www.compphys.cn south china university of technology growth of cluster xiaobao yang department of...
Post on 14-Dec-2015
217 Views
Preview:
TRANSCRIPT
www.compphys.cn
South China University of Technology
www.compphys.cn
Growth of Cluster
Xiaobao Yang
Department of Physics
www.compphys.cn
Clusters experimentally observed
Nature 318,162(1985); 407,60(2000);
C60
C20
Graphene fragments
ACS nano,6,8203(2012)
www.compphys.cnScience 299, 96 (2003);PRL, 103,047402(2009)
Diamond fragments
www.compphys.cn
http://www.chem.brown.edu/research/LSWang/http://www.tsinghua.edu.cn/publish/chemen/2141/2011/20110413125809079217484/20110413125809079217484_.html
B20B19
www.compphys.cnNature 425, 593-595. 9 October 2003
http://blog.sciencenet.cn/home.php?mod=space&uid=279992&do=blog&id=509211
The Geometry of Universe
Platonic and Archimedean solids
www.compphys.cn
How Clusters are found?
Wulf’s construction Dense Packing and Symmetry in Small Clusters of Microspheres
Science 301,403(2003)
www.compphys.cn
The growth and evolution of Clusters
Nucleation / Diffusion / ReconstructionSubstrate / bonding / Temperature
www.compphys.cn
Cluster growth model
No diffusion vs Full diffusion
www.compphys.cn
Cluster Growth Model►Diffusion Limited
Aggregation (DLA) Model
►Discussion on programming
►Fractals? The dimensionality of clusters?
1. Choose initial position of a walker at random on a circle r0.
2. If the walker wanders too far from the cluster (say, >1.5r0), start a new walker.
3. As the cluster grows, r0 should be increased. (say, keep r0=5Rcluster)
4. When the walker is far from the cluster, a greater step size may be adopted. Refer to DLA.m
www.compphys.cn
DLA.mclear; clf;
M = 300; N=1000; % MxM grid; N particles;A=zeros(M,M);for i=1:M; for j=1:M; if(abs(i-M/2)<M/40 & j==round(M/2)) A(i,j)=-1; end; end; end;
%imagesc(A); hold on
nparticle=0;
while(nparticle < N) %generate a random walker within the belt of 3/5*M/2<R<4/5*M/2 r=(rand/5+3/5)*M/2; theta=rand*pi*2; x=round(r*cos(theta)+M/2); y=round(r*sin(theta)+M/2); check = 1; R=M*9/20; % checker for walkers wandering too far i.e., |walker-center|>R. % if not meet the seed, continue wandering while( A(x-1,y) ~= -1 & A(x+1,y) ~= -1 & A(x,y-1) ~= -1 & A(x,y+1) ~= -1) x=x+sign(rand-0.5); y=y+sign(rand-0.5); if( abs(x-M/2)>M*9/20 || abs(y-M/2)>M*9/20 ); check=0; break; end; %walker elimated if wandering too far! end if(check==1); A(x,y) = -1; nparticle = nparticle + 1; end; %meets seed; seed updated. %if(mod(nparticle,100)==0); imagesc(A); endend;
colormap(winter); imagesc(A);axis([0 M 0 M],'square','equal');
1)Place the seed2)Atomic random walk towards the seed3) Check if the atom is attached4) Neglect the atom far from the seed5) Draw the cluster
www.compphys.cn
Matlab: illustrationcolormap
image()imagesc()pcolor()
www.compphys.cn
Cluster Growth Model►Eden Model
Eden Cluster
Consider a two dimensional lattice of points (x, y).
Placing a seed particle at the origin (x = 0, y = 0).
Growing by the addition of particles to its perimeter.
unoccupied near-neighbor sites as the perimeter sites of the cluster.
Choose one of these perimeter sites at random and place a particle at the chosen
location.
This process is then repeated; update perimeter and particle.
Continue this process until a cluster of the desired size is obtained.
www.compphys.cn
clearx=[0 0]; %initialnearest=[0 1 0 -1 1 0 -1 0];edge=nearest;%perimeter for ii=1 growsite=ceil(length(edge)*rand); tmp=ones(4,1)*edge(growsite,:)+nearest; x=[x edge(growsite,:)]; tmp=[tmp edge]; edge=setdiff(tmp,x,'rows');end
x
nearest
x
edge
tmp
www.compphys.cn
Dimensionality of the cluster
where r is small enough.
For a straight line:
( ) 2m r r
2( )m r r
www.compphys.cn
Eden vs. DLA cluster
www.compphys.cn
Morphology of a Class of Kinetic Growth Models
PRL,55,2515(1985)
(a) p = l, (b) p = 0.9, (c) p = 0.8, (d) p = 0.7,(e) p = 0.6, (f) p = 0.545.
If we modify our model so that all perimeter sites are active growth sites for all time, then we approach the Eden model in the limit p->0.
Place a seed particle at a site on a two dimensional square lattice.1)Check the four neighbors of the seed and occupy each one, independently, with a probability p2)Sample the nearest neighbors of the second generation and fill these sites independently with a probability p.
sites which are not filled are blocked and cannot be filled at a later time.
www.compphys.cn
clearx=[0 0]; %initialedge=[0 1 0 -1 1 0 -1 0];nearest=edge;p=0.6 % propobality unactive=[];
for ii=1:200%generation growsite=rand(size(edge,1),1)<p; if max(mod(growsite+1,2))>0 unactive=[unactive edge((growsite==0),:)]; end x=[x edge(growsite,:)]; tmp=edge(growsite,:);
for jj=1:size(tmp,1) edge=[edge repmat(tmp(jj,:),size(nearest,1),1)+nearest]; end
edge=setdiff(edge,x,'rows'); if length(unactive)>0 edge=setdiff(edge,unactive,'rows'); endend
plot(x(:,1),x(:,2),'*')hold onplot(edge(:,1),edge(:,2),'o')axis equalplot(unactive(:,1),unactive(:,2),'ro')
1) Select the active sites with p2) Record the unactive sites3) Add atoms and update edge4) Delete edge from cluster and unactive sites
www.compphys.cn
The role of energy
Shapes in square lattice:1) Diamond2) Square3) Triangular
www.compphys.cn
Simulation of cluster
1) Adding atoms2) Atom diffusion allowed Adding and deleting to conserve
the number of atoms3) Energy estimation4) Accept or reject the configuration
www.compphys.cn
Spanning clusters
www.compphys.cn
Percolation Problems
1. Porous rock (Original percolation problem,
Broadbent and Hammersley, 1957)
2. Forest fires, etc
Suppose a large porous rock is submerged under water for a long time, will the water reach the center of the stone?
p=0.48
How far from each other should trees in a forest (orchard) be planted in order to minimize the spread of fire (blight)?
www.compphys.cn
What is percolation
2-dimension percolation
6x6
120x120
Infinite x infinite (critical coverage)
2x2 lattice
Percolated system if a spanning cluster exist (connects top and bottom exists)
What is the probability for a system to be percolated for a given coverage?
?
www.compphys.cn
Simulation of Percolation
www.compphys.cn
Percolation clear, clf, colormap gray;
M=24; p=0.7;A=rand(M,M);
for i=1:M; for j=1:M;if(A(i,j)<p ) A(i,j)=0; else A(i,j)=1; end end; end
imagesc(A)
for ii=1:size(A,1) for jj=1:size(A,2) text(jj,ii,num2str(A(ii,jj))); hold on endend
www.compphys.cn
Divide the data into two groups
A=[ 0 0 1 1 0 1 1 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 0 0 1 0 0 1 0 0 1 1 0 1 0 1 0 0 0 1 0 0 0 0 0 0 1 0 1 1 1 0 0 0 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0];
rock=[];hole=[];for ii=1:size(A,1) for jj=1:size(A,2) if A(ii,jj)==1 rock=[rock ii jj]; else hole=[hole ii jj]; end endend
www.compphys.cn
Find the spanning clustersfor ii=3%1:size(hole,1) tp1=hole(ii,:); tp2=[tp1 repmat(tp1,size(nearest,1),1)+nearest]; tp2=intersect(tp2,hole,'rows'); while size(tp1,1)<size(tp2,1) tp1=tp2; for jj=1:size(tp1,1) tp2=[tp2 repmat(tp1(jj,:),size(nearest,1),1)+nearest]; end tp2=intersect(tp2,hole,'rows'); end cluster(ii,:) =[min(tp2(:,1)) max(tp2(:,1)) min(tp2(:,2)) max(tp2(:,2))];end
www.compphys.cn
Homework
For lecture notes, refer to http://www.compphys.cn/~xbyang/
Sending to 17273799@qq.com when ready
Apply the model in PRL,55,2515(1985)for triangular and hexagonal lattice.
top related