betaFeedback

WorksPhysics and Astronomy

Many-body interactions in the dielectric theory of stopping power of solids for classical and quantum projectiles

Abstract

We take account of the many-body dynamic interactions by including the wave-vector and frequency-dependent exchange and correlations (xc) kernel f_{xc}(\qv,ω) in the framework of the dielectric theory of the stopping power of crystals for moving charges. The cases of classical and quantum projectiles are considered. We find that (I) the role of the xc effects in slowing of charges in crystalline solids is more pronounced than it is within the jellium model and (II) For velocities below the stopping maximum, inclusion of xc leads to the improvement of the comparison of the theory with experiment over a range of solid targets. On the other hand, in the high-velocity regime, the role of the dynamic xc proves negligible, which does not contradict experiment, and which we substantiate analytically within the jellium model of the target. The input to our theory is the microscopic dielectric matrix ε_{\Gv \Gv'}(\qv,ω) within the random phase approximation (i.e., with neglect of fxcf_{xc}), the calculation of which matrix is implemented in the existing solid-state codes.

Introduction

Interest in slowing of ions travelling through matter dates back to the beginning of the last century [1, 2]. The modern theory of the stopping power (SP) (energy loss by a projectile per unit pass length) had started from the works by Ritchie [3, 4], wherein the target materials were modelled with the homogeneous electron gas (HEG), and the dielectric approach had been employed. With the advent of the density-functional theory (DFT) [5, 6] and the time-dependent DFT (TDDFT) [7, 8], the limitations of the HEG model, as well as those of the first Born approximation, the latter underlying the dielectric approach, have been overstepped [9, 10, 11].

The central quantities in the real-time TDDFT and in its linear-response variant are the time-dependent exchange-correlation (xc) potential vxc(r,t)v_{\mathrm{xc}}(\mathbf{r},t) and the xc kernel fxc(r,r′,t−t′)f_{\mathrm{xc}}(\mathbf{r},\mathbf{r}',t-t'), respectively [7, 8]. These quantities are never known exactly, necessitating construction of approximations. Understandably, the more complicated is the system under study, the simpler and less accurate approximations are generally used for vxcv_{\mathrm{xc}} and fxcf_{\mathrm{xc}}. As a result, in the studies of SP of real solids, the adiabatic local density approximation (ALDA) to TDDFT is used almost exclusively, which choice suffers from the disadvantage of dropping out the effects due to the spatial and temporal non-locality of the electronic response of target systems [12, 13].

At the other end of the spectrum is an approach which employs the non-uniform jellium model and linear-response TDDFT, while striving for more accurate xc kernels [11, 14, 15]. This method does preserve the non-locality, but it is applicable in the limit of zero velocity of the projectile only, the quantity sought for being the friction coefficient Q=lim⁡v→01vdE(v)dxQ=\lim_{v\to0}\frac{1}{v}\frac{dE(v)}{dx} [16].

In the dielectric theory of SP, which will be further detailed below, the central assumption is the adoption of the perturbative treatment of the projectile-target interaction, to the first order in the latter (first Born approximation). In other respects, this theory is well fit to account for the more sophisticated features of the particle-solid interactions: (i) the static and dynamic xc can be naturally included in the dielectric formalism by means of the linear response TDDFT; (ii) the theory is designed for the case of finite velocities of the projectile; (iii) the theory does not rely on the mean-field approximation of the Ehrenfest dynamics [17], but it can be formulated totally quantum mechanically, not only with respect to the target system’s electrons, but to the projectile particle as well (see Appendix § 6); (iv) the theory is conceptually simple, requiring only the evaluation of the dielectric function (DF) of the target system. On the other hand, while possessing an important advantage of the non-linearity, other approaches to the problem of SP usually suffer from the lack of one or more of the features (i)–(iv).

The above observations make it essential to establish the bounds on the applicability of the dielectric approach for the SP problem, the ones which do not result from simplifications used in the construction of DF, but are rather inherent to the approach itself. In this article, we undertake to implement the dielectric theory of SP in its unabridged form, verifying results against available experimental data. At the level of the theory based on the accurate evaluation of the dielectric matrix of a periodic solid accompanied by the inclusion of the dynamic many-body (MB) interactions by means of the non-adiabatic linear-response TDDFT, we generally observe an improved comparison with experiment. Exceptions are also found and discussed.

This work is organized as follows. In § 2, for the jellium solid model, we formulate, implement, and discuss the dielectric approach to the SP problem, which includes the quantum mechanical effects on the part of the projectile and the dynamical xc effects on the part of the target. In § 3, we perform the generalization of the results of § 2 to real solids. Our conclusions are collected in Sec. IV. More involved derivations and technical details are collected in Appendices A-C.

Jellium solid

Within the dielectric approach, SP for a projectile of the charge ZZ and infinite mass, moving with the velocity vv in a medium characterized by the wave-vector and frequency-dependent DF ε(q,ω)\varepsilon(q,\omega) is given by [18] [3]

dEdx=−2Z2πv2∫0∞dqq∫0vqωIm⁡1ε(q,ω) dω,(1)\frac{dE}{dx}=-\frac{2Z^{2}}{\pi v^{2}}\int_{0}^{\infty}\frac{dq}{q}\int_{0}^{vq}\omega\operatorname{Im}\frac{1}{\varepsilon(q,\omega)}\,d\omega, \tag*{(1)}

wherein we use atomic units (e2=ℏ=me=1e^{2}=\hbar=m_{e}=1). If the projectile has a finite mass MM, then the kinematic constraint modifies Eq. (1) to [19, 20]

dEdx=−2Z2πv2∫02Mvdqq∫0vq−q22MωIm⁡1ε(q,ω) dω.(2)\frac{dE}{dx}=-\frac{2Z^{2}}{\pi v^{2}}\int_{0}^{2Mv}\frac{dq}{q}\int_{0}^{vq-\frac{q^{2}}{2M}}\omega\operatorname{Im}\frac{1}{\varepsilon(q,\omega)}\,d\omega. \tag*{(2)}

From the outset, it is conceptually important to differentiate between the classical and quantum mechanical features in Eqs. (1) and (2). As written in atomic units, Eq. (2) disguises the distinction between the two kinds of effects. This distinction can be revealed by re-introducing the world constants in Eq. (2), while ensuring the proper dimensionalities

dEdx=−2Z12e2πv2∫02Mvℏdqq∫0vq−ℏq22MωIm⁡1ε(q,ω) dω.(3)\frac{dE}{dx}=-\frac{2Z_{1}^{2}e^{2}}{\pi v^{2}}\int_{0}^{\frac{2Mv}{\hbar}}\frac{dq}{q}\int_{0}^{vq-\frac{\hbar q^{2}}{2M}}\omega\operatorname{Im}\frac{1}{\varepsilon(q,\omega)}\,d\omega. \tag*{(3)}

From Eq. (3), it becomes clear that the differences in the integration limits between Eqs. (1) and (2) are due to quantum effects exclusively, since they disappear in the classical limit ℏ→0\hbar\to0. In Appendix A we give an independent derivation of the expression (6) for SP of a periodic crystal, of which Eqs. (2)–(3) are a specific case.

For the jellium model of a solid, Eqs. (1) and (2) have been extensively used in the literature within the random phase approximation (RPA) to DF, which amounts to the neglect of xc effects [18, 20] [3, 4]. Our interest lying in the role of the latter effects, we start from considering HEG of interacting electrons (electron liquid), taking into account the xc kernel fxch(q,ω)f_{xc}^{h}(q,\omega), which enters the expression for DF through the relations [8]

1ε(q,ω)=1+4πq2χh(q,ω),(4)\frac{1}{\varepsilon(q,\omega)}=1+\frac{4\pi}{q^{2}}\chi^{h}(q,\omega), \tag*{(4)}
1χh(q,ω)=1χsh(q,ω)−4πq2−fxch(q,ω),(5)\frac{1}{\chi^{h}(q,\omega)}=\frac{1}{\chi_{s}^{h}(q,\omega)}-\frac{4\pi}{q^{2}}-f_{xc}^{h}(q,\omega), \tag*{(5)}

where χh(q,ω)\chi^{h}(q,\omega) and χsh(q,ω)\chi_{s}^{h}(q,\omega) are the interacting physical and the auxiliary independent-electron (Kohn–Sham) density response functions, respectively, the latter being known exactly by the Lindhard formula [21]. The property of viscosity is describable in terms of the frequency-dependent xc kernel fxch(q,ω)f_{xc}^{h}(q,\omega) [13], which is especially important in the context of the transport phenomena [22, 23].

With Eqs. (2), (4)–(5), we have conducted calculations of SP of jellium of various densities for protons and positrons, within RPA (fxch=0f_{xc}^{h}=0) and with xc effects included. For fxchf_{xc}^{h}, we were using the ALDA and the recent constraint-based revised modified Constantin–Pitarke (rMCP07) xc kernel [24], which is currently considered accurate at all densities of the fluid phase of HEG. In Figs. 1 and 2, results for rs=2r_{s}=2, 6 a.u. and rs=10r_{s}=10, 20 a.u., respectively, where rs=[3/(4πn)]1/3r_{s}=[3/(4\pi n)]^{1/3} is the density parameter of HEG of the density nn, are shown. We observe that, at lower velocities, RPA systematically underestimates SP, while ALDA overestimates it, as compared with the calculation based on the rMCP07 parametrization. This discrepancy grows with the decrease of the electronic density (increase of rsr_{s}). We note, and this applies to all the results below, that for proton SP is numerically indistinguishable from that for an infinite mass particle.

A notable feature in Figs. 1 and 2 is the negligible contribution of fxcf_{xc}, both rMCP07 and ALDA ones, in the higher-velocity regime. While this trend in the velocity-dependence of SP is far from obvious a priori, in Appendix C we prove that this holds for an arbitrary causal xc kernel.

In Figs. 3–4, we compare the jellium model calculation results with experimental SP data on aluminium and silicon targets for protons. On the left slope of the peaks (lower velocities), we observe an improved agreement with experiment due to the account for the xc effects. We note that ALDA performs slightly better than rMCP07, which, however, is an artefact of neglecting the crystallinity effects, as will be seen below.

To the right from the maximum (higher velocities) the theory largely underestimates the experimental SP. It is known that, at the growing velocities, SP becomes strongly affected by the excitation of the core electrons [20, 30], which effect the jellium model cannot, obviously, capture.

Figure 2. Same as Figure 1, but for rs=10r_s = 10 (left) and 20 a.u.

Figure 1. Stopping power of jellium of the density parameter rs=2r_s = 2 (left) and 6 a.u. versus the velocity of the projectile in units of the Fermi velocity vFv_F. Results for proton and positron, with the inclusion of the xc effects on the level of rMCP07 and ALDA (solid and dashed lines, respectively) and without xc (RPA, dotted lines) are shown.

The role of xc in the SP of HEG for classical projectiles has been earlier addressed in the extensive literature [36, 37, 38, 39]. Those studies, accordingly, used earlier available parametrizations for the local field correction function (or, equivalently, xc kernel) [40, 41, 42, 43], or have used Mermin’s dielectric function, as in Ref. [38]. These have been surpassed in the more recent developments [24, 44]. We also note a promising first-principles (not constraint-based) exchange-only kernel, derived from a time-dependent variational principle [45, 46]. Here we are using the state-of-the-art exchange and correlation parametrization rMCP07 [24].

Figure 4. Same as Figure 3, but for jellium of the density parameter rs=2.05r_s = 2.05 a.u. Symbols are experimental data for the silicon target: MeV2021 [31], Tra2019 [32], HB2006 [33], Fm2002 [34], and Ep1994a [35], as compiled in Ref. [29].

Figure 3. Stopping power of jellium of the density parameter rs=2.07r_s = 2.07 a.u. for protons calculated within the rMCP07 (solid line), ALDA (dashed line), and RPA (dotted line) approximations. Symbols are experimental data for the aluminium target: Mo2004 [25], Pr2001b [26], MT1996 [27], and Ep1992 [28], as compiled in Ref. [29]. The datasets’ namings are those from the latter reference.

Crystals

We refine the theory by including the crystallinity effects in real solids. The generalization of Eq. (2) for a spatially periodic system and random SP is (see Appendix

A)

dEdx=−Z2π2v∫q∈ΩBdq∑Gv⋅(q+G)−(q+G)22M∣q+G∣2×Θ[v⋅(q+G)−(q+G)22M]Im⁡ϵGG−1[q,v⋅(q+G)−(q+G)22M],(6)\frac{dE}{dx}=-\frac{Z^2}{\pi^2v}\int_{q\in\Omega_B}d\mathbf{q}\sum_{\mathbf{G}}\frac{\mathbf{v}\cdot(\mathbf{q}+\mathbf{G})-\frac{(\mathbf{q}+\mathbf{G})^2}{2M}}{|\mathbf{q}+\mathbf{G}|^2}\times\Theta\left[\mathbf{v}\cdot(\mathbf{q}+\mathbf{G})-\frac{(\mathbf{q}+\mathbf{G})^2}{2M}\right]\operatorname{Im}\epsilon^{-1}_{\mathbf{G}\mathbf{G}}\left[\mathbf{q},\mathbf{v}\cdot(\mathbf{q}+\mathbf{G})-\frac{(\mathbf{q}+\mathbf{G})^2}{2M}\right], \tag*{(6)}

where Θ\Theta is the Heaviside step function. In Eq. (6) the summation is over the infinite set of the reciprocal vectors G\mathbf{G} of the lattice, the integration in q\mathbf{q} is over the first Brillouin zone, and ϵGG′−1(q,ω)\epsilon^{-1}_{\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega) is the inverse microscopic dielectric matrix of the system, of which the diagonal elements only are needed. Equation (6) is the quantum mechanical generalization of the corresponding classical SP formula for periodic solids [30, 47].

Within the linear-response TDDFT of crystals, the matrix version of the relations (4)–(5) reads

ϵGG′−1(q,ω)=1+4π∣q+G∣∣q+G′∣χGG′(q,ω),(7)\epsilon^{-1}_{\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega)=1+\frac{4\pi}{|\mathbf{q}+\mathbf{G}||\mathbf{q}+\mathbf{G}'|}\chi_{\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega), \tag*{(7)}
χGG′−1(q,ω)=χs,GG′−1(q,ω)−4π∣q+G∣2δGG′−fxc,GG′(q,ω).(8)\chi^{-1}_{\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega)=\chi^{-1}_{s,\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega)-\frac{4\pi}{|\mathbf{q}+\mathbf{G}|^2}\delta_{\mathbf{G}\mathbf{G}'}-f_{xc,\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega). \tag*{(8)}

We were calculating χs,GG′(q,ω)\chi_{s,\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega) with the all-electron full-potential linearised augmented-plane wave code ELK [48]. At larger projectiles’ velocities, the excitation of the core states of the target plays the major role [20, 30], which, in the calculations, necessitates the promotion of the core to valence electrons and, as a consequence, the use of large dielectric matrices. Further particulars of the calculations are collected in Appendix § 7.

The spatial and temporal non-locality of fxcf_{xc} was accounted for within the approximation of HEG with the average, over the unit cell, density of valence electrons, i.e.,

fxc,GG′(q,ω)≈fxcHEG(q+G,ω)δGG′,(9)f_{xc,\mathbf{G}\mathbf{G}'}(\mathbf{q},\omega)\approx f^{\mathrm{HEG}}_{xc}(\mathbf{q}+\mathbf{G},\omega)\delta_{\mathbf{G}\mathbf{G}'}, \tag*{(9)}

where for fxcHEGf^{\mathrm{HEG}}_{xc} the parametrization rMCP07 was used [24].

In Figure 5, SP of crystalline aluminium, calculated with the use of Eqs. (6)–(9), is plotted against the velocity of protons and compared with experimental data. To the left of the peak, inclusion of the crystallinity together with xc effects makes a considerable difference, moving theoretical results closer to the measured data. To the right from the peak, xc effects play a negligible role, while the real-solid calculation catches partially excitations from the deeper electron levels. The waviness of the spectra is due to the comparatively sparse qq-points grid used for integration, and it was confirmed to be smoothed out by making the grid denser, however, at the cost of much more expensive computations.

Figure 5. Stopping power of aluminium crystal for protons with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data: Mo2004 [25], Pr2001b [26], MT1996 [27], and Ep1992 [28], as compiled in Ref. [29].

To take account of the excitations from the deeper electronic levels, a large number of G\mathbf{G}-vectors must be used in Eq. (6), owing to the localized character of those levels, which requires many terms in their Fourier transform. In Appendix § 7 we list the parameters used in the calculations. The further increase of the computational complexity was beyond our technical means, which contributes to the remaining discrepancies with experiment at the higher projectiles’ velocities.

Results for diamond, Figure 6, silicon, Figure 7, and beryllium, Figure 8, are qualitatively similar to those for aluminium, clearly confirming the importance of the combined crystallinity and xc effects at and to the left from the spectra maxima. The older experimental data for Be are, however, in dissonance with our theoretical, as well as with the newer experimental, results.

Figure 6. Stopping power of diamond crystal for protons with (solid line) and without (RPA, dashed line) inclusion of the xc effects. Symbols are experimental data [49], as compiled in Ref. [29].

Figure 7. Stopping power of silicon crystal for proton projectiles with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data: MeV2021 [31], Tra2019 [32], HB2006 [33], Fm2002 [34], and Ep1994a [35], as compiled in Ref. [29].

Figure 8. Stopping power of beryllium crystal for protons with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data: Bn1980 [50], Mr1968 [51], Ka1953 [52], and Wa1949 [53], as compiled in Ref. [29].

Alkali metals lithium and rubidium, Figure 9 and Figure 10, respectively, present notable exceptions to the agreement between the theory and experiment at and to the left from the main maxima. The dielectric theory of SP, obviously, breaks down for those metals in this velocity range. This can be understood as follows: In the low velocity regime and at the lower densities of HEG, a bound state of the pseudo-atom is formed [9], which, for the proton projectile, happens at rs≳2r_s \gtrsim2 a.u. The first Born approximation for the inelastic scattering cannot, obviously, capture this process, resulting in the failure of the dielectric approach for Li (rs=3.25r_s = 3.25 a.u.) and Rb (rs=5.32r_s = 5.32 a.u.). On the other hand, although for proton in the Al (rs=2.07r_s = 2.07 a.u.) and Si (rs=2.05r_s = 2.05 a.u.) hosts the bound state develops as well, it is marginally shallow, practically not affecting the SP.

Figure 9. Stopping power of lithium crystal for protons with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data: Ep1995 [54], Bd1956 [55], as compiled in Ref. [29]. The jellium model results are also shown in order to contrast a difference with the real-solid calculation for this low-conduction-electron density metal.

Figure 10. Stopping power of rubidium crystal for protons with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data: Ep1994b [56], as compiled in Ref. [29]. The calculated curves have been smoothed to eliminate a strong noise. The jellium model results are also shown in order to contrast a drastic difference with the real-solid calculation for this low-conduction-electron density metal.

To the right from the peaks the comparison is fairly good. We, further, note that Rb turns out an example of a simple metal for which the jellium model is irrelevant at all to the SP calculation, as seen from Figure 10. Our choice of alkali targets was determined by the availability of experimental data [29].

In Figs. 11 and 12, the dependence of the stopping power of Al and Si on the velocity is plotted for electron projectiles. The factor

gx=1+∣q+G∣4v4−∣q+G∣2v2,(10)g_x = 1 + \frac{\lvert q + G \rvert^4}{v^4} - \frac{\lvert q + G \rvert^2}{v^2}, \tag*{(10)}

which describes the exchange between the impinging electron and those of the target [57], was included under the sum in Eq. (6). The overall agreement between the theory and experiment is fairly good, although xc in the response of the target does not play a major role in the case of the electron projectile. We note that the rapid disappearance of the SP for electron at lower velocities in Figs. 11 and 12 is in accord with the results for the friction coefficient in the zero-velocity limit within the nonlinear theory [58].

Conclusions

We have enhanced the dielectric theory of the stopping power of solids for moving charges by simultaneously accounting for the crystallinity and the many-body interaction effects. The many-body effects have been included by means of the linear-response TDDFT via the dynamic exchange and correlation kernel as an ingredient to the dielectric response of a solid.

We have found the role of the many-body effects, in combination with the crystallinity ones, quite important. Moreover, the account for those effects brings the theory into an improved comparison with experiment for the majority of considered targets. At the same time, we have observed a discrepancy between the theory and experiment for alkali metals (Li and Rb) at lower velocities of

Figure 11. Stopping power of aluminium crystal for electron with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data from Ref. [59].

projectiles, which signifies the breakdown of the dielectric theory of the stopping power for those targets in the low velocity regime, prompting for the resolution to be sought beyond the linear theory.

Acknowledgments

V.U.N. acknowledges the hospitality of the Donostia International Physics Center, San Sebastián, Spain. V.M.S. acknowledges financial support by Grant PID2022-139230NB-I00 funded by MCIN/AEI/10.13039/501100011033.

Data availability – The data that support the findings of this article are openly available [60].

Quantum mechanical derivation of the formula for the stopping power within the dielectric formalism

We consider a target system of NeN_e electrons and NnN_n nuclei. A projectile charge ZZ is travelling through this system. We consider electrons of the target and the projectile as quantum particles, while nuclei of the target are point charges pinned at their equilibrium positions. The time-dependent Schrödinger equation for the wave-function Π(r,R,t)\Pi(\mathbf{r},\mathbf{R},t) of the whole system, where r\mathbf{r} stands for the set of the electronic coordinates ri\mathbf{r}_i, reads

i∂Π(r,R,t)∂t=H^Π(r,R,t),(11)i\frac{\partial\Pi(\mathbf{r},\mathbf{R},t)}{\partial t}=\hat{H}\Pi(\mathbf{r},\mathbf{R},t), \tag*{(11)}

where

H^=H^0+P^22M+H^1,(12)\hat{H}=\hat{H}_0+\frac{\hat{\mathbf{P}}^2}{2M}+\hat{H}_1, \tag*{(12)}

H^0\hat{H}_0 is the Hamiltonian of the target, P^=−i∇R\hat{\mathbf{P} }=-i\nabla_{\mathbf{R}} is the momentum operator of the projectile, MM being its mass,

H^1=∑i=1Ne−Z∣ri−R∣+∑j=1NnZjZ∣Rj−R∣,(13)\hat{H}_1=\sum_{i=1}^{N_e}-\frac{Z}{\lvert\mathbf{r}_i-\mathbf{R}\rvert}+\sum_{j=1}^{N_n}\frac{Z_jZ}{\lvert\mathbf{R}_j-\mathbf{R}\rvert}, \tag*{(13)}

where Rj\mathbf{R}_j and ZjZ_j are the coordinates and charges of nuclei, respectively, and R\mathbf{R} is the coordinate of the projectile.

Let, at t→−∞t\to-\infty, the target and the projectile wave-functions be well separated, and the target to be in its ground-state Ψ0(r)\Psi_0(\mathbf{r}). Then we can write to the first Born approximation in H^1\hat{H}_1

Π~(r‾,R,t→+∞)=Ψ0(r‾)Φ~(R,t→−∞)+1i∫−∞∞dteiH^0teitP^2/2MH^1e−iH^0te−itP^2/2MΨ0(r‾)Φ~(R,t→−∞),(14)\widetilde{\Pi}(\underline{r},R,t\to+\infty)=\Psi_0(\underline{r})\widetilde{\Phi}(R,t\to-\infty)+\frac{1}{i}\int_{-\infty}^{\infty}dt e^{i\hat{H}_0t}e^{it\hat{P}^{2}/2M}\hat{H}_1e^{-i\hat{H}_0t}e^{-it\hat{P}^{2}/2M}\Psi_0(\underline{r})\widetilde{\Phi}(R,t\to-\infty), \tag*{(14)}

where the overhead tilde denotes the corresponding wave-function in the interaction picture. Furthermore,

⟨Ψn≠0(r‾)∣Π~(r‾,R,t→+∞)⟩r‾=1i∫−∞∞dt ei(En−E0)teitP^2/2M⟨Ψn(r‾)∣H^1∣Ψ0(r‾)⟩r‾e−itP^2/2MΦ~(R,t→−∞),(15)\left\langle\Psi_{n\ne0}(\underline{r})\left|\widetilde{\Pi}(\underline{r},R,t\to+\infty)\right.\right\rangle_{\underline{r}}=\frac{1}{i}\int_{-\infty}^{\infty}dt\,e^{i(E_n-E_0)t}e^{it\hat{P}^{2}/2M}\left\langle\Psi_n(\underline{r})\left|\hat{H}_1\right|\Psi_0(\underline{r})\right\rangle_{\underline{r}}e^{-it\hat{P}^{2}/2M}\widetilde{\Phi}(R,t\to-\infty), \tag*{(15)}

where Ψn\Psi_n and EnE_n are the complete set of the eigenfunctions of the target and its corresponding eigenenergies, respectively, and ⟨…⟩r‾\langle\ldots\rangle_{\underline{r}} denotes integration over electronic coordinates. Substituting Eq. (13) into Eq. (15), introducing the electron density operator

n^(r)=∑i=1Neδ(ri−r),(16)\hat{n}(\mathbf{r})=\sum_{i=1}^{N_e}\delta(\mathbf{r}_i-\mathbf{r}), \tag*{(16)}

and noting that the second term in Eq. (13) does not contribute to the matrix element, we can rewrite Eq. (15) as

⟨Ψn≠0(r‾)∣Π~(r‾,R,t→+∞)⟩r‾=Zi∫⟨Ψn(r‾)∣n^(r‾)∣Ψ0(r‾)⟩r‾ei(En−E0)teitP^2/2M∣r−R∣e−itP^2/2MΦ~(R,t→−∞) drdt,(17)\left\langle\Psi_{n\ne0}(\underline{r})\left|\widetilde{\Pi}(\underline{r},R,t\to+\infty)\right.\right\rangle_{\underline{r}}=\frac{Z}{i}\int\left\langle\Psi_n(\underline{r})\left|\hat{n}(\underline{r})\right|\Psi_0(\underline{r})\right\rangle_{\underline{r}}\frac{e^{i(E_n-E_0)t}e^{it\hat{P}^{2}/2M}}{\lvert\mathbf{r}-R\rvert}e^{-it\hat{P}^{2}/2M}\widetilde{\Phi}(R,t\to-\infty)\,d\mathbf{r}dt, \tag*{(17)}

or, after the Fourier transform

Φ~(P,t)=(2π)−3/2∫Φ~(R,t)e−iP⋅RdR,(18)\widetilde{\Phi}(\mathbf{P},t)=(2\pi)^{-3/2}\int\widetilde{\Phi}(R,t)e^{-i\mathbf{P}\cdot R}dR, \tag*{(18)}
⟨Ψn≠0(r‾)∣Π~(r‾,R,t→+∞)⟩r‾=Ziπ(2π)3/2∫⟨Ψn(r‾)∣n^(r‾)∣Ψ0(r‾)⟩r‾ei(En−E0)teit(P−q)2/2Mq2eiq⋅re−itP2/2Mdq Φ~(P,t→−∞)ei(P−q)⋅RdP dr dt.(19)\left\langle\Psi_{n\ne0}(\underline{r})\left|\widetilde{\Pi}(\underline{r},R,t\to+\infty)\right.\right\rangle_{\underline{r}}=\frac{Z}{i\pi(2\pi)^{3/2}}\int\left\langle\Psi_n(\underline{r})\left|\hat{n}(\underline{r})\right|\Psi_0(\underline{r})\right\rangle_{\underline{r}}\frac{e^{i(E_n-E_0)t}e^{it(\mathbf{P}-\mathbf{q})^2/2M}}{q^2}e^{i\mathbf{q}\cdot\mathbf{r}}e^{-it\mathbf{P}^2/2M}dq\,\widetilde{\Phi}(\mathbf{P},t\to-\infty)e^{i(\mathbf{P}-\mathbf{q})\cdot R}d\mathbf{P}\,d\mathbf{r}\,dt. \tag*{(19)}

Integrating in Eq. (19) over time explicitly, we can write

⟨Ψn≠0(r‾)∣Π~(r‾,R,t→+∞)⟩r‾=Ziπ(2π)3/2∫⟨Ψn(r‾)∣n^(r‾)∣Ψ0(r‾)⟩r‾δ[En−E0+(P−q)22M−P22M]eiq⋅rdqq2Φ~(P,t→−∞)ei(P−q)⋅RdP dr,(20)\left\langle\Psi_{n\ne0}(\underline{r})\left|\widetilde{\Pi}(\underline{r},R,t\to+\infty)\right.\right\rangle_{\underline{r}}=\frac{Z}{i\pi(2\pi)^{3/2}}\int\left\langle\Psi_n(\underline{r})\left|\hat{n}(\underline{r})\right|\Psi_0(\underline{r})\right\rangle_{\underline{r}}\delta\left[E_n-E_0+\frac{(\mathbf{P}-\mathbf{q})^2}{2M}-\frac{\mathbf{P}^2}{2M}\right]e^{i\mathbf{q}\cdot\mathbf{r}}\frac{dq}{q^2}\widetilde{\Phi}(\mathbf{P},t\to-\infty)e^{i(\mathbf{P}-\mathbf{q})\cdot R}d\mathbf{P}\,d\mathbf{r}, \tag*{(20)}

and

∫∣⟨Ψn≠0(r‾)∣Π~(r‾,R,t→+∞)⟩r‾∣2dR=Z28π5∫δ[En−E0+(P−q)22M−P22M]×eiq⋅rei(P−q)⋅R⟨Ψn(r‾)∣n^(r‾)∣Ψ0(r‾)⟩r‾⟨Ψn(r‾′)∣n^(r′)∣Ψ0(r‾′)⟩r‾∗δ[En−E0+(P′−q′)22M−P′22M]×e−iq′⋅r′e−i(P′−q′)⋅RΦ~(P,t→−∞)Φ~∗(P′,t→−∞)dq dq′q2q′2dP dr dP′ dr′ dR,(21)\begin{aligned} \int\left|\left\langle\Psi_{n\ne0}(\underline{r})\left|\widetilde{\Pi}(\underline{r},R,t\to+\infty)\right.\right\rangle_{\underline{r}}\right|^2dR=\frac{Z^2}{8\pi^5}\int\delta\left[E_n-E_0+\frac{(\mathbf{P}-\mathbf{q})^2}{2M}-\frac{\mathbf{P}^2}{2M}\right] \\ &\times e^{i\mathbf{q}\cdot\mathbf{r}}e^{i(\mathbf{P}-\mathbf{q})\cdot R}\left\langle\Psi_n(\underline{r})\left|\hat{n}(\underline{r})\right|\Psi_0(\underline{r})\right\rangle_{\underline{r}}\left\langle\Psi_n(\underline{r}')\left|\hat{n}(\mathbf{r}')\right|\Psi_0(\underline{r}')\right\rangle_{\underline{r}}^{*}\delta\left[E_n-E_0+\frac{(\mathbf{P}'-\mathbf{q}')^2}{2M}-\frac{\mathbf{P}'^2}{2M}\right] \\ &\times e^{-i\mathbf{q}'\cdot\mathbf{r}'}e^{-i(\mathbf{P}'-\mathbf{q}')\cdot R}\widetilde{\Phi}(\mathbf{P},t\to-\infty)\widetilde{\Phi}^{*}(\mathbf{P}',t\to-\infty)\frac{dq\,dq'}{q^2q'^2}d\mathbf{P}\,d\mathbf{r}\,d\mathbf{P}'\,d\mathbf{r}'\,dR, \tag*{(21)} \end{aligned}

or, integrating further over R\mathbf{R} and P′\mathbf{P}',

∫∣⟨Ψn≠0(r)|Π~(r,R,t→+∞)⟩r∣2dR=Z2π2∫δ[En−E0+(P−q)22M−P22M]×⟨Ψn(r)|n^(r)|Ψ0(r)⟩r⟨Ψn(r)|n^(r′)|Ψ0(r)⟩r∗δ[P22M−(P−q+q′)22M]×eiq⋅r−iq′⋅r′Φ~(P,t→−∞)Φ~∗(P−q+q′,t→−∞)dqdq′q2q′2dPdrdr′.(22)\begin{aligned} \int\left|\left\langle\Psi_{n\ne0}(\mathbf{r})\middle|\widetilde{\Pi}(\mathbf{r},\mathbf{R},t\to+\infty)\right\rangle_{\mathbf{r}}\right|^2 d\mathbf{R} = \frac{Z^2}{\pi^2}\int\delta\left[E_n-E_0+\frac{(\mathbf{P}-\mathbf{q})^2}{2M}-\frac{\mathbf{P}^2}{2M}\right] \\ &\times\left\langle\Psi_n(\mathbf{r})\middle|\hat{n}(\mathbf{r})\middle|\Psi_0(\mathbf{r})\right\rangle_{\mathbf{r}}\left\langle\Psi_n(\mathbf{r})\middle|\hat{n}(\mathbf{r}')\middle|\Psi_0(\mathbf{r})\right\rangle_{\mathbf{r}}^* \delta\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q}+\mathbf{q}')^2}{2M}\right] \\ &\times e^{i\mathbf{q}\cdot\mathbf{r}-i\mathbf{q}'\cdot\mathbf{r}'}\widetilde{\Phi}(\mathbf{P},t\to-\infty)\widetilde{\Phi}^*(\mathbf{P}-\mathbf{q}+\mathbf{q}',t\to-\infty)\frac{d\mathbf{q}d\mathbf{q}'}{q^2q'^2}d\mathbf{P}d\mathbf{r}d\mathbf{r}'. \tag*{(22)} \end{aligned}

For the energy gain by the target after the interaction we can write

ΔE=∑n≠0(En−E0)∫∣⟨Ψn(r)|Π~(r,R,t→+∞)⟩r∣2dR,(23)\Delta E=\sum_{n\ne0}(E_n-E_0)\int\left|\left\langle\Psi_n(\mathbf{r})\middle|\widetilde{\Pi}(\mathbf{r},\mathbf{R},t\to+\infty)\right\rangle_{\mathbf{r}}\right|^2d\mathbf{R}, \tag*{(23)}

which, with the use of Eq. (22), yields

ΔE=Z2π2∫∑n≠0[P22M−(P−q)22M]δ[P22M−(P−q+q′)22M]δ[En−E0+(P−q)22M−P22M]×⟨Ψn(r)|n^(r)|Ψ0(r)⟩r⟨Ψ0(r)|n^(r′)|Ψn(r)⟩reiq⋅r−iq′⋅r′Φ~(P,t→−∞)Φ~∗(P−q+q′,t→−∞)dqdq′q2q′2dPdrdr′.(24)\begin{aligned} \Delta E=\frac{Z^2}{\pi^2}\int\sum_{n\ne0}\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right]\delta\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q}+\mathbf{q}')^2}{2M}\right]\delta\left[E_n-E_0+\frac{(\mathbf{P}-\mathbf{q})^2}{2M}-\frac{\mathbf{P}^2}{2M}\right] \\ &\times\left\langle\Psi_n(\mathbf{r})\middle|\hat{n}(\mathbf{r})\middle|\Psi_0(\mathbf{r})\right\rangle_{\mathbf{r}}\left\langle\Psi_0(\mathbf{r})\middle|\hat{n}(\mathbf{r}')\middle|\Psi_n(\mathbf{r})\right\rangle_{\mathbf{r}}e^{i\mathbf{q}\cdot\mathbf{r}-i\mathbf{q}'\cdot\mathbf{r}'}\widetilde{\Phi}(\mathbf{P},t\to-\infty)\widetilde{\Phi}^*(\mathbf{P}-\mathbf{q}+\mathbf{q}',t\to-\infty)\frac{d\mathbf{q}d\mathbf{q}'}{q^2q'^2}d\mathbf{P}d\mathbf{r}d\mathbf{r}'. \tag*{(24)} \end{aligned}

On the other hand, for the density response function χ(r,r′,ω)\chi(\mathbf{r},\mathbf{r}',\omega) of the same target one can write [13]

χ(r,r′,ω)=∑n≠0[⟨Ψn(r)|n^(r)|Ψ0(r)⟩r⟨Ψ0(r)|n^(r′)|Ψn(r)⟩rE0−En+ω+iη+⟨Ψn(r)|n^(r′)|Ψ0(r)⟩r⟨Ψ0(r)|n^(r)|Ψn(r)⟩rE0−En−ω−iη],(25)\chi(\mathbf{r},\mathbf{r}',\omega)=\sum_{n\ne0}\left[\frac{\left\langle\Psi_n(\mathbf{r})\middle|\hat{n}(\mathbf{r})\middle|\Psi_0(\mathbf{r})\right\rangle_{\mathbf{r}}\left\langle\Psi_0(\mathbf{r})\middle|\hat{n}(\mathbf{r}')\middle|\Psi_n(\mathbf{r})\right\rangle_{\mathbf{r}}}{E_0-E_n+\omega+i\eta}+\frac{\left\langle\Psi_n(\mathbf{r})\middle|\hat{n}(\mathbf{r}')\middle|\Psi_0(\mathbf{r})\right\rangle_{\mathbf{r}}\left\langle\Psi_0(\mathbf{r})\middle|\hat{n}(\mathbf{r})\middle|\Psi_n(\mathbf{r})\right\rangle_{\mathbf{r}}}{E_0-E_n-\omega-i\eta}\right], \tag*{(25)}

and, therefore,

Θ(ω)Im⁡χ(r,r′,ω)=−π∑n≠0⟨Ψn(r)|n^(r)|Ψ0(r)⟩r⟨Ψ0(r)|n^(r′)|Ψn(r)⟩rδ(En−E0−ω),(26)\Theta(\omega)\operatorname{Im}\chi(\mathbf{r},\mathbf{r}',\omega)=-\pi\sum_{n\ne0}\left\langle\Psi_n(\mathbf{r})\middle|\hat{n}(\mathbf{r})\middle|\Psi_0(\mathbf{r})\right\rangle_{\mathbf{r}}\left\langle\Psi_0(\mathbf{r})\middle|\hat{n}(\mathbf{r}')\middle|\Psi_n(\mathbf{r})\right\rangle_{\mathbf{r}}\delta(E_n-E_0-\omega), \tag*{(26)}

where Θ(ω)\Theta(\omega) is the Heaviside step-function. Combining Eqs. (24) and (26), we have

ΔE=−Z2π3∫[P22M−(P−q)22M]Θ[P22M−(P−q)22M]δ[P22M−(P−q+q′)22M]Im⁡χ[r,r′,P22M−(P−q)22M]×eiq⋅r−iq′⋅r′Φ~(P,t→−∞)Φ~∗(P−q+q′,t→−∞)dqdq′q2q′2dPdrdr′.(27)\begin{aligned} \Delta E=-\frac{Z^2}{\pi^3}\int\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right]\Theta\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right]\delta\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q}+\mathbf{q}')^2}{2M}\right]\operatorname{Im}\chi\left[\mathbf{r},\mathbf{r}',\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right] \\ &\times e^{i\mathbf{q}\cdot\mathbf{r}-i\mathbf{q}'\cdot\mathbf{r}'}\widetilde{\Phi}(\mathbf{P},t\to-\infty)\widetilde{\Phi}^*(\mathbf{P}-\mathbf{q}+\mathbf{q}',t\to-\infty)\frac{d\mathbf{q}d\mathbf{q}'}{q^2q'^2}d\mathbf{P}d\mathbf{r}d\mathbf{r}'. \tag*{(27)} \end{aligned}

In the case of a crystal the density response function χ(r,r′,ω)\chi(\mathbf{r},\mathbf{r}',\omega) can be written in terms of the matrix χGG′(q,ω)\chi_{GG'}(\mathbf{q},\omega) indexed with the reciprocal vectors of the crystal lattice

χ(r,r′,ω)=1(2π)3∫q∈ΩBdq∑G,G′χGG′(q,ω)eiq⋅(r−r′)eiG⋅r−iG′⋅r′,(28)\chi(\mathbf{r},\mathbf{r}',\omega)=\frac{1}{(2\pi)^3}\int_{\mathbf{q}\in\Omega_B}d\mathbf{q}\sum_{G,G'}\chi_{GG'}(\mathbf{q},\omega)e^{i\mathbf{q}\cdot(\mathbf{r}-\mathbf{r}')}e^{i\mathbf{G}\cdot\mathbf{r}-i\mathbf{G}'\cdot\mathbf{r}'}, \tag*{(28)}

where ΩB\Omega_B is the first Brillouin zone. Therefore,

ΔE=−Z2π3(2π)3∫q′∈ΩBdq′′∑G,G′∫[P22M−(P−q)22M]Θ[P22M−(P−q)22M]δ[P22M−(P−q+q′)22M]eiq⋅r−iq′⋅r′×Φ~(P,t→−∞)Φ~∗(P−q+q′,t→−∞)dqdq′q2q′2dPdrdr′Im⁡χGG′[q′′,P22M−(P−q)22M]eiq′′⋅(r−r′)eiG⋅r−iG′⋅r′.(29)\begin{aligned} \Delta E=-\frac{Z^2}{\pi^3(2\pi)^3}\int_{\mathbf{q}'\in\Omega_B}d\mathbf{q}''\sum_{G,G'}\int\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right]\Theta\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right]\delta\left[\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q}+\mathbf{q}')^2}{2M}\right]e^{i\mathbf{q}\cdot\mathbf{r}-i\mathbf{q}'\cdot\mathbf{r}'} \\ &\times\widetilde{\Phi}(\mathbf{P},t\to-\infty)\widetilde{\Phi}^*(\mathbf{P}-\mathbf{q}+\mathbf{q}',t\to-\infty)\frac{d\mathbf{q}d\mathbf{q}'}{q^2q'^2}d\mathbf{P}d\mathbf{r}d\mathbf{r}'\operatorname{Im}\chi_{GG'}\left[\mathbf{q}'',\frac{\mathbf{P}^2}{2M}-\frac{(\mathbf{P}-\mathbf{q})^2}{2M}\right]e^{i\mathbf{q}''\cdot(\mathbf{r}-\mathbf{r}')}e^{i\mathbf{G}\cdot\mathbf{r}-i\mathbf{G}'\cdot\mathbf{r}'}. \tag*{(29)} \end{aligned}

In Eq. (29) we first integrate in r′r' over the whole space

ΔE=−Z2π3∫dq′′∑q′′∈ΩB∑G,G′∫[P22M−(P−q)22M]Θ[P22M−(P−q)22M]δ[P22M−(P−q−q′′−G′)22M]ei(q+q′′+G)⋅rΦ~(P,t→−∞)Φ~∗(P−q−q′′−G′,t→−∞)dq dPq2∣q′′+G′∣2dr Im⁡χGG′[q′′,P22M−(P−q)22M].(30)\Delta E=-\frac{Z^2}{\pi^3}\int dq''\sum_{q''\in\Omega_B}\sum_{G,G'}\int\left[\frac{P^2}{2M}-\frac{(P-q)^2}{2M}\right]\Theta\left[\frac{P^2}{2M}-\frac{(P-q)^2}{2M}\right]\delta\left[\frac{P^2}{2M}-\frac{(P-q-q''-G')^2}{2M}\right]e^{i(q+q''+G)\cdot r} \widetilde{\Phi}(P,t\to-\infty)\widetilde{\Phi}^{*}(P-q-q''-G',t\to-\infty)\frac{dq\,dP}{q^2\lvert q''+G'\rvert^2}dr\,\operatorname{Im}\chi_{GG'}\left[q'',\frac{P^2}{2M}-\frac{(P-q)^2}{2M}\right]. \tag*{(30)}

From this point, we consider a film of the crystal, infinite in the xyxy plane and confined to the interval [−L/2,L/2][-L/2,L/2] in the zz-coordinate. Integrating in rr over the film, we have

ΔE=−8Z2π∫dq′′∑q′′∈ΩB∑G,G′∫[P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz−qz)22M]×Θ[P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz−qz)22M]×δ[P22M−(Px+Gx−Gx′)22M−(Py+Gy−Gy′)22M−(Pz−qz−qz′′−Gz′)22M]sin⁡[(qz+qz′′+Gz)L/2]qz+qz′′+Gz×Φ~(P,t→−∞)Φ~∗(Px+Gx−Gx′,Py+Gy−Gy′,Pz−qz−qz′′−Gz′,t→−∞)×dqz dP[(qx′′+Gx)2+(qy′′+Gy)2+qz2]∣q′′+G′∣2Im⁡χGG′[q′′,P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz−qz)22M].(31)\begin{aligned} \Delta E={}&-\frac{8Z^2}{\pi}\int dq''\sum_{q''\in\Omega_B}\sum_{G,G'}\int\left[\frac{P^2}{2M}-\frac{(P_x+q''_x+G_x)^2}{2M}-\frac{(P_y+q''_y+G_y)^2}{2M}-\frac{(P_z-q_z)^2}{2M}\right]\\ &\times\Theta\left[\frac{P^2}{2M}-\frac{(P_x+q''_x+G_x)^2}{2M}-\frac{(P_y+q''_y+G_y)^2}{2M}-\frac{(P_z-q_z)^2}{2M}\right]\\ &\times\delta\left[\frac{P^2}{2M}-\frac{(P_x+G_x-G'_x)^2}{2M}-\frac{(P_y+G_y-G'_y)^2}{2M}-\frac{(P_z-q_z-q''_z-G'_z)^2}{2M}\right]\frac{\sin[(q_z+q''_z+G_z)L/2]}{q_z+q''_z+G_z}\\ &\times\widetilde{\Phi}(P,t\to-\infty)\widetilde{\Phi}^{*}(P_x+G_x-G'_x,P_y+G_y-G'_y,P_z-q_z-q''_z-G'_z,t\to-\infty)\\ &\times\frac{dq_z\,dP}{[(q''_x+G_x)^2+(q''_y+G_y)^2+q_z^2]\lvert q''+G'\rvert^2}\operatorname{Im}\chi_{GG'}\left[q'',\frac{P^2}{2M}-\frac{(P_x+q''_x+G_x)^2}{2M}-\frac{(P_y+q''_y+G_y)^2}{2M}-\frac{(P_z-q_z)^2}{2M}\right]. \tag*{(31)} \end{aligned}

Let the projectile wave-packet be Gaussian

Φ~(P)=(2π)3/4σ3/2e−σ2(P−P0)2,(32)\widetilde{\Phi}(P)=\left(\frac{2}{\pi}\right)^{3/4}\sigma^{3/2}e^{-\sigma^2(P-P_0)^2}, \tag*{(32)}

where σ\sigma is large. In the limit σ→∞\sigma\to\infty Eq. (32) ensures that ∣Φ~(P)∣2→δ(P−P0)\lvert\widetilde{\Phi}(P)\rvert^2\to\delta(P-P_0). Then, due to Φ~∗\widetilde{\Phi}^{*} in Eq. (31), only Gx′=GxG'_x=G_x and Gy′=GyG'_y=G_y contribute, which allows to rewrite Eq. (31) as

ΔE=−8Z2π∫dq′′∑q′′∈ΩB∑G,G′∫[P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz−qz)22M]×Θ[P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz−qz)22M]×δ[Pz22M−(Pz−qz−qz′′−Gz′)22M]sin⁡[(qz+qz′′+Gz)L/2]qz+qz′′+Gz×Φ~(P,t→−∞)Φ~∗(Px,Py,Pz−qz−qz′′−Gz′,t→−∞)×dqz dz[(qx′′+Gx)2+(qy′′+Gy)2+qz2][(qx′′+Gx)2+(qy′′+Gy)2+(qz′′+Gz′)2]dP×Im⁡χGG′[q′′,P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz−qz)22M].(33)\begin{aligned} \Delta E={}&-\frac{8Z^2}{\pi}\int dq''\sum_{q''\in\Omega_B}\sum_{G,G'}\int\left[\frac{P^2}{2M}-\frac{(P_x+q''_x+G_x)^2}{2M}-\frac{(P_y+q''_y+G_y)^2}{2M}-\frac{(P_z-q_z)^2}{2M}\right]\\ &\times\Theta\left[\frac{P^2}{2M}-\frac{(P_x+q''_x+G_x)^2}{2M}-\frac{(P_y+q''_y+G_y)^2}{2M}-\frac{(P_z-q_z)^2}{2M}\right]\\ &\times\delta\left[\frac{P_z^2}{2M}-\frac{(P_z-q_z-q''_z-G'_z)^2}{2M}\right]\frac{\sin[(q_z+q''_z+G_z)L/2]}{q_z+q''_z+G_z}\\ &\times\widetilde{\Phi}(P,t\to-\infty)\widetilde{\Phi}^{*}(P_x,P_y,P_z-q_z-q''_z-G'_z,t\to-\infty)\\ &\times\frac{dq_z\,dz}{[(q''_x+G_x)^2+(q''_y+G_y)^2+q_z^2][(q''_x+G_x)^2+(q''_y+G_y)^2+(q''_z+G'_z)^2]}dP\\ &\times\operatorname{Im}\chi_{GG'}\left[q'',\frac{P^2}{2M}-\frac{(P_x+q''_x+G_x)^2}{2M}-\frac{(P_y+q''_y+G_y)^2}{2M}-\frac{(P_z-q_z)^2}{2M}\right]. \tag*{(33)} \end{aligned}

Integrating in Eq. (33) over qzq_z, we have due to the δ\delta-function

ΔE=−8MZ2π∫q′′∈ΩBdq′′∑G,G′∫1Pz[P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz+qz′′+Gz′)22M]×Θ[P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz+qz′′+Gz′)22M]×sin⁡[(Gz−Gz′)L/2]Gz−Gz′∣Φ~(P,t→−∞)∣2dP[(qx′′+Gx)2+(qy′′+Gy)2+qz2][(qx′′+Gx)2+(qy′′+Gy)2+(qz′′+Gz′)2]×Im⁡χGG′(q′′,P22M−(Px+qx′′+Gx)22M−(Py+qy′′+Gy)22M−(Pz+qz′′+Gz′)22M).(34)\Delta E=-\frac{8MZ^{2}}{\pi}\int_{q^{\prime\prime}\in\Omega_{B}}d q^{\prime\prime}\sum_{G,G^{\prime}}\int\frac{1}{P_{z}}\left[\frac{\mathbf{P}^{2}}{2M}-\frac{(P_{x}+q_{x}^{\prime\prime}+G_{x})^{2}}{2M}-\frac{(P_{y}+q_{y}^{\prime\prime}+G_{y})^{2}}{2M}-\frac{(P_{z}+q_{z}^{\prime\prime}+G_{z}^{\prime})^{2}}{2M}\right] \\ \times\Theta\left[\frac{\mathbf{P}^{2}}{2M}-\frac{(P_{x}+q_{x}^{\prime\prime}+G_{x})^{2}}{2M}-\frac{(P_{y}+q_{y}^{\prime\prime}+G_{y})^{2}}{2M}-\frac{(P_{z}+q_{z}^{\prime\prime}+G_{z}^{\prime})^{2}}{2M}\right] \\ \times\frac{\sin[(G_{z}-G_{z}^{\prime})L/2]}{G_{z}-G_{z}^{\prime}}|\tilde{\Phi}(\mathbf{P},t\to-\infty)|^{2}\frac{dP}{[(q_{x}^{\prime\prime}+G_{x})^{2}+(q_{y}^{\prime\prime}+G_{y})^{2}+q_{z}^{2}][(q_{x}^{\prime\prime}+G_{x})^{2}+(q_{y}^{\prime\prime}+G_{y})^{2}+(q_{z}^{\prime\prime}+G_{z}^{\prime})^{2}]} \\ \times\operatorname{Im}\chi_{GG^{\prime}}\left(q^{\prime\prime},\frac{\mathbf{P}^{2}}{2M}-\frac{(P_{x}+q_{x}^{\prime\prime}+G_{x})^{2}}{2M}-\frac{(P_{y}+q_{y}^{\prime\prime}+G_{y})^{2}}{2M}-\frac{(P_{z}+q_{z}^{\prime\prime}+G_{z}^{\prime})^{2}}{2M}\right). \tag*{(34)}

The leading term in Eq. (34), proportional to LL, is at Gz′=GzG_{z}^{\prime}=G_{z}, all other terms vanishing after the division by LL. Then, after the integration over P\mathbf{P}, we have

lim⁡L→∞ΔEL=−4MZ2πP0z∫q′′∈ΩBdq′′∑G∫[P022M−(P0+q′′+G)22M]×Θ[P022M−(P0+q′′+G)22M]1∣q′′+G∣4Im⁡χGG[q′′,P022M−(P0+q′′+G)22M].(35)\lim_{L\to\infty}\frac{\Delta E}{L}=-\frac{4MZ^{2}}{\pi P_{0z}}\int_{q^{\prime\prime}\in\Omega_{B}}d q^{\prime\prime}\sum_{G}\int\left[\frac{P_{0}^{2}}{2M}-\frac{(P_{0}+q^{\prime\prime}+G)^{2}}{2M}\right] \\ \times\Theta\left[\frac{P_{0}^{2}}{2M}-\frac{(P_{0}+q^{\prime\prime}+G)^{2}}{2M}\right]\frac{1}{|q^{\prime\prime}+G|^{4}}\operatorname{Im}\chi_{GG}\left[q^{\prime\prime},\frac{P_{0}^{2}}{2M}-\frac{(P_{0}+q^{\prime\prime}+G)^{2}}{2M}\right]. \tag*{(35)}

Finally, taking account of the relation between the density-response matrix and the dielectric matrix

ϵGG′−1(q,ω)=δGG′+4π∣q+G∣∣q+G′∣χGG′(q,ω),(36)\epsilon^{-1}_{GG^{\prime}}(q,\omega)=\delta_{GG^{\prime}}+\frac{4\pi}{|q+G||q+G^{\prime}|}\chi_{GG^{\prime}}(q,\omega), \tag*{(36)}

we have

lim⁡L→∞ΔEL=−MZ2π2P0z∫q′′∈ΩBdq′′∑G∫[P022M−(P0+q′′+G)22M]×Θ[P022M−(P0+q′′+G)22M]1∣q′′+G∣2Im⁡ϵGG−1[q′′,P022M−(P0+q′′+G)22M].(37)\lim_{L\to\infty}\frac{\Delta E}{L}=-\frac{MZ^{2}}{\pi^{2}P_{0z}}\int_{q^{\prime\prime}\in\Omega_{B}}d q^{\prime\prime}\sum_{G}\int\left[\frac{P_{0}^{2}}{2M}-\frac{(P_{0}+q^{\prime\prime}+G)^{2}}{2M}\right] \\ \times\Theta\left[\frac{P_{0}^{2}}{2M}-\frac{(P_{0}+q^{\prime\prime}+G)^{2}}{2M}\right]\frac{1}{|q^{\prime\prime}+G|^{2}}\operatorname{Im}\epsilon^{-1}_{GG}\left[q^{\prime\prime},\frac{P_{0}^{2}}{2M}-\frac{(P_{0}+q^{\prime\prime}+G)^{2}}{2M}\right]. \tag*{(37)}

Using the symmetry property of the dielectric matrix ϵ−G′,−G−1(−q,ω)=ϵG,G′−1(q,ω)\epsilon^{-1}_{-G^{\prime},-G}(-q,\omega)=\epsilon^{-1}_{G,G^{\prime}}(q,\omega), we see that Eq. (37) proves Eq. (6). In the specific case of the HEG dielectric function ϵ(q,ω)\epsilon(q,\omega), by integrating separately over the radial and angular parts of qq, Eq. (6) is reduced to Eq. (2).

Details of calculations for crystals

We were carrying out calculations for real-solids with the all-electron full-potential linearised augmented-plane wave code ELK [48]. For the realistic band-structures of semiconductors, the Tran-Blaha xc potential TB09 [61] was used, and, for the uniformity, it was also used in the case of metals.

To evaluate χs,GG′(q,ω)\chi_{s,GG^{\prime}}(q,\omega), we were first calculating ϵs,GG′−1(q,ω)\epsilon^{-1}_{s,GG^{\prime}}(q,\omega) (task 180 of the ELK code), inverting the latter matrix, then taking use of the relation

ϵs,GG′(q,ω)=δGG′−4π∣q+G∣∣q+G′∣χs,GG′(q,ω).\epsilon_{s,GG^{\prime}}(q,\omega)=\delta_{GG^{\prime}}-\frac{4\pi}{|q+G||q+G^{\prime}|}\chi_{s,GG^{\prime}}(q,\omega).

Finally, Eqs. (6)–(9) were applied.

To account for the core-states’ excitation, the promotion of the core to valence states was done. Ideally, all electrons should be considered valence, which, however, was not always feasible due to the computational limitations. The division of electrons into the two categories used is shown in Table I. The large number of empty bands is needed at larger velocities of the projectiles as well. In general, all the parameters of the calculations were chosen such as to aim at the convergence at larger velocities, which, however, was not always possible. At lower velocities of the projectile results are well converged with respect to the parameters of the calculations.

Crystalkk- and qq-pointsEmpty bandsCore electronsValence electronsDielectric matrix
Al16×16×1616 \times16 \times16512211941×941941 \times941
Si12×12×1212 \times12 \times122562121471×14711471 \times1471
Diamond16×16×1616 \times16 \times1635606645×645645 \times645
Be16×16×1616 \times16 \times1651204761×761761 \times761
Li16×16×1616 \times16 \times1661203887×887887 \times887
Rb8×8×88 \times8 \times851212252731×27312731 \times2731

Table I. Parameters of the ground-state and the dielectric matrix calculations used in producing results in Figure 5–Figure 12.

Figure 12. Stopping power of silicon crystal for electron with (rMCP07 solid line, ALDA dashed line) and without (RPA, dotted line) inclusion of the xc effects. Symbols are experimental data from Ref. [59].

All possible symmetries have been used to reduce the calculation time to acceptable values.

High velocity

Here we prove that any causal fxcf_{xc} does not affect the high-velocity asymptotics of the stopping power of HEG. We write down the difference between SP with inclusion of some fxcf_{xc} and the RPA one. According to Eq. (2), (4), and (5)

dEdx−(dEdx)RPA=−8Z2v2×∫02Mvdqq3∫0vq−q22Mdω ωIm⁡{χs(q,ω)1−χs(q,ω)[4πq2+fxc(q,ω)]−χs(q,ω)1−χs(q,ω)4πq2}\frac{dE}{dx}-\left(\frac{dE}{dx}\right)_{RPA}=-\frac{8Z^{2}}{v^{2}}\times\int_{0}^{2Mv}\frac{dq}{q^{3}}\int_{0}^{vq-\frac{q^{2}}{2M}}d\omega\,\omega\operatorname{Im}\left\{\frac{\chi_{s}(q,\omega)}{1-\chi_{s}(q,\omega)\left[\frac{4\pi}{q^{2}}+f_{xc}(q,\omega)\right]}-\frac{\chi_{s}(q,\omega)}{1-\chi_{s}(q,\omega)\frac{4\pi}{q^{2}}}\right\}

or

dEdx−(dEdx)RPA=8Z2v2×∫02Mvdqq3∫0vq−q22Mdω Im⁡χs2(q,ω)fxc(q,ω){1−χs(q,ω)[4πq2+fxc(q,ω)]}[1−χs(q,ω)4πq2].\frac{dE}{dx}-\left(\frac{dE}{dx}\right)_{RPA}=\frac{8Z^{2}}{v^{2}}\times\int_{0}^{2Mv}\frac{dq}{q^{3}}\int_{0}^{vq-\frac{q^{2}}{2M}}d\omega\,\operatorname{Im}\frac{\chi_{s}^{2}(q,\omega)f_{xc}(q,\omega)}{\left\{1-\chi_{s}(q,\omega)\left[\frac{4\pi}{q^{2}}+f_{xc}(q,\omega)\right]\right\}\left[1-\chi_{s}(q,\omega)\frac{4\pi}{q^{2}}\right]}.

Taking v→∞v\to\infty in the upper limit of the inner integral and considering that the integrand is an even function of ω\omega, we can write asymptotically

dEdx−(dEdx)RPA→v→∞4Z2v2×∫02Mvdqq3Im⁡∫−∞∞dω ωχs2(q,ω)fxc(q,ω){1−χs(q,ω)[4πq2+fxc(q,ω)]}[1−χs(q,ω)4πq2](38)\frac{dE}{dx}-\left(\frac{dE}{dx}\right)_{RPA}\xrightarrow{v\to\infty}\frac{4Z^{2}}{v^{2}}\times\int_{0}^{2Mv}\frac{dq}{q^{3}}\operatorname{Im}\int_{-\infty}^{\infty}d\omega\,\frac{\omega\chi_{s}^{2}(q,\omega)f_{xc}(q,\omega)}{\left\{1-\chi_{s}(q,\omega)\left[\frac{4\pi}{q^{2}}+f_{xc}(q,\omega)\right]\right\}\left[1-\chi_{s}(q,\omega)\frac{4\pi}{q^{2}}\right]} \tag*{(38)}

Furthermore, since, due to causality, the integrand in (C1) is an analytical function of ω\omega in the upper complex half-plane, since χs(q,ω)∼∣ω∣→∞1∣ω∣2\chi_{s}(q,\omega)\underset{|\omega|\to\infty}{\sim}\frac{1}{|\omega|^{2}}, and since ∣fxc(q,ω)∣|f_{xc}(q,\omega)| is bounded in this half-plane [13], by Cauchy integral theorem we conclude that the inner integral in (C1) evaluates to zero. This proves that SP with inclusion of a fxcf_{xc} has the same high-velocity asymptotic behavior as the RPA one.

References

  1. [5]P. Hohenberg and W. Kohn, Inhomogeneous electron gas, Phys. Rev. 136, B864 (1964).
  2. [6]W. Kohn and L. J. Sham, Self-consistent equations including exchange and correlation effects, Phys. Rev. 140, A1133 (1965).DOI
  3. [7]E. Runge and E. K. U. Gross, Density-functional theory for time-dependent systems, Phys. Rev. Lett. 52, 997 (1984).DOI
  4. [8]E. K. U. Gross and W. Kohn, Local density-functional theory of frequency-dependent linear response, Phys. Rev. Lett. 55, 2850 (1985).
  5. [9]P. M. Echenique, R. M. Nieminen, J. C. Ashley, and R. H. Ritchie, Nonlinear stopping power of an electron gas for slow ions, Phys. Rev. A 33, 987 (1986).pubmed.ncbi.nlm.nih.gov/9896701
  6. [10]J. M. Pruneda, D. Sánchez-Portal, A. Arnau, J. I. Juarristi, and E. Artacho, Electronic stopping power in LiF from first principles, Phys. Rev. Lett. 99, 235501 (2007).pubmed.ncbi.nlm.nih.gov/18233379
  7. [11]V. U. Nazarov, J. M. Pitarke, C. S. Kim, and Y. Takada, Time-dependent density-functional theory for the stopping power of an interacting electron gas for slow ions, Phys. Rev. B 71, 121106(R) (2005).DOI
  8. [12]E. K. U. Gross and W. Kohn, Density functional theory of many-fermion systems, in Advances in Quantum Chemistry, Vol. 21, edited by S. B. Trickey (Academic, San Diego, 1990).
  9. [13]G. F. Giuliani and G. Vignale, Quantum Theory of the Electron Liquid (Cambridge University Press, Cambridge, 2005).
  10. [14]V. U. Nazarov, J. M. Pitarke, Y. Takada, G. Vignale, and Y.-C. Chang, Including nonlocality in the exchange-correlation kernel from time-dependent current density functional theory: Application to the stopping power of electron liquids, Phys. Rev. B 76, 205103 (2007).DOI
  11. [15]V. U. Nazarov, J. M. Pitarke, Y. Takada, G. Vignale, and Y.-C. Chang, Time-dependent current-density functional theory for the friction of ions in an interacting electron gas, Int. J. Mod. Phys. B 22, 3813 (2008).DOI
  12. [16]D. Roth, C. Celedon, D. Goebl, E. Sanchez, B. Bruckner, R. Steinberger, J. Guimpel, N. Arista, and P. Bauer, Systematic analysis of different experimental approaches to measure electronic stopping of very slow hydrogen ions, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 437, 1 (2018).DOI
  13. [17]P. Ehrenfest, Bemerkung über die angenäherte gültigkeit der klassichen mechanik, innerhalb der quantenmechanik, Z. Physik 45, 455–457 (1927).
  14. [18]J. Lindhard and A. Winther, Stopping power of electron gas and equipartition rule, K. Dan. Vidensk. Selsk. Mat.-Fys. Medd. 34, 1 (1964).
  15. [19]C. D. Archubi and N. R. Arista, General formulation of interactions and energy loss of particles in plasmas: Quantum-wave-packet model versus a semiclassical approach, Phys. Rev. A 105, 032806 (2022).DOI
  16. [20]C. D. Archubi, C. C. Montanari, and N. R. Arista, Dielectric approaches for interactions of protons, positrons, and electrons in cold matter and plasmas, Phys. Rev. A 108, 012822 (2023).DOI
  17. [21]J. Lindhard, On the properties of a gas of charged particles, K. Dan. Vidensk. Selsk. Mat.-Fys. Medd. 28, 1 (1954).
  18. [22]V. U. Nazarov, T. N. Todorov, and E. K. U. Gross, Viscous current-induced forces, Phys. Rev. Lett. 133, 026301 (2024).DOI
  19. [23]V. U. Nazarov, T. N. Todorov, and E. K. U. Gross, Nanomotors driven by viscous ac currents, Phys. Rev. Lett. 136, 117002 (2026).DOI
  20. [24]A. D. Kaplan, N. K. Nepal, A. Ruzsinszky, P. Balone, and J. P. Perdew, First-principles wave-vector- and frequency-dependent exchange-correlation kernel for jellium at all densities, Phys. Rev. B 105, 035123 (2022).DOI
  21. [25]S. P. Møller, A. Csete, T. Ichikawa, H. Knudsen, U. I. Uggerhøj, and H. H. Andersen, Stopping power in insulators and metals without charge exchange, Phys. Rev. Lett. 93, 042502 (2004).pubmed.ncbi.nlm.nih.gov/15323754
  22. [26]D. Primetzhofer, S. Rund, D. Roth, D. Goebl, and P. Bauer, Electronic excitations of slow ions in a free electron gas metal: Evidence for charge exchange effects, Phys. Rev. Lett. 107, 163201 (2011).pubmed.ncbi.nlm.nih.gov/22107378
  23. [27]G. Martínez-Tamayo, J. C. Eckardt, G. H. Lantschner, and N. R. Arista, Energy loss of H+ and He+ in Al, Zn, and Au in the very low- to intermediate-energy range, Phys. Rev. A 54, 3131 (1996).pubmed.ncbi.nlm.nih.gov/9913832
  24. [28]C. Eppacher and D. Semrad, Dependence of proton and helium energy loss in solids upon plasma properties, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 69, 33 (1992).DOI
  25. [29]Stopping Power of Matter for Ions, Graphs, Data, Comments and Programs, https://www-nds.iaea.org/stopping/; C. Montanari and P. Dimitriou, The IAEA stopping power database, following the trends in stopping power of ions in matter, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 408, 50 (2017).
  26. [30]A. A. Shukri, F. Bruneval, and L. Reining, Ab initio electronic stopping power of protons in bulk materials, Phys. Rev. B 93, 035128 (2016).
  27. [31]M. Mery, J. D. Uribe, M. Flores, N. R. Arista, V. A. Esaulov, and J. E. Valdés, Electronic energy loss and straggling in low energy H+ and H2+ interaction with silicon films, Radiation Effects and Defects in Solids 176, 73 (2021).DOI
  28. [32]T. T. Tran, L. Jablonka, B. Bruckner, S. Rund, D. Roth, M. A. Sortica, P. Bauer, Z. Zhang, and D. Primetzhofer, Electronic interaction of slow hydrogen and helium ions with nickel-silicon systems, Phys. Rev. A 100, 032705 (2019).DOI
  29. [33]G. Hobler, K. Bourdelle, and T. Akatsu, Random and channeling stopping power of H in Si below 100keV, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 242, 617 (2006).DOI
  30. [34]M. Fāmān, G. H. Lantschner, J. Eckardt, N. Arista, J. Gayone, E. Sanchez, and F. Lovey, Energy loss and angular dispersion of 2–200 keV protons in amorphous silicon, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 193, 91 (2002).
  31. [35]C. Eppacher, Ph.D. thesis, Schriften der Johannes-Kepler-Universität Linz, Universitätsverlag Rudolf Trauner (1995).
  32. [36]S. Tanaka and S. Ichimaru, Stopping power of degenerate electron liquid at metallic densities, Journal of the Physical Society of Japan 54, 2537 (1985).DOI
  33. [37]N.-P. Wang and I. Nagy, Electronic energy loss of helium ions in aluminum, Phys. Rev. A 55, 2083 (1997).DOI
  34. [38]M. D. Barriga-Carrasco, Mermin dielectric function versus local field corrections on proton stopping in degenerate plasmas, Laser and Particle Beams 26, 389–395 (2008).
  35. [39]M. D. Barriga-Carrasco, Dynamical local field corrections on energy loss in plasmas of all degeneracies, Phys. Rev. E 79, 027401 (2009).
  36. [40]S. Ichimaru and K. Utsumi, Analytic expression for the dielectric screening function of strongly coupled electron liquids at metallic and lower densities, Phys. Rev. B 24, 7385 (1981).DOI
  37. [41]K. Utsumi and S. Ichimaru, Dielectric formulation of strongly coupled electron liquids at metallic densities. VI. Analytic expression for the local-field correction, Phys. Rev. A 26, 603 (1982).DOI
  38. [42]S. Ichimaru, S. Mitake, S. Tanaka, and X.-Z. Yan, Theory of interparticle correlations in dense, high-temperature plasmas. I. general formalism, Phys. Rev. A 32, 1768 (1985).pubmed.ncbi.nlm.nih.gov/9896264
  39. [43]B. Dabrowski, Dynamical local-field factor in the response function of an electron gas, Phys. Rev. B 34, 4989 (1986).pubmed.ncbi.nlm.nih.gov/9940320
  40. [44]A. Ruzsinszky, N. K. Nepal, J. M. Pitarke, and J. P. Perdew, Constraint-based wave vector and frequency dependent exchange-correlation kernel of the uniform electron gas, Phys. Rev. B 101, 245135 (2020).DOI
  41. [45]V. U. Nazarov, Time-dependent effective potential and exchange kernel of homogeneous electron gas, Phys. Rev. B 87, 165125 (2013).DOI
  42. [46]V. U. Nazarov and V. M. Silkin, Exchange kernel f_x^h(q,ω) of electron liquid from the variational principle of MacLachlan, Phys. Rev. B 110, 104310 (2024).DOI
  43. [47]I. Campillo, J. M. Pitarke, and A. G. Eguiluz, Electronic stopping power of aluminum crystal, Phys. Rev. B 58, 10307 (1998).DOI
  44. [48]http://elk.sourceforge.net.
  45. [49]W. Käferböck, W. Rössler, V. Necas, P. Bauer, M. Peñalba, E. Zarate, and A. Arnau, Comparative study of the stopping power of graphite and diamond, Phys. Rev. B 55, 13275 (1997).DOI
  46. [50]K. Brunner, W. Hink, and M. Roth, Stopping power for H+ in Be (20–120 keV), Nuclear Instruments and Methods 173, 357 (1980).DOI
  47. [51]K. Morita, H. Akimune, and T. Suita, Energy loss of low energy protons and deuterons in evaporated metallic films, Journal of the Physical Society of Japan 25, 1525 (1968).DOI
  48. [52]D. Kahn, The energy loss of protons in metallic foils and mica, Phys. Rev. 90, 503 (1953).DOI
  49. [53]S. D. Warshaw, The stopping power for protons in several metals, Phys. Rev. 76, 1759 (1949).DOI
  50. [54]C. Eppacher, R. Díez Muño, D. Semrad, and A. Arnau, Stopping power of lithium for hydrogen projectiles, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 96, 639 (1995), the Interaction of Swift Particles and Electromagnetic Fields with Matter.DOI
  51. [55]M. Bader, R. E. Pixley, F. S. Mozer, and W. Wahling, Stopping cross section of solids for protons, 50-600 keV, Phys. Rev. 103, 32 (1956).DOI
  52. [56]C. Eppacher, G. Zemsauer, and D. Semrad, How good are interpolations in stopping power tabulations? The case of Rb and of Sr, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 90, 92 (1994).
  53. [57]V. I. Ochkur, Soviet Phys. J. E. T. P. 47, 1766 (1964).
  54. [58]V. U. Nazarov and E. K. U. Gross, Stopping power of electron liquid for slow quantum projectiles, Contributions to Plasma Physics https://onlinelibrary.wiley.com/doi/pdf/10.1002/ctpp.70132.
  55. [59]D. C. Joy, S. Luo, R. Gauvin, P. Hovington, and N. Evans, Experimental measurements of electron stopping power at low energies, Scanning Microsc. 10, 4 (1996).
  56. [60]V. U. Nazarov and V. M. Silkin, Many-body interactions in the dielectric theory of stopping power of solids for classical and quantum projectiles, DOI: 10.6084/m9.figshare.32089053.
  57. [61]F. Tran and P. Blaha, Accurate band gaps of semiconductors and insulators with a semilocal exchange-correlation potential, Phys. Rev. Lett. 102, 226401 (2009).pubmed.ncbi.nlm.nih.gov/19658882

Paper details

vlnc:W_y8h95fd5
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