Non-equilibrium formulation of helicity-dependent thermal field for ultrafast magnetization dynamics
Ezio Iacocca · 2026
Read Non-equilibrium formulation of helicity-dependent thermal field for ultrafast magnetization dynamics by Ezio Iacocca (2026) 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-2603-06929.txt.
Introduction
Iacocca proposes a non-equilibrium heat field for ultrafast magnet motion, with features tied to atomic spin flip odds. Femtosecond bursts from lasers or currents drive far-from-equilibrium motion, modeled usually by atomistic spin models, while micromagnetic models fail to reproduce ultrafast demagnetization on any grid. Treating each spin flip as one quantum of angular momentum gives matching temperatures in the thousands of Kelvin. Demagnetization is reproduced in numbers for several cell sizes, and the approach can be refined toward grid-independent multiscale work.
Catalog record
| CREATOR | Ezio Iacocca |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 22007 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.06929
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Non-equilibrium formulation of helicity-dependent thermal field for ultrafast magnetization dynamics Ezio Iacocca Abstract Far-from-equilibrium magnetization dynamics can be accessed when a magnetic material is subject to a femtosecond excitation, such as an optical laser or an electric current. Numerically, the demagnetization of magnetic materials is typically modeled by atomistic spin dynamics. Micromagnetic models generally fail to reproduce ultrafast demagnetization in a grid independent manner. Here, we propose a non-equilibrium thermal field whose features depend on atomic spin flip probabilities. Under the assumption that each spin flip is equivalent to a quantum of angular momentum, equivalent temperatures on the order of thousands of Kelvin are achieved. Demagnetization is quantitatively reproduced for several cell sizes. The presented approach can be further refined and extended towards a grid-independent and multiscale modeling of ultrafast magnetization dynamics. The effect of temperature on the magnetization of a single-domain particle can be expressed as a thermal field Brown 1963a with a magnitude that obeys the fluctuation-dissipation theorem d’Aquino et al. 2006 . This approach is strictly valid when the particle is at thermal equilibrium. The field magnitude is proportional to V^{-1/2} , where V is the particle’s volume, indicating that the magnetization of a large magnetic volume fluctuates less, on average. Within a micromagnetic framework Brown 1963b , where a magnetic particle is discretized in cells, the thermal field is applied on each cell. Each of these cells has dimensions on the order of the exchange length, implying that the fluctuations are large compared to the total particle’s volume but the average remains low due to the uncorrelated field. This approach has been extensively used in the research of spin-torque nano-oscillators, see e.g. Refs. Xiao et al. 2005 ; Thadani et al. 2008 ; Carpentieri and Torres 2010 ; d’Aquino et al. 2011 ; Dumas et al. 2013 ; Puliafito et al. 2019 . Even though the thermal field was derived in thermal equilibrium, it has been applied to model ultrafast magnetism Kirilyuk et al. 2010 in atomistic spin dynamics Eriksson et al. 2017 , returning excellent agreement with experiments Radu et al. 2011 ; Ostler et al. 2012 ; Iacocca et al. 2019 . In fact, it has been shown that purely thermal effects can drive ultrafast magnetization dynamics by optical lasers, e.g., Ostler et al. 2012 ; Mangin et al. 2014 , and even electrical pulses Yang et al. 2017 ; Wilson et al. 2017 ; Jhuria et al. 2020 . However, these ultrafast effects are generally not possible to reproduce quantitatively with micromagnetic simulations where the cells are composed by hundreds of atoms. This imposes a tremendous limitation in the modeling of ultrafast magnetization dynamics that require a full micromagnetic approach, e.g., domain patterns Jangid et al. 2023 and solitons Turenne et al. 2022 , as well as the full profile of optical lasers. It must be noted that descriptions based on a temperature-dependent dynamic equation, the Landau-Lifshitz-Bloch equation Evans et al. 2012 ; Atxitia et al. 2010 ; Jakobs and Atxitia 2022 , have been extended to a micromagnetic description with excellent agreement with ASD in the case of ferrimagnets Raposo et al. 2022 . In this letter, we derive a non-equilibrium contribution to the thermal field that is grid-independent and suitable for ultrafast magnetization dynamic simulations. The non-equilibrium field has two main contributions. First, the magnitude of the field includes the energy contribution from spin flips, proportional to the number of spins and their spin flip probability. Second, the distribution of the thermal field is determined by the probability distribution of spin flips. In contrast to the uniformly distributed thermal field expected from uncorrelated noise, the derived contribution has a mean and standard deviation, converging to a Gaussian for a large number of spins. A key advance of our work is the quantification of the magnon energy as a function of cell-size, which allows for a consistent and grid-independent determination of the non-equilibrium field magnitude. Our starting point is a statistical model for spin flips. This approach was introduced in Ref. Gorchon et al. 2016 to model multishot switching in ferromagnets. In this spirit, we consider a left circularly polarized laser pulse, with helicity \sigma . This laser favors a down spin orientation which is modeled in Ref. Gorchon et al. 2016 as a differential temperature between up ( T_{\uparrow} ) and down ( T_{\downarrow} ) spins, with T_{\uparrow}>T_{\downarrow} . Here, we will consider the probabilities of switching for each spin \begin{cases}P\quad\text{if spin-up}\\ B\quad\text{if spin-down}\end{cases}. (1) For our selected helicity, P>B meaning that it is more likely for an up spin to switch due to the optical laser. In this framework, a small probability implies that the laser fluence is low, while P=B implies linear polarization. Let us consider a micromagnetic cell with N spins. We represent the number of up spins as u and the number of down spins as d . Therefore, N=u+d . The normalized magnetization of the cell is given by m_{z}=(u-d)/N , where we set the z direction as our spin basis. In other words, we assume the material has perpendicular magnetic anisotropy. This definition is simply a quantization of the m_{z} component that limits to a continuum as N\rightarrow\infty . The pertinent question is, given an initial m_{z}^{i} , what is the probability for m_{z}^{f} to occur? To solve this problem, we consider a number of switching events, n for the up spins and m for the down spins. Therefore, the final magnetization is m_{z}^{f}=\frac{U-D}{N}=\frac{2U-N}{N}, (2) where U=u-n+m and D=d+n-m are the final up and down spins, respectively. The probability of achieving m_{z}^{f} from m_{z}^{i} is given by the sum of probabilities P_{ud,nm} for all the available combinations of n and m that achieve U and D given u and d , P_{ud,nm}=\frac{u!d!P^{n}(1-P)^{(u-n)}B^{m}(1-B)^{(d-m)}}{n!m!(u-n)!(d-m)!}. (3) For example, given N=4 , the initial magnetization m_{z}^{i}=0.5 is obtained by any permutation of the spin arrangement \uparrow\uparrow\uparrow\downarrow . A transition to m_{z}^{f}=m_{z}^{i} is possible if n=m=0 (no switching events) or if n=m=1 , resulting in the probability P_{31,00}+P_{31,11} . In the limit of large N , the central limit theorem ensures that the distribution of m_{z}^{f} will be Gaussian. Therefore, we can define its mean and standard deviation: \displaystyle\mu_{PB} \displaystyle= \displaystyle\sum_{U,D}\frac{U-D}{N}P_{ud,nm}, (4a) \displaystyle\sigma_{PB}^{2} \displaystyle= \displaystyle\left[\sum_{U,D}\left(\frac{U-D}{N}\right)^{2}P_{ud,nm}\right]-\mu_{PB}^{2}, (4b) where the sums are performed over all U,D given an initial u and d subject to switching events n and m . We now connect this description with that of a magnetic field. The main assumption in this letter is that the distribution quantified by Eqs. ( 4 ) determines the distribution of a non-equilibrium field. We define the non-equilibrium field as \mathbf{H}_{n-e}=H_{n_{e}}\mathbf{\hat{h}}_{n-e} . The unit vector is given by \displaystyle\mathbf{\hat{h}}_{n-e}\cdot\hat{\mathbf{x}} \displaystyle= \displaystyle\arcsin{\eta}\cos{\phi}, (5a) \displaystyle\mathbf{\hat{h}}_{n-e}\cdot\hat{\mathbf{y}} \displaystyle= \displaystyle\arcsin{\eta}\sin{\phi}, (5b) \displaystyle\mathbf{\hat{h}}_{n-e}\cdot\hat{\mathbf{z}} \displaystyle= \displaystyle\eta, (5c) where \eta is a random number obeying a Gaussian distribution with mean \mu_{PB} and standard deviation \sigma_{PB} while \phi is a uniformly distributed random number in the range [0,2\pi] . This is the main difference with the traditional thermal field which is uniformly distributed on a sphere. The magnitude of the non-equilibrium thermal field, H_{n-e} , is determined by the number of spin flips expected from the distribution. Given the initial magnetization m_{z}^{i} with u and d spins, the final magnetization is