490 likes | 623 Views
Dynamical heterogeneity and relaxation time very close to dynamical arrest. L. Cipelletti LCVN, Université Montpellier 2 and CNRS Collaborators: L. Berthier (LCVN), V. Trappe (Fribourg) Students: P. Ballesta, G. Brambilla, A. Duri, D. El Masri Postdocs: S. Maccarrone, M. Pierno.
E N D
Dynamical heterogeneity and relaxation time very close to dynamical arrest L. Cipelletti LCVN, Université Montpellier 2 and CNRS Collaborators: L. Berthier (LCVN), V. Trappe (Fribourg) Students: P. Ballesta, G. Brambilla, A. Duri, D. El Masri Postdocs: S. Maccarrone, M. Pierno
Soft glassy/jammed materials E. Weeks
Dynamical heterogeneity is ubiquitous! Repulsive disks Granular matter Colloidal Hard Spheres Keys et al. Nat. Phys. 2007 Weeks et al. Science 2000 A. Widmer-Cooper Nat. Phys. 2008
Outline • Average dynamics and dynamical hetrogeneity of supercooled colloidal HS • Temporal fluctuations of the dynamics in jammed/glassy materials • Real-space measurements of correlations: ultra-long range spatial correlations and elasticity
Equilibrium dynamics of supercooled HS Multispeckle DLS • PMMA in decalin/tetralin • a ~ 110 nm • polidispertsity 12.2% (TEM) fs(qa=2.75,t) t (sec) Brambilla et al., PRL 2009
j dependence of the relaxation time MCT : ta~ (jc-j)-g jc= 0.59, g = 2.6 VFT: ta~ exp[(j0-j)-d] j0= 0.614, d = 1 j0= 0.637, d = 2 Brambilla et al., PRL 2009
Glass/jamming transition in colloidal HS Fluid Glass (out of equilibrium) MCT jc ~ .57-.59 jrcp ~ .64 Fluid Jamming jc ~ .57-.59 jj = jrcp ~ .64 Fluid Glass Thermodynamic jc ~ .57-.59 j0 jrcp ~ .64
Dynamical heterogeneity x Weeks et al. Science 2000
Dynamical heterogeneity x Weeks et al. Science 2000 var[f(t)] c4(t) dynamical susceptibility
Dynamical susceptibility in glassy systems Supercooled liquid (Lennard-Jones) Lacevic et al., Phys. Rev. E 2002 c4=N var[Q(t)]
Dynamical susceptibility in glassy systems Spatial correlation of the dynamics x Weeks et al. Science 2000
Dynamical heterogeneity: the theoreticians’ trick Goal: calculate 4-point dynamical susceptibility c4 ~ size of rearranged region For colloidal HS at high j Berthier et al., J. Phys. Chem. 2007
DH in colloidal HS A useful relationship ~(x/a)3 • equilibrium • OK at high j j ~ 0.20 – 0.58 t (sec) dynamical susceptibility c4(t) Berthier et al., J. Phys. Chem.2007 Berthier et al., Science 2005
j dependence of the max of c4 MCT ~ (jc-j)-2 * High j : deviation from MCT Brambilla et al., PRL 2009
Outline • Average dynamics and dynamical hetrogeneity of supercooled colloidal HS • Temporal fluctuations of the dynamics in jammed/glassy materials • Real-space measurements of correlations: ultra-long range spatial correlations and elasticity
lag t time t Time Resolved Correlation (TRC) degree of correlationcI(t,t) = - 1 < Ip(t) Ip(t+t)>p < Ip(t)>p<Ip(t+t)>p Cipelletti et al. J. Phys:Condens. Matter 2003, Duri et al. Phys. Rev. E 2006
Average over t intensity correlation function g2(t) - 1 degree of correlationcI(t,t) = - 1 < Ip(t) Ip(t+t)>p < Ip(t)>p<Ip(t+t)>p Average dynamics g2(t) - 1 ~ f(t)2
Average over t intensity correlation function g2(t) - 1 degree of correlationcI(t,t) = - 1 < Ip(t) Ip(t+t)>p < Ip(t)>p<Ip(t+t)>p fixed t, vs.t fluctuations of the dynamics Average dynamics g2(t) - 1 ~ f(t)2 var[cI(t)] ~ c(t)
(Polydisperse) colloids near rcp • PVC in DOP • a ~ 5 µm • polidispertsity ~33% • slightly soft • jclose to rcp • multiple scattering (DWS) • L ~ 10 nm << a Ballesta et al., Nature Physics 2008
Non-monotonic behavior of c*(j) Fit: g2(t)-1 ~ exp[-(t/t0)b] relaxation time stretching exponent peak of dynamical susceptibility non monotonic!
Non-monotonic behavior of c*(j) • Competition between: • increasing x (c* as j ) • increasingly restrained displacement • jump size decreases • (many events over t0, c* as j ) • Model: • random events of size x • Poissonian statistics • Quasi-ballistic motion • ( <dr2> ~np, p = 1.65)
Simulating the model cell thickness! inverse jump size volume fraction
Outline • Average dynamics and dynamical hetrogeneity of supercooled colloidal HS • Temporal fluctuations of the dynamics in jammed/glassy materials • Real-space measurements of correlations: ultra-long range spatial correlations and elasticity
sample object plane Dq Dq lens focal plane diaphragm image plane CCD Measuring x by Photon Correlation Imaging 2.3 mm Duri et al., Phys. Rev. Lett. 2009 q = 90° 1/q ~ 45 nm
Local, instantaneous dynamics: cI( t, t,r) < Ip(t) Ip(t+t)>p(r) cI(t, t , r) = - 1 < Ip(t)>p<Ip(t+t)>p(r) Note: <<cI(t, t , r)>t>r = g2(t)-1 [g2(t)-1] ~ f(t)2 2.3 mm
Dynamic Activity Maps Brownian particles g2(t)-1~ exp[-t/tr], tr = 40 s Colloidal gel g2(t)-1~ exp[-(t/tr)1.5], tr = 5000 s cI (t0,tr/200 , r) Movie accelerated 10x 2 mm cI (t0,tr/10 , r) Movie accelerated 500x 2 mm
Spatial correlation of the dynamics: x ~ system size in jammed soft matter! Maccarrone et al., Soft Matter 2010
When does x infinity? • G' > G'' : strain propagation in a predominantly solid-like material • Magnitude of G'unimportant (G'onions/G'colloidal gel ~ 6x104 !!) • Origin of elasticity: • entropic(e.g. hard spheres) x very small • enthalpic (attractive gels, squeezed particles...) xsystem-size
Evidence of internal stress relaxation Dynamics of actin/fascin networks strain field @ late stages of network formation J. Kaiser, O. Lielig,, G. Brambilla, LC, A. Bausch, submitted age average strain field microcopic dynamics
Conclusions • DH a general feature of glassy/jammed dynamics • Supercooled hard spheres: • equilibrium dynamics above the (apparent) MCT divergence • xlimitedto 5-10a • Jammed materials: • x and cmay decouple! • x ~system size • role of the origin of elasticity
Thanks to... L. Berthier (LCVN), V. Trappe (Fribourg) Hard spheres G. Brambilla, M. Pierno, D. El Masri (LCVN), A. Schofield (Edinburgh), G. Petekidis (FORTH) TRC A. Duri, P. Ballesta (LCVN), D. Sessoms, H. Bissig (Firbourg) PCI A. Duri, S. Maccarrone (LCVN), D. Sessoms (Fribourg), E. Pashkowski (Unilever) Funding CNES, ACI, ANR, PICS, Unilever
Open questions... DH and aging? Origin of the dynamics in (undriven) soft materials
Determining the volume fraction j dependence of the short-time diffusing coefficient t = 1/Dssq2 j = 0.2 Absoluteuncertainty: ~4% Relative uncertainty: ~ 10-4
PS gel: time-averaged dynamics g2(q,t) - 1 ~ [f(q,t)]2 • Fast dynamics: overdamped vibrations • (~ 500 nm) Krall & Weitz PRL 1998 • Slow dynamics: rearrangements
PS gels: q dependence of tf and p « ballistic » motion « compressed » exponential
PS gel: temporally heterogeneous dynamics cI(tw,t) @ fixed t : temporal fluctuations <cI(tw,t)>tw:average dynamics
Length scale dependence of c = var(cI) PS gel Increasing q 1.1 +/- 0.1 Duri & LC, EPL 76, 972 (2006)
Scaling of c* c* ~ var(n)/<n> ~ 1/<n> <n> ~ tf ~ 1/q c* ~ q
Dynamical susceptibility in glassy systems Supercooled liquid (Lennard-Jones) Lacevic et al., Phys. Rev. E 2002 c4=N var[Q(t)]
Dynamical susceptibility in glassy systems N regions c4=N var[Q(t)] c4 dynamics spatially correlated
Dynamical susceptibility in glassy systems Spatial correlation of the dynamics x Weeks et al. Science 2000
The smart trick applied to colloidal HS cf(t) F(t) Df 10-3 t Define
Dynamical heterogeneity: the theoreticians’ trick Goal: calculate 4-point dynamical susceptibility c4 ~ size of rearranged region For colloidal HS at high j Berthier et al., J. Phys. Chem. 2007
Evidence of a growing dynamic length scale j f ~ 0.20 – 0.58 Berthier et al., Science 2005
j dependence of the max of c4 * High j : deviation from MCT MCT Brambilla et al., PRL 2009
Photon Correlation Imaging (PCIm) Dq object plane focal plane image plane q CCD q sample lens annular slit 2.3 mm q = 6.4° q = 1 mm-1. Duri et al., Phys. Rev. Lett. 2009
Other PCIm geometries object plane sample beam splitter object plane sample Dq Dq lens focal plane diaphragm Dq Dq image plane lens CCD focal plane diaphragm image plane CCD a) b)