monte carlo simulations-surface barriers in nanoporous materials
DESCRIPTION
For caculate the surface forcesTRANSCRIPT
![Page 1: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/1.jpg)
U i ki ti M t C l i l ti Using kinetic Monte Carlo simulations for investigating
surface barriers in nanoporous materials surface barriers in nanoporous materials
L H i kLars HeinkeFritz-Haber-Institute of the Max-Planck-Society, Berlin
Jörg KärgerUniversity Leipzigy p g
Com-Phys-09 Workshop,Leipzig 26.11.2009
![Page 2: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/2.jpg)
Introduction – nanoporous materials
Nanoporous crystals: Nanoporous crystals: • Periodic, regular structure• Pore size: few Å - few nm• Zeolites and metal-organic frameworks (MOF)
Applications:• Adsorption (e.g. of water)• Ion exchange (e.g. water softener)• Molecular sieve (mass separation gas purification)• Molecular sieve (mass separation, gas purification)• Catalysts (e.g. in petrochemistry)
Mass transport crucial!
![Page 3: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/3.jpg)
Experimental data - recorded concentration profiles
Best way for studying mass transfer concentration profiles!(Pressure step new equilibrium concentration mass transfer)
Investigated system: propane in MOF Zn(tbip) (one-dimensional pore structure)
equilibriumconcentration ceq
initialconcentration c00
![Page 4: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/4.jpg)
Experimental data - recorded concentration profiles
surface barrier
intracrystalline diffusion
Surface barrier describedby surface permeability α.
equilibriumconcentration c
i.e. thin layer with permeability α
boundary concentration c (t)
ceqsurf eq surf( )j c c
csurf(t)
Mass transfer often limited by surface barriers! (not only by diffusion in the bulk crystal)
![Page 5: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/5.jpg)
Transport, Maxwell-Stefan and self-diffusivity and surface permeability of propane
Transport parameters of propane in MOF Zn(tbip)
10-12s-1 m s
-1
p , y p y p p
10-8
10-13
10
usiv
ity D
/ m
2 s
erm
eabi
lity
/
0.0 0.2 0.4 0.6 0.8 1.0
10-9
0.0 0.2 0.4 0.6 0.8 1.0
concentration c/ molecules per segment
mean boundary concentration (ceq+csurf)/2
/ molecules per segment
diffu
sur
face
pe
/ molecules per segment
Diffusivity and surface permeability same concentration dependence!
h h d b /Furthermore: Ethane, propane and n-butane same α / D
Both facts can not be explained by a smaller pore diameter at surface
or by entropic effects!
![Page 6: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/6.jpg)
Model of surface barriers
• Crystal structure lattice (1 pore segment = 1 lattice site)• Most pore entrances blocked, molecules enter only via open pore entrances
(percentage of open pores popen)P ll l h ti t l d f t• Parallel pores have cross connections = crystal defects(percentage of cross connections py = pz)
• Gas phase = reservoir with fixed concentration (ceq)
![Page 7: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/7.jpg)
yz
Kinetic Monte-Carlo simulations
Details for simulations:• Lattice jump model• Every time step nt every molecule makes jump attempt
xz
• Restriction:– Each site can be occupied by only one molecule (follows from isotherm)– jumps in y- or z-direction can only occur when adjacent pores are cross
connectedconnected
• No interaction between the particles (hard-sphere model) or nearest-neighbour interaction (Reed-Ehrlich model)
• fixed concentration in the reservoir ceq
ceq > c0 adsorption experiment < d ti i tceq < c0 desorption experiment
ceq = c0 self-diffusion experiment
Assumption: Jump rate in x, y and z are equal.Assumption: Jump rate in x, y and z are equal.
![Page 8: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/8.jpg)
Simulated concentration profiles
nt = 10,000time nt
Concentration profiles along xcorrespond to solution of diffusion equation with surface barriers
Concentration profiles perpendicular to xare curved near the surface and flat in the bulk crystalequation with surface barriers.
(D and α = const.)and flat in the bulk crystal.
![Page 9: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/9.jpg)
Resulting surface permeability
Surface permeability• constant during transfer process(if diffusivity constant; α ~ jump rate ~ D)α jump rate D)
• independent of equilibrium concentration • independent of length of pore system
![Page 10: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/10.jpg)
Surface permeability depends on
Resulting surface permeability
p y p• fraction of open pores (popen)• probability of cross connections (py = pz)
( ) d / D t α (popen, py) and α / D = const.
pp5
y
y
y
y
pp
pp
421
415
~open~ p
![Page 11: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/11.jpg)
Now: Interaction between particles Diffusivity concentration dependent
Interaction between particles – Reed-Ehrlich model
Now: Interaction between particles Diffusivity concentration dependent
Nearest-neighbour interaction between particles on the lattice
UJump rate ~
(Reed-Ehrlich approach)
TkU
b
NeighbourNearest exp
concentration dependence 5
concentration dependence similar diffusivity
(value of propane in MOF Zn(tbip))
All properties of the surface barriers presented! All properties of the surface barriers presented!
![Page 12: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/12.jpg)
Self-diffusion simulations with concentration c = 0.3 and c = 0.5 (Φ=5)
Determining popen in MOF Zn(tbip)
Ensemble average
Self diffusion simulations with concentration c 0.3 and c 0.5 (Φ 5)
g
Time average
Experimental data:Propane in MOF Zn(tbip)
p ≈ 0 05 py ≈ 0.05
![Page 13: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/13.jpg)
p can be adjusted to fit the experimental data
Determining py in MOF Zn(tbip)
popen can be adjusted to fit the experimental data (py = 0.05)
Propane in MOF Zn(tbip)
Fraction of open surface pores is about 0.033 % , i.e. one among 55 × 55 pores is open!
(Under the assumption of equal jump rates in all directions!)
![Page 14: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/14.jpg)
Conclusion
• Surface barrier represented by a high percentage of totally blocked pores Surface barrier represented by a high percentage of totally blocked pores which are cross-connected.
• Influence of different parameters in the model studied by means of kinetic M t C l i l tiMonte-Carlo simulations.
• All properties of the surface barriers of alkanes in MOF Zn(tbip) are very well represented!p
It can be concluded that the surface barriers are caused by this structure!
![Page 15: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/15.jpg)
Acknowledgment
Department of Interface Physics, University Leipzig
IRTG “Diffusion in Nanoporous Materials”
Studienstiftung des deutschen Volkes
Atomic-Force-Microscopy group and Chemical Physics department at the Fritz-Haber-Institutedepartment at the Fritz Haber Institute
Many thanks for your attention!
![Page 16: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/16.jpg)
![Page 17: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/17.jpg)
Variation of jump rates
Results are only valid under the assumptions of same jump rate everywhere.
What happens if jump rate in surface plane is smaller?
popen=0.033%py=0.05
Jump rate in surface up to 1 order of magnitude smaller no difference!
![Page 18: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/18.jpg)
Comparison with analytical results for aD
Simulated results vs. Continuum approach
Comparison with analytical results for continuum system (Dudko, Berezhkovskii, Weiss, J. Phys. Chem, 2005.))
If lattice can be approximated as continuum
22LaD
If lattice can be approximated as continuum, good approximation.
nx × ny
![Page 19: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/19.jpg)
Periodic or random distribution of open pores
![Page 20: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/20.jpg)
Surface barrier described by surface permeability. (inversely proportional) 1.0n
c
ceq
Mesoscopic description of the surface barriers
All quantities follow from concentration profilesSurface permeability experimentally determined!
surf eq surf( )j c c 0.4
0.6
0.8
con
cent
ratio csurf
eq
surf eq surf( )j
0 10 20 30 40 50
0.0
0.2surface
centre
( , )dc y t y
norm
aliz
ed
y / µm
10-6
left-hand sideright-hand side/ m
s-1
y / µm
10-7
right hand side
rmea
bilit
y
/
surface
centresurf
d ( )dd
( )c t y
tc
0.0 0.2 0.4 0.6 0.8 1.010-8
surfa
ce p
erboundary concentration c
surfeq surf
( )c c
boundary concentration csurf
![Page 21: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/21.jpg)
130 µm
Crystal structure of MOF Zn(tbip)
Metal organic framework (MOF) Zn(tbip) 25 µm
y
Parallel chains of pore segments
xz
Guest molecule: propane(complemented by ethane and n-butane)
![Page 22: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/22.jpg)
Experimental setup of interference microscopy
• Basic principle: optical path length depends on adsorbate concentration.• Spatial resolution: 0.5 µm × 0.5 µm
![Page 23: Monte Carlo Simulations-Surface Barriers in Nanoporous Materials](https://reader033.vdocuments.net/reader033/viewer/2022042702/55cf9919550346d0339b8c9c/html5/thumbnails/23.jpg)
Sticking probability Pst
Sticking probability
Sticking probability Pst= probability that a gas molecule that hits the crystal surface continues its trajectory in the pore space,i.e.
injP
surf eq surf in out( )j c c j j in eqj c
stgas on surface
Pj
• no concentration
N
dependence• desorption to
vacuum: ceq=0, jin=0
Sti ki b bilit
From kinetic theory of gases gas on surface 2AN pjR T M
Sticking probability:• Pst ≈ 6 × 10-6 for methanol in ferrierite• Pst ≥ 0.01 for isobutane in silicalite
Varies over many order of magnitude!
L. Heinke et al., Phys. Rev. Lett. 99 (2007) 228301.