Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Lecture 4Microstates, Multiplicity of a macrostate
29.08.2019
Fys2160 2019 1
Microstates and macrostates
• Microstates:• MD, ideal gas: particular xi and vi
(not countable)• RW, particular xi and vi (countable
on grid)• Algorithmic: single particle states
si=1 (left), si=0 (right)
• Macrostates• n = number of particles on left
side• n = 0, 1, 2, ..., N• Algo: n=Si si
Fys2160 2018 2
Microstates and macrostates• Box with left and right side• Example: 𝑁 = 5• Particles can be distinguished (i=1,
2,.. 5)• Particle states left:
• si=1• right: si=0
• macrostates n=Si si (= 1,2, ..5)• Total number of microstates: 25=32• List the possible microstates of n=1
Fys2160 2018 3
• 00001• 00010• 00100• 01000• 10000
5 microstates => multiplicity 𝛀(n,N) of macrostate n=1 is 𝛀(1,5) = 5
Multiplicity of macrostates
4
n=200011001010100110001001100101001100100101010011000
𝛀(0,5)=10
n=10000100010001000100010000
𝛀(0,5)=5
n=311100110101011001110110011010110011011010101100111
𝛀(0,5)=10
n=41111011101110111011101111
𝛀(4,5)=5
n=511111
𝛀(5,5)=1
n=000000
𝛀(0,5)=1
Number of possible microstates: Ω. = ∑0123 Ω 𝑛 = 32(= 23)
General formula for multiplicity: Ω 𝑛,𝑁 = :!:<0 !0!
Ω 2,5 = 3!=!>!
= 10
Probability of macrostates: P n, N = Ω 𝑛,𝑁 /2: = >EF:!:<0 !0!
Plots of multiplicity and probability
Fys2160 2018 5
Summary: multiplicity of a system with Nparticles in a macrostate with n particles onthe left side
Number of microstates with n on left side: Ω 𝑛,𝑁 = :!0! :<0 !
Total number of microstates
G012
:
Ω 𝑛,𝑁 = G012
:𝑁!
𝑛! 𝑁 − 𝑛 ! = 2:
Probability of a macrostate with n on left side:
𝑃 𝑛,𝑁 =Ω 𝑛,𝑁
∑0Ω 𝑛,𝑁=
𝑁! 2<:
𝑛! 𝑁 − 𝑛 !
Fys2160 2018 6
n N-n
Apply this type of combinatorics to
1. Paramagnetic systems
2. Random walks
3. Thermal vibrations in crystals
Fys2160 2018 7
Two-state paramagnet model
Paramagnetic solid:
• A system of N independent, localised particles with spin 𝑠 = ±1 in a constantmagnetic field 𝑩
• Energy of a single spin 𝜖 = −𝑠𝜇𝐵
• For 𝑁 spins , we have 𝑁 = 𝑁↓ + 𝑁↑• Net magnetization
𝑴 = 𝝁G𝒊1𝟏
𝑵
𝒔𝒊 = 𝝁 𝑵↑ − 𝑵↓ = 𝝁(𝑵 − 𝟐𝑵↑)
• Average magnetization⟨𝑴⟩ = 𝝁(𝑵 − 𝟐⟨𝑵↑⟩)
Fys2160 2018 8
𝐁
𝜖+𝜇𝐵
0
−𝜇𝐵
𝑠 = −1
𝑠 = +1
Two-state paramagnet model, B=0
• Consider a paramagnet with N spins at zero applied field
• Spins ↑ or ↓ have the same energy
• Microstate is a particular configuration of spins ↑ and ↓
What is the multiplicity of macrostate with 𝑵↑ out of 𝑵 spins?
𝛀 𝐍,𝐍↑ = ?
Fys2160 2018 9
What is the multiplicity of macrostate with 𝑵↑ out of 𝑵 spins?
Ω 𝑁,𝑁↑ =𝑁!
𝑁↑! 𝑁 − 𝑁↑ !
Total number of microstates
G:↑12
:
Ω(𝑁↑) = G012
:𝑁!
𝑁↑! 𝑁 − 𝑁↑ != 2:
Probability of a macrostate with 𝑁↑ spins up
𝑃 𝑁↑ = 2<:Ω 𝑁,𝑁↑
Fys2160 2018 10
What is the average number of 𝑁↑?
𝑁↑ = G:↑12
:
𝑁↑ 𝑃 𝑁,𝑁↑
Fys2160 2018 11
𝑁↑ = 2<: G:↑12
:
𝑁↑𝑁!
𝑁↑! 𝑁 − 𝑁↑ !
𝑁↑ =𝑁2
Macrostate with𝑁↑ = 𝑁↑ has the largest multiplicity
• 𝑁↑ = :>
• ⟨𝑴⟩ = 𝝁(𝑵 − 𝟐⟨𝑵↑⟩)
• ⟨𝑴⟩ = 𝟎
• In the absence of an external magneticfield, spins are randomly oriented with a zero net magnetization
Fys2160 2018 12
1D Random walk
• Random motion of a walker along a line• Discrete time steps 𝑁 = 0,1,2⋯ in units of Δ𝑡 = 1• Discrete space: lattice index 𝑗 = 0,±1,±2⋯ with increments Δ𝑥 = 1
• At each timestep, the walker has probability 𝑝 = f>
to the right 𝑗 →𝑗 + 1
and probability 𝑞 = f>
to the left 𝑗 → 𝑗 − 1
• What is the probability distribution for R steps to the rightN steps, 𝑃(𝑁, 𝑅)?
• What is the mean displacement ⟨𝑆⟩ after 𝑁 steps?
Fys2160 2018 13
𝑝 =12
𝑞 =12
𝑗 + 1𝑗𝑗 − 1
1D Random Walk
Fys2160 2018 14
𝑝 =12𝑞 =
12
𝑗 + 1𝑗𝑗 − 1
1D Random Walk
Fys2160 2018 15
𝑝 =12𝑞 =
12
𝑗 + 1𝑗𝑗 − 1
1D Random walk
After 𝑁 steps, we have 𝑅 steps to the right and 𝐿 steps to the right
𝑅 + 𝐿 = 𝑁, 𝑆 = 𝑅 − 𝐿 (net displacement)
• Number of configurations in which we have 𝑅 right steps out of 𝑁 steps
Ω 𝑁,𝑅 =𝑁!
𝑅! 𝑁 − 𝑅 !
• Probability for 𝑅 steps to the right out of 𝑁 steps
𝑃 𝑁,𝑅 = Ω 𝑁,𝑅12
l 12
:<l= 2<:
𝑁!𝑅! 𝑁 − 𝑅 !
Fys2160 2018 16
𝑝 =12𝑞 =
12
𝑗 + 1𝑗𝑗 − 1
1D Random walk
• Probability for 𝑅 steps to the right out of 𝑁 steps
𝑃 𝑁, 𝑅 = 2<:𝑁!
𝑅! 𝑁 − 𝑅 !
• Normalization condition: probability for 𝑁 steps
Gl12
:
𝑃(𝑁, 𝑅) = 2<: Gl12
:𝑁!
𝑅! 𝑁 − 𝑅 != 1
Fys2160 2018 17
𝑝 =12𝑞 =
12
𝑗 + 1𝑗𝑗 − 1
Average displacement from the origin
𝑅 =𝑁2
𝑆 = 𝑅 − (𝑁 − 𝑅 ) = 2 𝑅 − 𝑁
𝑆 = 0
Fys2160 2018 18
𝑝 =12𝑞 =
12
𝑗 + 1𝑗𝑗 − 1
Thermal vibrations in solids: Einstein crystal
• Collection of identical harmonicoscillators
• Each atomin3Dhas3one-dimensional harmonic oscillators
• Classical harmonic oscillator
• 𝑈f =z{
>|+ f
>𝑚𝜔>(𝑥 − 𝑥2)>
• Frequency 𝜔 = �|
Fys2160 2018 19
𝑈 =𝑝>
2𝑚+12𝜅𝑥>
𝑉 = 0, 𝑇 =𝑝>
2𝑚
𝑉 =12𝜅𝑥>, 𝑇 = 0𝑈
Einstein crystal modelOne quantum harmonic oscillator
�𝐻 =�̂�>
2𝑚+12𝜅�𝑥> =
�̂�>
2𝑚+12𝑚𝜔> �𝑥>
Quantized energy levels 𝜖 = 𝑛 + f>ℏ𝜔
Energy level relative to the ground state
Δ𝜖 = 𝑛ℏ𝜔
Total energy of N harmonic oscillators
𝑈: = ∑�1f: 𝜖� = ∑�1f: 𝑛� ℏ𝜔 +:>ℏ𝜔
Energy units (integer numbers) for N quantum harmonic oscillators at frequency 𝝎
𝐪 =𝑼𝑵 −
𝑵𝟐 ℏ𝝎
ℏ𝝎 Fys2160 2018 20
Find multiplicity Ω(𝑞, 𝑁) of a macrostate with 𝑁 oscillators and 𝑞 units of energy distributed between them
𝑞 = 1, 𝑁 = 4
Ω 1,4 = 4
Fys2160 2018 21
#1 #2 #3 #4
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
Find multiplicity Ω(𝑞, 𝑁) of a macrostate with 𝑁 oscillators and 𝑞 units of energy distributed between them
𝑞 = 2, 𝑁 = 4
Ω 2,4 =?
Fys2160 2018 22
Ω 2,4 = 10
1 oscillators has 2 energy quanta, and 3 oscillators are in the ground state
2 oscillators have 1 energy quanta each, and 2 oscillators are in the ground state
2 0 0 0
1 1 0 0
2000 11000200 10100020 10010002 0101
00110110
Trick: Call balls 0 and walls 1
𝑞 = 2, 𝑁 = 4
Ω 2,4 =?
Fys2160 2018 23
Ω 2,4 =𝑁 − 1 + 𝑞 !𝑞! 𝑁 − 1 ! = 10
1 oscillators has 2 energy quanta, and 3 oscillators are in the ground state
2 oscillators have 1 energy quanta each, and 2 oscillators are in the ground state
0 0 1 1 1
0 1 0 1 1
00111011101110001011011011101010110101011001101011
N-1 wallsq balls=> 2 states, N-1+q particles, N-1
General formula for multiplicity formacrostate• n in state 1 (N-1)• in system with 2 states• N particles (N-1+q)
Ω 𝑛,𝑁 =𝑁!
𝑁 − 𝑛 ! 𝑛!
Find multiplicity Ω(𝑞, 𝑁) of a macrostate with 𝑁 oscillators and 𝑞 units of energy distributed between them
Ω 3,4 =?
Fys2160 2018 24
Ω 3,4 = 20
3 oscillators are in an excited state, and the 4th is in the ground state
1 oscillators has 2 energy quanta, 1 oscillator has 1 energyquanta and 2 oscillators are in the ground state
1 oscillators has 3 energy quanta, and 3 oscillators are in the ground state
Weakly-coupled Einstein crystals
Ω� 𝑞�,𝑁� =𝑁� − 1 + 𝑞� !𝑞�! 𝑁� − 1 !
Ω 𝑞�,𝑁� =𝑁� − 1 + 𝑞� !𝑞�! 𝑁� − 1 !
Composite system: 𝑞 = 𝑞� + 𝑞�, 𝑁 = 𝑁� + 𝑁�
Multiplicity of a macrostate of the composite systems
𝜴𝐭 = 𝛀𝑨 𝒒𝑨,𝑵𝑨 ⋅ 𝛀𝑩(𝒒𝑩,𝑵𝑩)
Fys2160 2018 25
CrystalA𝑁�, 𝑞�
CrystalB𝑁�, 𝑞�
Weakly-coupled Einstein crystals
Multiplicity of a macrostate with 𝑞�and 𝑞� = 𝑞 − 𝑞� for two coupledEinstein solids
𝜴𝐭(𝐪𝐀, 𝐪, 𝐍)= 𝛀𝑨 𝒒𝑨,𝑵𝑨 ⋅ 𝛀𝑩(𝒒𝑩,𝑵𝑩)
What the macrostate with themaximum multiplicity ?
Matlab tests..
Fys2160 2018 26
CrystalA𝑁�, 𝑞�
CrystalB𝑁�, 𝑞�
Weakly-coupled Einstein crystals
Multiplicity of a macrostate with 𝑞� and 𝑞� = 𝑞 − 𝑞� for two coupled Einstein solids
𝜴𝐭(𝐪𝐀, 𝐪, 𝐍) = 𝛀𝑨 𝒒𝑨,𝑵𝑨 ⋅ 𝛀𝑩(𝒒𝑩,𝑵𝑩)
What the macrostate with the maximummultiplicity ?
Matlab tests..
Fys2160 2018 27
CrystalA𝑁�, 𝑞�
CrystalB𝑁�, 𝑞�
Take home---• Multiplicity in two-state systems (paramagnets, random walk)
Ω 𝑛,𝑁 =𝑁!
𝑛! 𝑁 − 𝑛 !
has a maximum peaked around the average value ⟨𝑛⟩
• Multiplicity in Einstein crystal by analogy with q «identical balls» (energy units) and N «identical bins» (oscillators)
Ω 𝑞,𝑁 =(𝑁 − 1+ 𝑞)!(𝑁 − 1)! 𝑞!
• Multiplicity of two-coupled Einstein crystal
Ω� = Ω� ⋅ Ω�
has also a maximum around the average value
• Macrostates with maximum multiplicity are the most likely and they correspond to the average values
Fys2160 2018 28
Fys2160 2018 29
NA = 300; NB = 100; % Nr of oscillatorsqA = 500; qB = 500; % Initial energyq = qA + qB; % Total energyN = NA + NB; % Total oscillatorsstate = zeros(N,1); % Microstate of the system% state(1:NA) is part A, state(NA+1:NA+NB) is part B% Generate initial, random microstateplaceA = randi(NA,qA,1);figure(1)for ip = 1:qA;
state(placeA(ip)) = state(placeA(ip)) + 1;endplaceB = randi(NB,qB,1);for ip = 1:qB;
i = NA + placeB(ip);state(i) = state(i) + 1;
end% Simulate random state developmentnstep = 100000; % Number of stepsEA = zeros(nstep,1); EB = zeros(nstep,1);for istep = 1:nstep
i1 = randi(N,1,1); % Select random osc. i1if (state(i1)>0) % Does it have any energy ?
i2 = randi(N,1,1); % Select random osc. i2state(i2) = state(i2) + 1; % Transferstate(i1) = state(i1) - 1; % energy
endEA(istep) = double(sum(state(1:NA)))/NA;EB(istep) = double(sum(state(NA+1:end)))/NB;if rem(istep,100)==1
subplot(2,1,1) % Microstate of systemplot((1:NA),state(1:NA),'b',(1:NB)+NA,state(NA+1:end),'r');xlabel('i'); ylabel('n_i'); drawnowsubplot(2,1,2); % Avg energy in each systemplot((1:istep),EA(1:istep),'-b',(1:istep),EB(1:istep),'-r');xlabel('t'); ylabel('q_A/N_A , q_B/N_B'), drawnow
endendline([0 nstep],[q/N q/N])
• NA=3NB => this means that the crystal is 3 times bigger• qA=qB => this means they have the same total energy• Temperature is related to energy per oscillator• Therefore temperature of crystal B is TB=3TA• Random transfer of energy => towards thermal equilibrium• Thermal equilibrium = equal temperature
Fys2160 2018 30
% Two Einstein solids NA & NB oscillators% in contact to exchange energyNA = 30; % Number of oscillators in crystal ANB = 10; % Number of oscillators in crystal Bq = 50; % qA+qB total energy of crystals A and B% Find multiplicity of all macrostatesnstate = q+1; % number of statesqA = 0:q; % all possible values of qAqB = q - qA; % all possible values of qBOmegaA = zeros(nstate,1);OmegaB = OmegaA;FontSize=20;% Loop through all macrostates and find multiplicityfor i = 1:nstate
OmegaA(i) = nchoosek(qA(i)+NA-1,qA(i)); %binomialOmegaB(i) = nchoosek(qB(i)+NB-1,qB(i)); %binomial
endOmegaTOT= OmegaA.*OmegaB; % Total multiplicityfigure(1)subplot(1,2,1), plot(qA,OmegaTOT/max(OmegaTOT),'-o');subplot(1,2,2), semilogy(qA,OmegaTOT/max(OmegaTOT),'-o');% pretty plotting...maxO=sprintf('%0.3g',max(OmegaTOT));title(['max(\Omega_{Tot}) = ',maxO],'FontSize',FontSize)for i=1:2
subplot(1,2,i)xlabel('q_A','FontSize',FontSize)ylabel('\Omega_{Tot}(N_A, N_B, q_A) / max(\Omega_{Tot})','FontSize',FontSize)h=gca;h.FontSize=FontSize;
end%%figure(2)P=OmegaTOT/sum(OmegaTOT);subplot(1,2,1), plot(qA,P,'-o');subplot(1,2,2), semilogy(qA,P,'-o');% pretty plotting...FontSize=20;for i=1:2
subplot(1,2,i)xlabel('q_A','FontSize',FontSize)ylabel('P(N_A, N_B, q_A)','FontSize',FontSize)h=gca;h.FontSize=FontSize;
end[Pmax,qmax]=max(P);subplot(1,2,1), text(60, Pmax+0.01,['q_A(P_{max})=',num2str(qmax)],'FontSize',FontSize-2)%%qA0=q/2; % (Hypothetical) initial energy of crystal A[Pmax,qmax]=max(P);qmax=qmax-1; % Matlab counts from 1.., q goes from 0..figure(2)subplot(1,2,1), text(qA0-10, Pmax/2,['q_{A,0}=',num2str(qA0)],'FontSize',FontSize-2)subplot(1,2,1), hold onplot([qA0 qA0],[0 max(OmegaTOT/sum(OmegaTOT))],'k-')plot([qA0 qA0],[0 max(OmegaTOT/sum(OmegaTOT))],'k-')hold offsubplot(1,2,2), hold on, semilogy([qA0 qA0],[1e-50 1],'k-'), hold offPmaxoverP0=sprintf('%0.5g',Pmax/P(qA0));subplot(1,2,2), text(30,1e-20,['P_{max}/P(q_{A0})=',PmaxoverP0],'FontSize',FontSize-2)