cond-mat0412681/PK.tex
1: 
2: \documentclass[
3: %preprint, 
4: aps,prb,
5: twocolumn,
6: groupedaddress]{revtex4}
7: \input{epsf}
8: \usepackage{epsfig} 
9: \usepackage{amsfonts}
10: \usepackage{pifont}
11: \usepackage{epstopdf}
12: \usepackage{epsfig}
13: \usepackage{amsfonts}
14: 
15: \def\mts2f{${\cal M}({\cal T}^{*(2F)})$}
16: \def\mtsa4{${\cal M}({\cal T}^{*(A_4)})$}
17: \def\ms{${\cal M}^{s}$}
18: \def\msr{${\cal M}^{s_r}$}
19: \def\msf{${\cal M}^{s_f}$}
20: 
21: \def\mf{$M_F$}
22: \def\mi{$M_I$}
23: 
24: \def\tp{${\cal T}^{(P)}$}
25: \def\ts2f{${\cal T}^{*(2F)}$}
26: \def\tsq{${\cal T}^{*(z)}$}
27: \def\cstsa4{${\cal C}^s_{{\cal T}^{*(A_4)}}$}
28: \def\tsa4{${\cal T}^{*(A_4)}$}
29: 
30: \def\ep{$\mathbb{E}_\parallel$}
31: \def\es{$\mathbb{E}_\perp$}
32: \def\zsix{${\mathbb{Z}}_6$}
33: 
34: \def\qdsix{$q_{D_6}$}
35: \def\dsix{$D_6$}
36: \def\d*six{$D^*_6$}
37: \def\ih{ $I_h$ }
38: 
39: \usepackage{times}
40: \usepackage{pifont}
41: 
42: \def\zfo{\ding{193}}
43: %Achse zwei
44: \def\dfo{\ding{194}}
45: %Achse drei
46: \def\ffo{\ding{196}}
47: %Achse fuenf
48: %%%
49: 
50: 
51: \begin{document}
52: 
53: 
54: \title{ 
55: Thick atomic layers of maximum density as 
56: bulk terminations of quasicrystals 
57: }
58: 
59: \author{Zorka Papadopolos}
60: \email[]{zorka.papadopolos@uni-tuebingen.de}
61: 
62: \affiliation{Institut f\"ur Theoretische Physik, 
63: Univ. T\"ubingen, D-72076 T\"ubingen, Germany}
64: 
65: \author{Gerald Kasner}
66: \affiliation{Institut f\"ur Theoretische Physik, 
67: Univ. Magdeburg, PF 4120, D-39016
68:   Magdeburg, Germany}
69: 
70: 
71: \date{\today}
72: 
73: \begin{abstract}
74: The clean surfaces of quasicrystals, 
75: orthogonal to the directions of the main symmetry axes, 
76: have a terrace-like appearance. 
77: We extend the Bravais' rule for crystals 
78: to quasicrystals, allowing that instead of a
79: \emph{ single} atomic plane 
80: a \emph{ layer} of atomic planes may form a bulk 
81: termination.
82: 
83: \end{abstract}
84: 
85: % insert suggested PACS numbers in braces on next line
86: \pacs{
87: %Quasicrystals:
88:       61.44.Br, 
89: %Structure of clean surfaces (reconstruction):
90:       68.35.Bs, 
91: %Scanning tunneling microscopy (including chemistry 
92: %induced with STM):
93:       68.37.Ef, 
94: %Electron diffraction and scattering (for 
95: %electron diffractometers, see 07.78.+s):
96:       61.14.-x 
97:   }
98: 
99: \maketitle
100: 
101: 
102: 
103: We consider clean surfaces of quasicrystals
104: by comparing their STM (scanning tunneling microscopy)
105: and SEI (secondary electron imaging) images 
106: to the bulk terminations in the deterministic atomic 
107: models. 
108: A bulk model is defined as an ideal quasiperiodic 
109: tiling with atomic decoration on a finite number of 
110: tiles. In Section~\ref{sec:decagonal}, 
111: an STM image of a decagonal quasicrystal is compared 
112: to the Burkov model\cite{Bur93} \mtsa4, 
113: based on a decagonal tiling \tsa4, 
114: see Ref.\cite{Baake90}. 
115: In section~\ref{sec:ico}, STM and SEI images of an 
116: icosahedral quasicrystal are compared to the 
117: model~\cite{Aper97}
118: \mts2f, based on icosahedral tiling  
119: \ts2f, see Refs~\cite{
120: ts2f,
121: PHK}. 
122: The clean surfaces, that we study, are considered not 
123: to be reconstructed.
124: 
125: In our previous work on  terrace-like clean surfaces 
126: of icosahedral quasicrystals we tried to define 
127: the 5fold terminations over the most dense layers of 
128: Bergman polytopes~\cite{KPKB},
129: which are the icosahedral motifs on a 
130: quasilattice \ts2f.  
131: Later, we noticed that the
132: 5fold terminations should better be defined by the 
133: 0.48~{\AA} thin, plane-like atomic layers, all of equal, 
134: maximum density~\cite{LGPK2}. 
135: In Ref.~\cite{PP2} we adopted  Bravais' rule
136: of maximum density, 
137: allowing that also the densities of thin 
138: ($< 0.6$~{\AA} broad), plane-like atomic layers in 
139: quasicrystals should be treated as the planes. 
140: The idea we developed in Ref. \cite{PP2}, and gave 
141: a possible explanation of the experimental 
142: fact, that although the 2fold planes in the model 
143: \mts2f\ are more dense than the 5fold planes, 
144: the 5fold surfaces are the most stable in 
145: icosahedral quasicrystals \cite{Shen00}.
146: In Ref.~\cite{PP2}, in which we used the ``thin'' 
147: layer concept in order to determine the terminations
148: in all principle symmetry directions, we announced the 
149: ``thick'' layer concept as well, that we elaborate in 
150: the present manuscript.
151: 
152: 
153: 
154: \section{Surfaces of decagonal quasicrystals
155: \label{sec:decagonal}}
156: 
157: 
158: Decagonal Al$_{65}$Cu$_{15}$Co$_{20}$ (d-AlCuCo) is 
159: periodic in $z$-direction~\cite{Kor90}, 
160: with periodicity t=4.13~{\AA}. The  $z$-direction is
161: orthogonal to the quasiperiodic decagonal $x-y$ plane. 
162: The phase crystallizes in a shape of long thin 
163: decagonal prisms, with the 2fold surfaces considerably 
164: larger than the 10-fold surfaces, 
165: see Fig.~\ref{fig:deca}(a).
166: %
167: %
168: % Fig. 1
169: %
170: \begin{figure}[]
171: \includegraphics[width=0.27\textwidth]
172: {PK-Fig1a.eps}
173: \includegraphics*[width=0.30\textwidth]
174: {PK-Fig1b.eps}
175: \includegraphics*[width=0.40\textwidth]
176: {PK-Fig1c.eps}
177:  \caption{
178:  (a)~(top) Long, thin decagonal prisms~\cite{K0} 
179:            of the d-AlCuCo;
180:  (b)~(middle) Estimated positions of 2fold surfaces 
181:         on an STM image of the 10fold terrace-like 
182:          surface of d-AlCuCo;
183:  (c)~(bottom) 2fold bulk terminations in \mtsa4\  
184:          marked on a 10fold one~\cite{Kram95}. 
185:       Note the $1.23$~{\AA} gaps on both sides of 
186:       a termination. 
187: \label{fig:deca}
188: }
189: \end{figure}
190: The size of the surface area is evidently a parameter 
191: of it's stability. 
192: In Fig.~\ref{fig:deca}(b), over the image of the 10fold 
193: surface \cite{Kor90}, we determine the positions of 
194: possible 2fold surfaces, orthogonal to the 10fold 
195: surface. 
196: In the model \mtsa4\ (t=4.18~{\AA})~\cite{Bur93} we 
197: investigate the densities of the ``thin'', 2fold layers 
198: containing 2 atomic planes. 
199: Among these, the most dense one is a $0.47$~{\AA}-layer 
200: of the density $\rho_{2}^{0.47A}=0.124$~{\AA}$^{-2}$.
201: Comparing it to the much smaller 10fold surface, 
202: which is of the density $\rho_{10}=0.146$~{\AA}$^{-2}$, 
203: we conclude that the $0.47$~{\AA}-layer can not 
204: represent the 2fold termination. But, on some positions
205: in the bulk these layers appear in 
206: pairs, $0.29$~{\AA} apart. Such, a rather ``thick'' 
207: $1.23$~{\AA}-layer of 4 planes on mutual
208: small distances (see Fig.~\ref{fig:deca}(c)) is a 
209: candidate for a terminating layer. These layers appear 
210: on distances mutually scaled by the factor 
211: $\tau=(1+\sqrt{5})/2$ ($12.3$~{\AA}, $19.9$~{\AA}, 
212: $32.2$~{\AA}, $52.1$~{\AA}\ldots, 
213: as in Fig.~\ref{fig:deca}(c)), in excellent agreement 
214: with those found on the STM image ($21$~{\AA}, 
215: $31$~{\AA}, $54$~{\AA}, see Fig.~\ref{fig:deca}(b)).  
216: 
217: 
218: 
219: 
220: 
221: \section{ Surfaces of icosahedral quasicrystals
222: \label{sec:ico}}
223: 
224: A feature of the surface of \mbox{i-AlPdMn}, not
225: accounted for by the ``thin'' layer analysis \cite{PP2}, 
226: is that not all types of maximally dense layers appear 
227: as surfaces:
228: for example, $(q,b)$ layers~\cite{classes}, 
229: $0.48$~{\AA} apart, are seen in 5fold
230: surfaces but equally dense $(b,q)$ layers, 
231: also $0.48$~{\AA} apart, 
232: are not. If both kind of layers were possible 
233: terminations, the sequence of much 
234: shorter terrace heights, than observed, could appear. 
235: %
236: % Fig. 2
237: %
238: \begin{figure}[]
239: \includegraphics*[width=0.48\textwidth]
240: {PK-Fig2a.eps}
241: \includegraphics*[width=0.48\textwidth]
242: {PK-Fig2b.eps}
243: \caption{
244: (a)~(top)
245: Density graph $\rho_{5f}(z_{\perp})$ of the ``thick'' 
246: 5fold layers $(q,b,b,q)$ (full line). The plateau 
247: (maximum density) of the graph defines the terminations. 
248: It is compared to the density graph of the $0.48$~{\AA} 
249: thin, plane-like layer $(q,b)$ (dotted line).
250: The symbol \ffo\ is the standard distance along 
251: a 5fold axis $z_{\perp}$ in the coding space \es.
252: (b)~(bottom) Density graph of the 
253: 5fold layers $(b,a,q,b,q,a,b)$, with spacings as in the
254: image. 
255: The bottom of the 
256: cavity
257: defines a sequence of the 
258: {\em minimum density} layers in \mts2f, situated above 
259: a {\em subsequence} of the
260:  5fold terminations.  
261: \label{fig:5f-thick-term}
262: }
263: \end{figure}
264: If one chooses to define a
265: termination incorporating the neighboring planes too,
266: as we did in decagonal case 
267: (in Section~\ref{sec:decagonal}),
268: one could introduce a ``thick'' layer as a bundle of
269: high density planes (or thin, plane-like layers).  
270: A 5fold termination can be considered to
271: be a ``thick'' layer consisting of a 
272: $(q,b)$  layer and a $(b,q)$ layer, 
273: each with the spacing $0.48$~{\AA}.
274: Such a layer contains 4 planes with
275: spacings:
276: $q$-plane, $0.48$~{\AA}, $b$-plane, $1.56$~{\AA}, 
277: $b$-plane, $0.48$~{\AA}, $q$-plane.  
278: For a bundle we define an effective
279: (averaged) density of within contained ``thin'' 
280: layers/planes
281: 
282: \vspace{0.3cm}
283: \noindent
284: { \small
285: $\rho_{5f}(z_{\perp})=({\rho}_{q_1}(z_{\perp})+
286: {\rho}_{b_1}(z_{\perp}))/2 +
287: ({\rho}_{q_2}(z_{\perp})+
288: {\rho}_{b_2}(z_{\perp}))/2$
289: }.
290: 
291: \vspace{0.3cm}
292: \noindent
293: As we see in Fig.~\ref{fig:5f-thick-term}(a), 
294: whereas for the thin-layer concept the width of the
295: support of the plateau is approximately 
296: $\frac{2{\tau}^2 }{\tau+2}$\ffo\ broad,  
297: and consequently~\cite{PP2} encodes the 
298: Fibonacci sequence of terrace heights $S=4.08$~{\AA} 
299: and $L=\tau S=6.60$~{\AA} 
300: ($\tau=(1+\sqrt{5})/2$),
301: in the thick-layer concept the width is 
302: exactly
303: $\frac{2{\tau} }{\tau+2}$\ffo\  
304: (see  Fig.~\ref{fig:5f-thick-term}(a)) 
305: and encodes the Fibonacci sequence of by factor $\tau$ 
306: larger terrace heights, i.e. $L=6.60$~{\AA} and 
307: $L+ S=\tau L=10.68$~{\AA}. 
308: Whereas on the clean surfaces, obtained at lower 
309: annealing temperature, even the terrace height 
310: $\tau^{-1}S=2.52$~{\AA} appears~\cite{Shen99}, 
311: on the 
312: surfaces obtained at the highest 
313: annealing temperature, on the contrary, 
314: a terrace height $L+ S$ appears   
315: (see Fig.~1 in Ref.~\cite{KPKB}). 
316: 
317: 
318: The height of the plateau  of the graph 
319: $\rho_{5f}$ (see Fig.~\ref{fig:5f-thick-term}(a)) 
320: defines the densities of the ``thick'' layer 
321: terminations to be $0.134$~{\AA}$^{-2}$, 
322: see table~\ref{tab:densities}.
323: The 5fold layers intertwining the terminations are 
324: of densities not higher than 
325: $0.072$~{\AA}$^{-2}(\ll 0.134$~{\AA}$^{-2}$). 
326: It is also a fact, that the density graphs of the 
327: ``thin'' and the ``thick'' layers have a strong overlap,
328: see Fig.~\ref{fig:5f-thick-term}(a).
329: Hence, almost any  $(q,b)$ ``thin'' layer termination 
330: occurs within such a $(q,b,b,q)$ ``thick'' layer 
331: termination. 
332: The same holds true for any icosahedral quasicrystal 
333: described by the \mts2f\ model. 
334: 
335: 
336: 
337: Above each termination, there is a $2.04$~{\AA} gap,
338: if we dare to neglect an $a$-plane of a density
339: smaller than $0.013$~{\AA}$^{-2}$. 
340: But, if each gap of $2.04$~{\AA} 
341: in the model \mts2f\ would be declared as 
342: a criterion of a termination to appear below it, 
343: as in Ref.~\cite{Y04}, the $2.04$~{\AA} terrace heights, 
344: that were not observed, should appear as well. 
345: However in the model \mts2f\ there is a 
346: low density 5fold layer $(b,a,q,b,q,a,b)$ 
347: (see Fig.~\ref{fig:5f-thick-term}(b)), 
348: $4.08$~{\AA} broad. 
349: The width of the
350: cavity 
351: on the density graph of these layers, 
352: $W=(2/(\tau +2))$\ffo, encodes a Fibonacci
353: sequence with the intervals 
354: $\tau L=10.68$~{\AA} and  $\tau^2 L=17.28$~{\AA}. 
355: These are the {\em minimum density} layers
356: of equal, $0.041$~{\AA}$^{-2}$ density, 
357: placed in the model over a {\em subsequence} of 
358: the terminations.
359: Hence, the minimum density layer sequence 
360: alone can not define
361: the terminations, because 
362: it does not reproduce 
363: the pairs of large terraces $6.60$~{\AA} 
364: apart, which were frequently observed.
365: 
366: 
367: 
368: 
369: 
370: %
371: % Fig. 3
372: %
373: \begin{figure}[]
374: \includegraphics*[width=0.48\textwidth]
375: {PK-Fig3a.eps}
376: \includegraphics*[width=0.48\textwidth]
377: {PK-Fig3b.eps}
378:   \caption{
379: (a)~(top) 
380: Density graph $\rho_{2f}(z_{\perp})$ 
381: of the ``thick'' 2fold
382: layers (full line). The plateau (maximum density) 
383: of the graph defines the terminations. It is compared 
384: to the ``thin'' layer termination (dotted line), 
385: which is a single, $abq$-plane termination. 
386: The symbol \zfo\ is the standard distance 
387: along a 2fold axis $z_{\perp}$ in the coding space
388: \es. 
389: (b)~(bottom) 
390: Density graph of the 
391: 2fold layer $(bq,bq)$. The cavity
392: defines a sequence of the {\em  minimum density} layers
393: in \mts2f,
394: of which a {\em subsequence} is situated above 
395: {\em all} the 2fold terminations.  
396: \label{fig:2f-thick-term}
397: }
398: \end{figure}
399: In the case of 2fold surfaces, we may replace a single 
400: dense 2fold terminating atomic
401: $abq$-plane~\cite{PP2} by a layer of 4 atomic planes
402: with spacings:
403: $abq$-plane, $1.48$~{\AA}, $bq$-plane, $0.92$~{\AA},
404: $bq$-plane, $1.48$~{\AA}, $abq$-plane. 
405: For a bundle we define an effective
406: (averaged) density of planes
407: 
408: \noindent
409: \vspace{0.3cm}
410: \noindent
411: {\small$\rho_{2f}(z_{\perp})=
412: (1/4)[{\rho}_{abq_1}(z_{\perp})+
413:  {\rho}_{bq_1}(z_{\perp})+
414:  {\rho}_{bq_2}(z_{\perp})+
415:  {\rho}_{abq_2}(z_{\perp})$]},
416: 
417: \vspace{0.3cm}
418: \noindent
419: For this ``thick'' 2fold layer
420: the peak of $\rho_{2f}(z_{\perp})$ is a perfectly 
421: flat plateau, see Fig.~\ref{fig:2f-thick-term}(a).
422: The height of the plateau defines the effective
423: densities of terminations to be $0.086$~{\AA}$^{-2}$, 
424: see Table~\ref{tab:densities}. 
425: The support of the width of the plateau equals  
426: $W=(1/2)$\zfo\ and encodes the 
427: Fibonacci sequence of 2fold terminations with the 
428: terrace heights $S=6.3$~{\AA} and 
429: $L=\tau S=10.2$~{\AA}. 
430: The standard distances along 5, 2 and 3fold axes 
431: in icosahedral structures are 
432: \ffo~/$\sqrt{\tau +2}={}$~\zfo~/2$(=1/\sqrt{2(\tau +2)})
433: ={}$~\dfo~$/\sqrt{3}$, where $\tau=(1+\sqrt{5})/2$.
434: The heights of the larger 2fold terraces were 
435: measured to be
436: $S=
437: %(\tau^2/2)$~\zfo${}=
438: 6.2$~{\AA} and $L=9.5$~{\AA}
439: (see Fig.~3(a) and 3(c) in Ref.~\cite{PP2}),
440: in good agreement with the predicted values.
441: The 2fold layers intertwining the terminations are 
442: of densities not higher than 
443: $0.079$~{\AA}$^{-2}(<0.086$~{\AA}$^{-2}$). 
444: The small 2fold terraces, or rather the 
445: pits within the big terraces  
446: (see Fig.~3(a) and 3(c) in Ref. \cite{PP2})
447: may be explained by the comparatively large distances 
448: between the atomic planes inside of the ``thick''
449: terminating layer ($1.5$~{\AA}, $2.4$~{\AA} and 
450: $3.9$~{\AA}).
451: These excellently reproduce the measured values
452: ($2.4$~{\AA} and $3.6$~{\AA}), 
453: see Fig.3(a) and 3(c) in Ref. \cite{PP2}.
454: 
455: In the model \mts2f\ there is a low density 
456: 2fold layer 
457: $(bq,bq)$: 
458: $0.92$~{\AA} gap, $bq$-plane, $0.57$~{\AA},
459: $bq$-plane, $0.92$~{\AA} gap
460: (see Fig.~\ref{fig:2f-thick-term}(b)). 
461: The width of the
462: cavity
463: on the graph of these layers, 
464: $W=(\tau/2)$\zfo, encodes a Fibonacci
465: sequence with the intervals 
466: $\tau^{-1} S=3.9$~{\AA} and $S=6.3$~{\AA}. 
467: These are the {\em minimum density} layers
468: of almost equal density, 
469: somewhat above $0.063$~{\AA}$^{-2}$.
470: A member of a {\em subsequence} of these layers 
471: is placed over  {\em each} 2fold termination.
472: %that allow also a $4.08$~{\AA} terrace heights.
473: Nevertheless, in the 2fold case the 
474: minimum density layer sequence {\em alone} 
475: can not define the terminations, because it predicts 
476: by $\tau^{-1}$ shorter terrace heights between the 
477: large 2fold terminations, than observed. 
478: 
479: 
480: The 3fold terminations could also be modeled
481: as ``thick'' layers of atomic planes in \mts2f, 
482: see table~\ref{tab:densities}.
483: But, inspecting the intertwining 3fold layers,
484: we see that these are of the densities {\em comparable} 
485: to the ``terminating'' ones.  
486: We also know that the 3fold surfaces facet readily 
487: \cite{Shen00}, 
488: and some correlated STM measurements, 
489: (as those in Fig.3(a) and 3(c) of Ref. \cite{PP2})
490: for the 3fold surfaces do not exist so far.
491: %
492: % Table I
493: %
494: \begin{table}[h]
495: \begin{center}
496: \caption{ 
497: Relative and absolute densities of the planes
498: and layer terminations orthogonal to 5, 2 and 3fold 
499: symmetry axes in \mts2f\  of \mbox{i-AlPdMn}.
500: The corresponding data for \mbox{i-AlCuFe} are similar. 
501: \label{tab:densities}
502: }
503: \begin{tabular}{l|lll}
504: \hline
505:    & 5fold & 2fold & 3fold\\
506: \hline
507: Densest planes (abs.) & 0.086\,{\AA}$^{-2}$&
508: 0.101\,{\AA}$^{-2}$&0.066\,{\AA}$^{-2}$ \\%[1mm]
509: Densest ``thin'' layers (abs.) &0.133\,{\AA}$^{-2}$&
510: 0.101\,{\AA}$^{-2}$&0.066\,{\AA}$^{-2}$\\
511: Densest ``thin'' layers (rel.) & 1 & 0.76 & 0.50\\
512: Densest ``thick'' layers (abs.) &0.134\,{\AA}$^{-2}$&
513: 0.086\,{\AA}$^{-2}$&0.058\,{\AA}$^{-2}$\\
514: Densest ``thick'' layers (rel.) & 1 & 0.64 & 0.44\\
515: \hline
516: \end{tabular}
517: \end{center}
518: \end{table}
519: 
520: 
521: On the STM measurements it is in general 
522: hard to judge whether ``thin'' layer or ``thick''  
523: layer terminations best model the physical surfaces.  
524: However, in the 5fold case, the ``thick'' layer concept 
525: removes the contradiction with respect to the 
526: Bravais' rule, that some, equally dense layers 
527: do not appear on the surfaces, see Ref.~\cite{PP2}.
528: In the 2fold case the ``thick'' layer concept is 
529: evidently better, it treats differently
530: the large terraces compared to the small pits inside. 
531: That the effective densities of the 2fold terminations 
532: are somewhat lower than the densities of some single 
533: 2fold planes is not contradictory, because these are 
534: included in the most dense layers.   
535: 
536: 
537: Fig.~\ref{fig:compSEI}(a)
538: shows the secondary-electron pattern 
539: %\cite{erbu,SEI} 
540: \cite{B04}
541: obtained from the clean pentagonal surface of a 
542: quasicrystalline 
543: Al$_{70}$Pd$_{20}$Mn$_{10}$ 
544: sample~\cite{K0}. 
545: %
546: % Fig. 4
547: %
548: \begin{figure}
549: \includegraphics[width=0.465\textwidth]
550: {PK-Fig4.eps}
551:   \caption{
552:    (a)~(left) Secondary-electron pattern obtained from 
553:        the pentagonal surface of a single icosahedral 
554:        Al$_{70}$Pd$_{20}$Mn$_{10}$ quasicrystal. 
555:        The center of the pattern is obscured by the 
556:        shadow of the electron gun used for the 
557:        excitation. The edge of the screen corresponds 
558:        to~\cite{B04} $\theta=52^\circ$.  
559: (b)~(right) Calculated secondary-electron pattern based 
560:     on the single scattering approximation of electrons 
561:     using model~\cite{E03} \mts2f.
562: \label{fig:compSEI}
563: }
564: \end{figure}
565: Secondary-electron images 
566: (SEI) represent an orthogonal 
567: projection to the sphere of the symmetry directions 
568: below a near-surface region 
569: of the sample. 
570: Apart from the icosahedral symmetry of the pattern, 
571: some groups of bright patches are seen to lie 
572: within bands, similar to Kikuchi bands 
573: %\cite{stern}, 
574: \cite{B04} connecting the 
575: 2fold-, 3fold-, and 5fold-symmetry directions.
576: In case of crystals these
577: bands are a direct consequence of well-defined dense 
578: {\it planes} 
579: of atoms.
580: The bands observed by a quasicrystal 
581: should be also a consequence of dense planes or  
582: plane-like layers within the quasicrystal, which, 
583: according to the secondary-electron pattern, 
584: lie  perpendicular to the principal directions of the 
585: icosahedron. 
586: Hence, the pattern in Fig.~\ref{fig:compSEI}(a) 
587: carries an
588: information of the long-range order. 
589: We note that there are bands 
590: perpendicular to 2fold- and 5fold-symmetry directions, 
591: but not to 3fold-symmetry directions. 
592: 
593: 
594: 
595: 
596: A quantum mechanical single-site scattering calculation 
597: %\cite{fadley} 
598: \cite{B04} is a faithful representation of the SEI 
599: pattern because it accounts for the wave nature of the 
600: secondary electrons. 
601: Fig.~\ref{fig:compSEI}(b) illustrates the results of 
602: the calculation~\cite{E03} using the coordinate set 
603: of the model \mts2f. This approximation overestimates 
604: the scattering intensity along chains of atoms 
605: %\cite{egel}, 
606: \cite{B04}
607: but suffices for 
608: purposes, since our interest is 
609: mainly in the presence or absence of bands on the 
610: screen. 
611: As in case of crystals \cite{B04} the band width 
612: is inversely proportional to the spacing of 
613: crystallographic planes, 
614: and the band width is related to the distance 
615: of interatomic planes 
616: %\cite{stern},
617: \cite{B04}.
618: The observed band widths reveal that the 
619: inter-planar distances of the highly dense 
620: 2fold planes are broader than these of the 
621: 5fold planes/plane-like layers 
622: by a factor  
623: 1.6.
624: In 5fold case, from the model we predict the distance  
625: $d_5=\tau/(\tau+2)$\ffo$=2.04$~{\AA}, which is the 
626: distance between the highly dense $(q,b)$ and the 
627: $(b,q)$ plane-like layers in the terminating 
628: thick layer $(q,b,b,q)$: 
629: $2.04$~{\AA}$=0.48$~{\AA}$+1.56$~{\AA}, 
630: see also Ref.~\cite{Z04}.
631: In the 2fold case, between the highly dense planes 
632: in the above defined ``thick'' layer, appears
633: once the distance of 0.92~{\AA} and twice the distance 
634: of 1.48~{\AA}. 
635: Hence, an average distance between the highly dense 
636: planes is 
637: $d_2=\frac{\tau}{3\sqrt{\tau+2}}$\zfo$=1.29$~{\AA}. 
638: And their ratio is $d_5/d_2=3/\sqrt{\tau+2}
639: \approx 1.6$. 
640: 
641: The SEI method does not determine the bulk termination, 
642: (it is testing the bulk circa 30~{\AA} below 
643: the surface). However, 
644: in the case of the ordinary 
645: crystals, the Kikuchi bands are related to the most 
646: dense atomic planes, and we show that the same holds 
647: true in the case of quasicrystals as well. 
648: %see also~\cite{B04}. 
649: Hence we may claim that SEI images are supporting the 
650: thick atomic layers of the high effective 
651: density to be the bulk terminations, if the Bravais' 
652: rule should be valid in quasicrystals as well.  
653: The not existing 3fold Kikuchi bands  are also 
654: supporting 
655: the model \mts2f, in which we find that the 
656: atoms collected by the 3fold  planes are almost 
657: uniformly distributed among these, without the 
658: notable repetitive layers of higher densities.     
659: 
660: \begin{thebibliography}{99}
661: 
662: 
663: \bibitem{Bur93} S.~E.~Burkov,
664:  %Enforcement of matching rules by chemical ordering 
665:  %in the decagonal AlCuCo quasicrystals,
666:  Phys.~Rev.~B,
667:  {\bf 47}, R12325
668:  %-12328,
669:  (1993).
670: 
671:  
672:  \bibitem{Baake90} M.~Baake,
673:           P.~Kramer, M.~Schlottmann 
674:           and D.~Zeidler,
675:  %Planar Patterns with Fivefold  Symmetry as Sections 
676:  %of Periodic  Structures in 4-space,
677:  Int.~J.~Mod.~Phys.~B {\bf 4}, 2217
678:  %-2268
679: (1990).
680: 
681: 
682: \bibitem{Aper97}  Z.~Papadopolos, 
683:                 P.~Kramer and  W.~Liebermeister,
684:  %{\em Atomic Positions for the Icosahedral F-Phase 
685:  %Tiling},
686:  in {\em Proc. of the Int. Conf. on 
687:                Aperiodic Crystals, Aperiodic 1997},
688:                %173-181,
689:   edited by M.~de~Boissieu,
690: J.-L.~Verger-Gaugry and 
691:                R.~Currant  
692:               (World Scientific, Singapore, 
693:                1998),  p.~173.
694: % series  ={AIP Conference Proceedings},
695: % number  = 505,
696: 
697: 
698: \bibitem{ts2f} P.~Kramer, 
699:               Z.~Papadopolos and D.~Zeidler, 
700:                in {\em Proc. of the Symp. 
701:               ``Symmetries in Science V: Algebraic 
702:                 structures, 
703: %\ldots''
704: their representations, 
705:      realizations and physical applications''
706: },
707: %Symmetries of icosahedral quasicrystals
708:   %pages = {395-427},
709:   edited by  B.~Gruber, 
710: and L.~C.~Biedenharn,
711:  (Plenum, New York, 1991), p.~395.
712: 
713: 
714: \bibitem{PHK} Z.~Papadopolos, 
715:   C.~Hohneker and P.~Kramer,
716:  %Tiles-inflation rules for the class of canonical 
717:  %tilings ${\cal T}^{*(2F)}$  derived by the projection 
718:  %method,
719:  Discrete Math., {\bf 221}, 101
720:  %-112,
721:  (2000).
722: 
723: \bibitem{KPKB} G.~Kasner,
724:                Z.~Papadopolos, P.~Kramer and 
725:                D.~E.~B\"urgler,
726: %Surface structure of   i-Al$_{68}$Pd$_{23}$Mn$_{9}$:
727: %An analysis based on the ${\cal T}^{*(2F)}$ tiling 
728: %decorated by Bergman polytopes,
729:  Phys.~Rev.~B {\bf 60}, 3899
730:  %-3907,
731:  (1999).
732: 
733: 
734: \bibitem{LGPK2} Z.~Papadopolos,
735:                 G.~Kasner, J.~Ledieu,  
736:                 E.~J.~Cox, N.~V.~Richardson,  Q.~Chen,  
737:                 R.~D.~Diehl, T.~A.~Lograsso, 
738:                 A.~R.~Ross and R.~McGrath,
739:  %Bulk termination of the quasicrystalline fivefold 
740:  %surface of  Al$_{70}$Pd$_{21}$Mn$_{9}$}},
741:  Phys.~Rev.~B {\bf 66}, 184207
742: %-(1-13), 
743: (2002).
744: 
745: \bibitem{PP2} Z.~Papadopolos,
746:               P.~Pleasants, G.~Kasner, 
747:               V.~Fourn\'{e}e, C.~J.~Jenks, J.~Ledieu 
748:               and R.~McGrath,
749:  %A maximum density rule for bulk terminations of 
750:  %quasicrystals,
751:  Phys.~Rev.~B {\bf 69}, 224201
752: %-(1-7), 
753: (2004).
754: 
755: \bibitem{Shen00}  Z.~Shen, 
756:                  W.~Raberg, M.~Heinzig,
757:                  C.~J.~Jenks, V.~Fourn\'{e}e,
758:                  M.~A.~Van~Hove, T.~A.~Lograsso,
759:                  D.~Delaney, T.~Cai,  P.~C.~Canfield,
760:               I.~R.~Fisher, A.~I.~Goldman, 
761: M.~J.~Kramer and P.~A.~Thiel,
762: %A LEED comparison of structural  stability of the 
763: %three high-symmetry surfaces of Al-Pd-Mn bulk 
764: %quasicrystals,
765:  Surf.~Sci. {\bf 450}, 1
766:  %-11,
767:  (1999).
768: 
769: 
770: 
771: 
772: \bibitem{Kor90} A.~R.~Kortan, 
773:                 R.~S.~Becker, F.~A.~Thiel 
774:                 and H.~S.~Chen,
775:  %Real-Space Atomic Structure of a Two-Dimensional 
776:   %Decagonal Quasicrystal,
777:  Phys.~Rev.~Lett., {\bf 64}, 200
778:  %-203,
779:  (1990).
780: 
781: \bibitem{K0}  due to A.~R.~Kortan. 
782: 
783: 
784: 
785: 
786: \bibitem{Kram95} P.~Kramer, 
787:                  A.~Quandt, M.~Schlottmann, 
788:                  and T.~Schneider,
789:                  %{\em Atomic clusters and electrons in 
790:                  %the Burkov model of AlCuCo},
791:                  Phys.~Rev.~B {\bf 51}, 8815 (1995). 
792: 
793: 
794: \bibitem{classes} The labels ``$q$'', ``$b$'' and 
795:                   ``$a$'' denote the three classes of 
796:                   atomic points in the model \mts2f.
797: 
798: \bibitem{Shen99}   Z.~Shen, 
799:               C.~R.~Stoldt, C.~J.~Jenks,
800:                    T.~A.~Lograsso, and P.~A.~Thiel,
801:              Phys.\ Rev.\ B {\bf 60}, 14688 (1999).
802: 
803: 
804: 
805: \bibitem{Y04} H.~R.~Sharma, 
806:             V.~Fourn\'ee, M.~Shimoda, 
807:             A.~R.~Ross, T.~A.~Lograsso, A.~P.~Tsai, 
808:             and A.~Yamamoto,
809:            % {\em Structure of the Fivefold Surface 
810:            % of the Icosahedral Al-Cu-Fe Quasicrystal:
811:            %  Experimental Evidence of Bulk Truncations 
812:            %  at Larger Interlayer Spacings},
813:              Phys.~Rev.~Lett.~{\bf 93}, 165502 (2004). 
814: 
815: 
816: 
817: \bibitem{E03}  due to M.~Erbudak.
818: 
819: 
820: \bibitem{B04} B.~Bolliger, Diss. Naturwissenschaften 
821:               ETH Z\"urich, Nr. 13562, 2000,
822:             %``Time-resolved structural
823:             %investigation of quasicrystal surfaces'' 
824: on http://e-collection.ethbib.ethz.ch/
825: 
826: 
827: 
828: 
829: \bibitem{Z04} Jin-Cheng~Zheng, 
830:             C.~H.~A.~Huan, 
831:             A.~T.~S.~Wee, M.~A.~Van~Hove, C.~S.~Fadley, 
832:             F.~J.~Shi, E.~Rotenberg, S.~R.~Barman, 
833:             J.~J.~Paggel, K.~Horn, Ph.~Ebert, 
834:             and K.~Urban,
835:            % {\em Atomic-scale structure of the fivefold 
836:            % surface of an AlPdMn quasicrystals: 
837:            % A quantitative x-ray photoelectron 
838:            % diffraction analysis},
839:             Phys.~Rev.~B {\bf 69}, 134107 (2004). 
840: 
841: \end{thebibliography}
842: 
843: 
844: \end{document}
845: