Subscriber access provided by READING UNIV
Article
Deposition Dynamics of Rod-shaped Colloids during Transport in Porous Media under Favorable Conditions ke li, and Huilian Ma Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b03983 • Publication Date (Web): 05 Feb 2018 Downloaded from http://pubs.acs.org on February 6, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
1
Deposition Dynamics of Rod-shaped Colloids during Transport in
2
Porous Media under Favorable Conditions
3 4
Ke Li, Huilian Ma∗
5
Department of Geology and Geophysics
6
University of Utah, Salt Lake City, UT 84112
7 8
Abstract
9
A 3D computational modeling study of the deposition dynamics of rod-shaped colloids during
10
transport in porous media under favorable conditions (no energy barrier to deposition) was
11
presented. The objective was to explore the influences of particle shape on colloid transport
12
and retention. During the simulation, both translation and rotation of ellipsoidal particles were
13
tracked and evaluated based on an analysis of all the forces and torques acting on the particle.
14
We observed that shape was a key factor affecting colloid transport and attachment. Rod
15
particles exhibited enhanced retention compared with spheres of equivalent volume at size
16
range greater than ~ 200 nm. The shape effect was most pronounced for particles around 200
17
nm to 1 µm under simulated conditions. Shape effect was also strongly dependent upon fluid
18
velocity; it was most significant at high velocity, but not so at very low velocity. The above-
19
described shape effect on retention was directly related to particle rotation dynamics, due to
20
the coupled effects from rotational diffusion and flow hydrodynamics. Rotational diffusion
21
changed particle orientation randomly, which caused rod particles to drift considerably across
∗
Corresponding author. E-mail: huilian.
[email protected]; Tel: (801)585-5976; Fax: (801)581-7065.
1 ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
22
flow streamlines for attachment at size range 200 nm ~ 1µm. The hydrodynamic effect
23
induced periodic particle rotation and oscillation, which rendered large size rod particles to
24
behave like “spinning bodies” prescribed by their long axes so to easily intercept with
25
collector surface for retention. Our findings demonstrated that the practice of using equivalent
26
spheres to approximate rods is inadequate in predicting the transport fate and adhesion
27
dynamics of rod-shaped colloids in porous media.
28 29
1. Introduction
30
The transport and deposition of colloids in saturated porous media is essential to manage and
31
remediate various environmental processes. For example, an accurate understanding of
32
colloid transport is necessary to assess surfactant loss in enhanced oil recovery,1 to track
33
contaminant fate,2 and especially to produce safe drinking water.3 Almost all colloids in the
34
environment are non-spherical in shape, such as clay particles, different types of bacteria (e.g.
35
Bacilli), human blood cells, polymer molecules, agglomerates and asbestos particles.
36
However, existing filtration theories often simplify natural colloids as spheres of certain
37
effective diameters, which have been demonstrated by increasing evidence4-9 to be insufficient
38
in describing their transport and retention behaviors.
39 40
Unlike spheres in a force field, for each different orientation, the forces acting on non-
41
spherical particles are different. Various researchers have proved that ellipsoidal particles
42
have different transport and retention behaviors compared with spheres.10-11 For example, rod-
43
shaped colloids exhibited elevated retention in column experiments compared to their 2 ACS Paragon Plus Environment
Page 2 of 56
Page 3 of 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
44
equivalent-volume spheres under otherwise identical conditions.12-13 A few studies also
45
showed that the non-spherical colloid shape affected not only the extent, but the pattern of
46
colloidal retention.10,
47
straining kinetics and found that as those particles approached pore-space constrictions, a
48
preferred orientation was adopted and theirs major axes tended to align with the local flow
49
direction. Yoo et al.17 used synthesized PLGA particles with spherical and elliptical disk shape
50
to investigate the rate of particle endocytosis and their distribution in endothelial cells. They
51
found that elliptical disks were endocytosed slower compared to spheres, and that disks
52
oriented tangentially along the nuclear membrane. Zhao et al.10,
53
simulation work of ellipsoids that those particles aligned in the stream-wise direction in the
54
near-wall region during channel flow, and that the particles affected the flow by altering the
55
turbulence structure near and around the kinetic energy peak. In the process of particles
56
depositing at the edge of evaporating suspension, Yunker et al.15 found that spheres deposited
57
randomly with surface diffusion, while ellipsoids induced strong capillary attraction on the
58
air-water interface and the deposition appeared to be ‘sticky’, i.e., clustering and creating void
59
regions and particle-rich regions. The contrasting behaviors between non-spherical and
60
spherical particles reflect different mechanisms by which non-spherical and spherical particles
61
transport and attach.
14-17
Xu et al.16 explored the effects of peanut-shaped particles on
14
concluded in their
62
63
Jeffrey (1922) predicted that rigid ellipsoidal particles tended to rotate with a certain period in
64
a linear shear field.18 This prediction was validated by Zia via his experiments,19 and has been
65
generally accepted. Although spheres also rotated in Jeffrey’s prediction, homogeneous 3 ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
66
spheres tended to follow streamlines under conditions when external forces or Brownian
67
diffusion were absent.18,
68
rotation of non-spherical particles, especially inertial particles, may also affect significantly
69
its translational motion.20 Broday21 calculated that inertial particles in vertical shear flow
70
migrated across the streamlines. Huang and Joseph22 also showed that elliptical cylinders
71
settling vertically in a channel drifted laterally due to the orientation-dependent hydrodynamic
72
interactions. These pioneer studies into shape effect introduced a valuable perspective on non-
73
spherical particle transport and retention. However, the above-mentioned studies did not
74
consider diffusion process, which is especially important for micro- and nano-meter size
75
ranged particles.
20
However, this is not the case for non-spherical particles. The
76 77
With respect to homogeneous sphere transport, Brownian rotation may be irrelevant; hence, it
78
is generally overlooked in general filtration theory. In contrast, rotational diffusion is a
79
fundamental characteristic describing the tumbling of natural anisotropic (non-spherical)
80
particles in solution. Rotational diffusion (or angular diffusion) influences non-spherical
81
particle transport through the orientation dependency of drag forces.23-24 Dresher et al.25
82
observed in their study that rotational diffusion diminished the hydrodynamic interactions
83
between two E. coli bacteria and altered their travelling paths. Kuzhir et al.11 found that the
84
magnetically induced rotational diffusion process caused the misalignment of particle
85
aggregates from the streamlines, which further illustrated that the long-range non-
86
hydrodynamic interactions of rotary diffusion process was also indispensable in studying
87
spheroid transport phenomenon. However, most of these studies were limited to particle sizes 4 ACS Paragon Plus Environment
Page 4 of 56
Page 5 of 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
88
in the Brownian regime, and they did not consider the influences of hydrodynamic
89
interactions induced by non-spherical shape.
90 91
Realizing the unique physical properties of non-spherical particles that can regulate transport
92
process, increasing attention has been attracted to the utilization of such particles in
93
biomedical field.26-28 For instances, many studies in drug-delivery systems showed that non-
94
spherical particles (e.g., rod-shaped or disc-shaped) exhibited better migration propensity or
95
adhesion onto blood vessel walls than spheres (including nano-spheres).29-35 But most of these
96
investigations, experimental or simulation, were carried out in parallel plate flow chambers
97
(PPFC).32-35 In contrast, studies on the influences of particle shape on transport and retention
98
in colloidal filtration field have been very limited.5, 36-37 Those few existing studies were all
99
from experimental investigations; and to the best knowledge of the authors, there is no
100
simulation work available yet from literature on non-spherical particle transport in porous
101
media.
102 103
In this work, we develop a three-dimensional particle transport model that incorporates the
104
non-spherical shape (e.g., rod-shaped particles as our model colloids here). This model will
105
simulate both the translation and orientation of rod-shaped particles in representative flow
106
fields based on an analysis of all the forces and torques acting the particle. From this
107
trajectory model analysis, we will obtain: i) 3D translation and rotation trajectory of
108
individual particles; ii) retention probability and distribution; iii) attachment orientation; iv)
109
transient particle position distribution; v) transient linear and angular velocities. Simulating 5 ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
110
both translation and rotation is very challenging, because non-spherical colloids may have
111
several different axes of rotation and also because its translation may depend upon its real-
112
time orientation. That is probably why quite a number of existing numerical studies on non-
113
spherical particle transport from biomedical field chose to simplify the problem with a 2D
114
model.20, 38-40 However, due to the complexities involved, we have two main assumptions in
115
our model that we believe are reasonable: i) since what we obtain here was clean single
116
collector efficiency, particles injected into our model systems were of dilute concentration
117
such that particle-particle collisions and aggregate formations could be neglected; ii) fluid
118
field disturbance due to the presence of particles was ignored. The model thus developed for
119
non-spherical particle transport can be modified and applied to various different geometric
120
flow models, because the essential simulation methodology stays the same. In this work, we
121
will present results obtained during our systematic investigations of rod-shaped colloids in
122
Happel sphere-in-cell model (i.e., a well-known model to represent porous media in classic
123
colloid filtration theory) under favorable attachment conditions (i.e., particle-collector
124
repulsion is absent). The simulation parameters spanned a wide range of ellipsoidal particle
125
sizes and shapes flowing under different pore water velocity. The main goals from these
126
simulations are to answer the following questions: i) under what conditions, non-spherical
127
particles can be treated as “effective spheres” and under what conditions, they can’t; ii) how
128
the shape of a particle affects its transport and retention? By what mechanisms?
129
130
2. Simulating Transport of Ellipsoidal Particles
131
Particle Transport Model. A Lagrangian trajectory approach was developed in this work to 6 ACS Paragon Plus Environment
Page 6 of 56
Page 7 of 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
132
simulate the time evolution of particle translation and rotation in representative porous media,
133
e.g., the well-known Happel sphere-in-cell model. The Happel model represents porous media
134
as a collection of spherical collectors each enclosed in a concentric spherical fluid shell. The
135
flow field within the Happel model can be described analytically.41-42 The general equations
136
of motion to account for the translation and rotation of a rigid particle are given by
137
138
139 140
= ∑
=
∙ = ∑ =
where t is time, m is the mass of the particle, is the moment of inertia of the particle; x and u
are the linear position and velocity vectors of the particle, respectively; Ω and are the
141
respective angular position and velocity vectors of the particle; and ∑ and ∑ represent the
142
total forces and torques acting on the particle, respectively. Our previous work simulated the
143
linear translations (i.e., eqs 1-2) of spherical particles in representative porous media models
144
extensively,4, 43-47 but did not explicitly account for particle orientation excepting converting
145
rotational motion of a sphere to linear translation.48 Here, we will focus on how to incorporate
146
both the rotational and linear motions of particles into simulation, using prolate ellipsoid
147
particles as model colloids. Due to symmetry for ellipsoids, the coupling tensor between
148
translation and rotation in eqs 1 and 3 vanishes based on Brenner’s research.49 Also in this work,
149
we treat Brownian motions differently from our previous work, where Brownian forces were
150
included into eq 1 to account for Brownian translation but Brownian rotation was neglected.
151
Here, we first determined the linear and rotational displacements from the total forces and
152
torques in eqs 1 and 3 that included only the deterministic forces and torques (i.e., excluding 7 ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 56
153
those from Brownian motions). We then added Brownian translation and rotation, which were
154
modeled in each degree of freedom as a stochastic process using Einstein’s equations, to those
155
resulted from the deterministic motions, respectively. Below, we shall describe in details how
156
to numerically solve the above equations of motion for an ellipsoid moving in shear flow.
157 158
Coordinate Systems. In order to describe the three-dimensional rotation and translation of
159
ellipsoidal particles, the following coordinate systems were employed (Figure 1): i) an inertial
160
frame of reference, = x, y, z, with its origin located at the collector (e.g., the origin of the
161
Happel sphere-in-cell model); ii) a particle frame, = x, y, z, with its origin at the center-of-
162
mass of the particle and its axes being the particle principal axes; and iii) a co-moving frame,
163
= x, , z, with its origin at the center-of-mass of the particle and its axes parallel to the
164
corresponding axes of the inertial frame. The translational motions of the ellipsoid were
165
described by tracking the center-of-mass of the particle in the inertial frame. The rotational
166
motions were described by following the three Euler angles (Figure 1) or the four Euler
167
parameters (i.e., quaternions) in the particle frame.
168 169
Transformation of the coordinates between the particle and co-moving frames is given by 50-51
= ,
170
(5)
171
where the orthonormal transformation matrix A can be expressed in terms of either Euler
172
angles or the four Euler parameters (i.e., q0, q1, q2, q3). Below, matrix A is given using Euler
173
parameters as
8 ACS Paragon Plus Environment
Page 9 of 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
1 − 2 $%% + $'% 2 $( $% + $' $) 2 $( $' − $% $) = 2 $% $( − $' $) 1 − 2 $'% + $(% 2 $% $' + $( $) *. 2 $' $( + $% $) 2 $' $% − $( $) 1 − 2 $(% + $%%
174
(6)
175
The four Euler parameters are subject to the following constraint to preserve the three degrees
176
of freedom for a rotating rigid body: $)% + $(% + $%% + $'% = 1.
177 178 179 180
(7)
Euler angles are related to the Euler parameters as $) = cos .
/01 3 2 cos . % 2 , $( %
/01 3 2 cos . % 2. %
sin .
= cos .
/41 3 2 sin .% 2, $% %
= sin .
/41 3 2 sin .% 2, $' %
=
(8)
181
The ranges for Euler angles used in this study are ϕ ∈[-π,π], θ ∈[0,π], and ψ ∈[-π,π].
182
Conversion from Euler parameters to Euler angles is provided in the Supporting Information.
183 184
Rotation Dynamics of an Ellipsoid in Shear Flow. In this work, we chose the Euler four
185
parameters to describe the time evolution of particle angular displacements, because the use
186
of Euler angles would involve a large number of trigonometric calculations, which are
187
computationally expensive, and also because there is a singularity in Euler angles (see
188
Supporting Information). However, Euler angles were used to assign the initial orientation of
189
a particle (most of simulation data here with random initial orientations). The initial Euler
190
parameters were then determined from eq 8. At the subsequent time steps, the Euler
191
parameters evolved with time as51
192
193 194
−$(