You are on page 1of 10

Article

pubs.acs.org/Macromolecules

Theory of Localization and Activated Hopping of Nanoparticles in


Cross-Linked Networks and Entangled Polymer Melts
Zachary E. Dell and Kenneth S. Schweizer*,

Department of Physics and Department of Materials Science, University of Illinois, Urbana, Illinois 61801, United States
ABSTRACT: We develop a microscopic, force-level statistical dynamical theory for
the localization and activated hopping dynamics of dilute spherical particles, both
hard and soft, in polymer networks and entangled melts. The main factor
controlling localization is the connement a particle experiences from polymer
entanglements and/or chemical cross-links as characterized by the ratio of the
eective nanoparticle diameter to the mechanical mesh length, 2Reff/dT. Dynamical
localization occurs when the latter ratio is of order unity, and the activated hopping
mobility drastically decreases as the connement parameter even modestly increases
past threshold. Local packing correlations slightly enhance the tendency to localize
for hard particles and dramatically enhance mobility for soft repulsive particles. The
concept of an eective nanoparticle size largely, but not completely, collapses the
diverse dynamical results for hard and soft repulsive particles. For hard spheres in
entangled polymer melts we nd the exponentially slow hopping diusivity is
generally much smaller than transport via reptation-driven entanglement network
dissolution, except for a narrow window of connement parameters of 2R/dT 1.52 and long enough chains.

I. INTRODUCTION
Nanoparticle motion in diverse crowded environments is of
broad interest, with applications in physics, material science,
cell biology, and medicine. Understanding the motion of a
tagged nanoparticle can provide length-scale-dependent information about the properties of structurally and/or
viscoelastically complex matrix materials.14 In polymer
nanocomposites and lled elastomers, nanoparticles can
provide massive reinforcement of elastic and other mechanical
properties.57 In biology and medicine, molecular transport
underlies many cell functions and drug delivery applications.8,9
The problem has begun to be addressed theoretically, albeit
largely at a tentative and single particle diusion level.1016
Of specic interest in this article is how nanoparticles diuse
in cross-linked polymer networks and entangled polymer melts
in the dilute tracer limit. Yamamoto and Schweizer have
previously developed a microscopic statistical mechanical
theory at the level of forces for spherical one10 and two11
particle diusivity in unentangled and entangled polymer melts
built on a combination of Brownian motion, polymer physics,
and mode coupling theory (MCT) ideas. By adopting a
constraint release dynamical picture, appropriate when particles
are not too small compared to the entanglement tube diameter,
single particle mobility is driven solely by length-scaledependent collective relaxation of polymer density uctuations.
Their primary nding was nanoparticle diusivity can be
massively enhanced relative to the continuum StokesEinstein
prediction due to a gradual coupling of the particle to the
transient entanglement network constraints as its size increases.
A second, complementary advance extends the self-consistent
general Lengevin equation (GLE) framework developed in refs
17 and 18 to the case of small nanoparticles which decouple
2013 American Chemical Society

from the transient entanglement network. The resultant theory


is in good agreement with simulations17 of hard particle
diusion in both unentangled and lightly entangled melts.
The present article builds on the above work but addresses a
dierent, third regime where nanoparticle motion is via
thermally induced activated barrier hopping. Such a rare
event becomes the only option in chemically cross-linked
polymer networks when the eective mechanical mesh size
becomes smaller than the particle diameter. It also may be
important over a narrow range of particle to entanglement
mesh diameter ratios in melts. To address hopping transport
requires going beyond the dynamically Gaussian GLE approach
previously developed.10,17 We employ the nonlinear Langevin
equation (NLE) theory, widely utilized for colloidal glasses and
gels and supercooled polymer melts.1922 It is based on the
concept of a dynamic free energy and is a quantitative,
microscopic, force-level description. Three questions are
addressed. First, under pure excluded volume conditions
(athermal, nonadsorbing polymers), how does the onset of
transient hard sphere localization, the localization length,
activated hopping time, and diusivity depend on cross-linked
network and nanoparticle variables? Second, under what
(subtle) conditions can hopping control particle diusion in
entangled polymer melts? Third, what is the inuence on
diusion of nonexcluded volume forces, specically polymer
particle attractions and soft repulsions? The latter is a toy
model of sterically stabilized particles, nanogels, vesicles, and
other soft objects.
Received: October 17, 2013
Revised: November 25, 2013
Published: December 17, 2013
405

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

Section II presents the models and dynamical theories


employed. Analytic results for the athermal hard sphere system
are derived and discussed for networks and melts in sections III
and IV, respectively. Section V establishes the role of variable
polymerparticle interactions and local packing correlations on
diusivity. The paper concludes in section VI with a brief
summary and future outlook.

II. THEORY
A. Models. We consider spherical nanoparticles of core
radius Rcore larger than the polymer monomer size. The
polymer is described at the lightly coarse-grained level of a
freely jointed chain of hard sphere sites of diameter d (typically
nm).23,24 Monomers interact with the nanoparticle via a hard
core repulsion plus an exponential tail pair potential:

r < R core + d /2

Upn(r ) =
r (R core + d /2)
r R core + d /2
pn exp

pn

Figure 1. A schematic depiction of a nanoparticle (hard core of radius


Rcore) in a cross-linked polymer network. Soft particles are modeled by
a repulsive polymerparticle tail potential (range pn) from which one
can dene an eective particle radius Reff via eq 2. Polymers are
composed of segments of size d. Both chemical cross-linking and
physical entanglements are mapped to a single length scale, dT,
characterizing monomer spatial connement.
= 2
F ( t )F (0)
p

(1)

(4)

where r is the center-to-center separation of the particle and


monomer, pn is the contact energy, and pn is the spatial range.
This potential models (i) athermal (pn = 0), (ii) attractive (pn
< 0), and (iii) soft repulsive (pn > 0) interactions. For the
latter, it is convenient to dene an eective size using standard
liquid state theory ideas:25
R eff + d /2 =

dr(1 eUpn(r))

Here, p is the segment number density, Spp(k) is the polymer


s
static (density uctuation) structure factor, and nn
(k,t)
(cpp(k,t)) is the normalized (at t = 0) dynamic structure
factor, or propagator, for the nanoparticle (polymer melt).
Eective particlemonomer forces enter via Cpn(k) which is the
Fourier transform of the segmentnanoparticle direct
correlation function dened via the polymer reference
interaction site model (PRISM) integral equation theory
as2730

(2)

For hard and attractive particles we set Reff Rcore.


Dynamically, the polymer network is described by Rouse
dynamics on length scales less than the entanglement or
mechanical mesh diameter, dT, and an arrested soft solid on
larger length scales. We shall use the tube diameter jargon
(literally relevant for polymer melts) to denote a mean total
dynamical connement length scale including (in networks)
both topological entanglement and chemical cross-link eects.
Experimentally, this length scale can be dened in terms of the
rubbery plateau modulus23 Gx pkBT/Nx where dT Nx1/2
and here is assumed to be unchanged in the presence of the
nanoparticle. We note that the latter simplication is rigorous
in the dilute particle limit at the ensemble-averaged level we
work at. However, physically the polymer packing structure
near a nanoparticle will not be the same as in the bulk, and
hence the rubbery modulus is expected to be locally perturbed.
How to microscopically include this higher order eect is an
unsolved problem at present, but we do not expect it to
qualitatively modify any of our results. A schematic depiction of
the model is given in Figure 1.
B. Dynamical Theories. In ideal MCT, the onset of
(transient) nanoparticle localization is characterized by a nite
value of the long time mean-square displacement20,26 (MSD),
(r( t ) r( 0))2 rL2. We compute this quantity using
mode coupling theory (NMCT) selfthe well-established naive
consistent equation:20,26
3
1

= 2F ( t ) F (0)
2
6
2rL

(2dk)3 k2Cpn2(k)Spp(k)nns(k , t ) cpp(k , t )

hpn(k) = Cpn(k)Spp(k)

(5)

where hpn(k) is the Fourier transform of hpn(r) = gpn(r) 1, and


gpn(r) is the monomerparticle pair correlation function.
To dynamically close the theory for entangled polymer melts,
it was previously proposed10
k 2D t
0
cpp(k , t ) = exp

S
k
(
pp )
+

t
S0
k 2d 2

exp 2 T exp
Nx
rep
3 Spp(k)

(6)

where D0 is the segmental short time diusion constant, rep is


the reptation (disentanglement) time, and Spp(k = 0) S0 =
pkBTT = pkBT/KB is the dimensionless compressibility which
quanties the amplitude of long wavelength density uctuations; T is the isothermal compressibility, and KB is the bulk
modulus.31 On short length scales (k1 dT/2), diusive
Rouse dynamics apply as encoded in the rst term on the righthand side of eq 6. The second term captures the longer (tube
and beyond) scale (k1 dT/2) physics where transient
localization occurs due to entanglement eects. Three aspects
enter:10 (i) a Gaussian tube spatial localization (ala neutron
spin-echo experiments32) including a de Gennes narrowing
factor, (ii) long time disentanglement via reptation, and (iii) an
overall amplitude, S0/Nx = Gx/KB 1, reecting the coupling
of density and stress uctuations. Cross-linked networks are
solids at long times, and thus eq 6 becomes

(3)

where 1 is the thermal energy, F(t) is the total force on the


nanoparticle at time t, and
406

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

cpp(k , t ) =

S0
k 2d 2
exp 2 T
Nx
3 Spp(k)

0 r < R core
gpn(r ) =
1 r R core

(7)

Localization of the nanoparticle is modeled in the standard


Gaussian manner:20,26
kr
s
nn
(k , t ) = exp L
6

3
hpn(k) = 4R core

(8)

3 3
dk
Cpn 2(k)Spp(k)
ln 2
kBT
2 2r
(2 )3 p

2
S
d T 2
r

0 exp k 2 +

Nx
3 2Spp(k)
6

(10)

If the thermal noise in eq 9 is dropped, then the ideal NMCT


result of strict localization of eq 3 is recovered. Examples of the
dynamic free energy are shown in Figure 2, and the localization
length, rL, barrier position, rB, and barrier height, FB, are
dened. The average hopping time is computed using the
general Kramers theory for stochastic processes:33
hop = 2s

rB

dr e Fdyn(r)

dr eFdyn(r )

2
e FB

KLKB

(15)

III. ANALYTIC ATHERMAL RESULTS IN POLYMER


NETWORKS
The random structure model considers only hard spheres and
hence Reff = Rcore R. Given Cpn(k) is proportional to a
spherical Bessel function (eqs 5 and 14), standard Gaussian
integration techniques can be used to analytically evaluate eq 10
to obtain

(11)

For small barriers, FB 2kBT, hopping events become less welldened; however, eq 11 still accurately captures the average
time it takes the particle to move a jump distance, rB rL.
When the barrier is large, FB 2kBT, eq 11 reduces to the
classic Kramers theory result:33
hop

(14)

PRISM is a theory for melts. To the extent the pair correlations


in dry networks are melt-like it is applicable. There are multiple
reasons this situation can hold. First, lled networks are
typically formed experimentally by melt mixing and then crosslinking. Second, we will show that particle localization occurs
on a relatively short length scale of the order of dT or much less.
Third, at zeroth order one generally thinks of a network as a
melt below the chemical cross-link scale.23 The latter is largely
relevant in the lightly cross-linked systems of primary interest
here where the entanglement length is smaller than the mean
distance between chemical cross-links. To mimic a rubber
network in numerical PRISM calculations, we set N = 1000
since this is large enough such that the asymptotic limit is
reached for the polymerparticle pair correlations.

(9)

kR core

Cij(r ) = (1 e Uij(r))gij(r )

where s is the nanoparticle short time friction constant, fs(t) is


the corresponding thermal random force, and Fdyn(r) is the
dynamic free energy. For networks the latter becomes
Fdyn(r )

j1 (kR core)

where j1(x) is the spherical Bessel function of the rst kind.


Second, the polymer collective density uctuations are
approximated by their long wavelength limit, Spp(k) Spp(0)
S0.
To capture local packing eects and variable nanoparticle
monomer interactions, we employ PRISM theory and model
the polymer as a freely jointed chain of N hard spheres of
diameter d and persistence (bond) length = 4d/3.10,27 The
monomer volume fraction is chosen to mimic melts or dry
networks, p = p4d3/3 = 0.4.2729 For all correlations we
adopt the PercusYevick sitesite closure30,31 outside the
distance of closest approach:

Equations 3, 4, 7, and 8 dene a self-consistent Gaussian


theory for rL2. For particles small relative to dT we expect
unbounded diusion at long times, and hence rL2 . For
large enough nanoparticles the cross-links should trap the
particle, and hence the localization length will be nite.
NMCT describes localization but not activated hopping. The
latter is a strongly non-Gaussian process and is the focus of the
nonlinear Langevin equation (NLE) theory1922 for the scalar
dynamical displacement of the particle from its initial position,
r(t), which obeys the stochastic evolution equation-of-motion:

s r(t ) = Fdyn(r ) + fs (t )
t
r

(13)

2R 3
3 3

= ln

dT
2 2r 2
6 p
kBT
dT
[b1/2{2b1 3 + (1 2b1)eb} + erf( b )]
6(2R /dT )2
b=
(2R /dT )2 r 2 + 4/ 2S0
(16)

Fdyn(r )

where r r/R. The rst ideal-entropy-like term favors the


liquid state, while the second term favors localization due to
polymerparticle forces. This dynamic free energy is
determined by three dimensionless parameters: (i) particle
size relative to tube diameter, 2R/dT, a measure of connement,
(ii) the tube diameter relative to the polymer packing length,
dT/p, where p1 (p2), and (iii) the dimensionless
compressibility, S0. At the analytic random structure level the
particle size enters only via the ratio 2R/dT.
For polymer melts, the packing length and tube diameter are
proportional, and experiments nd p2dT 18.34 In crosslinked systems this value can be smaller (tighter eective tube).

(12)

where s s(2Rcore)2 is the particle short time scale and K i =


Ki (2Rcore)2 is the dimensionless absolute value of the
curvature of Fdyn.
C. Structural Models. Equilibrium pair correlations are
input to the dynamical theories. We adopt two models: (i) an
analytically tractable structural (random) continuum model10
and (ii) numerical PRISM theory.2730 The former is applicable
when dT, 2Rcore d and all interactions are hard core, which
motivates the two dening simplications of model (i). First,
monomers pack randomly around the nanoparticle:
407

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

walk (space curve, d 0). For long chains it is known


S0 (p3)2 (p/)2, and since in melts p dT, one has S0
(dT/)2. Now, from Table 1 for p2dT = 18 we numerically
nd (2R/dT)c S01/6. Since dT S1/2
0 , the particle size at
localization scales as Rc S01/3, restoring the intuition that
particles are harder to localize in more compressible polymeric
media.
A related trend in Table 1 is that with decreasing p2dT the
localization onset shifts to larger connement values, again
seemingly counterintuitive. However, this also is a manifestation of assuming polymer properties are independently
adjustable. In real cross-linked systems, the ratio p/dT is
reduced compared to the melt because the eective tube
diameter is smaller. If this is taken into account, dT will decrease
by a factor of 2 for p2dT varying from 18 to 9, implying the
particle size 2Rc at localization is 0.6 times smaller for p2dT
= 9 compared to 18. Thus, we do predict nanoparticles localize
more easily as p2dT is decreased, as expected physically.
The tightness of particle trapping is quantied by the
localization length, rL. Figure 3 shows the two extrema lengths

Figure 2. Dynamic free energy proles, in units of the thermal energy,


for various nanoparticle sizes based on the analytic hard sphere model
with p2dT = 18 and S0 = 0.25. The important length and energy
scales are indicated: rL, rB, FB.

Here we perform calculations for p2dT = 18,9, and four values


of S0 = 0.1, 0.25, 0.4, and 1.0 where the rst three span melt
values27 and the last corresponds to ignoring the de Gennes
narrowing factor in eq 7. For xed polymer parameters, particle
dynamics are studied as a function of the connement
parameter, 2R/dT.
A. Localization Phenomena. NMCT predicts the onset of
localization, or within NLE theory the crossover of nanoparticle
motion from diusive to activated hopping corresponding to
when the dynamic free energy rst acquires a minimum.20
Figure 2 shows representative results for p2dT = 18 and S0 =
0.25. For 2R/dT 1.32, the nanoparticle is not trapped by the
entanglement network, and beyond this value it is. The
critical value for localization depends on polymer properties
and is denoted as (2R/dT)c; numerical results are summarized
in Table 1. Note that in all cases the onset of localization is a

Figure 3. The two extrema length scales (units of particle radius) of


the dynamic free energy as a function of the connement parameter
for several dierent dimensionless compressibilities in the analytic hard
core model. p2dT = 18 for all curves. The lower (upper) branch
represents the localization length rL (barrier position rB). The blue
crosses are results using the PRISM theory structural input and should
be compared to the S0 = 0.25 curve.

Table 1. Localization Transition for the Analytic Athermal


Structure Modela
p2dT

S0

(2R/dT)c

18

0.10
0.25
0.40
1.00
0.25

1.58
1.32
1.22
1.08
1.53

of the dynamic free energy as a function of reduced particle


size; the lower (upper) branches of these curves are rL (barrier
location, rB). In all cases near the localization transition rL dT
2R, but as the connement parameter grows, the localization
length signicantly shrinks. For small increases of 2R/dT >
(2R/dT)c, it decreases approximately as a critical power law:

Values of the dimensionless nanoparticle connement ratio at the


onset of localization (2R/dT)c are given for several polymer systems.
Here p2dT is the ratio of the tube diameter to the polymer packing
length, and S0 is the dimensionless compressibility of the polymer
system.

rL , c rL
R

priori predicted (not assumed) to occur when 2R/dT 1,


consistent with physical intuition. Variations in (2R/dT)c reect
specic properties of the polymer network. One trend is that a
higher dimensionless compressibility reduces (2R/dT)c, a
seemingly counterintuitive result. However, the latter deduction
assumes that S0 can be changed independent of all other
polymer properties. In real systems this is not true since the
compressibility is related to polymer chemistry and density27
and hence the tube diameter. To illustrate this point, we recall
the Gaussian thread eld theoretic version of PRISM theory27
where a polymer is an uncrossable continuous ideal random

0.24 0.09

2R
2R


dT
dT c

(17)

The exponent variability reects both a weak dependence on S0


and uncertainties in tting our numerical calculations to the
form of eq 17. Concerning the former aspect, the apparent
exponent slightly decreases as S0 increases, implying localization
tightens more slowly for more compressible systems. Equation
17 remains a reasonable representation only up to 2R/dT
(2R/dT)c + 0.3. From 2R/dT (2R/dT)c + 0.3 to (2R/dT)c +
3, there is a transition regime with no discernible simple
analytic behavior. For very strong connement, 2R/dT
(2R/dT)c + 3, an asymptotic limit is reached that can be derived
408

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

analytically from eq 16 for 2R/dT 1 and rL/R 1. This


result is

1/2
2
rL
72
3
1/4 dT

(3
S
)
=

0
2

2R
R
p dT

(18)

The localization length is related to dT as expected, since the


polymer network is driving nanoparticle connement; however,
for large particles rL is much smaller than the tube diameter.
Even under strong localization conditions, where hopping will
not be physically relevant in real systems (see section III.B), the
trends in eq 18 may be experimentally observable.
B. Dynamic Barrier and Activated Hopping. Three
quantities mainly characterize the particle hopping process: (i)
the jump distance rB rL, (ii) barrier height FB, and (iii) mean
hopping time hop. Figure 3 implicitly shows results for the
jump distance which one sees grows as connement increases.
Near threshold, 2R/dT (2R/dT)c, an approximate critical
power law ts our calculations:
0.48 0.03

2R
rB rL
2R


dT
R
dT c

Figure 4. Mean barrier hopping time in units of the particle short time
scale, s s(2R)2, as a function of the distance of the connement
parameter from its critical value for the analytic hard sphere model.
p2dT = 18 for all curves. The black, dot-dot-dashed curve is a t of
the low connement regime to the exponential form of eq 21 . (Inset)
Barriers for the same systems. The blue crosses are the barriers using
the PRISM structural model input and should be compared to the S0 =
0.25 curve.

(19)

other nonactivated motions. However, we can clearly conclude


that in experiments, hopping will only be possible to observe in
the smaller connement regime.
To explicitly illustrate the massive reduction in mobility, the
S0 = 0.25 system was analyzed further. By choosing dT = 8 nm
to mimic a polystyrene melt,34 and taking s = 1 ms as a typical
nanoparticle short time constant, we nd for 2R 2Rc = 10 nm
that hop 1 ms s. As the particle size is increased by a
modest amount to 2R = 18 nm, the hopping time increases to
100 000 years! Thus, there is a massive reduction in hopping
mobility with small changes in particle size.
In between 2R/dT (2R/dT)c + 0.4 and 2R/dT (2R/dT)c +
10 a transition regime is observed, where no simple analytic
behavior is found. Above 2R/dT (2R/dT)c + 10, we nd an
asymptotic limit can be derived

which is robust to variations in the dimensionless compressibility except for S0 = 1.0. Equation 19 applies until 2R/dT
(2R/dT)c + 1. For 2R/dT (2R/dT)c, an analytic result can
again be derived. Since rL 0, the hopping distance is
dominated by the barrier location which obeys rB dT, and in
this limit we nd
2
1
1/3 2R
rB rL
r
B (6 )1/6 p 2dT
48
dT
dT
dT

(20)

The hopping distance grows very rapidly as the connement


parameter is increased, independent of the compressibility. As a
caveat, NLE theory is less reliable when the jump distance is
large. 19,20 Moreover, as a practical matter hopping is
experimentally irrelevant in the large connement parameter
regime since barriers are enormous as shown below.
Figure 4 shows results for the barrier height (inset) and mean
hopping time (main) for the p2dT = 18 systems. Here the
connement parameter is expressed as the distance from its
value at the onset of localization, a format that results in a near
collapse of all the curves with the exception of S0 = 0.1. It is
evident that very small changes in the degree of connement
induce a dramatic growth of the barrier and hopping time. For
modest values of 2R/dT (2R/dT)c, a power law t to the
barrier calculations works well:
1.7 0.1

2R
FB
2R


kBT
dT c
dT

FB


2R
2R

= 0.08 exp 16

dT
dT c

(23)

where Vn is the particle volume and Gx is the network shear


plateau modulus. Equation 23 is reminiscent of elastic models
for ow and relaxation in deeply supercooled liquids.35,36 Here,
it reects that in order to move, the particle must isotropically
deform the surrounding network on a scale of order its size. A
practical caveat is that at the onset of the asymptotic scaling the
barrier is already FB 104kBT, and thus in experiment or
simulation such a R3 scaling will never be observed.
Finally, one can ask how hopping transport is aected by
polymer properties? As mentioned above, when connement is
viewed relative to the onset of localization, the dimensionless
compressibility of the polymer liquid, S0, has little eect on
hopping (e.g., see Figure 4). Decreasing p2dT weakens the
rate of growth of the barrier and hopping time. However, even
with this downward shift, the conclusion is the same: small
increases in the ratio 2R/dT above its critical value results in a
drastic decrease of particle mobility.

(21)

This result holds to 2R/dT (2R/dT)c + 0.4, at which point the


barrier is already rather high, FB 810kBT. Likewise, the
hopping time can be t as an exponential law:
hop

1
2
(2R )3 kBTp 2 = VnGx
6
dT

(22)

IV. ACTIVATED HOPPING IN POLYMER MELTS


In cross-linked systems, once the particle becomes localized the
only channel for motion is hopping. On the other hand, in

as shown in Figure 4. This works until 2R/dT (2R/dT)c + 0.3,


corresponding to FB 3kBT. We emphasize that when barriers
are small, hopping may become dicult to distinguish from
409

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

tively accurate only if the time scales are suciently separated


so the two processes can be distinguished.
Given the reptation time sets the time scale for constraint
release motion, and the particle short time scale follows from its
Rouse friction constant, the hopping time relative to the
constraint release time is
hop
rep

N 3 2R 5 2
3
(p 2dT )
e FB

18
Nx dT KLKB

(24)

When hop/rep 1, hopping will dominate, and when hop/rep


1, constraint release will dominate. Figure 5 shows
calculations using p2dT = 18 and S0 = 0.25 for several
degrees of entanglement. Results are shown for FB = 220kBT
since for low barriers hopping is not well-dened, and for FB >
20kBT hopping is astronomically slow. One sees from Figure 5
that if N/Nx = 5, then particle hopping is slower than polymer
disentanglement, suggesting it is not important, consistent with
recent simulations of lightly entangled melts.17 As the degree of
entanglement increases, the window of relevant connement
ratios, 2R/dT, grows. However, for experimentally accessible
values of N/Nx 200, the window remains very small, and we
estimate 2R/dT 1.51.8.
One can also phrase the above question as for a given particle
size and tube diameter, what is the minimum chain length
necessary to see hopping? By dening the threshold as when
hop = rep, one can solve eq 24 for Nc:

Figure 5. Ratio of the mean hopping time to the reptation time in


melts of various degrees of entanglement N/Nx, with p2dT = 18 and
S0 = 0.25.

2d 1/3 5/3
1/3
T
2
R
2

F

e B
Nc = Nx

18 dT KLKB

(25)

The main frame of Figure 6 shows hopping is relevant only for


2R/dT 1.52.0 based on this criterion, and N/Nx 10 is
necessary to observe hopping.
As an alternative, but related, measure of the relevance of
hopping, we compare the particle diusivity achievable based
on activation to that of the constraint release mechanism. A
simple and natural choice for the hopping diusion constant is

Figure 6. Minimum value of N/Nx for which hopping is important in


melts based on the four indicated criteria. Here p2dT = 18 and S0 =
0.25 are chosen to mimic melt conditions. (Inset) Hopping diusivity
(purple, solid) compared to its constraint release analogue in melts.
The horizontal lines correspond to (from top to bottom) N/Nx = 5,
10, 25, 100, and 200.

entangled polymer melts the tube constraints have a nite


lifetime, and two classes of competing, system-specic channels
of motion can be envisioned: (i) nanoparticle active motion
to escape dynamical constraints and (ii) passive motion via
polymer constraint release. A recent theory17,18 has combined
these two mechanisms by solving the GLE self-consistently for
the particle mean square displacement as a function of time.
However, the GLE theory was dynamically closed at the
Gaussian level, which does not account for large amplitude
activated hopping.
Hence, there remains the question of under what conditions
hopping contributes to nanoparticle diusion in entangled
polymer melts? Intuitively, we expect two conditions must
hold. First, reptation must be slow enough for hopping to be
relevant; otherwise, constraint release would dominate. This
will only be true for long enough chains. Second, the particle
to tube diameter ratio must fall in a narrow window near unity
to ensure that the particle becomes transiently localized but
that the barrier height is suciently low so that hopping can
occur quickly enough. At zeroth order, we will quantify these
conditions based on comparing the time scales for hopping and
constraint release. A caveat is that calculating hopping and
constraint release mobilities independently will be quantita-

D hop =

(rB rL)2
6hop

(26)

In the analytic model the short time diusivity, Ds, sets the
mobility scale. Since hopping is associated with escape from
entanglement constraints, we assume all relevant Rouse modes
rst relax, and thus obtain the N-independent result10

1
kBT
8 R3

= kBT

Ds =
3 R g 2 R
Rouse

(27)

where the Rouse viscosity is R = 0N and 0 is the segmental


melt viscosity. Expressing the constraint release diusivity from
ref 10 in terms of the same units, one nds
1

DCR

= 1 + ent
Ds
Rouse

3
ent
d
1 N
=
3 3S0 T
Rouse
12 Nx
2R

2
2
2
2

+1
e3 S0(R / dT ) 1 +
2
2
2
2

3 S0(R /dT )
3 S0(R /dT )

(28)

410

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

to make a fair comparison, the nite size of monomers is


incorporated into the analytic model by taking R R + d/2 in
eq 16, which ensures that the contact separations in gpn(r)
match. Additionally, we match the polymer parameters of both
models by choosing N = 1000 and p = 0.4, which xes S0 =
0.25. Finally, the packing parameter for the analytic model was
p2dT = 18, which implies dT = 27d/16p = 13.25d.
Results for the extrema lengths are shown in Figure 3 as
crosses. Comparing to the S0 = 0.25 case, the major dierence
between the two results is the precise onset of localization; the
PRISM-based calculations predict (2R/dT)c = 1.05, modestly
smaller than the analytic result of (2R/dT)c = 1.32. This
dierence is consistent with the intuitive idea that the conning
forces on the nanoparticle are enhanced when local packing
correlations (layering) is included. When viewed relative to the
localization transition onset, the behavior of the extrema length
scales are essentially identical in the two calculations. Barrier
height calculations are shown in Figure 4 (inset), expressed
relative to (2R/dT)c. The PRISM-based calculations grow
modestly more quickly with increasing particle size, and hence
so will the hopping time, consistent with the idea that local
packing correlations quantitatively reduce mobility.
Overall, the dynamical results based on the continuum model
agree remarkably well with the fully numerical results. This
seems understandable as a consequence of the short-range
nature of the structural correlations and the fact they are
oscillatory in a roughly symmetric manner about the random
value of unity. As a result, their dynamical consequences are
largely averaged out, as a priori done by the analytic structure
model.
B. Repulsive Interactions. A largely open question is the
role of particlemonomer interactions on nanoparticle
diusion. Recently, there has been much interest in composite
systems of soft nanoparticles, such as nanogels, microgels, hard
colloids with grafted polymer layers, vesicles, and micelles.37,38
To crudely model the eects of particle softness, we study
particles that interact with the polymers via a hard core plus soft
repulsive exponential tail as in eq 1. Using the PRISM structural
correlations as input, the eects of a soft interfacial interaction
on particle dynamics in cross-linked polymers are explored.
Consider rst the inuence of repulsion strength pn at
constant 2Rcore = pn = 5d for tube diameters in the range dT =
220d. System parameters are summarized in the rst section
of Table 2 along with the localization transition tube diameter
ratio dT,c/d for each case. Note that as pn increases, the
localization transition tube diameter, and hence (dT/2Rcore)c,
increases. This eect is dramatic since dT,c varies from d for
the hard core case to 20d for the very soft sphere case of pn
= 5. There is no seemingly obvious relationship between
particle core size and the transition value. However, since
increasing pn results in an eectively larger particle diameter as
it appears to the monomers, this trend is qualitatively expected.
In an attempt to collapse the above calculations, we scale dT
by the eective particle size 2Reff of eq 2. The resulting values of
2Reff/d and (2Reff/dT)c are given in Table 2. This normalization
approximately collapses the localization onset to (2Reff/dT)c
12. To further investigate this, extrema length scale proles
were calculated as a function of both dT/2Rcore and dT/2Reff, as
shown in Figure 8. When expressed as a function of dT/2Rcore
(inset), there is a huge variance in the curves mainly associated
with the shifts to higher tube diameters. This variance can be
largely collapsed if dT is scaled by the eective particle size. The
extrema length scales can then be compared to Reff as suggested

A direct comparison of the diusivities can now be made, as a


function of N/Nx, 2R/dT, and S0. The inset of Figure 6 shows
results for barriers again spanning the range FB 220kBT.
Note that for N/Nx = 5, the polymer is only slightly entangled,
and we expect (and nd) hopping is not relevant. As N/Nx
grows, the window in 2R/dT where hopping is faster than
constraint release diusion does widen. But, even for the most
heavily entangled N/Nx = 200 system, hopping is important
only for 2R/dT 1.51.8, consistent with our conclusion based
on the time scale comparisons. Figure 6 also shows the value of
N/Nx where Dhop = DCR, along with lower and upper bounds
on this threshold corresponding to when Dhop = DCR/10 and
Dhop = 10 DCR, respectively. We again conclude that in all cases
long chains are necessary to see hopping (N/Nx 10), and it is
only relevant for a narrow range of connement parameters of
order unity. Our conclusion that hopping is the primary mode
of transport only for strongly entangled polymers over a small
window of connement parameter appears to be consistent
with the qualitative arguments of Cai and Rubinstein.12,13

V. ROLE OF VARIABLE NANOPARTICLEPOLYMER


INTERACTIONS
To explicitly study the inuence of equilibrium local structure
and interactions on our dynamical results, we employ PRISM
theory as described in section II. Typical monomerparticle
pair correlations functions, gpn(r), are shown in Figure 7 for

Figure 7. Equilibrium monomerparticle pair correlation functions for


attractive (dashed red), soft repulsive (dotted black, dash-dotted green,
dash-double dotted yellow), and hard core (solid blue) systems. The
particle core diameter is 2Rcore = 10d, and the attraction/repulsion
range is pn = d. (Inset) Fourier transform of the nonrandom part of
the pair distribution function, hpn(k), for the same conditions of the
main frame and attractive (solid red and dashed blue), hard sphere
(dotted black), and soft repulsive (dot-dashed green) systems.

2Rcore = 10d with a tail potential range of pn = d for three types


of systems: (i) hard spheres, (ii) attractive spheres, and (iii) soft
repulsive spheres. Compared to the hard sphere case,
attractions enhance local correlations of monomers around
the tagged particle, while repulsions decrease the correlations.
For the highly repulsive systems, a signicant correlation hole
develops, ala a dewetting of the polymerparticle interface.
A. Hard Spheres. By comparing the dynamic free energy
extrema length scales and barrier for hard spheres based on
PRISM theory structural input to that deduced based on the
analytic structural continuum model, the role of local
nonrandom packing correlations can be investigated. In order
411

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

Table 2. Localization Transition for PRISM Structural Modelsa


system

2Rcore/d

pn/d

2Reff/d

dT,c/d

(2Reff/dT)c

repulsive: constant
core size

pn
0
1
2
4

5
5
5
5

5.00
12.97
18.29
24.67

4.25
7.25
12.25
19.25

1.18
1.79
1.48
1.28

repulsive: constant
eective size

0.1
1
2.5
5
10

5
5
5
5
0

2.26
2.06
1.71
1.14
0.00

10

6.25
6.45
6.75
7.60
9.10

1.60
1.55
1.48
1.31
1.10

attractive

10

0.0
0.5
0.5
1.0
1.0

10

9.10
8.25
7.60
7.75
7.10

1.10
1.21
1.32
1.29
1.41

0
2
4
2
4

Calculations of the tube diameter at the onset of localization, dT,c, are shown for several dierent nanoparticle diameters, 2Rcore, and monomer
nanoparticle interfacial interaction strengths (ranges), pn (pn) (dened in eq 1). It is convenient to dene an eective particle size (via eq 2) and
compare it to the transition tube diameter (2Reff/dT)c. All length scales are expressed in terms of the polymer segment diameter d.

Figure 9. The two extrema lengths (main) and barrier height (inset)
of the dynamic free energy for soft particles of constant eective size
2Reff = 10d with various core sizes. For all cases pn = 5 and pn/d is
determined by the constant eective size condition of eq 2. The
2Rcore/d = 10 case represents a hard particle, and as the core size is
decreased the softness of the ller is increased. The barrier height
results are shown as a function of the inverse mesh length dT1
rendered dimensionless by the mesh length at the onset of localization
dT,c.

Figure 8. The two extrema lengths of the dynamic free energy for soft
repulsive systems with 2Rcore = 5d and pn = 5d and several dierent
repulsion strengths pn. Both the tube diameter and extrema lengths
are normalized by the eective particle size. (Inset) To exhibit the
extent of data collapse in the main frame, the uncollapsed results are
shown where dT is normalized by the particle core size and the extrema
lengths are normalized by the monomer diameter.

by the analytic model. Doing so, we nd the massive (but not


perfect) collapse shown in the main panel of Figure 8.
We now investigate the role of softness on hopping
dynamics. In order to accomplish this while holding everything
else constant, a collection of systems of xed 2Reff = 10d were
studied. By choosing specic values of 2Rcore and pn, the
repulsion range pn was then appropriately determined. Varying
the core size from 2Rcore = 0.110d spans cases from a nearly
fully soft particle to a hard sphere. The system parameters and
localization transition values for these systems are summarized
in the center section of Table 2. First note the onset of
localization is shifted to smaller dT,c values for softer particles.
This is an intuitive trend and also applies to the extrema length
scale proles in Figure 9. Additionally, it is evident from the
latter that as 2Rcore 0, the results tend toward the same limit
of a fully soft particle. Barrier calculations are shown in the inset
of Figure 9. In all cases, small changes of the tube diameter
yield large changes of the barrier height. The main eect of

particle softness is slowing down how quickly the barrier grows.


For softer llers, more connement relative to the transition
value is necessary to produce the same barrier (and hence
hopping time). All these results are intuitive as softer particles
should be able to squeeze through the network more easily.
The above studies show that the main eect driving
localization in cross-linked nanocomposites is the connement
a (soft) particle experiences from the network as characterized
by the value of 2Reff/dT. When scaled in this manner, the largest
variations in dynamics are collapsed, and all localization
transitions occur at 2Reff/dT 1. There are smaller, secondorder perturbations due to particle softness. We nd that for
two particles with the same eective size the softer the particle
is, the more dicult it becomes to induce localization (smaller
dT,c). The nanoparticle mobility drastically decreases in all cases
as 2Reff/dT grows; however, the decrease is slower as particles
become softer.
412

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

C. Attractive PolymerFiller Interactions. Chemistry


can be chosen to favor monomer adsorption, creating an
attractive interfacial potential. We study this case for hard
particles with short-range attractions modeled via eq 1 with pn
< 0, 2Rcore/d = 10, and pn/d = 0.5 and 1.0. The attraction
strength pnwas varied from zero to the strong value of 5kBT.
The bottom panel of Table 2 summarizes the system
parameter choices and the computed localization onset values.
Short-range attractions shift the localization transition to
smaller tube diameters for both attraction ranges studied; the
same trend is found in the full extrema length scale proles in
Figure 10. Based on a local dynamical perspective, this result is

trends could be a consequence of the core approximation of


MCT and NLE theory which replace real forces by eective
forces determined by pair structure. This approach works well
for repulsive systems. However, when there is a combination of
attractive and repulsive forces, concerns have been raised in the
context of supercooled liquids as to whether such a MCT-like
force renormalization accurately captures the dynamical
consequences of competing forces.39,40 Our results for the
attractive systems may be related to this issue, and future
research and simulations are needed to systematically address
this open issue.

VI. DISCUSSION
We have developed a microscopic, force-level statistical
dynamical theory for the localization and activated hopping
dynamics of dilute spherical particles, both hard and soft, in
cross-linked networks and entangled polymer melts. Our
primary ndings and predictions are as follows. First and
foremost, the main factor controlling localization is the
connement the particle experiences from polymer entanglements and cross-links as characterized by the ratio of the
eective nanoparticle diameter to the mechanical mesh length,
2Reff/dT. Dynamic localization occurs when 2Reff dT, and the
hopping mobility drastically decreases as the connement
parameter even modestly increases past threshold. In addition
to this main eect, local packing correlations modestly enhance
the tendency to localize for hard particles, and for repulsive
particles increasing softness, even at xed eective particle
diameter, can enhance mobility. While the eects of soft
interfacial repulsions are small compared to those of particle
size (connement), they still have dramatic consequences for
the onset of localization and in turn the nanoparticle mobility.
For entangled melts, we nd the exponentially slow activated
hopping diusivity is ineective relative to particle transport
controlled by entanglement network dissolution for most cases.
The exception appears to be a narrow window of connement
parameters of 2R/dT 1.52 for suciently long chains.
Future experiments and simulations to test the regimes where
hopping is relevant for networks and melts would be very
valuable. Additionally, experimental evidence for the scaling of
the localization length and the hopping time with the
connement ratio would be an important test of our
predictions.
Our approach relies on two key theoretical assumptions
inherent to both MCT and NLE theory. First, we assume
particle hopping and polymer network uctuations are
isotropic. This, of course, is true in the context of ensemble
averaged results, but how much error it might incur at the level
of an individual hopping event is unknown. Second, we do not
allow the nanoparticle to actively (and perhaps nonlinearly)
modify the local network structure as part of the activated
hopping motion.13 Relaxing this assumption involves taking
into account higher order (greater than two body) static and
dynamical correlations in MCT and NLE theory, an open
problem for microscopic force-level approaches.
The theory presented here can be straightforwardly
generalized to treat particles in semidilute polymer solutions.
Nonspherical particles can also be studied, including the role of
translational versus rotational degrees of freedom in the
localization and hopping process. The present work provides
a foundation for attacking the very dicult nondilute
nanoparticle regime of high interest for polymer nanocomposites. However, this requires addressing several compli-

Figure 10. The two extrema lengths (main) and barrier height (inset)
of the dynamic free energy for particles with core 2Rcore = 10d and
various attraction ranges and strengths. The hard sphere case (solid
red) is shown for reference. The barrier heights are plotted as a
function of the inverse mesh length dT1 rendered dimensionless by
the mesh length at the onset of localization dT,c.

counterintuitive; as attractions are added, one might expect it


should be more dicult for the particle to move. On the other
hand, we are concerned with localization driven by physical
eects on a mesoscopic length scale, the tube diameter, not the
very short local scale characteristic of the attraction range. The
prior work of ref 10 observed a similar trend for nanoparticle
diusion based on the polymer constraint release mechanism.
For highly entangled systems, they found that nanoparticle
diusivity increases as the interfacial attraction was increased, a
trend, however, that is reversed in unentangled melts where the
physics is more local.
Hence, our results are consistent with the prior study10 and
can be explained, in the context of our approximate theory, in a
similar manner. Specically, the eective forces on the particle
are quantied by the structural correlations on the relevant
length scales dictated by dynamical relaxation. In Figure 7, one
sees that as the attraction is increased the local correlations
grow (adsorption), indicating stronger forces locally. But,
considering hpn(k) on the larger mesoscopic length scale, there
is a reduction of correlation (and hence forces) since |hpn(k)|
decreases with attraction strength. This raises the question as to
whether the local or global structure is more relevant to
nanoparticle motion. Given that dT 2Rcore near the transition
and that the cuto wavevector in the polymer propagator (eq
7) is kcut = (32S0)1/2/dT, we nd kcutd 0.27. Hence, the force
correlations are controlled to leading order by the mesoscopic
regime, thereby mathematically explaining our results.
A physical caveat to the above discussion is that while the
theory is internally consistent, the seemingly counterintuitive
413

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

Macromolecules

Article

(30) Chandler, D.; Andersen, H. C. J. Chem. Phys. 1972, 57, 1930


1937.
(31) Hansen, J. P.; McDonald, I. R. Theory of Simple Liquids, 3rd ed.;
Academic Press: London, UK, 2006.
(32) Richter, D.; Monkenbusch, M.; Arbe, A.; Colmenero, J. Adv.
Polym. Sci. 2005, 174, 1221.
(33) Zwanzig, R. Nonequilibrium Statistical Mechanics; Oxford
University Press: Oxford, UK, 2001.
(34) Fetters, L. J.; Lohse, D. J.; Richter, D.; Witten, T. A.; Zirkel, A.
Macromolecules 1994, 27, 46394647.
(35) Dyre, J. C. Rev. Mod. Phys. 2006, 78, 953972.
(36) Dyre, J. C. J. Non-Cryst. Solids 1998, 235, 142149.
(37) Bonnecaze, R. T.; Cloitre, M. Adv. Polym. Sci. 2010, 236, 117
161.
(38) Mackay, M. E.; Tuteja, A.; Duxbury, P. M.; Hawker, C. J.; Van
Horn, B.; Guan, Z.; Chen, G.; Krishnan, R. S. Science 2006, 311,
17401743.
(39) Berthier, L.; Tarjus, G. Eur. Phys. J. E 2011, 34, 96.
(40) Bhling, L.; Veldhorst, A. A.; Ingebrigtsen, T. S.; Bailey, N. P.;
Hansen, J. S.; Toxvaerd, S.; Schrder, T. B.; Dyre, J. C. J. Phys.:
Condens. Matter 2013, 25, 032101.

cations, a primary one being the assumption that polymer


dynamics is unperturbed by the presence of nanoparticles.

AUTHOR INFORMATION

Corresponding Author

*E-mail kschweiz@illinois.edu (K.S.S.).


Notes

The authors declare no competing nancial interest.

ACKNOWLEDGMENTS
This work was funded by Michelin-France. We thank Umi
Yamamoto and Sanat Kumar for helpful discussions.
REFERENCES

(1) Mackay, M. E.; Dao, T. T.; Tuteja, A.; Ho, D. L.; Van Horn, B.;
Kim, H.; Hawker, C. J. Nat. Mater. 2003, 2, 762766.
(2) Mason, T. G. Rheol. Acta 2000, 39, 371378.
(3) Pryamitsyn, V.; Ganesan, V. Macromolecules 2006, 39, 844856.
(4) Squires, T. M.; Mason, T. G. Annu. Rev. Fluid Mech. 2010, 42,
413438.
(5) Hamed, G. R. Rubber Chem. Technol. 2000, 73, 524533.
(6) Kessler, M. R. Polym. Rev. 2012, 52, 229233.
(7) Shubhra, Q. T. H; Alam, A. K. M. M.; Quaiyyum, M. A. J.
Thermoplast. Compos. Mater. 2013, 26, 362391.
(8) Ellis, R. J. Curr. Opin. Struct. Biol. 2001, 11, 114119.
(9) Zimmerman, S. B.; Minton, A. P. Annu. Rev. Biophys. Biomol.
Struct. 1993, 22, 2765.
(10) Yamamoto, U.; Schweizer, K. S. J. Chem. Phys. 2011, 135,
224902.
(11) Yamamoto, U.; Schweizer, K. S. J. Chem. Phys. 2013, 139,
064907.
(12) Cai, L.; Panyukov, S.; Rubinstein, M. Macromolecules 2011, 44,
78537863.
(13) Cai, L. Structure and Function of Airway Surface Layer of the
Human Lungs & Mobility of Probe Particles in Complex Fluids. Ph.D.
Thesis, University of North Carolina, Chapel Hill, NC, 2012.
(14) Ganesan, V.; Pryamitsyn, V.; Surve, M.; Narayanan, B. J. Chem.
Phys. 2006, 124, 221102.
(15) Egorov, S. A. J. Chem. Phys. 2011, 134, 084903.
(16) Brochard Wyart, F.; de Gennes, P. G. Eur. Phys. J. E 2000, 1,
9397.
(17) Kalathi, J. T.; Yamamoto, U.; Schweizer, K. S.; Grest, G. S.;
Kumar, S. K. Nanoparticle Diusion in Polymer Nanocomposites.
Phys. Rev. Lett., submitted for publication.
(18) Yamamoto, U.; Schweizer, K. S. Anomalous Transport of a
Single Nanoparticle in Entangled Polymer Liquids, manuscript in
preparation.
(19) Schweizer, K. S. J. Chem. Phys. 2005, 123, 244501.
(20) Schweizer, K. S.; Saltzman, E. J. J. Chem. Phys. 2003, 119, 1181
1196.
(21) Schweizer, K. S. Curr. Opin. Colloid Interface Sci. 2008, 12, 297
306.
(22) Chen, K.; Saltzman, E. J.; Schweizer, K. S. Annu. Rev. Condens.
Matter Phys. 2010, 1, 277300.
(23) Rubinstein, M.; Colby, R. H. Polymer Physics; Oxford University
Press: Oxford, UK, 2003.
(24) Doi, M.; Edwards, S. F. The Theory of Polymer Dynamics;
International Series of Monographs on Physics; Oxford University Press:
Oxford, UK, 1990.
(25) Barker, J. A.; Henderson, D. J. Chem. Phys. 1967, 47, 4714
4721.
(26) Kirkpatrick, T. R.; Wolynes, P. G. Phys. Rev. A 1987, 35, 3072
3080.
(27) Schweizer, K. S.; Curro, J. G. Adv. Chem. Phys. 1997, 98, 1142.
(28) Hooper, J. B.; Schweizer, K. S. Macromolecules 2005, 38, 8858
8869.
(29) Hall, L. M.; Schweizer, K. S. J. Chem. Phys. 2008, 128, 234901.
414

dx.doi.org/10.1021/ma4021455 | Macromolecules 2014, 47, 405414

You might also like