by almost \(10\) meV under out-of-plane rotation, the transmission is more than \(20\%\) lower for spin along [001], and the in-plane tunnel anisotropy is about \(14\%\), with [110] and \([\bar110]\) inequivalent [1201.1439].

Finite scattering regions with spin-orbit coupling provide another transmission-based GAMC mechanism. In a 2DEG with SOC and an in-plane magnetic field applied only in a finite central region of length \(L\), the conductance becomes angle dependent because the finite geometry creates a scattering problem with Fabry–Pérot-like interference between multiple reflections inside the SOC+field region [2103.05901]. The paper found that the longitudinal conductance \(G_{xx}\) is \(\pi\)-periodic and the transverse conductance \(G_{yx}\) is \(2\pi\)-periodic in the field angle \(\phi\). The transverse oscillation amplitude is enhanced in patches of the \((\alpha,b)\)-plane, and \(G_{yx}\) can show one-fold, three-fold, five-fold, or seven-fold oscillations for large \(\alpha\) and \(b\). A central conclusion was that SOC in a material is sufficient to observe PHE and anisotropic magnetoconductance without anisotropic magnetic ordering or nontrivial band topology [2103.05901].

The weak-link proposal based on the Aharonov–Casher phase isolates the role of geometric phase in an especially compact form [2201.10845]. For a one-dimensional Rashba weak link of length \(d\), the AC phase is \(\phi_{\rm AC}=k_{\rm so}d\). At zero field it is hidden in a diagonal propagator and does not affect conductance, but a Zeeman field component perpendicular to the SOI pseudo-field mixes the spin channels and turns the phase into a measurable conductance anisotropy. In the ballistic regime the paper derived
\[
\frac{G(B,\theta)-G(0,0)}{G(B,0)-G(0,0)} = 1-\frac{1}{2}\sin^2\theta\,\sin^2(\phi_{\rm AC}),
\]
making the anisotropy maximal at \(\theta=\pi/2\) and \(\phi_{\rm AC}=\pi/2\), and zero when \(\theta=0,\pi\) or \(\phi_{\rm AC}=0,\pi\). This establishes GAMC as a route to calibrating Rashba coupling through an angular magnetoconductance measurement rather than a ring-interference experiment [2201.10845].

## 4. Finite thickness, weak antilocalization, and topological transport channels

In Bi\(_2\)Se\(_3\) thin films, GAMC is defined by the contrast between out-of-plane and in-plane magnetoconductance and is attributed to the finite penetration depth of topological surface states rather than to a generic crystallographic anisotropy [2106.07698]. For an ideal strictly 2D state of zero thickness, an in-plane field would not thread flux through interference loops, but real topological surface states occupy a slab-like region of thickness \(\lambda\). The in-plane weak-antilocalization correction is modeled as
\[
\Delta \sigma_{xx} = \frac{e^2}{2\pi h}\tilde{A}_1 \ln \left(1+\frac{B^2}{B_{||}^2} \right),
\qquad
B_{||} = \sqrt{\frac{2\hbar B_{\phi,1}}{e\lambda^2}},
\]
and the paper extracted \(\lambda \sim 7 \text{ nm at } 1.75\text{ K}\). The total anisotropy was analyzed by combining out-of-plane HLN-type WAL/WL terms with the in-plane finite-thickness term, while also subtracting electron-electron interaction corrections. The anisotropy \(\sigma_{xx}^{\parallel}-\sigma_{xx}^{\perp}\) shows a strong peak near \(\theta=\pm 90^\circ\), and the fitted surface-channel prefactors were \(|\tilde{A}_1|\sim 1.6\) for c-Al\(_2\)O\(_3\)/Bi\(_2\)Se\(_3\) and \(|\tilde{A}_1|\sim 1.8\) for SiO\(_2\)/Bi\(_2\)Se\(_3\), consistent with stronger top-bottom decoupling in the more disordered SiO\(_2\)-grown films [2106.07698].

In InSb quantum wells, in-plane magnetoconductance mapping was used to extract anisotropic disorder and \(g\)-factor parameters from the suppression of weak antilocalization [1902.07570]. Hall bars along \([\bar{1}10]\) and \([\bar{1}00]\) were measured in a dilution refrigerator with a 4-1-1 vector magnet, and the in-plane response was found to track crystal direction rather than Hall-bar orientation. The in-plane suppression is predominantly Zeeman-driven, with disorder-induced orbital coupling as a secondary contribution. The best fit for one orientation yielded \(g m^*/m_e = 0.91\) and \(\gamma_D = 0.0029\ \text{T}^{-1}\), implying \(|g| \approx 45\) for \(m^*/m_e \approx 0.02\). Averaged over angles, \(\langle g m^*/m_e \rangle \approx 1.009\), corresponding to \(|g| \approx 48\), while \(\gamma_d\) varies about \(\pm 20\%\). These measurements show GAMC as a diagnostic of anisotropic electronic structure in a 2DEG with strong Zeeman response [1902.07570].

A topological-channel version of giant anisotropy appears in the quantum anomalous Hall insulator based on an 8-quintuple-layer Cr-doped (Bi,Sb)\(_2\)Te\(_3\) film [1503.03556]. When magnetization is out of plane, conduction proceeds through a single massless one-dimensional chiral edge state and \(\rho_{xx}\) is small; tilting the magnetization toward the plane reduces the magnetic gap, increases the edge-state penetration depth
\[
\lambda = \frac{\hbar v_F}{\Delta},
\]
and promotes intermixing with dissipative channels. The resulting AMR is modeled with a four-terminal Landauer–Büttiker form in which the edge transmission coefficient is
\[
\eta = 1 - e^{-\Delta/\Delta_0}, \qquad \Delta = \Delta_m \cos\theta.
\]
The anisotropic magnetoresistance ratio,
\[
\text{AMR} = \frac{\rho_{xx,\max}-\rho_{xx,\min}}{\rho_{xx,\min}},
\]
reaches approximately \(140\%\) for device A, \(75.6\%\) for device B, and \(100.2\%\) for device C, whereas the purely in-plane AMR is only about \(10.7\%\) and follows a conventional \(\cos^2\phi\) form. The giant out-of-plane anisotropy is therefore interpreted as a QAH edge-state to diffusive crossover rather than as ordinary ferromagnetic AMR [1503.03556].

## 5. Ballistic trajectories, edge roughness, and geometry-selected orbit families

Sinai billiards make the trajectory-selection aspect of GAMC explicit. In normal and Andreev two-dimensional open cavities with a central antidot, particles move along magnetic-field-induced circular arcs, and the cavity geometry determines whether these arcs hit the antidot, the walls, or the leads [2012.02440]. For the normal case the conductance is
\[
G_{\rm N}=\frac{2e^2}{h}\,T_{\rm e},
\]
whereas in the Andreev case it is
\[
G_{\rm S}=\frac{2e^2}{h}\,(T_{\rm e}+R_{\rm h}).
\]
The paper compared four geometries—square centered antidot, square displaced antidot, rectangular centered antidot, and circular centered antidot—and emphasized that the critical field \(B_c\), above which the outgoing hole current vanishes in the Andreev case, depends only on geometry. The rectangular cavity gives the smallest \(G_N\) and \(G_S\), because the reduced vertical size increases electron reflection and hole transmission in ways that suppress conductance, while the circular cavity is exceptional in that \(G_N \approx G_S\) for \(B < B_c\). The general conclusion is that geometry selects the dominant orbit families—wall-bouncing trajectories, rosette-like orbits, or skipping orbits—and thereby shapes the magnetoconductance curve [2012.02440].

Graphene nanoribbons with rough edges exhibit a different trajectory-selected signature: an edge roughness induced dip (ERID) in magnetoconductance when the cyclotron radius approaches the ribbon width [1112.1806]. The conductance is computed from Landauer–Büttiker transport, \(G=\frac{2e^2}{h}T(E_F)\), with
\[
r_c=\frac{\hbar k_F}{eB}.
\]
For armchair ribbons of width \(W=50\,\mathrm{nm}\) and length \(L=220\,\mathrm{nm}\), the dip occurs at \(B_{\min}\approx 5\,\mathrm{T}\), corresponding to \(W/r_c\approx 0.8\), and \(B_{\min}\) is nearly inversely proportional to the ribbon width. The anisotropy arises because armchair and zigzag edges couple very differently to roughness: armchair ribbons show a clear ERID, whereas zigzag ribbons require roughness amplitudes of about \(\delta \gtrsim 6\) before the dip becomes visible. The microscopic explanation is that zigzag ribbons have strongly reduced current density within about \(\approx 5\,\mathrm{nm}\) from each edge, so edge roughness initially perturbs a region that carries little current. The dip survives only when bulk disorder is sufficiently weak [1112.1806].

These ballistic and semiclassical examples sharpen a common point: geometry-induced anisotropy need not mean a simple angular modulation of a uniform medium. It can also mean that geometry changes which orbit families are even available, and therefore changes the transmission/reflection partition itself. This suggests that GAMC is often most transparent in systems where channel counting or classical trajectory families can be identified directly [2012.02440] [1112.1806].

## 6. Nonlinear and quantum-geometric extensions

A nonlinear descendant of GAMC appears as giant magnetochiral anisotropy in topological-insulator nanowires [2109.05188]. In thin bulk-insulating \((\mathrm{Bi}_{1-x}\mathrm{Sb}_x)_2\mathrm{Te}_3\) nanowires, a top gate artificially breaks inversion symmetry and splits quantum-confined topological surface-state subbands. A magnetic field then shifts these spin-polarized, strongly non-parabolic subbands, producing non-reciprocal transport described by
\[
R = R_0(1+\gamma B I).
\]
The paper reported \(|\gamma|\sim 6\times 10^4\ \mathrm{A^{-1}T^{-1}}\) in one device and values approaching \(\sim 1\times 10^5\ \mathrm{A^{-1}T^{-1}}\) across devices, and described this as the largest ever reported MCA rectification coefficient in a normal conductor. The angular dependence follows approximately \(\gamma \approx \gamma_0 \cos\alpha\) for rotation in the \(zx\)-plane and is essentially zero for rotation in the \(xy\)-plane, consistent with the vector form \(\gamma \propto P\cdot(\hat{B}\times\hat{I})\) [2109.05188].

Tellurene extends the geometry concept from real-space structure to Bloch-state geometry itself. In noncentrosymmetric polar media, the resistance is written as
\[
R(I,P,B)=R_0\left[1+\beta B^2+\gamma^{\pm}{\bf I}\cdot({\bf P}\times{\bf B})\right],
\]
where the eMChA coefficient \(\gamma^{\pm}\) is determined by the quantum metric dipole and the polarization \({\bf P}\) [2502.09141]. The underlying semiclassical picture adds a geodesic correction built from Christoffel symbols of the quantum metric to the carrier velocity, and the normalized nonlinear coefficient obeys the predicted universal scaling \(\gamma^{\pm}(V)\sim V^{-5/2}\), verified by phase-sensitive second harmonic transport measurements on 2D tellurium films [2502.09141].

The valence-band study of \(p\)-type tellurene generalized this result by showing that the nonlinear response is not purely chiral [2602.13721]. The paper distinguishes the chiral form
\[
\frac{\Delta R(B)}{R(0)} = 2\,\chi\,\mathbf{I}\cdot\mathbf{B}
\]
from the polar form
\[
\frac{\Delta R(B)}{R(0)} = 2\,\gamma\,\mathbf{I}\cdot(\mathbf{P}\times\mathbf{B}),
\]
and argues that the valence band supports both, with \(\chi\neq 0\) and \(\gamma\neq 0\). The key mechanism is multiband: an isolated low-energy valence block has finite quantum metric but vanishing metric dipole, whereas Löwdin downfolding from remote Weyl-node-containing bands generates nonzero Christoffel-symbol components and thus a finite polar coefficient \(\gamma\). The measured second-harmonic response \(V^{2\omega}_\parallel(\mu)\) and the shifted angular dependence \(V^{2\omega}_\parallel(\phi)\) are reproduced quantitatively by this mixed chiral-plus-polar picture [2602.13721].

Taken together, these nonlinear studies suggest a broader interpretation of GAMC: the “geometry” responsible for anisotropic magnetoconductance need not be confined to device shape or boundary conditions. It can also be the geometry of wave functions in Hilbert space, encoded by subband curvature asymmetry, quantum confinement, or the quantum metric dipole. Across linear and nonlinear regimes, the recurring principle is the same: geometry converts field orientation into a transport anisotropy by controlling which states exist, how they interfere, and how effectively they couple to current [2109.05188] [2502.09141] [2602.13721].

Source: https://www.emergentmind.com/topics/geometry-induced-anisotropic-magnetoconductance-gamc