cond-mat0104064/rr.tex
1: \documentstyle[aps,pre,epsf,multicol]{revtex}
2: \begin{document}
3: \draft
4: \title{Reconstructed Rough Growing Interfaces; \\
5:   Ridgeline Trapping of Domain Walls}
6: 
7: \author{ Chen-Shan Chin and Marcel den Nijs}
8: 
9: 
10: \address{ Department of Physics, University of Washington, P.O. Box 351560, \\
11:   Seattle, Washington 98195-1560}
12: 
13: \date{\today}
14: 
15: \maketitle
16: 
17: \begin{abstract}
18:   We investigate whether surface reconstruction order exists in
19:   stationary growing states, at all length scales or only below a
20:   crossover length, $l_{\rm rec}$.  The later would be similar to
21:   surface roughness in growing crystal surfaces; below the equilibrium
22:   roughening temperature they evolve in a layer-by-layer mode within a
23:   crossover length scale $l_{\rm R}$, but are always rough at large
24:   length scales.  We investigate this issue in the context of KPZ type
25:   dynamics and a checker board type reconstruction, using the
26:   restricted solid-on-solid model with negative mono-atomic step
27:   energies.  This is a topology where surface reconstruction order is
28:   compatible with surface roughness and where a so-called
29:   reconstructed rough phase exists in equilibrium.  We find that
30:   during growth, reconstruction order is absent in the thermodynamic
31:   limit, but exists below a crossover length $l_{\rm rec}>l_{\rm R}$,
32:   and that this local order fluctuates critically.  Domain walls
33:   become trapped at the ridge lines of the rough surface, and thus the
34:   reconstruction order fluctuations are slaved to the KPZ dynamics.
35: \end{abstract}
36: \pacs{PACS number(s): 64.60.Cn, 02.50.Ey, 05.40.-a, 68.35.Rh }
37: 
38: \begin{multicols}{2}
39:   \narrowtext
40: 
41: \section{introduction}
42: 
43: Equilibrium surface phase transitions have been a topic of research
44: for several decades.  Various types of critical behaviors are well
45: established both in theoretical models and actual experiments.  This
46: includes surface roughening~\cite{vBN,Wortis}, surface
47: melting~\cite{Dash,Dietrich}, and surface reconstruction
48: ~\cite{King,VV,MdN-Pt}.  Moreover, the competition between these
49: phenomena leads to additional phases and phase transitions, like
50: disordered flat phases, preroughening transitions, and reconstructed
51: rough phases~\cite{MdN-PR1,MdN-PR2,MdN-King}.  Roughening induced
52: deconstruction in Pt(110) \cite{MdN-Pt,RVK} and preroughening induced
53: deconstruction in Si(110) type geometries are other examples of this
54: competition~\cite{MdN-Si}.
55: 
56: The theory of dynamic non-equilibrium processes like surface growth
57: has flourished during the last decade as well.  Several new types of
58: dynamic universality classes have been identified.  KPZ type growth is
59: one example~\cite{KPZ,HHZh,RSOS,KrSp,KANK,KK,CCS-MdN,Lassig}.
60: Unfortunately, in this area a gap seems to widen between theoretical
61: and experimental interests.  Theoretical oriented research tends to
62: focus on universal aspects of these processes, such as the large scale
63: properties of growing surfaces in the stationary growing state and on
64: how this state is approached in the asymptotic large time limit.
65: Experimental oriented research tends to focus on more microscopic
66: short distance aspects of growing surfaces, e.g., as encountered in
67: actual epitaxial growth.
68: 
69: One of the fundamental issues, relevant to both perspectives, is
70: whether any of the above equilibrium surface phase transitions persist
71: in the stationary state of growing interfaces.  In this paper we
72: address whether surface reconstruction order can exist during growth.
73: 
74: This issue is related to the absence of surface roughening transitions
75: in growing surfaces. Below the equilibrium roughening transition
76: temperature $T_{\rm R}$ the growing surface is rough at large length
77: scales, but remains flat and grows layer-by-layer at distances shorter
78: than a crossover length scale $l_{\rm R}$, which varies with
79: temperature and oversaturation.  We review this briefly in
80: section~\ref{nucleation} in the context of elementary nucleation
81: theory.
82: 
83: Consider a surface that is flat and reconstructed in equilibrium at
84: low temperatures.  Below $T_{\rm R}$ it appears to grow within $l_{\rm
85:   R}$ as flat in a layer-by-layer mode.  Moreover, below $T_{\rm rec}$
86: (if $T_{\rm rec}<T_{\rm R}$) it appears as reconstructed if the new
87: particles can find their proper reconstruction positions at times
88: scales that are short compared to the rate at which a new layer is
89: completed. Presume that this is indeed the case.  The next, more
90: intriguing question is whether $l_{\rm rec}$ can be larger than
91: $l_{\rm R}$; i.e., whether rough growing surfaces be reconstructed?
92: The compatibility of surface roughness with surface reconstruction was
93: addressed in the context of equilibrium phase transitions several
94: years ago.  The answer depends on intricate details of the surface
95: topology.  For example, in missing row reconstructed (MRR) (110)
96: facets in FCC crystals, like Au and Pt, roughness is incompatible with
97: reconstruction order, and the surface roughening transition must
98: destroy the reconstruction simultaneously~\cite{MdN-Pt}.  In such
99: geometries, the reconstruction order can not exist in growing surfaces
100: beyond the roughness length scale either, and $l_{\rm rec}\leq l_{\rm
101:   R}$.
102: 
103: Surface roughness and reconstruction are compatible with each other in
104: other crystal structures.  Simple cubic (SC) MR reconstructed (110)
105: facets are an example.  In equilibrium, they can roughen before the
106: reconstruction order deconstructs, $T_{\rm R}<T_{\rm rec}$.  The
107: intermediate phase is known as a reconstructed rough
108: phase~\cite{MdN-Pt,MdN-King}.  For those surfaces it might be possible
109: to observe genuine deconstruction type phase transitions in growing
110: surfaces. Or, if not, the surface reconstruction can at least persist
111: well beyond the roughness crossover length scale, $l_{\rm rec}>l_{\rm
112:   R}$, and will be limited by an independent mechanism.  These issues
113: are the topic of our research reported here.
114: 
115: In section~\ref{nucleation} we review rough versus layer-by-layer
116: growth in surfaces, and in section~\ref{recrough} the basic properties
117: of equilibrium reconstructed rough phases.  Next, in section
118: \ref{recroughgrowth}, we start to focus on the reconstruction versus
119: dynamic roughness issue, and then, in section~\ref{RSOSmodel}, we
120: choose a specific type of reconstruction and a specific type of
121: surface growth dynamics to study it quantitatively by means of Monte
122: Carlo (MC) simulations.  The model must be as simple as possible,
123: avoiding secondary effects that might obscure the central issue.  Our
124: choice is the so-called restricted solid-on-solid (RSOS) model with
125: negative step energies, which describes a simple cubic checker board
126: type reconstruction, and KPZ type growth.  The MC simulation results
127: are presented in section~\ref{MCdata}, and analyzed in
128: section~\ref{trappedloops}.  Finally, in section~\ref{conclude} we
129: summarize our results.
130: 
131: 
132: \section{Roughness in growing surfaces}\label{nucleation}
133: 
134: The topic of this paper is whether surface reconstruction order can
135: exist during growth, but as a start it is useful to review briefly the
136: related issue of dynamic surface roughness from a long and short
137: length scale perspective.  Elementary nucleation theory suffices for
138: this purpose.  Equilibrium crystal surfaces undergo well defined
139: roughening transitions from macroscopic flat to macroscopic rough.  On
140: the other hand, growing surfaces are theoretically ``always
141: rough''~\cite{Nozieres,Villain}.  This seems at odds with practical
142: reality, where surfaces appear to grow quite differently below and
143: above the equilibrium roughening temperature $T_{\rm R}$.  Above
144: $T_{\rm R}$ they are rough (dynamic roughness) while below $T_{\rm R}$
145: they seem flat (layer-by-layer step-flow growth).  Above $T_{\rm R}$
146: the growth velocity $v_{\rm g}$ is proportional to the oversaturation
147: $v_{\rm g}\sim\Delta \mu$, while below $T_{\rm R}$ it is inversely
148: proportional to a nucleation time scale $v_{\rm g}\sim \tau^{-1}$ with
149: $\tau^{-1} \sim \exp\left({-a \eta^2/ (\Delta \mu
150:     k_BT)}\right)$~\cite{Nozieres}.  $\eta$ is the equilibrium step
151: free energy.  As a result, crystal growth shapes have sharp angles, in
152: which many facets, including all that are above their $T_{\rm R}$, are
153: missing.  This apparent difference in growth mechanism is one of the
154: most useful experimental tools to locate equilibrium roughening
155: transitions in crystal facets.
156: 
157: The origin of the exponential factor in $\tau$ is the existence of a
158: nucleation barrier for creating a terrace of height $h\to h+1$ below
159: $T_{\rm R}$ .  The edge (step) free energy loss term (proportional to
160: $\eta$ times the circumference) competes with the surface energy gain
161: term (proportional to $\Delta \mu$ times the terrace area). The
162: nucleation barrier vanishes when the step free energy $\eta$ vanishes,
163: i.e., at $T_{\rm R}$.  After a new terrace larger than the nucleation
164: thresholds is nucleated with an exponential small probability, it
165: spreads out fast by particle adhesion at its edge into a macroscopic
166: domain, until it merges with other spreading terraces that have
167: nucleated in the mean time, and thus complete the new surface layer.
168: However, new terraces are nucleated on top of
169: spreading terraces as well.  This nesting effect, together with the spatial
170: fluctuations of nucleation events leads to a loss of a well defined
171: (length scale free) global reference surface level.  This means that
172: although at small enough length scales the surface looks flat and
173: seems to grow layer-by-layer, at large length scales it is rough.
174: 
175: There is no phase transition between the layer-by-layer and rough
176: growth regimes, only a characteristic crossover length scale.  The
177: latter is of order $l_{\rm R}= v_{\rm s}\tau$, with $v_{\rm s}$ the
178: step velocity (determined by the particle deposition rate at the step
179: edge) and $\tau$ the above time scale at which terrace nuclei are
180: being created.  Surface flatness cannot be maintained during growth
181: over large length scales, but at small oversaturations ($\Delta \mu$)
182: and sufficiently below $T_{\rm R}$ (large step free energies $\eta$)
183: the growing surface can appear to be flat for all practical purposes,
184: over any typical experimental length scale.
185: 
186: The same type of issues arise in our study concerning the
187: compatibility of surface reconstruction order with growth dynamics.
188: First we address whether surface reconstruction order can persist
189: during growth at macroscopic length scales (the thermodynamic limit);
190: and, if not, whether it might still exist in a practical sense within
191: a characteristic length scale, $l_{\rm rec}$ below the equilibrium
192: reconstruction temperature $T_{\rm rec}$.
193: 
194: \section{Reconstructed Rough Equilibrium Phases}\label{recrough}
195: 
196: Surface reconstruction is conventionally associated with flat
197: interfaces.  However, surface roughness not necessarily destroys the
198: reconstruction order.  A rough but still reconstructed surface is in a
199: so-called reconstructed rough (RR) phase.  The equilibrium versions of
200: RR phases were studied theoretically some years ago in the context of
201: the competition between surface roughening and reconstruction in MR
202: reconstructed FCC (110) facets~\cite{VV,MdN-Pt}.  The topological
203: details of those FCC surface prevent the existence of RR phases,
204: implying that in Pt(110) the surface roughness and deconstructs
205: simultaneously~\cite{MdN-Pt} as observed experimentally in
206: Pt(110)~\cite{RVK}.  This implies immediately that during growth
207: reconstruction order is limited to the roughness crossover length
208: scale, $l_{\rm rec}\leq l_{\rm R}$.  The same theoretical studies 
209: also identified other surface geometries where RR phases do exist.
210: For those $l_{\rm rec}$ is not limited by $l_{\rm R}$.  In this
211: section we review the basic properties of RR phases, using as examples
212: checkerboard and MR type reconstructed simple cubic stackings.
213: 
214: To avoid confusion, it is useful to distinguish between misplacement
215: and displacement type reconstruction~\cite{MdN-King}.  In
216: misplacement reconstructions, particles have moved to different
217: solid-on-solid type stacking positions, or are removed altogether,
218: compared to the unreconstructed flat surface structure.  The checker
219: board reconstruction in Fig.\ref{chkbrd} and also the more realistic
220: MR type reconstructions are examples of this.  The average surface
221: height has changed by half a unit, $h\to h-\frac{1}{2}$.  In
222: displacement reconstructions the atomic stacking does not change.
223: Instead, the atoms are merely elastically distorted at the surface
224: with a commensurate or incommensurate period compared to the bulk.
225: Misplacement type reconstructions are more likely to disorder at
226: temperatures near $T_{\rm R}$ than displacement type reconstructions.
227: For clarity we focuses here on misplacement reconstructions.
228: 
229: The definition of the reconstruction order parameter is at the core of
230: RR phases ~\cite{MdN-Pt,MdN-King}.  In checker board and MR reconstructed SC
231: (110) facets , the reconstruction order can be formulated in two
232: distinct ways.  One formulation keeps track of whether the black or
233: white fields (even or odd rows) are on top.  The other measures it in
234: terms of anti-ferromagnetic order in the parity type Ising variables
235: $S_r=\exp(i \pi h_r)$, with $h_r=0,\pm1,\pm2,\cdots$ the surface
236: height at site $r$, see Fig.\ref{scmr}.  These two formulations might
237: seem equivalent in flat surfaces, but they are not in the presence of
238: roughness.
239: 
240: The compatibility of surface reconstruction with surface roughness
241: depends on topological properties of step and domain wall excitations;
242: on how they affect the two versions of the order parameter.
243: Fig.\ref{scmr} shows in cartoon style a cross section of the
244: reconstructed surface, and also domain wall and step excitations.  The
245: domain wall in (b) does not change the surface height.  Notice that
246: both order parameters change sign.  Across the step in (c), from left
247: to right, the even-odd order parameter changes sign, but the parity
248: order is unaffected.  At the step in (d) the opposite happens.
249: 
250: These two types of steps are the only topologically distinct ones that
251: are possible; (c) couples only to the even-odd row type order
252: parameter and (d) only to to the parity version.  
253: 
254: It is possible to construct many more step and domain wall structures
255: that look locally different from the ones in the figure, but those
256: induce the same change in height and/or reconstruction order(s) and
257: therefore are from a topological point of view identical to the ones
258: in the figure.  Notice also that the excitations in (b)-(d) are
259: related to each other in the sense that any of the three can be
260: interpreted as a bound state the two others. Elastic surface
261: deformations in the actual atomic positions near the surface and
262: additional ones near the steps and domain walls, influence the local
263: internal structure of steps and domain walls, but do not affect those
264: topological features, and therefore need not explicitly be represented
265: in the following discussion.  (They certainly renormalize the step and
266: domain wall energies and the interactions between such surface
267: excitations.)
268: 
269: The fate of reconstruction versus roughness depends on the energies of
270: these steps and domain walls, including the kink energies.  They set
271: the scale of the meander type entropy and therefore the temperature
272: dependence of the step free energies.  If the domain wall free energy
273: vanishes first, the surface remains flat but the reconstruction
274: vanishes, $T_{\rm rec}<T_{\rm R}$.  In case the free energy of one of
275: the two types of steps vanishes first, the surface enters a
276: reconstructed rough phase, $T_{\rm R}<T_{\rm rec}$.  At the roughening
277: transition one of the two reconstruction order parameters vanishes,
278: but the other type of order remains.  So there exist two topologically
279: distinct types of RR phases.  (Notice that only the one with the
280: parity type order is readily observable by, e.g., conventional X-ray
281: diffraction.)
282: 
283: In the RSOS model below, the RR phase has parity order, i.e., the step
284: free energy of the (c) type steps is zero, but walls and (d) type
285: steps still have non-zero free energy.  We will refer to those
286: excitations as ``loops of zero's", because in the rough surface they
287: show up as contours across which the height change is zero, $dh=0$.
288: The deconstruction transition (inside the rough phase) takes place at
289: the temperature where the surface tension of the loops vanishes.  In
290: equilibrium that turns out to be an ordinary Ising transition.  This
291: conclude our brief review.  For more details we refer to
292: Refs.~\cite{MdN-Pt} and \cite{MdN-King}.
293: 
294: \section{reconstructed rough growth}\label{recroughgrowth}
295: 
296: Let's focus now on surface growth.  Only in surfaces where equilibrium
297: RR phases are topologically possible, can the surface reconstruction
298: length scale $l_{\rm rec}$ exceed the onset of dynamic roughness
299: length scale $l_{\rm R}$.  Moreover, it is quite possible that the
300: reconstruction order persists over all length scales ($l_{\rm rec}\to
301: \infty$), such that a genuine dynamic deconstruction phase transition
302: takes place in the stationary state of the growing surface, just like
303: in equilibrium.
304: 
305: For comparison, imagine a two dimensional (2D) lattice with on each
306: site an height variable and an Ising spin degree of freedom
307: (representing the reconstruction order). This leads to two coupled
308: master equations, one for surface growth, e.g., KPZ type dynamics, and
309: another for the reconstruction order, e.g., Glauber type Ising
310: dynamics.  In equilibrium surfaces, the coupling between the two
311: sectors is weak, to the extend that the reconstruction transition in
312: the Ising sector and the roughening transition in the height variable
313: sector do not interfere with each other~\cite{MdN-Pt,MdN-RSOS}.  The central
314: issue is whether and how this coupling changes during growth.
315: The Ising dynamics itself, is blind to the growth bias.  If the
316: coupling between the two sectors is remains weak, the Ising spins can
317: still reach the Gibbs equilibrium state and undergo a conventional
318: equilibrium reconstruction transition.
319: 
320: Coupled master equations of this type have been studied recently in
321: the context of specific 1D growth models.  Those display strong
322: coupling between the Ising and roughness degrees of freedom, such as
323: growth being pinned down by Ising domain
324: walls~\cite{Drossel,Kotrla,Noh}.  Pinning favors to spontaneous
325: facetting. In our 2D model, we observe different effects, besides the
326: obvious fact that in 1D equilibrium reconstruction order can not
327: exist.
328: 
329: \section{ Restricted solid-on-solid Model}\label{RSOSmodel}
330: 
331: The 2D restricted solid on solid (RSOS) is one of the work horses of
332: surface physics research. Integer valued height variables $h_r=0,\pm
333: 1,\pm2,\cdots$ are assigned to a square lattice and nearest neighbor
334: heights are restricted to differ by at most one unit, $dh=0,\pm1$. The
335: energy
336: \begin{equation}
337: E =  \frac{1}{2} K \sum_{\left<r, r^{\prime}\right>} (h_r-h_{r^{\prime}})^2
338: \end{equation}
339: depends only on nearest neighbor interactions.  We use dimensionless
340: units, $K=J/k_{\rm B}T$.  The $K>0$ side of the phase diagram contains
341: a conventional equilibrium surface roughening
342: transition~\cite{MdN-RSOS}.  Moreover, the non-equilibrium version has
343: been studied extensively for $K>0$ as well, because it is a natural
344: lattice realization of KPZ growth ~\cite{KPZ,HHZh,RSOS,KrSp,KANK,KK,CCS-MdN}.
345: 
346: For $K<0$, the model contains one of the simplest examples of an
347: equilibrium RR phase~\cite{MdN-RSOS}, and is probably the most compact
348: formulation of the coupling between Ising and surface degrees of
349: freedom.  The $dh=\pm1$ steps are more favorable than flat $dh=0$
350: segments.  At zero temperature, $K\to-\infty$, the $dh=0$ states are
351: frozen out, and the model reduces to the so-called body centered solid
352: on solid (BCSOS) model, but in this version it lacks step energies,
353: which means that the surface is rough even at zero temperature.  The
354: surface is rough, but since nearest neighbor heights must differ by
355: one, all heights on one sublattice are even and odd at the other, or
356: the other way around.  This two-fold degeneracy represents the checker
357: board type RR order.  The staggered magnetization, defined in terms of
358: the parity spin type variables $S_i = \exp(i\pi h_r)$, is non-zero.
359: 
360: The $dh=0$ excitations that appear at $T>0$ form closed loops and
361: behave like Ising type domain walls. The reconstruction order changes
362: sign across such loops.  Their sizes diverge at the equilibrium
363: deconstruction transition $K_{\rm c}=-0.9630$~\cite{MdN-RSOS}.
364: (Determined by transfer matrix finite size scaling techniques).  The
365: Ising and roughness variables couple only weakly.  Numerically, all
366: reconstruction aspects of the transition follow conventional Ising
367: critical exponents.  Moreover, the thermodynamic singularities in the
368: Ising sector affect only the temperature dependence of the surface
369: roughness parameter $K_{\rm G}$, defined in terms of the height-height
370: correlator,
371: \begin{equation}
372: \langle (h_{r+r_0}-h_{r_0})^2 \rangle \simeq (\pi K_{\rm G})^{-1} \ln(r).
373: \end{equation}
374: The continuum limit analysis confirms these numerical results.  The
375: point in the generalized phase diagram where the Gaussian (height) and
376: Ising degrees of freedom decouple is a stable renormalization type
377: fixed point~\cite{MdN-King}.
378: 
379: We study this same RSOS model in the presence of a KPZ type growth
380: bias.  In the MC simulation, we first select an update column and next
381: whether a particle deposition or evaporation event will be attempted.
382: The move is rejected if it would result in a violation of the RSOS
383: condition, $dh=0,\pm1$.  If allowed, it will take place with
384: probability $P =\min(p, p e^{-\Delta E_j})$ in case of deposition, and
385: with probability $P =\min(q, q e^{-\Delta E_j})$ for evaporation.
386: Without loss of generality we can choose $p+q=1$.  At infinite
387: temperature ($K=0$) and deposition only ($q=0$) the model reduces to
388: the well known Kim-Kosterlitz ~\cite{KK} model for KPZ type growth.
389: 
390: We will present only our MC results far from equilibrium, i.e., at
391: $q=0$ with deposition only.  We observe no qualitative differences
392: closer to equilibrium, $0<p<1$, but the interpretation of the data
393: becomes increasingly obscured (as expected) by (conventional)
394: crossover scaling from the equilibrium deconstruction phase
395: transition.
396: 
397: At low temperatures, $K \rightarrow -\infty$, the Metropolis dynamics
398: slows down considerably. The rejection rate becomes high and the
399: density of active sites becomes low.  Therefore we employ the
400: following rejection free algorithm.  During the MC simulation we keep
401: a list of active sites, i.e., sites where particles can deposit
402: without violating the RSOS condition.  They are grouped in
403: $j=1,\cdots,5$ sets, according to the five distinct energy changes
404: $\Delta E_j$ that can occur during deposition.  First we preselect one
405: of those 5 sets, with probability $(p_j N_j)/(\sum_j p_j N_j)$, where
406: $p_j =\min(1,e^{-\Delta E_j})$ and $N_j$ is the number of sites of
407: type $j$.  Next, a particle is randomly deposited at one of the sites
408: in that specific set $j$.  Rejection free procedures like this upset
409: the flow of time.  To restore proper time, we increase the MC time
410: during each update step by $1/p\times 1/N_j$.  We checked explicitly
411: that this reproduces the correct value for the KPZ dynamic exponent
412: $z=8/5$~\cite{KK,CCS-MdN,Lassig} at $K\simeq 0$; we find $z\simeq
413: 1.6\pm 0.1$.
414: 
415: The above algorithm resolves the slowing down problem in the actual MC
416: simulation, but does not address its origin.  In the limit $K
417: \rightarrow -\infty$ the RSOS model reduces to the BCSOS model, with
418: $dh=\pm1$ at all bonds.  The $dh=0$ loops are frozen out completely.
419: In BCSOS type KPZ growth dynamics, 2 particles are deposited at ones
420: in the form of vertically oriented bricks, otherwise a ``forbidden''
421: configuration with $dh=0$ would arise.  In the $K<0$ RSOS model at
422: very low temperatures the same event is achieved as a 2-step
423: two-particle process, by the deposition of a second particle at the
424: same site soon after the first one.  The probability for deposition of
425: the first particle is equal to $p= L^{-2}\exp(2K)$.  The second
426: particle deposition on top of it happens with probability $p=L^{-2}$).
427: This implies that the time clock in the RSOS model runs slower by a
428: factor $r=\exp(2K) ( 1 + 4 \exp(K) + ...)$.
429: 
430: A final remark about surface roughness.  In normal surfaces, the
431: equilibrium roughness increases with temperature; due to the fact that
432: meander type entropy renormalizes the step energy into a reduced step
433: free energy~\cite{MdN-King}.  In our model, surface roughness evolves
434: the opposite way; it decreases with temperature.  The surface is less
435: rough at infinite temperature $K=0$ than in the zero temperature limit
436: $K\to-\infty$.  A high temperature RSOS surface, with $dh=0,\pm1$ is
437: obviously less rough than a BCSOS surface, with only $dh=\pm1$.
438: Recall that this BCSOS model lacks step energies, such that it is just
439: as rough at $T=0$ as at $T\to\infty$.  From the BCSOS perspective the
440: thermally excited $dh=0$ loops stiffen the surface, and give rise to
441: an inverted roughness versus temperature profile. On the one hand,
442: this is an interesting phenomena in its own right.  Moreover we could
443: fine tune it by introducing next-nearest neighbor interactions, since
444: they represent BCSOS type step energies.  On the other hand, this
445: effect is unlikely to affect the central question we want to address
446: (how do roughness and reconstruction degrees couple during growth) and
447: therefore we choose not do so in this study.
448: 
449: \section{ reconstruction during growth}\label{MCdata}
450: 
451: We search for reconstruction order as function of temperature, for
452: $-\infty<K<0$.  The susceptibility type parameter~\cite{MC-suscep},
453: \begin{equation}
454: \chi = L^2(\langle m^2 \rangle - \langle | m | \rangle^2)
455: \end{equation}
456: of the reconstruction order parameter,
457: \begin{equation}
458: m=\left< (-1)^{x+y} ~e^{i \pi h(x,y)} \right>
459: \end{equation}
460: is shown in Fig.\ref{susClps} for the stationary state of the growing
461: surface, as function of $K$ for different system sizes $L^2$.  The
462: sharp maxima seem to confirm the existence of a dynamic surface
463: reconstruction transition into a RR phase.  However, several features
464: are very different from equilibrium. The peak height diverges as
465: $\xi\sim L^2$. i.e., stronger than at the equilibrium transition point
466: where it scales as $\chi\sim L^{\gamma/\nu}$.  This could be a signal
467: of a first order phase transition.  However, the peak position does
468: not converge to a specific critical point $K_c$. Instead it keeps
469: shifting with lattice size. It scales logarithmically, as $K_{\rm
470:   peak}(L)\simeq-A\ln(L/L_0)$ with $A=0.77\pm 0.05$ and $L_0= 2.2\pm
471: 0.2$.
472: 
473: Next, we monitor in detail the reconstruction order parameter $m$ near
474: and below the equilibrium $K_c$ as function of time.  It behaves
475: similar as in conventional spontaneously ordered phases, but
476: flip-flops more frequently than justifiable from finite size effects
477: alone. Moreover, the fluctuations in $m$ within each phase are too
478: strong.  Fig.\ref{mhstgm} quantifies this in terms of a histogram of
479: the number of times a specific value of $m$ appears in a typical time
480: series.  The distribution has two distinct peaks, suggesting the
481: presence of spontaneously broken reconstruction order, but the tails
482: have a power law shape instead of the exponential form mandatory for a
483: spontaneous broken symmetry.
484: 
485: Power laws are the hallmark of critical fluctuations.  So, quite
486: surprisingly, it appears as if the RR order is critical at low
487: temperatures for all $K<K_{\rm peak}$. Instead of an isolated critical
488: point, we seem to be dealing with a critical phase.
489: 
490: \section{ Loops trapped on ridge lines }\label{trappedloops}
491: 
492: The surprising critical fluctuations in the reconstruction order
493: parameter can be traced to the following loop dynamics.  Consider a
494: typical configuration at very low temperatures.  Fig.\ref{trappedloop}
495: shows an example~\cite{MdN-Korea}.  The surface is in an almost pure
496: BCSOS type dynamic rough stationary state (with $dh\pm 1$), and
497: contains only a few $dh=0$ loops separating surface areas of opposite
498: checker board type RR order.
499: 
500: The typical life cycle of such a loop runs as follows. It is nucleated
501: in a valley bottom. Next it runs up hill, grows in diameter,
502: encompassing the entire valley, until it becomes trapped on a ridge
503: line.  There it lingers until another loop annihilates it, or when the
504: KPZ surface fluctuations to which it is slaved shrink it back to zero.
505: 
506: Fig.\ref{1D_dyn}(a) represents a cross-section of the 2D rough surface
507: near a valley.  It shows a domain of opposite reconstruction inside an
508: otherwise perfectly reconstructed rough configuration.  The two flat
509: segments are the locations where the domain wall loop intersects the
510: cross section.  In equilibrium, the loop fluctuates with equal
511: probability up and down the slope because depositions and evaporations
512: are equally likely.  A growth bias breaks this symmetry, the loops
513: move more likely upwards than downwards, see Fig.\ref{1D_dyn}(a).  This
514: upward drift is the driving force responsible for the trapping of
515: loops at ridge lines, and thus creates a strong coupling between the
516: roughness and reconstruction degrees of freedom, unlike equilibrium
517: where they effectively decouple.
518: 
519: A few comments on the topology of ridge lines in rough surfaces might
520: be useful. Imagine a rolling ball in this landscape, like in the well
521: known analogy with renormalization flow in statistical physics.
522: Presume strong friction such that the velocity is proportional to the
523: force, i.e., the gradient of the slope, at all times.  The hill tops
524: are the completely unstable ``fixed points''.  The valleys are the
525: attractors.  The ridge lines form the water sheds between valleys.
526: Every ridge line runs from an hill top to a saddle point.  From each
527: hill top an arbitrary number of ridge lines can emerge, but only two
528: ridge lines can end at each saddle point (at opposite sides of the
529: single direction in which the saddle point attracts).  So the ridge
530: lines form a network, and since none of them can not stop in midair it is a
531: closed network.  The KPZ rough surface is scale invariant, which means
532: that this ridge line network has fractal properties.
533: 
534: Ignore for the time being the scale invariant aspects of the network.
535: Imagine a landscape consisting of deep smooth valleys surrounded by
536: ridge lines; unlike the real rough surface where every deep valley
537: consists of collections of sub valleys.  The life cycle of a
538: macroscopic loop in this surface starts with the nucleation of a new
539: seedling-loop at the floor of the valley and its rise along the
540: slopes, during which it grows into a macroscopic object.  The only
541: loops of interest are those nucleated at the valley bottom and then
542: run up-hill encompassing the entire valley.  Only those loops are
543: topologically trapped and stable.  Loops nucleated on the slopes
544: annihilate by stochastic fluctuations before becoming macroscopically
545: large. The same is true for loops nucleated out of the valley bottom
546: but running up-hill on one slope segment only.
547: 
548: The rise of a seedling loop out of the valley bottom into a
549: macroscopic object, is a very fast process.  Almost no MC moves that
550: make the loop grow and rise are rejected; energy barriers are rare,
551: because the length of the loop (its energy) increases uniformly.
552: Compared to this, the nucleation frequency in the valley bottom is
553: very small.  This means that the time scale at which a macroscopic
554: loop emerges out of the valley is limited by the nucleation time scale
555: $\tau_{\rm n}$ and independent of the valley size.
556: 
557: To measure $\tau_{\rm n}$ we prepared a surface in the BCSOS KPZ
558: stationary state and measure (at a very low temperature, $K \ll K_{\rm
559:   c}$) the intervals between macroscopic loop events.  Numerically we
560: find $\tau_n\sim\exp(- \alpha K)$ (measured in BCSOS time units) with
561: $\alpha=3.0\pm 0.1$.
562: 
563: This agrees qualitatively with the following estimate.  The deposition
564: of the first particle in the valley bottom occurs with probability
565: $p=L^{-2}e^{2K}$.  This creates a fledgling loop, but one that is
566: indistinguishable from the intermediate state in an elementary BCSOS
567: type growth event (where a second particle is dropped on top of it
568: with probability $p=L^{-2}$).  The loop grows when the next particle
569: is dropped not on top but next to the previous one.  That happens with
570: probability $p=L^{-2}e^{K}$.  The nucleation threshold diameter
571: $l_{\rm c}$ is reached when the loop growth and BCSOS growth become
572: distinguishable, i.e., when the annihilation of a loop requires the
573: creation of a new well distinguishable loop inside it.  That happens
574: at about $l_c^2 \simeq 7$, see Fig.\ref{1D_dyn}b.  The time scale at
575: which that stage is reached is approximately $t\simeq L^{-2}e^{-4K}$
576: (in BCSOS time units), which is of the same order of magnitude as the
577: above numerical nucleation time scale.
578: 
579: The loop rises out of the valley until it becomes trapped on the ridge
580: line that borders this valley to adjacent ones.  From
581: there on the loop is slaved to the growth fluctuations of the surface.
582: Valleys grow and shrink (without bias), open-up, fill-up and merge.
583: The loop has to follow this dance of the ridge line until a new loop
584: nucleates out of the valley and annihilates it, or when the encircled
585: terrain happens to shrink to zero (fills-up) by surface growth
586: fluctuations.
587: 
588: We expect that the life time $\tau_z(L)$ of a ridge line of size $L$
589: in a growing surface, scales as a power law, $\tau_z\sim L^z$, with
590: $z$ the dynamic exponent of the surface roughness degrees of freedom
591: (KPZ like in our model).  To test this, we measure the decay times of
592: large macroscopic defect loops (of about half the lattice
593: size) as function of $L$, at low temperatures $K\ll K_c$.  The data in
594: Fig.\ref{timeHis} collapse indeed onto one universal curve after a
595: rescaling of time by $\tau_z \approx L^z$.  The collapse fits best at
596: $z=1.7\pm 0.1$ (in BCSOS time units), which is consistent with the
597: known KPZ dynamic exponent $z=8/5$ \cite{KK,CCS-MdN,Lassig}.
598: 
599: The ridge line fluctuations are responsible for the power law tails in
600: the time distribution of RR order, Fig.\ref{mhstgm}.  Those critical
601: fluctuations only show up below a characteristic length scale $l_{\rm
602:   rec}$, where the nucleation time scale $\tau_{\rm n} \sim
603: \exp(-\alpha K)$ is larger than the surface growth time scale, $\tau_z
604: \sim L^z$.  A simple estimate for $l_{\rm rec}$ follows from equating
605: the two time scales, $l_{\rm rec}\sim \exp(\frac{\alpha}{z} K)$.
606: 
607: The peaks in the susceptibility, in Fig.\ref{susClps}, reflect this
608: crossover length $l_{\rm rec}$.  Recall that the peak shifts
609: logarithmically. By setting $\tau_n=\tau_z$ we obtain the same
610: logarithmic behavior, $K_{\rm c}=-\frac{z}{\alpha}\ln(L/L_0)$.  The
611: prefactor is too small by about 30\%, but this is not a surprise
612: because the estimate is rather simple minded.  It ignores for example
613: the self similarity of the rough surface.  Consider a sub valley
614: adjacent to an already trapped loop.  Suppose a new loop nucleates out
615: of this subvalley.  The loop segments annihilate each other in pairs.
616: The net effect of this nucelation event is therefore that the trapped
617: loop jumps across the sub valley.  It now follows the complementary
618: segment of the ridge line that encircles the sub valley.  Such events
619: renormalize $\tau_z$, in particular near $K_c$.
620: 
621: 
622: \section{ Conclusions }\label{conclude}
623: 
624: In this paper we study the compatibility of surface reconstruction and
625: surface roughness during growth.  There are several possibilities.
626: 
627: In surfaces where reconstructed rough (RR) phases are topologically
628: forbidden, like missing row reconstructed FCC(110) facets, the
629: reconstruction order can not exist on a global scale in the stationary
630: growing state.  It can appear only locally within the crossover
631: roughness length scale $l_{\rm R}$ within which the surface grows in a
632: layer-by-layer fashion, i.e., $l_{\rm rec}\leq l_{\rm R}$.
633: 
634: The reconstruction length scale $l_{\rm rec}$ can exceed $l_{\rm R}$,
635: only in surfaces where equilibrium reconstructed rough phases are
636: topologically possible, and those surfaces could in principle even
637: display genuine deconstruction type phase transitions in the
638: stationary growing state.
639: 
640: We address this issue in the context of KPZ type dynamics, in the RSOS
641: model with negative coupling constant $K<0$, which in equilibrium has
642: a checker board type RR phase and a true deconstruction phase
643: transition inside the rough phase.  We find that the stationary
644: growing rough state lacks true macroscopic RR order; $l_{\rm rec}$
645: remains finite.  Moreover, we identify the mechanism that sets the
646: temperature dependence of $l_{\rm rec}$.
647: 
648: The fundamental features are an upward drift of the reconstruction
649: domain wall loops and their trapping at the ridge lines of the
650: surface.  There, the loops are slaved to fluctuations of the surface growth
651: dynamics.  $l_{\rm rec}$ is set by the competition between two life
652: times: the nucleation time scale of a new loop out of the valleys
653: (annihilating existing trapped loops) and the time scale
654: $\tau_{KPZ}\sim L^z$ at which a ridge line of radius $L$ vanishes due
655: to surface growth fluctuations.
656: 
657: At length scales smaller than $l_{\rm rec}\sim \exp(\frac{\alpha}{z}
658: K)$, the surface appears as reconstructed rough, and the life time of the
659: loops is determined by the KPZ growth dynamic fluctuations.  The
660: latter follow power laws. This manifests itself in critical
661: fluctuations in the reconstruction order at length scales smaller than
662: $l_{\rm rec}$.  In x-ray diffraction from such a growing interface,
663: one would observe not only power law shaped peaks associated with the
664: surface roughness, but also at temperatures where $l_{\rm rec}$ is
665: larger than the coherence length of the surface, power law shaped
666: reconstruction diffraction peaks.
667: 
668: At length scales larger than $l_{\rm rec}$, the surface appears as
669: unreconstructed rough. Loops of that size die by nucleation of new
670: loops instead of KPZ surface fluctuations, and they are not trapped
671: anymore, because loop segments can hop across sub valleys of size $l >
672: l_{\rm rec}$ by means of nucleation of new loops in sub valleys.
673: 
674: In our study we chose to focus on KPZ type surface growth dynamics,
675: but we have good reasons to expect that the trapping of domain walls
676: on ridge lines is a common phenomenon.  In general, the quasi-critical
677: fluctuations will reflect the dynamic exponent of whatever growth
678: dynamics is applicable.  In recent studies of 1D models with KPZ and
679: Ising type coupled degrees of freedom the Ising defects become trapped
680: in valleys and canyons and thus pin-down the
681: growth~\cite{Drossel,Kotrla}.  We expect that a tendency towards
682: facetting instead of ridge line trapping can also be realized in our
683: 2D model by varying the local growth rates.  This research is
684: supported by the National Science Foundation under grant DMR-9985806.
685: 
686: \begin{references}
687: \bibitem{vBN} H.~van Beijeren and I.~Nolden, in {\it Structures} {\it
688:     and} {\it Dynamics} {\it of} {\it Surfaces}, W. Schommers and
689:   P.~von Blanckenhagen, eds. (Springer, Berlin, 1987).
690:   
691: \bibitem{Wortis} M.~Wortis, in {\it Chemistry} {\it and} {\it Physics}
692:   {\it of} {\it Solid} {\it Surfaces} {\it VII}, R.~Vanselow and
693:   R.F.~Hove, eds. (Springer, Berlin 1988).
694:   
695: \bibitem{Dash} J.G.~Dash, Contemp.~Phys. {\bf 30}, 89 (1989).
696:   
697: \bibitem{Dietrich} S.~Dietrich in {\it Phase Transitions and Critical
698:     Phenomena}, Vol.12, eds C.~Domb and J.~Lebowitz, (Academic Press,
699:   London, 1988).
700:   
701: \bibitem{King} {\em The Chemical Physics of Solid Surfaces and
702:     Heterogeneous Catalysis}, Vol. 7 edited by D.~King (Elsevier,
703:   Amsterdam, 1994).
704:   
705: \bibitem{VV} I.~Vilfan and J.~Villain, Surf.~Sci. {\bf 199}, L165
706:   (1988); Phys.~Rev.~Lett. {\bf 65}, 1830 (1990); Surf.~Sci. {\bf
707:     257}, 368 (1991).
708:   
709: \bibitem{MdN-Pt} M.~den Nijs, Phys.~Rev.~Lett. {\bf 66}, 907 (1991),
710:   and Phys.~Rev.~B {\bf 46}, 10386 (1992).
711:   
712: \bibitem{MdN-PR1} K.~Rommelse and M.~den Nijs, Phys.~Rev.~Lett. {\bf
713:     59}, 2578 (1987); Phys.~Rev. {\bf B 40}, 4709 (1989).
714:   
715: \bibitem{MdN-PR2} M.~den Nijs, Phys.~Rev.~Lett. {\bf 64}, 435 (1990).
716:   
717: \bibitem{MdN-King} M.~den Nijs, chapter 4 in the ref.[\cite{King}].
718:   
719: \bibitem{RVK} I.~K. Robinson, E. Vlieg, and K. Kern, Phys.~Rev.~Lett.
720:   {\bf 63}, 2578 (1989).
721:   
722: \bibitem{MdN-Si} Marcel den Nijs, J.~Phys.~A.{\bf 30}, 397-404 (1997).
723:   
724: \bibitem{Nozieres} P.~Nozi\`{e}res and F.~Gallet, J.~Phys.(Paris) {\bf
725:     48}, 353(1987).
726:   
727: \bibitem{Villain} A.~Pimpinelli and J.~Villain, {\it Physics of Crystal
728:     Growth} (Cambridge University Press, 1997).
729:   
730: \bibitem{KPZ} M.~Kardar, G.~Parisi, and Y.C.~Zhang, Phys.~Rev.~Lett.
731:   {\bf 56}, 889 (1986);
732:   
733: \bibitem{HHZh} T.~Halpin-Healy and Y.C.~Zhang, Phys.~Rep. {\bf 254},
734:   215 (1995).
735: 
736: \bibitem{RSOS} J.G.~Amar and F.~Family, Phys.~Rev.~Lett. {\bf 64}, 543
737:   (1990), {\it ibid.} {\bf 64}, 2334 (1990);
738:   
739: \bibitem{KrSp} J.~Krug and H.~Spohn, Phys.~Rev.~Lett. {\bf 64}, 2332 (1990).
740:   
741: \bibitem{KANK} J.~Kim, T.~Ala-Nissila and J.M.~Kosterlitz,
742:   Phys.~Rev.~Lett. {\bf 64}, 2333 (1990).
743: 
744: \bibitem{KK} J.M.~Kim and J.M~. Kosterlitz, Phys.~Rev.~Lett. {\bf 62},
745:   2289 (1989).
746: 
747: \bibitem{CCS-MdN} C.S.~Chin and M.~den Nijs, Phys.~Rev.~E.~{\bf 59} ,
748:   2633-2641 (1999).
749: 
750: \bibitem{Lassig} M.~L\"assig, Phys.~Rev.~Lett. {\bf80}, 2366 (1998).  
751: 
752: \bibitem{MdN-RSOS} M.~den Nijs, J.~Phys.~A {\bf 18}, L549 (1985).  
753: 
754: \bibitem{Drossel} B.~Drossel, and M.~Kardar, Phys.~Rev.~Lett. {\bf
755:     85}, 614 (2000).
756:   
757: \bibitem{Kotrla} M.~Kotrla, and M.~Predota, Europhys.~Lett. {\bf 39},
758:   251 (1997); M.~Kotrla, F.~Slanina and M.~Predota, Phys.~Rev.~B {\bf
759:     58}, 10003 (1998).
760:   
761: \bibitem{Noh} J.D.~Noh, H.~Park, and M.~den Nijs, Phys.~Rev.~Lett., in
762:   press.
763:   
764: \bibitem{MC-suscep} K.~Binder, and D.~W.~Heermann, {\em Monte Carlo
765:     Simulation in Statistical Physics} (Springer-Verlag, Heidelberg,
766:   1997).
767: 
768: \bibitem{MdN-Korea} A color version of this figure, can be found on
769:   our WEB page, and also in, Marcel den Nijs, Bulletin of the APCTP
770:   (Korea), Spring 2001 (in press).
771: 
772: \end{references}
773: 
774: \pagebreak
775: \begin{figure}
776:   \centerline {\epsfxsize=5 cm \epsfbox{chkbrd.eps}} \vskip 10 true pt
777: \caption{
778:   Checker board type misplacement surface reconstruction}
779: \label{chkbrd}
780: \end{figure}
781: 
782: \begin{figure}
783:   \centerline {\epsfxsize=5 cm \epsfbox{scmr.eps}} \vskip 10 true pt
784: \caption{
785:   (a) A perfect reconstructed surface. (b) A domain wall; both order
786:   parameters change sign. (c) A step where only the even-odd row order
787:   changes sign. (d) A step where only the parity order changes sign.
788:   }
789: \label{scmr}
790: \end{figure}
791: 
792: \begin{figure}
793:   \centerline{\epsfxsize=8cm \epsfbox{sus.eps}} \vskip 10 true pt
794: \caption{
795:   Reconstruction order susceptibility $\chi$ as function of
796:   temperature at system sizes $L=8$-$64$.  The data collapses onto a
797:   single curve by the shift $K^\prime = K-K_{\rm{peak}}(L)$, with
798:   $K_{\rm{peak}}(L)=-0.77\ln(L/2.2)$.}
799: \label{susClps}
800: \end{figure}
801: 
802: 
803: \begin{figure}
804:   \centerline {\epsfxsize=8 cm \epsfbox{his.eps}} \vskip 10 true pt
805: \caption{
806:   Histogram (insert) of the reconstruction order parameter, $m$, at
807:   $L=32$ and $K=-3.2$ from $2^{18}$ data points using $\Delta M=0.01$
808:   as bin width.  The tails about the peaks at $m=\pm 1$ scale as power
809:   laws (main frame) with exponent $-0.9\pm0.1$.}
810: \label{mhstgm}
811: \end{figure}
812: 
813: \begin{figure}
814:   \centerline{\epsfxsize = 8 cm \epsfbox{trappedloop.eps}} \vskip 10
815:   true pt
816: \caption{ A typical low temperature configuration of the
817:   growing surface with one large loop trapped at a ridge-line.  }
818: \label{trappedloop}
819: \end{figure}
820: 
821: \begin{figure}
822:   \centerline{\epsfxsize = 8 cm \epsfbox{domains.eps}} \vskip 10 true pt
823: \caption{ 
824:   (a) One dimensional cross section of the surface near a valley with
825:   two loop segments.  On the slope, {\it a} ({\it d\,}) are the only
826:   active adsorption (desorption) sites.  The domain walls always move
827:   upwards during adsorption. (b) A loop of size of $l_c$ nucleated at
828:   the bottom of a local valley. Gray and white sites have different
829:   surface reconstruction parity order.}
830: \label{1D_dyn}
831: \end{figure}
832: 
833: \begin{figure}
834:   \centerline{\epsfxsize = 8 cm \epsfbox{timeHis.eps}} \vskip 10 true
835:   pt
836: \caption{
837:   Histogram of the decay time of a trapped loop at $K=-6.0$.  The data
838:   collapses by rescaling time by a factor $L^{1.7}$.} .
839: \label{timeHis}
840: \end{figure}
841: 
842: \end{multicols}
843: \end{document}
844: