betaFeedback

WorksPhysics and Astronomy

Time-Dependent Self Consistent Harmonic Approximation: Anharmonic nuclear quantum dynamics and time correlation functions

Abstract

Most material properties of great physical interest are directly related to nuclear dynamics, e.g. the ionic thermal conductivity, Raman/IR vibrational spectra, inelastic X-ray, and Neutron scattering. A theory able to compute from first principles these properties, accounting for the anharmonicity and quantum fluctuations in the nuclear energy landscape that can be implemented in systems with hundreds of atoms is missing. Here, we derive an approximate theory for the quantum time evolution of lattice vibrations at finite temperature. This theory introduces the time dynamics in the Self-Consistent Harmonic Approximation (SCHA) and shares with the static case the same computational cost. It is nonempirical, as pure states evolve according to the Dirac least action principle and the dynamics of the thermal ensemble conserves both energy and entropy. The static SCHA is recovered as a stationary solution of the dynamical equations. We apply perturbation theory around the static SCHA solution and derive an algorithm to compute efficiently quantum dynamical response functions. Thanks to this new algorithm, we have access to the response function of any general external time-dependent perturbation, enabling the simulation of phonon spectra without following any perturbative expansion of the nuclear potential or empirical methods. We benchmark the algorithm on the IR and Raman spectroscopy of high-pressure hydrogen phase III, with a simulation cell of 96 atoms. Our work also explores the nonlinear regime of the dynamical nuclear motion, providing a paradigm to simulate the interaction with intense or multiple probes, as in pump-probe spectroscopy, or chemical reactions involving light atoms, as the proton transfer in biomolecules

Introduction

The computational power available for research dramatically increased in the last decades paving the way to the birth of a new field of science: material design. We can predict the physical properties of a material in silico, anticipating the experimental data and enabling for an automatic search of target compounds before their synthesis. An example of an impressive result is the prediction of record-breaking high-temperature superconductivity in hydrates, like hydrogen sulfide[1] and LaH10[2] that came together with experimental result[3, 4, 5], thanks to the success of ab initio crystal structure prediction[6, 7, 8].

Material science can also assist in interpreting experimental data, as it is possible, in principle, to anticipate the outcome of almost any experimental technique by calculating dynamical correlation functions. For example, IR and optical spectroscopy are obtained by simulating the dipole-dipole time correlation functions, Raman signal with the polarizability-polarizability dynamical correlation function, while Neutron and X-Ray scattering can be simulated calculating the dynamical structure factor.

However, first-principle calculations are still far from being perfect, and they often miss the required precision. Correctly describing ionic vibrations in ab initio simulation is challenging. Even if the energy of ionic vibration is orders of magnitude smaller than the typical energy involved in chemical bonds, the nuclear excited states are in the same range of energy as thermal excitation available at room temperature. Therefore, ionic motion is responsible for almost all properties of materials that depend on temperature as thermal expansion and thermal conductivity.

The development of time-dependent Hartree-Fock (TD-HF) and time-dependent density functional theory[9, 10] (TD-DFT) paved the way to the simulation of dynamical quantum correlation functions for electrons. Still, a full quantum theory of nuclear motion that can be routinely applied in realistic systems (with hundreds of atoms in the simulation cell, treating ab initio the electrons) is missing.

The usual approach to lattice dynamics is through perturbation theory. Here, the energy landscape expanded in the Taylor series around the static position (minimum of the total electronic energy of the system). This approach fails in systems strongly anharmonic or close to the second-order phase transition, where harmonic phonons are unstable.

Ionic time correlation functions can be computed also in the presence of strong anharmonicity with ab initio molecular dynamics (AIMD)[11, 12], but it neglects nuclear quantum dispersion. This is critical for molecular crystals or systems with light atoms, where the Debye temperature is comparable or above room temperature. In these cases, AIMD provides untrustable results.

The quantum equivalent of AIMD is path-integral molecular dynamics (PIMD), but, in its original imaginary time derivation, it is a static theory. PIMD has been extended in many ways to describe dynamical quantities, as an analytical continuation to real time[13], Fourier path-integral methods[14, 15]; but these methods are computationally heavy and possible only in prototypical cases, where the correlation function has few poles and if a force-field is available to clean stochastic noise[16]. To overcome the computational cost of the almost exact PIMD dynamical extension, approximate empirical methods have been developed, as centroid molecular dynamics[17, 18], Gaussian molecular dynamics[19], and ring-polymer molecular dynamics[20, 21, 22, 23]. These methods have been successfully applied to study the IR spectrum of water employing force-fields, but, still, their application is computationally much more expensive than AIMD, limiting the range of applicability when coupled with an ab initio description of the electronic degrees of freedom. Other semi-classical approaches have been developed to describe quantum dynamical properties, as Ehrenfest dynamics[24], Gaussian wave-packet dynamics[25], and time-dependent self-consistent field[26], however, the first two do not conserve energy during the wave-function propagation, and all of them are still far computationally more demanding than AIMD, preventing a systematic use in real cases.

Another method appreciated in the bio-chemistry community is the Nuclear Electronic-Orbital Method[27]. Here, nuclei are treated with the same level as electrons. This approach allows for almost exact treatment of quantum fluctuations even beyond the Born-Oppenheimer approximation, however, their computational cost is extreme compared to the aforesaid methods, and it is suited only to simulate small molecules, or when the quantum mechanical degrees of freedom are a small fraction of the systems (as in QM/MM methods).

Relevant cases where all these techniques fail are materials with light atoms or close to structural instability, where ionic fluctuations are sizable. For example, the simulation of high-pressure hydrates requires a full quantum and anharmonic treatment of the nuclear motion. These materials are attracting a lot of attention due to the discovery of room-temperature superconductivity[28]. On the other side, systems close to structural instabilities play a fundamental role in technological and industrial applications; among them, we have multiferroics, charge density waves in 2D layers, and ferroelectrics, where the huge anharmonic phonon scattering is exploited also to increase their thermoelectric efficiency. Moreover, quantum nuclear dynamics beyond the linear regime are fundamental for studying chemical reactions involving light atoms. A significant example is the proton transfer in biomolecules.

In this work, we formulate a new dynamical theory of quantum nuclear motion at finite temperature and derive an algorithm to calculate time-dependent correlation functions fully ab initio. This method has a computational cost of the same order as standard AIMD but fully accounts for the effect of quantum fluctuations non empirically. Our theory is a time-dependent (TD) extension of the Stochastic Self-Consistent Harmonic Approximation (SCHA), that already proved to be very efficient in describing equilibrium properties of many materials, as high-pressure hydrates[29, 30, 31, 32], hydrogen[33, 34, 35], charge density waves[36, 37, 38] and thermoelectric materials[39, 40, 41].

In Sec. II, we introduce the theoretical framework of this paper. We revise the general dynamical response to an external time-dependent perturbation and the relationship between the nuclear time-correlation functions to some common experimental techniques. The Time-Dependent SCHA (TD-SCHA) is introduced in Sec. III, where we derive the master equation for the density matrix by imposing that a Gaussian wave-packet must minimize the Dirac action. We also discuss generic features of the time-evolution as energy and entropy conservation, steady-state solutions, and the thermodynamic equilibrium (where the SCHA solution is recovered). In Sec. IV, we linearize the TD-SCHA equations of motion around the equilibrium solution. We derive the explicit formula to compute the quantum dynamical response functions and derive a new Lanczos-based algorithm to compute them. In Sec. VII, we benchmark the method both in an illustrative 1D toy model and in the realistic case of high-pressure hydrogen phase III, where quantum dispersion and anharmonicity strongly affect the vibrational spectra. We prove, comparing the simulation of IR and Raman spectra to experiments, how the TD-SCHA can easily handle such a complex system with 96 atoms in the simulation cell computing forces between ions fully ab initio within Density Functional Theory (DFT).

The quantum nuclear evolution

Almost all experimental techniques used to characterize materials involve an exchange in energy between the probe and the sample below hundreds of eV. At this energy scale, the only components of the material that interact with the probe are valence electrons and ions, which include the atomic nucleus plus the core electrons. These are the degrees of freedom we are interested in studying. In the Born-Oppenheimer (BO) approximation, electrons are fast moving with respect to ions. For this reason, the electrons feel the ionic lattice frozen, and relax to their ground state for each ionic configuration. Thus, the full electron-ion wave-function ∣Ψ⟩|\Psi\rangle can be factorized in a ionic and electronic part:

⟨R,r∣Ψ⟩=⟨R∣ψion⟩⟨r∣ψel[R]⟩(1)\langle\mathbf{R},\mathbf{r}|\Psi\rangle=\langle\mathbf{R}|\psi_{\mathrm{ion}}\rangle\langle\mathbf{r}|\psi_{\mathrm{el}}[\mathbf{R}]\rangle \tag*{(1)}

where r\mathbf{r} are the electron positions, ∣ψion⟩|\psi_{\mathrm{ion}}\rangle is the wave-function of ions and ∣ψel[R]⟩|\psi_{\mathrm{el}[\mathbf{R}]}\rangle is the electronic ground state with the ions fixed in the R\mathbf{R} position.

In the whole manuscript, we use bold fonts to indicate vectors or matrices; products between them are the standard rows-by-columns product. For quantum operators and wave-functions, we use the Dirac notation: we use a hat  ^\hat{\ } to distinguish quantum operators from real numbers.

In this work, we focus on physical properties that depend only on ions. For this reason, we will drop the ion index from the wave-function, and ∣ψ⟩|\psi\rangle always refers to the nuclear wave-function ∣ψion⟩|\psi_{\mathrm{ion}}\rangle. Within BO approximation, the ionic wave-function obeys the Schroedinger equation with a BO Hamiltonian H^\hat{H}:

H^=∑a=13Np^a22ma+V(R^)(2)\hat{H} = \sum_{a=1}^{3N} \frac{\hat{p}_a^2}{2m_a} + V(\hat{R}) \tag*{(2)}

where p^a\hat{p}_a is the momentum operator of nucleus aa, mam_a the mass of the aa-th atom. To use a compact notation, each index indicates both atomic and Cartesian components, so, if not differently specified, it ranges from 1 to 3N3N (NN is the number of atoms). The BO potential V(R)V(R) is the ionic energy landscape, obtained as the ground state energy of the electronic problem with ions fixed in positions RR. The BO energy landscape V(R)V(R) is usually obtained ab initio by solving the electronic problem with fixed nuclei within DFT or HF approximations.

This work aims to describe how (nuclear) physical properties change in time when the system in equilibrium is perturbed by an external probe. This is the typical setup for any experiment: when t<t0t < t_0, the system is in equilibrium. At t=t0t = t_0, it starts interacting with an external time-dependent perturbation Vext(R^,t)V^{\mathrm{ext}}(\hat{R},t). We are interested in the expectation value of a generic nuclear observable A^\hat{A} at time t>t0t > t_0.

At T=0T = 0 K, the nuclear wave-function in equilibrium satisfies the static Schroedinger equation:

H^∣ψ⟩=EGS∣ψ⟩(3)\hat{H} \lvert\psi\rangle= E_{\mathrm{GS}} \lvert\psi\rangle \tag*{(3)}

The presence of a time-dependent perturbation for t≥t0t \geq t_0 introduces a time dependency in the nuclear wave-function ∣ψ(t)⟩\lvert\psi(t)\rangle. Thus, the quantum expectation value of the A^\hat{A} observable at time tt is:

A(t)=⟨ψ(t)∣A^∣ψ(t)⟩(4)\mathcal{A}(t) = \langle\psi(t)\lvert\hat{A}\rvert\psi(t)\rangle \tag*{(4)}

The Schroedinger equation governs the time-evolution of ∣ψ(t)⟩\lvert\psi(t)\rangle:

iℏddt∣ψ(t)⟩=H^td(t)∣ψ(t)⟩.(5)i\hbar\frac{d}{dt}\lvert\psi(t)\rangle= \hat{H}_{\mathrm{td}}(t)\lvert\psi(t)\rangle. \tag*{(5)}

where the H^td(t)\hat{H}_{\mathrm{td}}(t) time-dependent Hamiltonian is:

H^td(t)=H^+Vext(R^,t),(6)\hat{H}_{\mathrm{td}}(t) = \hat{H} + V^{\mathrm{ext}}(\hat{R},t), \tag*{(6)}

and (5) is solved with the initial condition that the wave-function at t=t0t = t_0 is in the ground state:

∣ψ(t0)⟩=∣ψ⟩.(7)\lvert\psi(t_0)\rangle= \lvert\psi\rangle. \tag*{(7)}

At finite temperature (T>0T > 0 K), we just need to replace the wave-function with a density matrix ρ^\hat{\rho} that can describe also mixture of states. At equilibrium (when t≤t0t \leq t_0), the density matrix is ρ^\hat{\rho}:

ρ^=∑ipi∣ψi⟩⟨ψi∣.(8)\hat{\rho} = \sum_i p_i \lvert\psi_i\rangle\langle\psi_i\rvert. \tag*{(8)}

where the ∣ψi⟩\lvert\psi_i\rangle are the eigenstates of the H^\hat{H} Hamiltonian with EiE_i energy, and pip_i are the Boltzmann occupations (kbk_b is the Boltzmann constant and TT the temperature):

H^∣ψi⟩=Ei∣ψi⟩(9)\hat{H}\lvert\psi_i\rangle= E_i\lvert\psi_i\rangle \tag*{(9)}
pi=e−βEiZZ=∑ie−βEiβ=1kbT(10)p_i = \frac{e^{-\beta E_i}}{Z} \qquad Z = \sum_i e^{-\beta E_i} \qquad \beta= \frac{1}{k_bT} \tag*{(10)}

After t0t_0, the system interacts with the external perturbation, and the density matrix depends on time ρ^(t)\hat{\rho}(t). This time evolution is simply given by the evolution of each state in the mixture according to the Schroedinger equation:

ρ^(t)=∑ipi∣ψi(t)⟩⟨ψi(t)∣.(11)\hat{\rho}(t) = \sum_i p_i \lvert\psi_i(t)\rangle\langle\psi_i(t)\rvert. \tag*{(11)}

Where ∣ψi(t)⟩\lvert\psi_i(t)\rangle are the same eigenstates of the BO Hamiltonian H^\hat{H} (Eq. (9)) at t=t0t = t_0 evolving with the time dependent Hamiltonian H^td(t)\hat{H}_{\mathrm{td}}(t). In this work, we focus on isolated systems, therefore, pip_i does not depend on time, and the density matrix satisfies the Liouville-Von Neumann equation:

iℏddtρ^(t)=H^td(t)ρ^(t)−ρ^(t)H^td(t)(12)i\hbar\frac{d}{dt}\hat{\rho}(t) = \hat{H}_{\mathrm{td}}(t)\hat{\rho}(t) - \hat{\rho}(t)\hat{H}_{\mathrm{td}}(t) \tag*{(12)}

In this case, the average of the observable is:

A(t)=⟨A^⟩ρ^(t)=Tr⁡[ρ^(t)A^](13)\mathcal{A}(t) = \left\langle\hat{A}\right\rangle_{\hat{\rho}(t)} = \operatorname{Tr}\left[\hat{\rho}(t)\hat{A}\right] \tag*{(13)}

A case of particular interest is the linear regime (when the external perturbation does not change the status of the system, neither breaking chemical bonds, nor triggering a macroscopic rearrangement of atoms, nor heating the sample). In this case the (small) external perturbation can be splitted in a time-independent coupling B^=B(R^)\hat{B} = \mathcal{B}(\hat{R}) between the external field and the ionic positions, and the time-envelope of the perturbation V(t)\mathcal{V}(t):

Vext(R^,t)=V(1)(R^,t)=B(R^)V(t)(14)V^{\mathrm{ext}}(\hat{R},t) = V^{(1)}(\hat{R},t) = \mathcal{B}(\hat{R})\mathcal{V}(t) \tag*{(14)}
V(t)=0fort<t0(15)\mathcal{V}(t) = 0 \qquad\text{for} \qquad t < t_0 \tag*{(15)}

In this case, the time-dependence of an observable is given by the response function[42]:

A(t)=A(t0)+∫−∞∞χAB(t−t′)V(t′) dt′(16)\mathcal{A}(t) = \mathcal{A}(t_0) + \int_{-\infty}^{\infty} \chi_{\mathcal{A}\mathcal{B}}(t-t')\mathcal{V}(t')\,dt' \tag*{(16)}

The response function χAB(t)\chi_{\mathcal{A}\mathcal{B}}(t) is directly related to time-correlation function through the Kubo equation:

χAB(t)=−iℏ⟨[eiℏH^tA^e−iℏH^t,B^]⟩ρ^(0)ϑ(t)(17)\chi_{\mathcal{A}\mathcal{B}}(t) = -\frac{i}{\hbar}\left\langle\left[e^{\frac{i}{\hbar}\hat{H}t}\hat{A}e^{-\frac{i}{\hbar}\hat{H}t},\hat{B}\right]\right\rangle_{\hat{\rho}(0)}\vartheta(t) \tag*{(17)}

Here, e−iℏH^te^{-\frac{i}{\hbar}\hat{H}t} is the time evolution operator in absence of the external perturbation and ϑ(t)\vartheta(t) is the Heaviside function. In particular, the convolution of (16) between χAB\chi_{\mathcal{A}\mathcal{B}} and V\mathcal{V} becomes a simple product in the frequency domain:

A(ω)−A(t0)=χAB(ω)V(ω).(18)\mathcal{A}(\omega) - \mathcal{A}(t_0) = \chi_{\mathcal{A}\mathcal{B}}(\omega)\mathcal{V}(\omega). \tag*{(18)}

The A^\hat{A} and B^\hat{B} operators depend on the particular experiment we want to simulate.

For example, to simulate IR spectra, we need to compute the dynamical nuclear correlation function χMM(ion)(ω)\chi_{MM}^{(ion)}(\omega), where A^=B^=M(R^)\hat{A}=\hat{B}=M(\hat{R}), and M(R^)M(\hat{R}) is the dipole moment along the probe polarization of the system when nuclei are located in R\boldsymbol{R}. Analogously, the Raman signal is obtained measuring the energy exchanged between the incoming and outcoming radiation. This energy exchange occurs thanks to the polarizability tensor α(R)\alpha(\boldsymbol{R}) of the system caused by the displacements in the ionic position R\boldsymbol{R}, induced by the probe. In this case, A^=B^=αxy(R^)\hat{A}=\hat{B}=\alpha_{xy}(\hat{\boldsymbol{R}}), where the xyxy directions are the incoming and outcoming polarization of light[43]. In the same way, the dynamical structure factor and the thermal conductivity are dynamical nuclear correlation functions[44].

References

  1. [1]Y. Li, J. Hao, H. Liu, Y. Li, and Y. Ma, The Journal of Chemical Physics 140, 174712 (2014).
  2. [2]F. Peng, Y. Sun, C. J. Pickard, R. J. Needs, Q. Wu, and Y. Ma, Physical Review Letters 119 (2017), 10.1103/physrevlett.119.107001.
  3. [3]A. P. Drozdov, M. I. Eremets, I. A. Troyan, V. Ksenofontov, and S. I. Shylin, Nature 525, 73 (2015).
  4. [4]M. Somayazulu, M. Ahart, A. K. Mishra, Z. M. Geballe, M. Baldini, Y. Meng, V. V. Struzhkin, and R. J. Hemley, Physical Review Letters 122 (2019), 10.1103/physrevlett.122.027001.
  5. [5]A. Drozdov, P. Kong, V. Minkov, S. Besedin, M. Kuzovnikov, S. Mozaffari, L. Balakirev, F. Balakirev, D. Graf, V. Prakapenka, et al., Nature 569, 528 (2019).
  6. [6]Y. Wang, J. Lv, L. Zhu, and Y. Ma, Physical Review B 82 (2010), 10.1103/physrevb.82.094116.
  7. [7]A. R. Oganov and C. W. Glass, The Journal of Chemical Physics 124, 244704 (2006).
  8. [8]C. J. Pickard and R. J. Needs, Journal of Physics: Condensed Matter 23, 053201 (2011).
  9. [9]E. Runge and E. K. U. Gross, Physical Review Letters 52, 997 (1984).
  10. [10]M. A. Marques, C. A. Ullrich, F. Nogueira, K. Burke, A. Rubio, and E. K. U. Gross, Time-Dependent Density Functional Theory (Springer Science & Business Media, 2006).
  11. [11]R. Car and M. Parrinello, Physical review letters 55, 2471 (1985).
  12. [12]A. Cupo, D. Tristant, K. Rego, and V. Meunier, npj Computational Materials 5 (2019), 10.1038/s41524-019-0220-1.
  13. [13]G. Baym and N. D. Mermin, Journal of Mathematical Physics 2, 232 (1961).
  14. [14]B. Chen, M. A. Gomez, M. Sehl, J. D. Doll, and D. L. Freeman, The Journal of Chemical Physics 105, 9686 (1996).
  15. [15]D. Kim, J. D. Doll, and J. E. Gubernatis, The Journal of Chemical Physics 106, 1641 (1997).
  16. [16]G. Krilov and B. J. Berne, The Journal of Chemical Physics 111, 9147 (1999).
  17. [17]J. Cao and G. A. Voth, The Journal of Chemical Physics 101, 6168 (1994).
  18. [18]F. Paesani and G. A. Voth, The Journal of Chemical Physics 132, 014105 (2010).
  19. [19]I. Georgescu and V. A. Mandelshtam, Physical Review B 82 (2010), 10.1103/physrevb.82.094305.
  20. [20]J. A. Poulsen, G. Nyman, and P. J. Rossky, The Journal of Chemical Physics 119, 12179 (2003).
  21. [21]R. Hernandez and G. A. Voth, Chemical Physics 233, 243 (1998).
  22. [22]X. Sun, H. Wang, and W. H. Miller, The Journal of Chemical Physics 109, 7064 (1998).
  23. [23]S. Bonella, M. Monteferrante, C. Pierleoni, and G. Ciccotti, The Journal of Chemical Physics 133, 164105 (2010).
  24. [24]X. Li, J. C. Tully, H. B. Schlegel, and M. J. Frisch, The Journal of Chemical Physics 123, 084106 (2005).
  25. [25]E. J. Heller, The Journal of Chemical Physics 62, 1544 (1975).
  26. [26]R. B. Gerber, V. Buch, and M. A. Ratner, The Journal of Chemical Physics 77, 3022 (1982).
  27. [27]F. Pavošević, T. Culpitt, and S. Hammes-Schiffer, Chemical Reviews 120, 4222 (2020).
  28. [28]E. Snider, N. Dasenbrock-Gammon, R. McBride, M. Debessai, H. Vindana, K. Vencatasamy, K. V. Lawler, A. Salamat, and R. P. Dias, Nature 586, 373 (2020).
  29. [29]I. Errea, M. Calandra, and F. Mauri, Physical Review B 89 (2014), 10.1103/physrevb.89.064302.
  30. [30]I. Errea, M. Calandra, C. J. Pickard, J. R. Nelson, R. J. Needs, Y. Li, H. Liu, Y. Zhang, Y. Ma, and F. Mauri, Nature 532, 81 (2016).
  31. [31]R. Bianco, I. Errea, M. Calandra, and F. Mauri, Physical Review B 97 (2018), 10.1103/physrevb.97.214101.
  32. [32]I. Errea, F. Belli, L. Monacelli, A. Sanna, T. Koretsune, T. Tadano, R. Bianco, M. Calandra, R. Arita, F. Mauri, and J. A. Flores-Livas, Nature 578, 66 (2020).
  33. [33]M. Borinaga, P. Riego, A. Leonardo, M. Calandra, F. Mauri, A. Bergara, and I. Errea, Journal of Physics: Condensed Matter 28, 494001 (2016).
  34. [34]M. Borinaga, I. Errea, M. Calandra, F. Mauri, and A. Bergara, Phys. Rev. B 93, 174308 (2016).
  35. [35]L. Monacelli, I. Errea, M. Calandra, and F. Mauri, Nature Physics (2020), 10.1038/s41567-020-1009-3.
  36. [36]R. Bianco, I. Errea, L. Monacelli, M. Calandra, and F. Mauri, Nano Letters 19, 3098 (2019).
  37. [37]R. Bianco, L. Monacelli, M. Calandra, F. Mauri, and I. Errea, Physical Review Letters 125 (2020), 10.1103/physrevlett.125.106101.
  38. [38]J. S. Zhou, L. Monacelli, R. Bianco, I. Errea, F. Mauri, and M. Calandra, Nano Letters 20, 4809 (2020).
  39. [39]U. Aseginolaza, R. Bianco, L. Monacelli, L. Paulatto, M. Calandra, F. Mauri, A. Bergara, and I. Errea, Physical Review Letters 122 (2019), 10.1103/physrevlett.122.075901.
  40. [40]U. Aseginolaza, R. Bianco, L. Monacelli, L. Paulatto, M. Calandra, F. Mauri, A. Bergara, and I. Errea, arXiv preprint arXiv:1906.02047 (2019).
  41. [41]G. A. S. Ribeiro, L. Paulatto, R. Bianco, I. Errea, F. Mauri, and M. Calandra, Physical Review B 97 (2018), 10.1103/physrevb.97.014306.
  42. [42]G. D. Mahan, Many Particle Physics, Third Edition (Plenum, New York, 2000).
  43. [43]M. Lazzeri and F. Mauri, Physical Review Letters 90 (2003), 10.1103/physrevlett.90.036401.
  44. [44]M. Simonecelli, N. Marzari, and F. Mauri, Nature Physics 15, 809 (2009).
  45. [45]L. Monacelli, G. Batignani, G. Fumero, C. Ferrante, S. Mukamel, and T. Scopigno, The Journal of Physical Chemistry Letters 8, 966 (2017).
  46. [46]T. cheng Li and P. qing Tong, Physical Review A 31, 1950 (1985).
  47. [47]R. Bianco, I. Errea, L. Paulatto, M. Calandra, and F. Mauri, Physical Review B 96 (2017), 10.1103/physrevb.96.014111.
  48. [48]D. Huber and E. J. Heller, The Journal of Chemical Physics 87, 5302 (1987).
  49. [49]D. Huber, E. J. Heller, and R. G. Littlejohn, The Journal of Chemical Physics 89, 2003 (1988).
  50. [50]H. Pal, M. Vyas, and S. Tomsovic, Physical Review E 93 (2016), 10.1103/physreve.93.012213.
  51. [51]J. Haegeman, J. I. Cirac, T. J. Osborne, I. Pizorn, H. Verschelde, and F. Verstraete, Physical Review Letters 107 (2011), 10.1103/physrevlett.107.070601.
  52. [52]T. Guaita, L. Hackl, T. Shi, C. Hubig, E. Demler, and J. I. Cirac, Physical Review B 100 (2019), 10.1103/physrevb.100.094529.
  53. [53]L. Hackl, T. Guaita, T. Shi, J. Haegeman, E. Demler, and J. I. Cirac, SciPost Phys. 9, 48 (2020).
  54. [54]H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, 2007).DOI
  55. [55]W. Kohn and J. M. Luttinger, Physical Review 108, 590 (1957).
  56. [56]A. Pribram-Jones, P. E. Grabowski, and K. Burke, Physical Review Letters 116 (2016), 10.1103/physrevlett.116.233001.
  57. [57]D. Rocca, R. Gebauer, Y. Saad, and S. Baroni, The Journal of Chemical Physics 128, 154105 (2008).
  58. [58]O. Hellman, I. A. Abrikosov, and S. I. Simak, Phys. Rev. B 84, 180301 (2011).
  59. [59]T. Tadano, Y. Gohda, and S. Tsuneyuki, Journal of Physics: Condensed Matter 26, 225402 (2014).
  60. [60]J.-M. Lihm and C.-H. Park, arXiv preprint arXiv:2010.15725 (2020).
  61. [61]Y. Gillett, M. G. Giantomassi, and X. Gonze, Physical Review B 88 (2013), 10.1103/physrevb.88.094305.
  62. [62]H. P. C. Miranda, S. Reichardt, G. Froehlicher, A. Molina-Sánchez, S. Berciad, and L. Wirtz, Nano Letters 17, 2381 (2017).
  63. [63]A. Putrino and M. Parrinello, Physical Review Letters 88 (2002), 10.1103/physrevlett.88.176401.
  64. [64]P. L. Silvestrelli, M. Bernasconi, and M. Parrinello, Chemical Physics Letters 277, 478 (1997).
  65. [65]W. Windl, P. Pavone, K. Karch, O. Schütt, D. Strauch, P. Giannozzi, and S. Baroni, Physical Review B 48, 3164 (1993).
  66. [66]W. H. Press, Numerical Recipes: The Art of Scientific Computing, Third Edition, 3rd ed. (Cambridge University Press, 2017).
  67. [67]B. Miehlich, A. Savin, H. Stoll, and H. Preuss, Chemical Physics Letters 157, 200 (1989).
  68. [68]P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I. Dabo, A. D. Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F. Mauri, R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto, C. Sbraccia, S. Scandolo, G. Sclauzero, A. P. Seitsonen, A. Smogunov, P. Umari, and R. M. Wentzcovitch, Journal of Physics: Condensed Matter 21, 395502 (2009).
  69. [69]P. Giannozzi, O. Andreussi, T. Brumme, O. Bunau, M. B. Nardelli, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, M. Cococcioni, N. Colonna, I. Carnimeo, A. D. Corso, S. de Gironcoli, P. Delugas, R. A. DiStasio, A. Ferretti, A. Floris, G. Fratesi, G. Fugallo, R. Gebauer, U. Gerstmann, F. Giustino, T. Gorni, J. Jia, M. Kawamura, H.-Y. Ko, A. Kokalj, E. Küçükbenli, M. Lazzeri, M. Marsili, N. Marzari, F. Mauri, N. L. Nguyen, H.-V. Nguyen, A. O. de la Roza, L. Paulatto, S. Poncé, D. Rocca, R. Sabatini, B. Santra, M. Schliph, A. P. Seitsonen, A. Smogunov, I. Timrov, T. Thonhauser, P. Umari, N. Vast, X. Wu, and S. Baroni, Journal of Physics: Condensed Matter 29, 465901 (2017).
  70. [70]D. R. Hammann, Phys. Rev. B 88, 085117 (2013).
  71. [71]I. Errea, M. Calandra, C. J. Pickard, J. Nelson, R. J. Needs, Y. Li, H. Liu, Y. Zhang, Y. Ma, and F. Mauri, Physical Review Letters 114 (2015), 10.1103/physrevlett.114.157004.
  72. [72]I. Errea, The European Physical Journal B 89 (2016).
  73. [73]A. F. Goncharov, E. Gregoryanz, R. J. Hemley, and H. K. Mao, Proceedings of the National Academy of Sciences 98, 14234 (2001).
  74. [74]N. D. Drummond, B. Monserrat, J. H. Lloyd-Williams, P. L. Ríos, C. J. Pickard, and R. J. Needs, Nature Communications 6 (2015), 10.1038/ncomms8794.
  75. [75]I. B. Magdău and G. J. Ackland, Physical Review B 87 (2013), 10.1103/physrevb.87.171110.
  76. [76]C. Zhang, C. Zhang, M. Chen, W. Kang, Z. Gu, J. Zhao, C. Liu, C. Sun, and P. Zhang, Physical Review B 98 (2018), 10.1103/physrevb.98.144301.
  77. [77]H.-D. Meyer and G. A. Worth, Theoretical Chemistry Accounts 109, 251 (2003).

Paper details

vlnc:W_kgrfb5jd
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents