Bosonized theory of de Haas-van Alphen quantum oscillation in Fermi liquids
Yuxuan Wang · 2025
Read Bosonized theory of de Haas-van Alphen quantum oscillation in Fermi liquids by Yuxuan Wang (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-2506-20735.txt.
Introduction
A magnetic field makes the magnetization of a metal ripple as the field grows. This de Haas-van Alphen effect is well understood in three dimensions. In flat, two-dimensional Fermi liquids the theory has a stubborn gap. Part of the signal comes from the wobbling self-energy of the electrons, which has no closed-form answer. Yuxuan Wang sidesteps the problem by bosonizing the Fermi surface. The Landau parameters of the liquid then enter the theory directly.
Cyclotron resonance and specific heat come out matching known results. A mode expansion reduces the effect to a simple quantum mechanics problem. The low and high temperature behavior departs from the classic Lifshitz-Kosevich formula. In three dimensions Wang shows those departures shrink to negligible size.
Catalog record
| CREATOR | Yuxuan Wang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-20735
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Bosonized theory of de Haas-van Alphen quantum oscillation in Fermi liquids Yuxuan Wang Abstract The de Haas-van Alphen effect (dHvA) of a 2d Fermi liquid remains poorly understood, due to the ∼ 𝒪 ( 1 ) \sim\mathcal{O}(1) contribution to the oscillations of grand potential from the oscillatory part of the fermionic self energy, which has no known closed-form solution. In this work, we solve this problem via coadjoint-orbit bosonization of the Fermi surface. Compared with the fermionic formalism, the issue of the oscillatory self energy is circumvented. As an effective field theory, Landau parameters F n F_{n} directly enter the theory. We use the bosonized theory to derive the energies of cyclotron resonance and specific heat, which are consistent with Fermi liquid theory. Via a mode expansion, we show that the problem of dHvA is reduced to 0+1D quantum mechanics. We obtain analytic expressions for the behavior of dHvA at low and high temperatures, which deviate from the well-known Lifshitz-Kosevich formula. We contrast this behavior with that of 3d Fermi liquids, for which we show such deviations are parametrically small. We discuss the effects of disorder on dHvA within the bosonized theory. I Introduction In gapless fermionic systems, quantum oscillations in a weak external magnetic field, i.e., de Haas-van Alphen (dHvA) and Shubnikov-de Haas effects, encode important information about the low-energy properties of the system, and have been widely used in experiments to determine the size, shape, and topology of the Fermi surface (FS) [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 ] . For noninteracting particles, dHvA can be obtained by directly computing the grand potential \Omega by summing the Landau levels. This result is known as the Lifshitz-Kosevich (LK) formula [ 20 ] . In 2d, its oscillatory part is given by \Omega_{\rm osc}=N_{\Phi}T\sum_{k=1}^{\infty}\frac{(-)^{k}\cos(k\mathcal A_{\rm FS}/B + k\gamma)}{k\sinh(2\pi^2 k T/\omega_c)}, (1) where N_{\Phi}=BL^{2}/2\pi is the Landau level degeneracy, T the temperature, \mathcal{A}_{\rm FS} the area enclosed by the FS, B the magnetic field, \omega_{c} the cyclotron frequency (or the Landau level spacing), and \gamma the Berry phase [ 21 , 2 , 15 ] of the Bloch wave around the FS 1 1 1 For simplicity, we will not consider Berry phase effects in this work. Such effects have been treated within bosonization in Ref. [ 22 ] in the noninteracting limit. . In 3d, a similar LK formula can be obtained by integrating \Omega_{\rm osc} over the momentum along the B field (see Eq. ( 89 )). Beyond the non-interacting limit, the treatment of dHvA within a many-body field-theoretic framework is much less straightforward. As the magnetization contains terms such as \sim\cos(a/B) that have an essential singularity at B=0 , dHvA cannot be captured within the linear (or nonlinear) response theory, even if the system is in a weak magnetic field and is weakly interacting. Instead, one has to directly compute the grand potential \Omega with interactions in the Landau level basis [ 23 , 24 ] . This approach faces several challenges. First, even in the absence of interactions, the energies of the Landau levels implicitly depend both on the dispersion and the geometry of the Bloch function, and have to be computed outside the field theory using, e.g., semiclassical methods. Second, the fermionic self energy in the presence of a magnetic field, which \Omega depends on, differs from their results at B=0 and is oscillatory in 1/B , i.e., \Sigma(B)=\Sigma_{B=0}+\Sigma_{\rm osc}(B) . Unlike \Sigma_{B=0} , \Sigma_{\rm osc}(B) is generally difficult to evaluate. Fortunately, in 3d, due to destructive interference along the field direction, \Sigma_{\rm osc} is strongly suppressed as \Sigma_{\rm osc}/\Sigma_{B=0}=\mathcal{O}{\left[\left({\omega_{c}}/{E_{F}}\right)^{3/2}\right]} [ 23 , 24 , 25 , 26 ] , where E_{F} is the Fermi energy, and to compute \Omega_{\rm osc} to leading order one can simply use the zero-field self energy. Taking advantage of this suppression, one can show that the dHvA for a 3d Fermi liquid retains the form of the LK formula, except that the cyclotron frequency \omega_{c}=B/m is defined with m as the effective mass. However, the situation in 2d is much more complicated. First of all, for T\ll\omega_{c} , fixing chemical potential \mu and particle number N lead to different results, even for a non-interacting Fermi gas. Indeed, in the former case, the chemical potential is almost always (with a unity probability) between Landau levels, while in the latter case, a Landau level is almost always partially filled. For partially filled Landau levels, interactions are expected to play a much more prominent role, and the ground state is not adiabatically connected to the zero-field Fermi liquid, and can be e.g., charge-ordered [ 27 , 28 , 29 ] or topologically ordered [ 30 ] . Within this work, we will take \mu as fixed, i.e., the grand canonical emsemble. In experiments, this can be achieved by connecting the system to a large reservoir, e.g., to source and drain contacts used in transport measurements. Conspicuously, the dHvA in 2d Fermi liquids remains poorly understood, even with \mu fixed. This is not due to any fundamental reason – after all, dHvA measures low-energy properties around the Fermi surface, which should be completely captured by the Fermi liquid theory. Instead, the main issue is that for a 2d Fermi liquid in a B field, the oscillatory component in the self energy is less suppressed than in 3d, \Sigma_{\rm osc}/\Sigma_{B=0}=\mathcal{O}(\omega_{c}/E_{F}) . Such an oscillation contributes to the grand potential \Omega at the same order ( \sim\mathcal{O}\left[(\omega_{c}/E_{F})^{2}\right] ) as the oscillation in the non-interacting part of \Omega , and cannot be neglected even at weak field [ 31 , 32 , 25 ] . In general, different harmonics of \Sigma_{\rm osc}(B) in \exp(ik\mathcal{A}_{\rm FS}/B) are coupled to each other, and a closed-form solution is unknown. Although at high enough temperatures with T\gtrsim\omega_{c} , these oscillations are exponentially suppressed, they remain important in the dHvA amplitudes A_{k>1} , which are exponentially suppressed even more strongly. As a result, only the amplitude of the leading harmonic is known [ 25 ] at T\gtrsim\omega_{c} for a 2d Fermi liquid and follows the LK formula. Similar issues occur in the dHvA for a 3d non-Fermi liquids, and a special solvable case was recently discussed in Ref. [ 26 ] . We show in this paper that the difficulty of the oscillatory self energy can be avoided by using the bosonization approach. Higher-dimensional bosonization of FS’s has a long history [ 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 22 , 43 ] , but until very recently the issue of magnetic response remained unaddressed by bosonization. Indeed, in earlier theories of bosonization, one first divides the FS into patches, bosonizes the patches, and then assembles them together. In such a theory, the area \mathcal{A}_{\rm FS} enclosed by the FS, which dHvA crucially relies on, does not enter. Recently, bosonization has been reformulated [ 38 ] in terms of the coadjoint orbit of canonical transformations, which describes the deformation of the Fermi sea, resulting in an