1 #LyX 2.4 created this file. For more info see https://www.lyx.org/
5 \save_transient_properties true
8 \use_default_options true
9 \maintain_unincluded_children false
11 \language_package default
14 \font_roman "default" "TeX Gyre Pagella"
15 \font_sans "default" "default"
16 \font_typewriter "default" "default"
17 \font_math "auto" "auto"
18 \font_default_family default
19 \use_non_tex_fonts false
23 \font_typewriter_osf false
24 \font_sf_scale 100 100
25 \font_tt_scale 100 100
27 \use_dash_ligatures true
29 \default_output_format default
31 \bibtex_command default
32 \index_command default
33 \float_placement class
34 \float_alignment class
35 \paperfontsize default
38 \pdf_author "Marek Nečada"
40 \pdf_bookmarksnumbered false
41 \pdf_bookmarksopen false
42 \pdf_bookmarksopenlevel 1
50 \use_package amsmath 1
51 \use_package amssymb 1
54 \use_package mathdots 1
55 \use_package mathtools 1
57 \use_package stackrel 1
58 \use_package stmaryrd 1
59 \use_package undertilde 1
61 \cite_engine_type default
65 \paperorientation portrait
77 \paragraph_separation indent
78 \paragraph_indentation default
80 \math_numbering_side default
85 \paperpagestyle default
87 \tracking_changes false
97 Infinite periodic systems
98 \begin_inset FormulaMacro
99 \newcommand{\dlv}{\vect a}
103 \begin_inset FormulaMacro
104 \newcommand{\rlv}{\vect b}
110 \begin_layout Standard
111 Although large finite systems are where MSTMM excels the most, there are
112 several reasons that makes its extension to infinite lattices (where periodic
113 boundary conditions might be applied) desirable as well.
114 Other methods might be already fast enough, but MSTMM will be faster in
115 most cases in which there is enough spacing between the neighboring particles.
116 MSTMM works well with any space group symmetry the system might have (as
117 opposed to, for example, FDTD with cubic mesh applied to a honeycomb lattice),
119 application of group theory in mode analysis quite easy.
120 \begin_inset Note Note
123 \begin_layout Plain Layout
129 And finally, having a method that handles well both infinite and large
130 finite system gives a possibility to study finite-size effects in periodic
134 \begin_layout Subsection
138 \begin_layout Standard
139 TODO Fourier transforms, Delta comb, lattice bases, reciprocal lattices
143 \begin_layout Subsection
144 Formulation of the problem
147 \begin_layout Standard
148 Let us have a linear system of compact EM scatterers on a homogeneous background
150 \begin_inset CommandInset ref
152 reference "subsec:Multiple-scattering"
159 , but this time, the system shall be periodic: let there be a
160 \begin_inset Formula $d$
164 \begin_inset Formula $d$
167 can be 1, 2 or 3) lattice embedded into the three-dimensional real space,
169 \begin_inset Formula $\left\{ \dlv_{i}\right\} _{i=1}^{d}$
172 , and let the lattice points be labeled with an
173 \begin_inset Formula $d$
176 -dimensional integar multiindex
177 \begin_inset Formula $\vect n\in\ints^{d}$
180 , so the lattice points have cartesian coordinates
181 \begin_inset Formula $\vect R_{\vect n}=\sum_{i=1}^{d}n_{i}\vect a_{i}$
185 There can be several scatterers per unit cell with indices
186 \begin_inset Formula $\alpha$
190 \begin_inset Formula $\mathcal{P}_{1}$
193 and (relative) positions inside the unit cell
194 \begin_inset Formula $\vect r_{\alpha}$
197 ; any particle of the periodic system can thus be labeled by a multiindex
199 \begin_inset Formula $\mathcal{P}=\ints^{d}\times\mathcal{P}_{1}$
203 The scatterers are located at positions
204 \begin_inset Formula $\vect r_{\vect n,\alpha}=\vect R_{\vect n}+\vect r_{\alpha}$
208 \begin_inset Formula $T$
211 -matrices are periodic,
212 \begin_inset Formula $T_{\vect n,\alpha}=T_{\alpha}$
216 In such system, the multiple-scattering problem
217 \begin_inset CommandInset ref
219 reference "eq:Multiple-scattering problem"
229 \begin_layout Standard
232 \outcoeffp{\vect n,\alpha}-T_{\alpha}\sum_{\left(\vect m,\beta\right)\in\mathcal{P}\backslash\left\{ \left(\vect n,\alpha\right)\right\} }\tropsp{\vect n,\alpha}{\vect m,\beta}\outcoeffp{\vect m,\beta}=T_{\alpha}\rcoeffincp{\vect n,\alpha}.\quad\left(\vect n,\alpha\right)\in\mathcal{P}\label{eq:Multiple-scattering problem periodic}
240 \begin_layout Standard
241 Due to periodicity, we can also write
242 \begin_inset Formula $\tropsp{\vect n,\alpha}{\vect m,\beta}=\tropsp{\alpha}{\beta}\left(\vect R_{\vect m}-\vect R_{\vect n}\right)=\tropsp{\alpha}{\beta}\left(\vect R_{\vect m-\vect n}\right)=\tropsp{\vect 0,\alpha}{\vect m-\vect n,\beta}$
246 Assuming quasi-periodic right-hand side with quasi-momentum
247 \begin_inset Formula $\vect k$
251 \begin_inset Formula $\rcoeffincp{\vect n,\alpha}=\rcoeffincp{\vect 0,\alpha}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect n}}$
255 \begin_inset CommandInset ref
257 reference "eq:Multiple-scattering problem periodic"
264 will be also quasi-periodic according to Bloch theorem,
265 \begin_inset Formula $\outcoeffp{\vect n,\alpha}=\outcoeffp{\vect 0,\alpha}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect n}}$
270 \begin_inset CommandInset ref
272 reference "eq:Multiple-scattering problem periodic"
279 can be rewritten as follows
282 \outcoeffp{\vect 0,\alpha}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect n}}-T_{\alpha}\sum_{\left(\vect m,\beta\right)\in\mathcal{P}\backslash\left\{ \left(\vect n,\alpha\right)\right\} }\tropsp{\vect n,\alpha}{\vect m,\beta}\outcoeffp{\vect 0,\beta}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect m}} & =T_{\alpha}\rcoeffincp{\vect 0,\alpha}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect n}},\nonumber \\
283 \outcoeffp{\vect 0,\alpha}\left(\vect k\right)-T_{\alpha}\sum_{\left(\vect m,\beta\right)\in\mathcal{P}\backslash\left\{ \left(\vect n,\alpha\right)\right\} }\tropsp{\vect 0,\alpha}{\vect m-\vect n,\beta}\outcoeffp{\vect 0,\beta}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect m-\vect n}} & =T_{\alpha}\rcoeffincp{\vect 0,\alpha}\left(\vect k\right),\nonumber \\
284 \outcoeffp{\vect 0,\alpha}\left(\vect k\right)-T_{\alpha}\sum_{\left(\vect m,\beta\right)\in\mathcal{P}\backslash\left\{ \left(\vect 0,\alpha\right)\right\} }\tropsp{\vect 0,\alpha}{\vect m,\beta}\outcoeffp{\vect 0,\beta}\left(\vect k\right)e^{i\vect k\cdot\vect R_{\vect m}} & =T_{\alpha}\rcoeffincp{\vect 0,\alpha}\left(\vect k\right),\nonumber \\
285 \outcoeffp{\vect 0,\alpha}\left(\vect k\right)-T_{\alpha}\sum_{\beta\in\mathcal{P}}W_{\alpha\beta}\left(\vect k\right)\outcoeffp{\vect 0,\beta}\left(\vect k\right) & =T_{\alpha}\rcoeffincp{\vect 0,\alpha}\left(\vect k\right),\label{eq:Multiple-scattering problem unit cell}
290 so we reduced the initial scattering problem to one involving only the field
291 expansion coefficients from a single unit cell, but we need to compute
293 \begin_inset Quotes eld
296 lattice Fourier transform
297 \begin_inset Quotes erd
300 of the translation operator,
303 W_{\alpha\beta}(\vect k)\equiv\sum_{\vect m\in\ints^{d}}\left(1-\delta_{\alpha\beta}\right)\tropsp{\vect 0,\alpha}{\vect m,\beta}e^{i\vect k\cdot\vect R_{\vect m}},\label{eq:W definition}
308 evaluation of which is possible but quite non-trivial due to the infinite
309 lattice sum, so we explain it separately in Sect.
311 \begin_inset CommandInset ref
313 reference "subsec:W operator evaluation"
323 \begin_layout Standard
324 As in the case of a finite system, eq.
325 can be written in a shorter block-matrix form,
328 \left(I-WT\right)\outcoeffp{\vect 0}\left(\vect k\right)=\rcoeffincp{\vect 0}\left(\vect k\right)\label{eq:Multiple-scattering problem unit cell block form}
335 \begin_inset CommandInset ref
337 reference "eq:Multiple-scattering problem unit cell"
344 can be used to calculate electromagnetic response of the structure to external
345 quasiperiodic driving field – most notably a plane wave.
346 However, the non-trivial solutions of the equation with right hand side
348 the external driving) set to zero,
351 \left(I-WT\right)\outcoeffp{\vect 0}\left(\vect k\right)=0,\label{eq:lattice mode equation}
361 Non-trivial solutions to
362 \begin_inset CommandInset ref
364 reference "eq:lattice mode equation"
371 exist if the matrix on the left-hand side
372 \begin_inset Formula $M\left(\omega,\vect k\right)=\left(I-W\left(\omega,\vect k\right)T\left(\omega\right)\right)$
375 is singular – this condition gives the
379 for the periodic structure.
380 Note that in realistic (lossy) systems, at least one of the pair
381 \begin_inset Formula $\omega,\vect k$
384 will acquire complex values.
386 \begin_inset Formula $\outcoeffp{\vect 0}\left(\vect k\right)$
389 is then obtained as the right
390 \begin_inset Note Note
393 \begin_layout Plain Layout
400 \begin_inset Formula $M\left(\omega,\vect k\right)$
403 corresponding to the zero singular value.
406 \begin_layout Subsection
410 \begin_layout Standard
411 In practice, equation
412 \begin_inset CommandInset ref
414 reference "eq:Multiple-scattering problem unit cell block form"
421 is solved in the same way as eq.
423 \begin_inset CommandInset ref
425 reference "eq:Multiple-scattering problem block form"
432 in the multipole degree truncated form.
435 \begin_layout Standard
436 The lattice mode problem
437 \begin_inset CommandInset ref
439 reference "eq:lattice mode equation"
446 is (after multipole degree truncation) solved by finding
447 \begin_inset Formula $\omega,\vect k$
451 \begin_inset Formula $M\left(\omega,\vect k\right)$
454 has a zero singular value.
455 A naïve approach to do that is to sample a volume with a grid in the
456 \begin_inset Formula $\left(\omega,\vect k\right)$
459 space, performing a singular value decomposition of
460 \begin_inset Formula $M\left(\omega,\vect k\right)$
463 at each point and finding where the lowest singular value of
464 \begin_inset Formula $M\left(\omega,\vect k\right)$
467 is close enough to zero.
468 However, this approach is quite expensive, for
469 \begin_inset Formula $W\left(\omega,\vect k\right)$
472 has to be evaluated for each
473 \begin_inset Formula $\omega,\vect k$
476 pair separately (unlike the original finite case
477 \begin_inset CommandInset ref
479 reference "eq:Multiple-scattering problem block form"
487 \begin_inset Formula $\trops$
490 , which, for a given geometry, depends only on frequency).
491 Therefore, a much more efficient approach to determine the photonic bands
493 \begin_inset Formula $\vect k$
496 -space (a whole Brillouin zone or its part) and for each fixed
497 \begin_inset Formula $\vect k$
500 to find a corresponding frequency
501 \begin_inset Formula $\omega$
504 with zero singular value of
505 \begin_inset Formula $M\left(\omega,\vect k\right)$
508 using a minimisation algorithm (two- or one-dimensional, depending on whether
509 one needs the exact complex-valued
510 \begin_inset Formula $\omega$
513 or whether the its real-valued approximation is satisfactory).
514 Typically, a good initial guess for
515 \begin_inset Formula $\omega\left(\vect k\right)$
518 is obtained from the empty lattice approximation,
519 \begin_inset Formula $\left|\vect k\right|=\sqrt{\epsilon\mu}\omega/c_{0}$
522 (modulo lattice points; TODO write this a clean way).
523 A somehow challenging step is to distinguish the different bands that can
524 all be very close to the empty lattice approximation, especially if the
525 particles in the systems are small.
526 In high-symmetry points of the Brilloin zone, this can be solved by factorising
528 \begin_inset Formula $M\left(\omega,\vect k\right)$
531 into irreducible representations
532 \begin_inset Formula $\Gamma_{i}$
535 and performing the minimisation in each irrep separately, cf.
537 \begin_inset CommandInset ref
539 reference "sec:Symmetries"
546 , and using the different
547 \begin_inset Formula $\omega_{\Gamma_{i}}\left(\vect k\right)$
550 to obtain the initial guesses for the nearby points
551 \begin_inset Formula $\vect k+\delta\vect k$
557 \begin_layout Subsection
558 Computing the Fourier sum of the translation operator
559 \begin_inset CommandInset label
561 name "subsec:W operator evaluation"
568 \begin_layout Standard
569 The problem evaluating
570 \begin_inset CommandInset ref
572 reference "eq:W definition"
576 is the asymptotic behaviour of the translation operator,
577 \begin_inset Formula $\tropsp{\vect 0,\alpha}{\vect m,\beta}\sim\left|\vect R_{\vect b}\right|^{-1}e^{ik_{0}\left|\vect R_{\vect b}\right|}$
580 that does not in the strict sense converge for any
581 \begin_inset Formula $d>1$
584 -dimensional lattice.
585 \begin_inset Note Note
588 \begin_layout Plain Layout
592 \begin_layout Plain Layout
594 \begin_inset Formula $d$
597 here is dimensionality of the lattice, not the space it lies in, which
598 I for certain reasons assume to be three.
599 (TODO few notes on integration and reciprocal lattices in some appendix)
609 In electrostatics, this problem can be solved with Ewald summation [TODO
611 Its basic idea is that if what asymptoticaly decays poorly in the direct
612 space, will perhaps decay fast in the Fourier space.
613 We use the same idea here, but the technical details are more complicated
614 than in electrostatics.
617 \begin_layout Standard
618 Let us re-express the sum in
619 \begin_inset CommandInset ref
621 reference "eq:W definition"
625 in terms of integral with a delta comb
626 \begin_inset FormulaMacro
627 \renewcommand{\basis}[1]{\mathfrak{#1}}
633 \begin_layout Standard
636 W_{\alpha\beta}(\vect k)=\int\ud^{d}\vect r\dc{\basis u}(\vect r)S(\vect r_{\alpha}\leftarrow\vect r+\vect r_{\beta})e^{i\vect k\cdot\vect r}.\label{eq:W integral}
641 The translation operator
642 \begin_inset Formula $S$
645 is now a function defined in the whole 3d space;
646 \begin_inset Formula $\vect r_{\alpha},\vect r_{\beta}$
649 are the displacements of scatterers
650 \begin_inset Formula $\alpha$
654 \begin_inset Formula $\beta$
659 \begin_inset Formula $S(\vect r_{\alpha}\leftarrow\vect r+\vect r_{\beta})$
663 \begin_inset Quotes eld
666 translation operator for spherical waves originating in
667 \begin_inset Formula $\vect r+\vect r_{\beta}$
671 \begin_inset Formula $\vect r_{\alpha}$
675 \begin_inset Quotes erd
679 \begin_inset Formula $S$
682 is in fact a function of a single 3d argument,
683 \begin_inset Formula $S(\vect r_{\alpha}\leftarrow\vect r+\vect r_{\beta})=S(\vect 0\leftarrow\vect r+\vect r_{\beta}-\vect r_{\alpha})=S(-\vect r-\vect r_{\beta}+\vect r_{\alpha}\leftarrow\vect 0)=S(-\vect r-\vect r_{\beta}+\vect r_{\alpha})$
688 \begin_inset CommandInset ref
690 reference "eq:W integral"
697 W_{\alpha\beta}(\vect k)=\left(2\pi\right)^{\frac{d}{2}}\uaft{(\dc{\basis u}S(\vect{\bullet}-\vect r_{\beta}+\vect r_{\alpha}\leftarrow\vect 0))\left(\vect k\right)}
702 where changed the sign of
703 \begin_inset Formula $\vect r/\vect{\bullet}$
706 has been swapped under integration, utilising evenness of
707 \begin_inset Formula $\dc{\basis u}$
711 Fourier transform of product is convolution of Fourier transforms, so (using
713 \begin_inset CommandInset ref
715 reference "eq:Dirac comb uaFt"
719 for the Fourier transform of Dirac comb)
722 W_{\alpha\beta}(\vect k) & = & \left(\left(\uaft{\dc{\basis u}}\right)\ast\left(\uaft{S(\vect{\bullet}-\vect r_{\beta}+\vect r_{\alpha}\leftarrow\vect 0)}\right)\right)(\vect k)\nonumber \\
723 & = & \frac{\left|\det\recb{\basis u}\right|}{\left(2\pi\right)^{\frac{d}{2}}}\left(\dc{\recb{\basis u}}^{(d)}\ast\left(\uaft{S(\vect{\bullet}-\vect r_{\beta}+\vect r_{\alpha}\leftarrow\vect 0)}\right)\right)\left(\vect k\right)\nonumber \\
724 & = & \frac{\left|\det\rec{\basis u}\right|}{\left(2\pi\right)^{\frac{d}{2}}}\sum_{\vect K\in\recb{\basis u}\ints^{d}}\left(\uaft{S(\vect{\bullet}-\vect r_{\beta}+\vect r_{\alpha}\leftarrow\vect 0)}\right)\left(\vect k-\vect K\right)\label{eq:W sum in reciprocal space}\\
725 & = & \frac{\left|\det\rec{\basis u}\right|}{\left(2\pi\right)^{\frac{d}{2}}}\sum_{\vect K\in\recb{\basis u}\ints^{d}}e^{i\left(\vect k-\vect K\right)\cdot\left(-\vect r_{\beta}+\vect r_{\alpha}\right)}\left(\uaft{S(\vect{\bullet}\leftarrow\vect 0)}\right)\left(\vect k-\vect K\right)\nonumber
731 \begin_inset Note Note
734 \begin_layout Plain Layout
736 \begin_inset Formula $\left(2\pi\right)^{\frac{d}{2}}$
740 \begin_inset Formula $\left(2\pi\right)^{-\frac{d}{2}}$
743 factor appearing in the convolution/product formula in the unitary angular
750 As such, this is not extremely helpful because the the
755 \begin_inset Formula $S$
758 has singularities in origin, hence its Fourier transform
759 \begin_inset Formula $\uaft S$
766 \begin_layout Standard
767 However, Fourier transform is linear, so we can in principle separate
768 \begin_inset Formula $S$
772 \begin_inset Formula $S=S^{\textup{L}}+S^{\textup{S}}$
777 \begin_inset Formula $S^{\textup{S}}$
780 is a short-range part that decays sufficiently fast with distance so that
781 its direct-space lattice sum converges well;
782 \begin_inset Formula $S^{\textup{S}}$
785 must as well contain all the singularities of
786 \begin_inset Formula $S$
791 \begin_inset Formula $S^{\textup{L}}$
794 , will retain all the slowly decaying terms of
795 \begin_inset Formula $S$
798 but it also has to be smooth enough in the origin, so that its Fourier
800 \begin_inset Formula $\uaft{S^{\textup{L}}}$
804 (The same idea lies behind the Ewald summation in electrostatics.) Using
805 the linearity of Fourier transform and formulae
806 \begin_inset CommandInset ref
808 reference "eq:W definition"
813 \begin_inset CommandInset ref
815 reference "eq:W sum in reciprocal space"
820 \begin_inset Formula $W_{\alpha\beta}$
823 can then be re-expressed as
826 W_{\alpha\beta}\left(\vect k\right) & = & W_{\alpha\beta}^{\textup{S}}\left(\vect k\right)+W_{\alpha\beta}^{\textup{L}}\left(\vect k\right)\nonumber \\
827 W_{\alpha\beta}^{\textup{S}}\left(\vect k\right) & = & \sum_{\vect R\in\basis u\ints^{d}}S^{\textup{S}}(\vect 0\leftarrow\vect R+\vect r_{\beta}-\vect r_{\alpha})e^{i\vect k\cdot\vect R}\label{eq:W Short definition}\\
828 W_{\alpha\beta}^{\textup{L}}\left(\vect k\right) & = & \frac{\left|\det\rec{\basis u}\right|}{\left(2\pi\right)^{\frac{d}{2}}}\sum_{\vect K\in\recb{\basis u}\ints^{d}}\left(\uaft{S^{\textup{L}}(\vect{\bullet}-\vect r_{\beta}+\vect r_{\alpha}\leftarrow\vect 0)}\right)\left(\vect k-\vect K\right)\label{eq:W Long definition}
833 where both sums expected to converge nicely.
834 We note that the elements of the translation operators
835 \begin_inset Formula $\tropr,\trops$
839 \begin_inset CommandInset ref
841 reference "eq:translation operator"
848 can be rewritten as linear combinations of expressions
849 \begin_inset Formula $\ush{\nu}{\mu}\left(\uvec d\right)j_{n}\left(d\right),\ush{\nu}{\mu}\left(\uvec d\right)h_{n}^{(1)}\left(d\right)$
852 (TODO WRITE THEM EXPLICITLY IN THIS FORM), respectively, hence if we are
853 able evaluate the lattice sums sums
854 \begin_inset Note Note
857 \begin_layout Plain Layout
858 CHECK THE FOLLOWING EXPRESSION FOR CORRECT FUNCTION ARGUMENTS
866 \sigma_{\nu}^{\mu}\left(\vect k\right)=\sum_{\vect n\in\ints^{d}\backslash\left\{ \vect 0\right\} }e^{i\vect{\vect k}\cdot\vect R_{\vect n}}\ush{\nu}{\mu}\left(\uvec{R_{n}}\right)h_{n}^{(1)}\left(R_{n}\right),\label{eq:sigma lattice sums}
871 then by linearity, we can get the
872 \begin_inset Formula $W_{\alpha\beta}\left(\vect k\right)$
878 \begin_layout Standard
879 TODO ADD MOROZ AND OTHER REFS HERE.
881 \begin_inset CommandInset citation
883 key "linton_one-_2009"
888 offers an exponentially convergent Ewald-type summation method for
889 \begin_inset Formula $\sigma_{\nu}^{\mu}\left(\vect k\right)=\sigma_{\nu}^{\mu(\mathrm{S})}\left(\vect k\right)+\sigma_{\nu}^{\mu(\mathrm{L})}\left(\vect k\right)$
893 Here we rewrite them in a form independent on the convention used for spherical
894 harmonics (as long as they are complex
895 \begin_inset Note Note
898 \begin_layout Plain Layout
905 The short-range part reads (UNIFY INDEX NOTATION)
908 \sigma_{n}^{m(\mathrm{S})}\left(\vect{\beta}\right)=-\frac{2^{n+1}i}{k^{n+1}\sqrt{\pi}}\sum_{\vect R\in\Lambda}^{'}\left|\vect R\right|^{n}e^{i\vect{\beta}\cdot\vect R}Y_{n}^{m}\left(\vect R\right)\int_{\eta}^{\infty}e^{-\left|\vect R\right|^{2}\xi^{2}}e^{-k/4\xi^{2}}\xi^{2n}\ud\xi\\
909 +\frac{\delta_{n0}\delta_{m0}}{\sqrt{4\pi}}\Gamma\left(-\frac{1}{2},-\frac{k}{4\eta^{2}}\right)Y_{n}^{m},\label{eq:Ewald in 3D short-range part}
915 \begin_inset Note Note
918 \begin_layout Plain Layout
919 NEPATŘÍ TAM NĚJAKÁ DELTA FUNKCE K PŮVODNÍMU
920 \begin_inset Formula $\sigma_{n}^{m(0)}$
928 and the long-range part (FIXME, this is the 2D version; include the 1D and
929 3D lattice expressions as well)
932 \sigma_{n}^{m(\mathrm{L})}\left(\vect{\beta}\right)=-\frac{i^{n+1}}{k^{2}\mathscr{A}}\sqrt{\pi}2\left(\left(n-m\right)/2\right)!\left(\left(n+m\right)/2\right)!\times\\
933 \times\sum_{\vect K_{pq}\in\Lambda^{*}}^{'}Y_{n}^{m}\left(\frac{\pi}{2},\phi_{\vect{\beta}_{pq}}\right)\sum_{j=0}^{\left[\left(n-\left|m\right|/2\right)\right]}\frac{\left(-1\right)^{j}\left(\beta_{pq}/k\right)^{n-2j}\Gamma_{j,pq}}{j!\left(\frac{1}{2}\left(n-m\right)-j\right)!\left(\frac{1}{2}\left(n+m\right)-j\right)!}\left(\gamma_{pq}\right)^{2j-1}\label{eq:Ewald in 3D long-range part}
939 \begin_inset Formula $\xi$
943 \begin_inset Formula $\beta_{pq}$
947 \begin_inset Formula $\Gamma_{j,pq}$
951 \begin_inset Formula $\eta$
954 is a real parameter that determines the pace of convergence of both parts.
956 \begin_inset Formula $\eta$
960 \begin_inset Formula $\sigma_{n}^{m(\mathrm{S})}\left(\vect{\beta}\right)$
963 converges but the slower
964 \begin_inset Formula $\sigma_{n}^{m(\mathrm{L})}\left(\vect{\beta}\right)$
968 Therefore (based on the lattice geometry) it has to be adjusted in a way
969 that a reasonable amount of terms needs to be evaluated numerically from
971 \begin_inset Formula $\sigma_{n}^{m(\mathrm{S})}\left(\vect{\beta}\right)$
975 \begin_inset Formula $\sigma_{n}^{m(\mathrm{L})}\left(\vect{\beta}\right)$
979 Generally, a good choice for
980 \begin_inset Formula $\eta$
983 is TODO; in order to achieve accuracy TODO, one has to evaluate the terms
984 on TODO lattice points.
985 (I HAVE SOME DERIVATIONS OF THE ESTIMATES IN MY NOTES; SHOULD I INCLUDE
989 \begin_layout Standard
990 In practice, the integrals in
991 \begin_inset CommandInset ref
993 reference "eq:Ewald in 3D short-range part"
1000 can be easily evaluated by numerical quadrature and the incomplete
1001 \begin_inset Formula $\Gamma$
1004 -functions using the series TODO and TODO from DLMF.