Professional Documents
Culture Documents
Ochiai 2006
Ochiai 2006
1029/2006WR004961, 2006
Sources
through experiments allows investigators to examine colloid
deposition patterns, which may be used to make some
mechanistic interpretations of colloid-media interaction
[e.g., Tufenkji et al., 2003]. However, destructive sampling
does not preserve information about colloid location relative
to media grains or various interfaces, which is essential for
understanding pore-scale processes, and does not allow
observation of processes over time. Given these limitations
of column experiments, it is recognized that noninvasive
visualization methods can greatly enhance investigation of
mm to cm
cm to m
Sample
cm
cm
cm
cm
minutes
minutes
minutes
minutes
Time
yes
yes
yes
yes
no
yes
no
no
no
no
paramagnetic marker
radioactive tracer
dopins
dopins
10 mm
3 mm
mm
Figure 3. Typical setup of a pore-scale visualization system showing (1) flow cell on microscope stage,
(2) imaging system consisting of microscope with CCD camera, and (3) pump to generate fluid flow.
chromes. Sherwood et al. [2003] reported no change in Bettahar, 2006], at least in certain concentration ranges.
motility of bacteria which were labeled with magnetite Ideally, a visualization system should have a low enough
nanoparticles. In the case of Phytophthora spp. (water- detection limit to allow use of colloid concentrations that are
molds), labeling of motile zoospores with a fluorophore- observed in nature (105 –1014 particles mL 1 [Kim, 1991]),
conjugated lectin has been shown to cause early encystment and that minimize concentration-dependent artifacts.
and substantial changes in surface chemistry [Hardham and [12] The ability to visualize changes in time is very
Suzaki, 1986]. Thus care must be taken to select labeling informative of some transport processes, while other pro-
materials that have minimal effect on colloid properties cesses may require observation only at discrete time points.
relevant to their transport dynamics. The temporal resolution of a given visualization system,
[10] Visualization of colloids in natural media poses defined as the smallest time interval required between
several challenges that are difficult to overcome. In visible distinct visualizations, is dependent on the exposure time
light systems, colloids can be imaged through translucent necessary to capture images with sufficient contrast and
media that is several grain layers thick which allows for resolution. Currently, light-based microscopy provides the
even packing and reduction of edge effects, but this comes greatest temporal resolution, on the order of milliseconds,
at the cost of signal degradation due to increased attenuation and can be used to visualize processes such as advective
and scattering of both excitation and emitted light. Natural transport of individual colloids through a pore. Light-based
aquifer material is generally opaque, preventing excitation visualization at the mesoscale requires longer exposure, on
light from reaching colloids and emitted light from reaching the order of minutes, depending on the intensity of fluores-
the detector. Therefore visible light studies with natural cence. Exposure times of MRI and X-ray visualization sys-
media are limited to thin sections consisting of only one or tems are also on the order of tens of seconds to minutes, and
two grain layers. Even if fluorescence techniques are used, thus would be appropriate for studies that require only single
autofluorescence of media grains is likely to interfere with images or sequential images of time-averaged transport.
differentiation of colloids from the medium [Li et al., 2004].
While MRI is not affected by media opacity or autofluor- 3. Pore-Scale Visualization Systems
escence, its use to image natural materials may be limited
due to signal interference by paramagnetic and ferromag- [13] Pore-scale visualization systems generally consist of
netic components of the media, such as iron oxides or other three main components: (1) a flow cell, (2) a pump to
metals, as well as organic matter and exchangeable cations induce fluid motion within the flow cell, and (3) an imaging
[Hall et al., 1997]. system (Figure 3). The imaging system usually comprises a
[11] Detection limit and sensitivity are important con- conventional or epifluorescent microscope (or magnifying
straints of visualization systems which measure colloid lens) coupled to a video or still image capture device. Flow
concentrations. The detection limit refers to the lowest cells are generally placed horizontally on the existing
concentration of colloids that can be identified as being microscope stage and flow is generated by a syringe or
present and is typically reported as parts per million mass peristaltic pump at either the influent or effluent end of the
colloids per mass liquid (ppm), or in particles per unit flow cell. A typical experiment entails pumping a colloid
volume (e.g., particles mL 1). Sensitivity is defined as the suspension through the flow cell and observing colloids as
smallest change in concentration that can be detected as they transit through the pore network or as they interact with
different. Deposition of colloids in porous media has been various interfaces (water-solid, air-water, air-water-solid).
demonstrated to be concentration-dependent [Bradford and The following paragraphs describe methods of flow cell
4 of 14
W12S06
Table 2. Selected Applications of Micromodels for Visualization of Colloidal Transport in Porous Media
Type of
Source Micromodel Network Characteristics Pore Dimensions, mm Visualization Method Moisture Conditions Nature of Colloids (Size, mm) Objective
a b a b
Wan and Wilson etched glass regular polygon pillars 200 /50 ; 300 /20 – 100 ; dark-field; unsaturated hydrophilic microspheres (0.6, 1.05); attachment at the
[1994] and irregular shapes 4 – 400 epifluorescence hydrophobic microsphere (0.95); air-water interface
bacteria (1.2 0.8, 1.0 0.8);
clay (0.5)
Wan et al. [1996] etched glass two-domain: fracture-matrix range of widths 0 – 600c epifluorescence saturated microspheres (1.0) two-domain flow
Lanning and Ford etched glass regular cylindrical pillars 360 – 375 60 – 83c light scattering to saturated E. coli dispersion of bacteria
[2002] determine turbidity
Sirivithayapakorn etched silicon simulation of natural 2.4 – 30 15c epifluorescence saturated latex microspheres (0.05, 1.0, 2.0, 3.0); size exclusion and
and Keller [2003a] pore system stained bacteriophage colloid acceleration
Sirivithayapakorn etched silicon simulation of natural 2.4 – 30 15c epifluorescence unsaturated latex microspheres (0.05, 1.0, 2.0, 3.0); dissolution of air-water
and Keller [2003b] pore system stained bacteriophage interface
Auset and Keller molded silicon rounded square pillars 10, 20 12c epifluorescence saturated latex microspheres (2,3,5,7) colloid dispersion
[2004] (PDMSd)
5 of 14
Baumann and Werth etched silicon regular cylindrical pillars 173a/35b 50c epifluorescence saturated microspheres (0.691,1.96) comparison to LB model
[2004] and filtration theory
predictionse
Auset et al. [2005] molded silicon simulation of natural 2.4 – 30 12c epifluorescence cyclical infiltration DAPI stained bacteria (2 1); Intermittent filtration of
(PDMSd) pore system and draining latex microspheres (5) bacteria and colloids
Chen and Flury etched glass regular cylindrical pillars 70 bright field unsaturated natural colloids(0.348), retention of mineral colloid
[2005] modified colloids (0.368), in unsaturated media
kaolinite (0.332),
Na-montmorillonite (0.324)
a
Diameter or range of diameters for pore body, when specified in the paper.
b
Diameter or range of diameters for pore throat, when specified in the paper.
c
Depth or range of depths of pore system, when specified in paper.
d
Polydimethylsiloxane.
e
OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS
Table 3. Applications of Media-Packed Flow Cells for Visualization of Colloidal Transport Through Porous Media
Dimensions
Source (L W D), cm Grain Size, mm Type of Microscopy Moisture Conditions Nature of Colloids (Size, mm)
Crist et al. [2004] 26.0 4.8 0.5 430 – 600 CCD camera attached to unsaturated nonfluorescent blue colloids (0.3,0.8)
video microscopic lens
Crist et al. [2005] 26.0 4.8 0.5 850 – 1700 CCD camera attached to unsaturated nonfluorescent blue colloids (4.8,5.2)
video microscopic lens
Bradford et al. 7.0 2.0 0.2 150; 710/150 Epi-fluorescence saturated latex microspheres (1.1, 3.0)
[2005, 2006]
Zevi et al. [2005] 22.0 2.0 0.5 850 – 1700 bright field unsaturated hydrophilic latex microspheres (0.8,2.6,4.8);
hydrophobic latex microsphere (5.2)
Ochiai et al. [2005] 5.0 0.1 0.01 50 – 100 bright field and saturated, unsaturated fluorescently stained cysts (8 – 10),
epifluorescence latex microspheres (1.0)
Gao et al. [2006] 6.0 1.6 0.3 355 – 425 bright field and unsaturated, infiltration latex microspheres (2.0)
epifluorescence and draining GFP-transformed E. coli
fabrication and summarize investigations which have used application by Wan and Wilson [1994], which revealed the
pore-scale visualization. role of air-water interfaces as potential sites for colloid
[14] Two types of flow-through systems are currently retention. Subsequent investigations using micromodels
being employed to visualize movement of individual col- (summarized in Table 2) have led to a variety of advances
loids at the pore scale: micromodels and media-packed flow including verification of putative mechanisms such as size
cells. Characteristics of these systems are summarized in exclusion [Sirivithayapakorn and Keller, 2003a] or film
Tables 2 and 3. Glass micromodels are constructed by straining [Gao et al., 2006], and evaluation of predictive
etching a pore network pattern and its mirror image onto models based on fluid dynamics and transport theory [Auset
two glass plates which are fused face to face at high and Keller, 2004; Baumann and Werth, 2004]. Glass micro-
temperatures [Wan and Wilson, 1994]. Silicon micromodels models introduced by Wan et al. [1996] and etched silicon
have been constructed by photochemically etching a pore micromodels developed by Rangel-German and Kovscek
network pattern onto a silicon wafer and anodically bonding [2006] contain both fracture and matrix fields and are
a glass plate to the top of the wafer and an aluminum potential tools for studying two-domain flow. Employing
[Baumann and Werth, 2004] or a glass [Keller et al., 1997] glass micromodels with spatially periodic pore networks,
plate to the bottom for support. Molded silicone micro- Lanning and Ford [2002] studied bacterial dispersion dur-
models have been fabricated by pouring poly(dimethylsi- ing saturated transport and chemotaxis under stagnant fluid
loxane) (PDMS) into a mold consisting of an etched silicon conditions. Using silicon micromodels with realistic pore
wafer, followed by curing and polymerization with a slab of geometries (Figure 4a), Sirivithayapakorn and Keller visu-
precured PDMS to form a closed channel system [Auset and alized both accelerated transport of colloids relative to the
Keller, 2004]. Pore network patterns are transferred to bulk liquid due to size exclusion and the presence of
glass or silicon plates by first coating the plates with preferential pathways [Sirivithayapakorn and Keller,
ultraviolet (UV) light sensitive photoresist, placing a lithog- 2003a], and the fate of latex microspheres and bacteriophages
raphy mask (with a negative pattern) over the plate, and during dissolution of trapped air bubbles [Sirivithayapakorn
then exposing areas which are to be etched to UV light. and Keller, 2003b]. Baumann and Werth [2004] visualized
Computer controlled lithography enables both precise con- transport of 0.7 and 2.0 mm fluorescent colloids through
trol of the network geometry and accurate replication silicon micromodels with a regular pore pattern (Figure 4b)
between plates (Figure 4). Micromodel pore networks are to evaluate the accuracy of flow path and particle velocity
two-dimensional in a topological pore network sense, predictions based on a 2-D lattice Boltzmann (LB) model
though some three-dimensional features are expressed at and filtration efficiency predictions based on colloid filtra-
the pore scale, including pore wedge features that exist tion theory. Auset and Keller [2004] studied particle trajec-
where plates meet [Wan and Wilson, 1994]. Additionally, tories, residence times, and dispersion coefficients for
pores of different depths on the same plate can be achieved
by sequential etching [Wan et al., 1996]. The surface
chemistry of micromodels depends on the composition of
the material used, and is generally considered to be homo-
geneous. The glass used by Wan and Wilson [1994] con-
sisted of 72% SiO2, 14% Na2O, 3.8% CaO, and other trace
metal oxides, while the main components of silicon wafers
used by Sirivithayapakorn and Keller [2003a] were silica
and silicon hydroxides. When water wet, the surface was
covered with surface hydroxyl groups and negatively
charged at neutral pH. The PDMS is hydrophobic and has
an air-water contact angle of approximately 109° [Auset and Figure 4. Examples of micromodels with (a) pore pattern
Keller, 2004], but can be rendered hydrophilic by sodium copied from thin slice of a porous media [from
silicate treatment [Auset et al., 2005]. Sirivithayapakorn and Keller, 2003b] and (b) regular pore
[15] The utility of visualization techniques in the study of network (reprinted from Baumann and Werth [2005],
colloid transport is illustrated by the initial micromodel copyright 2005, with permission from Elsevier).
6 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Objective lens
22 Inlet port
0.5
22
25
15”
Light source
Outlet
Figure 5. Schematic of tilted, open-faced visualization chamber utilized by Crist et al. [2004, 2005] and
Zevi et al. [2005], showing chamber dimensions and positioning of video magnification lens and light
source. Adapted with permission from Zevi et al. [2005] Copyright 2005 American Chemical Society.
different sized colloids in idealized pore networks of PDMS [2005] designed a 0.2 2.0 7.0 cm flow cell (150 and
micromodels. Auset et al. [2005] employed PDMS models 710 mm) to visualize straining of 1.1 and 3.0 mm sulfate
with irregular pore geometry to investigate intermittent colloids at textural interfaces in homogeneous (150 mm) or
transport of bacteria and latex microspheres during cyclical layered (750/150 mm) sand, and to investigate the role of
infiltration and draining events. Using variably saturated straining in microbial deposition [Bradford et al., 2006].
glass micromodels, Chen and Flury [2005] observed the [17] A flow cell system under development by our group
lack of attachment of naturally occurring and wastewater- [Ochiai et al., 2005] (Figure 7) consists of a rectangular
treated colloids, pure kaolinite, or Na-montmorillonite to the glass capillary (#5010, Vitrocom, Mountain Lakes, NJ)
liquid-gas interface. packed with a monolayer of media grains. Capillaries
[16] Flow cells packed with granular media are also being are available in various dimensions (depths from 20 to
used to visualize colloids at the pore scale (summarized in
Table 3). Crist et al. [2004, 2005] developed an open-faced
infiltration chamber filled with quartz sand (850 to 1700 mm)
to a depth of 0.5 cm (Figure 5). A unique feature of this
flow chamber is that it is mounted at an angle between 15°
and 45°. This system was used to investigate retention of
hydrophilic and hydrophobic colloids at the air-water-solid
(AWS) and water-solid (WS) interfaces in unsaturated
media [Crist et al., 2004, 2005; Zevi et al., 2005] as well
as to image water flow through water films using confocal
microscopy [Zevi et al., 2005]. The possibility that the open
chamber face results in artifacts of evaporation remains a
point of contention [Steenhuis et al., 2005; Wan and
Tokunaga, 2005]. Gao et al. [2006] constructed flow cells
from transparent acrylic, which were fitted with stainless
steel membranes at the inlet and outlet ends to maintain
capillary pressure, and packed with quartz sand (grain size
range 355 to 425 mm) to a porosity of 0.4 (Figure 6). Using
this flow cell, they observed mobilization and retention
dynamics of 2.0 mm fluorescently dyed latex microspheres Figure 6. Quartz-sand packed visualization flow cell con-
in various compartments of partially saturated sand under structed from acrylic plates used in the work of Gao et al.
various moisture contents and flow regimes. Bradford et al. [2006].
7 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Figure 7. (top) Schematic of media-packed flow-cell under development by Ochiai et al. [2005],
showing rectangular glass capillary with internal dimensions 0.1 1.0 mm (A) connected to 1.6 mm
barbed nylon elbows (B) and mounted on a standard glass slide (C). (bottom) Examples of capillaries
packed with 90 mm diameter glass beads, quartz sand, and forest soil.
1000 mm). A nylon mesh with 35 mm openings is attached of pore surfaces makes micromodels particularly useful for
on either end of the capillary to prevent loss of media. investigations that focus on pore geometry and interfaces.
Capillaries are connected to other components of the flow At the same time, since the etched models are funda-
system via microbore tubing and 1.6 mm (0.0625 in) barbed mentally two-dimensional networks rather than three-
nylon connectors which also serve to mount the capillaries dimensional granular structures, many critical aspects of
to a microscope slide. The rectangular profile of the pore-scale geometry may not be well represented in com-
capillary provides two transparent flat faces through which parison to field conditions (e.g., isolated pores, pendant
media grains and colloids can be viewed without distortion water, dendritic flow paths, and heterogeneous surface
associated with wall curvature. Rectangular capillaries of composition). The use of media-packed flow cells enables
various depths (20 to 1000 mm) are available and can be the visualization of colloid behavior in more realistic 3D
adapted for flow cell use. We are using 0.1 1.0 50 mm pore spaces, and thus colloid interaction with more natural
capillaries with a 0.1 mm wall that approximates the thick- air-water interfacial configurations. These types of systems
ness of a standard microscope slide cover slip (0.17 mm), are therefore well suited to the study of pore-scale colloid
allowing use of high-magnification objectives with standard behavior under partially or variably saturated media. In
working distances. These capillary flow cells are currently addition, the potential to use any type of media grain in
being used to investigate transport of Phytophthora spp. flow cells introduces the possibility of investigating effects
zoospores and cysts in saturated and unsaturated media. of media heterogeneity (physical or chemical) on interac-
[18] Edge effects created by the flow cell walls are a tions of colloids, media, and air-related interfaces. To date,
concern for current pore-scale visualization systems. Thus however, investigations using media-packed flow cells have
materials should be chosen with careful consideration of been conducted using quartz sands with well-established
colloid characteristics to minimize system artifacts associ- characteristics. The utility of micromodels for studying chemi-
ated with the visualization system itself. The current sys- cal or physical (roughness) surface heterogeneity remains
tems are predominantly horizontally oriented and effects of to be explored and depends on the development of methods
gravity are assumed to be unimportant over the spatial and to manipulate surface properties with fairly high spatial
temporal scale of experiments. This assumption, however, resolution as demonstrated by Auset and Keller [2006].
may not be valid for larger or denser colloids, whose
settling behavior may have an impact on transport. These 4. Mesoscale Visualization Systems
potential shortcomings aside, the benefits gained from direct
observation of microscale processes warrant continued use [19] While pore-scale visualization methods enable in-
of these visualization systems. The ability to design and vestigation of mechanisms affecting individual or small
fabricate precise pore geometries and relative homogeneity populations of colloids, they have limited applicability at
greater length scales where larger-scale media features play
8 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Figure 8. Schematic of mesoscale epifluorescence imaging system developed by Zhang and Wang
[2006] showing illuminator (1), closed optical path image (COPI) chamber (2), capture system chamber
(3), and sample chamber (4).
a major role [Sun et al., 2001]. At the mesoscale (here oriented 20 20 1 cm flow cell packed with quartz sand,
defined as 10– 1000 times the pore scale) heterogeneities mounted within a light-tight box (Figure 8). Excitation light
such as nonrandom distribution of grain and pores shapes from a 150 Watt halogen lamp is filtered and directed into a
and sizes, textural inclusions, variability in saturation, closed optical path image chamber and through an optical-
gradients in pore water chemistry, local pore clogging, or grade transmittable acrylic viewing window where fluores-
presence/evolution of preferential flow paths are expected to cent latex particles in the flow cell are excited. Emitted light
affect overall colloid movement through and retention is reflected off a 45° angle mirror, passed through an
within a medium. Mesoscale visualization methods do not emission filter and collected by a 14-bit, cooled charge-
track individual colloids but enable observation of colloid coupled device (CCD) camera that resides in a separated
concentrations in a given volume of media. In addition to chamber of the light-tight box. Each pixel in the resultant
providing a means to validate upscaling of pore-scale 1024 1024 pixel array can be resolved to an area as small
mechanisms to behavior of colloid populations at the as 400 mm2. Placement of the source and detector on the
column length scale, mesoscale visualization can aid in same side of the flow cell is intended to reduce the amount
investigation of the influence of various media heterogene- of excitation light that enters the detector and the sensitivity
ities and features on colloids populations as they pass of the system is reported at 2 105 particles mL 1
through the medium. (approximately 0.1 ppm). The system has been used to
[20] The basic setup of mesoscale visualization systems observe nonexponentially decreasing deposition of colloids
is analogous to that of pore-scale systems, and entails as a function of pore water velocity near a radial injection
(1) a media-packed flow cell, (2) a means to generate fluid point. A second reflexive (epifluorescent) mesoscale imag-
flow, and (3) an imaging system. Physical dimensions and ing system has been developed [Bridge et al., 2006] which
configuration of flow cells vary widely from system to utilizes UV light to illuminate fluorescent colloids and
system and are critical considerations as they determine the conservative tracer as they are coadvected through a verti-
length scale of media features that can be investigated. cally oriented 200 100 6.7 mm chamber packed with
Imaging in mesoscale systems is also constrained by quartz sand. Pixel resolution of this system is 0.478 mm2,
parameters such as detection limit, sensitivity, and spatial and pixel data have a dynamic range of 256 brightness
or temporal resolution which may determine their applica- values. This system has been used to calculate attachment
bility to a given study. Briefly, the three types of visualiza- efficiencies with varying ionic strength.
tion systems currently under development discussed here are [22] The experimental system being developed by our
(1) epifluorescent systems developed by Zhang and Wang group [Kraft and Selker, 2005; Weisbrod et al., 2003b]
[2006] and Bridge et al. [2006], (2) a transmitted fluores- consists of parallel glass plates (45 60 cm) held vertically
cence system employed by Kraft and Selker [2005], and (3) in an air-tight manifold to form a 1 cm thick inner chamber
MRI systems utilized by Sherwood et al. [2003], Olson et al. which is filled with granular media (Figure 9). Colloid
[2004], and Baumann and Werth [2005]. suspensions or other fluids are introduced into the chamber
[21] A light-based mesoscale visualization method devel- by multiple inlets located at the top of the sand pack, and
oped by Zhang and Wang [2006] utilizes a horizontally effluent is collected at drainage points at the bottom of the
9 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Figure 9. Schematic of mesoscale light transmission visualization under development by Kraft and
Selker [2005], showing arrangement of optical components and media chamber. Note: not drawn to scale.
chamber. Fluorescent colloids in the porous media are 2001], imbibition of saline solutions [Weisbrod et al., 2002],
excited by an incandescent 100 Watt lamp on one side of air-water interfacial area [Niemet et al., 2002], vapor trans-
the chamber and light emitted from colloids is detected by a port [Parker et al., 2006; Weisbrod et al., 2003a], colloid
CCD camera located on the side opposite the light source. concentration [Weisbrod et al., 2003b], and microbial
The detection limit has been improved 100-fold [Kraft and growth [Rockhold et al., 2006; Uesugi et al., 2001; Yarwood
Selker, 2005] by placing narrow band-pass filters in front et al., 2002, 2006]. The Weisbrod et al. [2003b] study
of both the light source (filter 32054, Chroma Corp, showed results indicating that colloidal particles will be
Trenton, NJ) and the CCD (filter 41293), to exclude light retained at the capillary fringe (Figure 11). Investigation of
off-frequency from peak excitation or emission wave-
lengths. The system employs a 512 512 pixel scientific-
grade CCD camera (Princeton Instruments, Roper Scientific,
Trenton, NJ) with exceptionally large photodiodes (29
29 mm) to allow detection of low light intensities
(10 photons per pixel) over 260,000 individual points,
with a 16-bit linear dynamic range. The resulting system
detection limit, given a particle diameter of 1 mm and
density of 1.055 g/cm3, is on the order of 0.6 ppm
(Figure 10), or approximately 9 105 particles mL 1.
Further improvements in detection limit are expected
through installation of a light source with higher output in
the target spectral band (e.g., LEDs). Spatial resolution of
this system is on the order of 1 mm. Spatial resolution can
be increased using a different camera, but since light is
averaged across the 1 cm thickness of the chamber, spatial
resolution less than 1 mm provides no additional informa-
tion. Captured images are pseudocolorized to accen-
tuate differences in emitted light intensity, which are then
correlated to colloid concentrations. Time-lapse image cap-
ture enables investigation of mass transport of colloids
through the experimental plane over time.
[23] The porous media visualization system described Figure 10. Processed and pseudocolorized light transmis-
above has been used to investigate a wide range of variably sion image from the Kraft and Selker [2005] system of four
saturated transport processes. These investigations include colloid plumes of different concentration in 30/40 silica
the observation and/or quantification of the following: sand. The detection of the 0.6 ppm (9 105 particles
fingered flow [Selker et al., 1992a], wetting front instability mL 1) plume represents a hundredfold improvement on the
[Selker et al., 1992b], water content [Niemet and Selker, system’s detection limit as a result of improved signal-to-
noise ratio.
10 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Figure 11. Processed time series of colloid fluorescence light transmission images over 1480 min.
Plumes are 0.04% concentration of 0.02 mm (left) and 1.0 mm (right) latex microspheres encountering the
capillary fringe, where the black lines are water content contours. Reprinted with permission from
Weisbrod et al. [2003b]. Copyright 2003 American Chemical Society.
microbial growth in the chamber has resulted in a model of general, resolution should be optimized for the length scale
microbial density as a function of light emission that can of interest. The MRI systems presented here are limited in
predict biomass over four orders of magnitude, and obser- size and may be most appropriate for studies of small-scale
vation has been made of colony growth against the direction media heterogeneities such nonrandom distribution of dif-
of flow [Yarwood et al., 2002]. ferent grain types (size, shape, composition). MRI systems
[24] Visualization systems based on MRI have been with lower resolution but which allow for larger samples
developed by several groups [Potter et al., 1996; Sherwood may be suitable for investigation of higher-order media
et al., 2003; Baumann and Werth, 2005]. Flow cells consist features and heterogeneities. Detection limit and system
of cylindrical columns ranging in diameter from 1 to 1.5 cm sensitivity are particularly relevant to the discussion of
and in length from 7.5 to 8 cm. Columns have been packed mesoscale visualization systems which track colloid concen-
with 250 to 300 mm diameter glass-coated polystyrene trations. The minimum colloid concentrations detectable by
microbeads [Sherwood et al., 2003] or with silica gel the light-based systems described here (105 – 106 particles
ranging in size from 250 to 1000 mm [Baumann and Werth, mL 1) fall in the lower range of naturally occurring concen-
2005]. Imaging is accomplished by placing the column trations (105 – 1014 particles mL 1) reported by Kim [1991].
within a magnetic resonance spectrometer system, and Detection limits of the MRI systems are not as well docu-
requires use of paramagnetic colloids or labeling of colloids mented, but for the system developed by Baumann and Werth
with paramagnetic materials. Sherwood et al. [2003] were [2005], appear to be on the order of 108 particles mL 1
able to visualize diffusion in a water-saturated porous (approximately 55 ppm). Colloid deposition for 1 mm par-
medium of motile and nonmotile bacteria labeled with ticles has been shown to be concentration-dependent at
50– 60 nm diameter magnetite particles at a spatial resolu- concentrations of 9.7 106 to 7.7 107 particles mL 1
tion of 330 mm. Utilizing the same visualization setup, [Bradford and Bettahar, 2006] (approximately 5.4 to
Olson et al. investigated chemotactic response of magnetite- 42.5 ppm), suggesting that interpretation of results from
labeled Pseudomonas putida F1 [Olson et al., 2004] and mesoscale visualization studies should take into account
retarded diffusion of P. putida F1 in porous media relative to potential concentration dependence. Currently, both visible
a conservative solute and Escherichia coli NR50 [Olson et light and MRI visualization methods have limited utility for
al., 2005]. Baumann and Werth [2005] observed retarded imaging colloid transport in natural material due to media
transport of 1.28 mm diameter superparamagnetic colloids opacity and signal interference by media components such as
(density 1.726 g cm 3) through 250– 600 mm silica gel paramagnetic and ferromagnetic particles, clay, or organic
compared to 850 – 100 mm silica gel. The maximum voxel matter [Hall et al., 1997]. Development of engineered media,
resolution achieved by this system was 200 200 which better approximate features of natural media such as
1000 mm. An example of images generated by the MRI surface roughness and chemistry while retaining character-
system developed by Baumann and Werth [2005] is pre- istics necessary for visualization, may extend the utility of
sented in Figure 12. current light and MRI-based visualization systems.
[25] Mesoscale visualization systems span a range of
spatial scales. While methods development efforts often 5. Summary and Conclusions
focus on improving spatial resolution, these refinements
must be weighed against the concomitant reduction in [26] Over the past decade, visualization studies have
sample size and scale of features that can be studied. In contributed to the elucidation of several mechanisms rele-
11 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
12 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Corapcioglu, M. Y., and S. Jiang (1993), Colloid-facilitated groundwater Li, X., C. Lin, J. D. Miller, and W. P. Johnson (2006), Pore-scale
contaminant transport, Water Resour. Res., 29, 2215 – 2226. observation of microsphere deposition at grain-to-grain contacts over
Crist, J. T., J. F. McCarthy, Y. Zevi, P. Baveye, J. A. Throop, and T. S. assemblage-scale porous media domains using X-ray microtomogra-
Steenhuis (2004), Pore-scale visualization of colloid transport and reten- phy, Environ. Sci. Technol., 40, 3762 – 3768.
tion in partly saturated porous media, Vadose Zone J., 3, 444 – 450. Li, Y., W. A. Dick, and O. H. Tuovinen (2004), Fluorescence microscopy
Crist, J. T., Y. Zevi, J. F. McCarthy, J. A. Throop, and T. S. Steenhuis for visualization of soil microorganisms: A review, Biol. Fertil. Soils, 39,
(2005), Transport and retention mechanisms of colloids in partially sa- 301 – 311.
turated porous media, Vadose Zone J., 4, 184 – 195. Manz, B., F. Volke, D. Goll, and H. Horn (2003), Measuring local flow
Cumbie, D. H., and L. D. McKay (1999), Influence of diameter on particle velocities and biofilm structure in biofilm systems with magnetic reso-
transport in a fractured shale saprolite, J. Contam. Hydrol., 37, 139 – 157. nance imaging (MRI), Biotechnol. Bioeng., 84, 424 – 432.
de Jonge, L. W., C. Kjaergaard, and P. Moldrup (2004a), Colloids and McCarthy, J. F., and L. D. McKay (2004), Colloid transport in the subsur-
colloid-facilitated transport of contaminants in soils: An introduction, face: Past, present, and future challenges, Vadose Zone J., 3, 326 – 337.
Vadose Zone J., 3, 321 – 325. McCarthy, J. F., and J. M. Zachara (1989), Subsurface transport of con-
de Jonge, L. W., P. Moldrup, G. H. Rubaek, K. Schelde, and J. Djurhuus taminants, Environ. Sci. Technol., 23, 496 – 502.
(2004b), Particle leaching and particle-facilitated transport of phosphorus Niemet, M. R., and J. S. Selker (2001), A new method for quantification of
at field scale, Vadose Zone J., 3, 462 – 470. liquid saturation in 2D translucent porous media systems using light
DeNovio, N. M., J. E. Saiers, and J. N. Ryan (2004), Colloid movement in transmission, Adv. Water Resour., 24, 651 – 666.
unsaturated porous media: Recent advances and future directions, Vadose Niemet, M. R., M. L. Rockhold, N. Weisbrod, and J. S. Selker (2002),
Zone J., 3, 338 – 351. Relationships between gas-liquid interfacial surface area, liquid satura-
DiCarlo, D. A., Y. Zevi, A. Dathe, S. Giri, B. Gao, and T. S. Steenhuis tion, and light transmission in variably saturated porous media, Water
(2006), In situ measurements of colloid transport and retention using Resour. Res., 38(8), 1135, doi:10.1029/2001WR000785.
synchrotron X-ray fluorescence, Water Resour. Res., 42, W12S05, Ochiai, N., M. Dragila, and J. Parke (2005), Use of rectangular capillary
doi:10.1029/2005WR004850. chambers for nondestructive microscopic observation of biofilm on sand,
Gao, B., J. Saiers, and J. Ryan (2006), Pore-scale mechanisms of colloid Eos Trans. AGU, 86(52) Fall Meet. Suppl., Abstract B31A-0963.
deposition and mobilization during steady and transient flow through Olson, M. S., R. M. Ford, J. A. Smith, and E. J. Fernandev (2004), Quan-
unsaturated granular media, Water Resour. Res., 42, W01410, doi:10.1029/ tification of bacterial chemotaxis in porous media using magnetic reso-
2005WR004233. nance imaging, Environ. Sci. Technol., 38, 3864 – 3870.
Hall, L. D., M. H. G. Amin, E. Dougherty, M. Sanda, J. Votrubova, K. S. Olson, M. S., R. M. Ford, J. A. Smith, and E. J. Fernandez (2005), Analysis
Richards, R. J. Chorley, and M. Cislerova (1997), MR properties of water of column tortuosity for MnCl2 and bacterial diffusion using magnetic
in saturated soils and resulting loss of MRI signal in water content resonance imaging, Environ. Sci. Technol., 39, 149 – 154.
detection at 2 tesla, Geoderma, 80, 431 – 448. Parker, L., R. Yarwood, and J. Selker (2006), Observations of gas flow in
Hardham, A., and E. Suzaki (1986), Encystment of zoospores of the fungus, porous media using a light transmission technique, Water Resour. Res.,
Phytophthora cinnamomi, is induced by specific lectin and monoclonal 42, W05501, doi:10.1029/2005WR004080.
antibody binding to the cell surface, Protoplasma, 133, 165 – 173. Pesnyakevich, A., and G. Lyon (1998), Transformation of Erwinia caroto-
Harvey, R. W., L. H. George, R. L. Smith, and D. R. LeBlanc (1989), vora ssp. atroseptica with the green fluorescent protein gene from
Transport of microspheres and indigenous bacteria through a sandy aqui- Aequorea Victoria, J. Phytopathology, 146, 57 – 60.
fer: Results of natural- and forced-gradient tracer experiments, Environ. Potter, K., R. L. Kleinberg, F. J. Brockman, and E. W. McFarland (1996),
Sci. Technol., 23, 51 – 56. Assay for bacteria in porous media by diffusion-weighted NMR,
Harvey, R. W., N. E. Kinner, D. Macdonald, D. Metge, and A. Bunn J. Magn. Reson., Ser. B, 113, 9 – 15.
(1993), Role of physical heterogeneity on the interpretation of small- Puls, R. W., and R. M. Powell (1992), Transport of inorganic colloids
scale laboratory and field observations of bacteria, microbial-sized through natural aquifer material: Implications for contaminant transport,
microsphere, and bromide transport through aquifer sediments, Water Environ. Sci. Technol., 26, 614 – 621.
Resour. Res., 29, 2713 – 2721. Rangel-German, E. R., and A. R. Kovscek (2006), A micromodel investi-
Hsieh, H., G. Brown, and M. Stone (1998), Measurement of porous media gation of two-phase matrix-fracture transfer mechanisms, Water Resour.
component content and heterogeneity using gamma ray tomography, Res., 42, W03401, doi:10.1029/2004WR003918.
Water Resour. Res., 34, 365 – 372. Rockhold, M. L., R. R. Yarwood, M. R. Niemet, P. J. Bottomley, F. J.
Johnson, P., N. Sun, and M. Elimelech (1996), Colloid transport in geo- Brockman, and J. S. Selker (2006), Visualization and modeling of the
chemically heterogeneous porous media: Modeling and measurements, colonization dynamics of a bioluminescent bacterium in variably satu-
Environ. Sci. Technol., 30, 3284 – 3293. rated, translucent quartz sand, Adv. Water Resour., in press.
Keller, A. A., M. J. Blunt, and P. V. Roberts (1997), Micromodel observa- Saiers, J. E. (2002), Laboratory observations and mathematical modeling of
tion of the role of oil layers in three-phase flow, Transp. Porous Media, colloid-facilitated contaminant transport in chemically heterogeneous
26, 277 – 297. systems, Water Resour. Res., 38(4), 1032, doi:10.1029/2001WR000320.
Keller, A. A., S. Sirivithayapakorn, and C. V. Chrysikopoulos (2004), Early Saiers, J. E., and G. M. Hornberger (1996), The role of colloidal kaolinite in
breakthrough of colloids and bacteriophage MS2 in a water-saturated sand the transport of cesium through laboratory sand columns, Water Resour.
column, Water Resour. Res., 40, W08304, doi:10.1029/2003WR002676. Res., 32, 33 – 41.
Kersting, A. B., D. W. Efurd, D. L. Finnegan, D. J. Rokop, D. K. Smith, Saiers, J., and J. Ryan (2005), Colloid deposition on non-ideal media: The
and J. L. Thompson (1999), Migration of plutonium in ground water at influences of collector shape and roughness on the single-collector effi-
the Nevada Test Site, Nature, 397, 56 – 59. ciency, Geophys. Res. Lett., 32, L21406, doi:10.1029/2005GL024343.
Khalili, A., A. Basu, and U. Pietrzyk (1998), Flow visualization in porous Saiers, J. E., G. M. Hornberger, and C. Harvey (1994), Colloidal silica
media via positron emission tomography, Phys. Fluids, 10, 1031 – 1033. transport through structured, heterogeneous porous media, J. Hydrol.,
Kim, J. (1991), Actinide colloid generation in groundwater, Radiochim. 163, 271 – 288.
Acta, 52/53, 71 – 81. Saiers, J. E., G. M. Hornberger, D. B. Gower, and J. S. Herman (2003), The
Ko, C., and J. Chen (2000), Dynamics of silica colloid deposition and role of moving air-water interfaces in colloid mobilization within the vadose
release in packed beds of aminosilane-modified glass beads, Langmuir, zone, Geophys. Res. Lett., 30(21), 2083, doi:10.1029/2003GL018418.
16, 6906 – 6912. Schafer, A., P. Ustohal, H. Harms, F. Stauffer, T. Dracos, and A. J. B.
Kraft, E. L., and J. S. Selker (2005), Methodological developments in in- Zehnder (1998), Transport of bacteria in unsaturated porous media,
situ measurement and visualization of colloid transport processes in the J. Contam. Hydrol., 33, 149 – 169.
subsurface, paper presented at Geological Society of America Annual Selker, J. S., P. Leclerq, J.-Y. Parlange, and T. Steenhuis (1992a), Fingered
Meeting and Exposition, Salt Lake City, Utah. flow in two dimensions: 1. Measurement of matric potential, Water
Lanning, L. M., and R. M. Ford (2002), Glass micromodel study of bacter- Resour. Res., 28, 2513 – 2521.
ial dispersion in spatially periodic porous networks, Biotechnol. Bioeng., Selker, J. S., J. Y. Parlange, and T. Steenhuis (1992b), Wetting front in-
78, 556 – 566. stability in homogeneous sandy soils under continuous infiltration, Soil
Lenhart, J. J., and J. E. Saiers (2004), Adsorption of natural organic matter Sci. Soc. Am. J., 56, 1346 – 1350.
to air-water interfaces during transport through unsaturated porous media, Seymour, J. D., S. L. Codd, E. L. Gjersing, and P. S. Stewart (2004),
Environ. Sci. Technol., 38, 120 – 126. Magnetic resonance microscopy of biofilm structure and impact on trans-
port in a capillary bioreactor, J. Magn. Reson., 167, 322 – 327.
13 of 14
W12S06 OCHIAI ET AL.: COLLOID TRANSPORT VISUALIZATION METHODS W12S06
Sherwood, J. L., J. C. Sung, R. M. Ford, E. J. Fernandez, J. E. Maneval, and Wan, J. M., T. K. Tokunaga, C. F. Tsang, and G. S. Bodvarsson (1996),
J. A. Smith (2003), Analysis of bacterial random motility in a porous Improved glass micromodel methods for studies of flow and transport in
medium using magnetic resonance imaging and immunomagnetic label- fractured porous media, Water Resour. Res., 32, 1955 – 1964.
ing, Environ. Sci. Technol., 37, 781 – 785. Weisbrod, N., M. R. Niemet, and J. S. Selker (2002), Imbibition of saline
Silliman, S. E., R. Dunlap, M. Fletcher, and M. A. Schneegurt (2001), solutions into dry and prewetted porous media, Adv. Water Resour., 25,
Bacterial transport in heterogeneous porous media: Observations from 841 – 855.
laboratory experiments, Water Resour. Res., 37, 2699 – 2707. Weisbrod, N., M. R. Niemet, T. McGinnis, and J. S. Selker (2003a), Water
Sirivithayapakorn, S., and A. Keller (2003a), Transport of colloids in satu- vapor transport in the vicinity of imbibing saline plumes: Homogeneous
rated porous media: A pore-scale observation of the size exclusion effect and layered unsaturated porous media, Water Resour. Res., 39(6), 1145,
and colloid acceleration, Water Resour. Res., 39(4), 1109, doi:10.1029/ doi:10.1029/2002WR001539.
2002WR001583. Weisbrod, N., M. R. Niemet, and J. S. Selker (2003b), Light transmission
Sirivithayapakorn, S., and A. Keller (2003b), Transport of colloids in un- technique for the evaluation of colloidal transport and dynamics in por-
saturated porous media: A pore-scale observation of processes during the ous media, Environ. Sci. Technol., 37, 3694 – 3700.
dissolution of air-water interface, Water Resour. Res., 39(12), 1346, White, D., M. M. Leifert, and K. K. Ryder (1996), Lux gene technology: A
doi:10.1029/2003WR002487. strategy to optimize biological control of soil-borne diseases, New Phy-
Steenhuis, T. S., J. F. McCarthy, J. T. Crist, Y. Zevi, P. C. Baveye, J. A. tol., 133, 173 – 181.
Throop, R. L. Fehrman, A. Dathe, and B. K. Richards (2005), Reply Wildenschild, D., J. Hopmans, C. Vaz, M. Rivers, D. Rikard, and
to ‘‘Comments on ‘Pore-scale visualization of colloid transport and B. Christensen (2002), Using X-ray computed tomography in hydrology:
retention in partly saturated porous media’,’’ Vadose Zone J., 4, Systems, resolutions, and limitations, J. Hydrol., 267, 285 – 297.
957 – 958. Yarwood, R. R., M. L. Rockhold, M. R. Niemet, J. S. Selker, and P. J.
Sun, N., M. Elimelech, N. Z. Sun, and J. N. Ryan (2001), A novel two- Bottomley (2002), Noninvasive quantitative measurement of bacterial
dimensional model for colloid transport in physically and geochemically growth in porous media under unsaturated-flow conditions, Appl. Environ.
heterogeneous porous media, J. Contam. Hydrol., 49, 173 – 199. Microbiol., 68, 3597 – 3605.
Thieme, J., G. Schneider, and C. Knochel (2003), X-ray tomography of a Yarwood, R., M. Rockhold, M. Niemet, J. Selker, and P. Bottomley (2006),
microhabitat of bacteria and other soil colloids with sub-100 nm resolu- Impact of microbial growth on water flow and solute transport in un-
tion, Micron, 34, 339 – 344. saturated porous media, Water Resour. Res., 42, W10405, doi:10.1029/
Toran, L., and A. V. Palumbo (1992), Colloid transport through fractured 2005WR004550.
and unfractured laboratory sand columns, J. Contam. Hydrol., 9, 289 – Yoon, J. S., J. T. Germaine, and P. J. Culligan (2006), Visualization of
303. particle behavior within a porous medium: Mechanisms for particle fil-
Tufenkji, N., J. A. Redman, and M. Elimelech (2003), Interpreting deposi- tration and retardation during downward transport, Water Resour. Res.,
tion patterns of microbial particles in laboratory-scale column experi- 42, W06417, doi:10.1029/2004WR003660.
ments, Environ. Sci. Technol., 37, 616 – 623. Zevi, Y., A. Dathe, J. F. McCarthy, B. K. Richards, and T. S. Steenhuis
Uesugi, S. L., R. R. Yarwood, J. S. Selker, and P. J. Bottomley (2001), A (2005), Distribution of colloid particles onto interfaces in partially satu-
model that uses the induction phase of lux gene-dependent biolumines- rated sand, Environ. Sci. Technol., 39, 7055 – 7064.
cence in Pseudomonas fluorescens HK44 to quantify cell density in Zhang, P., and Y. Wang (2006), Epi-fluorescence imaging of colloid trans-
translucent porous media, J. Microbiol. Methods, 47, 315 – 322. port in porous media at decimeter scales, Environ. Sci. Technol., 40(19),
Wan, J. M., and T. K. Tokunaga (1997), Film straining of colloids in 6064 – 6069, doi:10.1021/es061104y.
unsaturated porous media: Conceptual model and experimental testing,
Environ. Sci. Technol., 31, 2413 – 2420.
Wan, J. M., and T. K. Tokunaga (2005), Comments on ‘‘Pore-scale visua-
lization of colloid transport and retention in partly saturated porous E. L. Kraft and J. S. Selker, Department of Biological and Ecological
media,’’ Vadose Zone J., 4, 954 – 956. Engineering, Oregon State University, Gilmore Hall, Room 116, Corvallis,
Wan, J., and J. L. Wilson (1994), Visualization of the role of the gas-water OR 97331-3609, USA. (krafte@engr.orst.edu)
interface on the fate and transport of colloids in porous media, Water N. Ochiai, Department of Crop and Soil Science, Oregon State
Resour. Res., 30, 11 – 23. University, Corvallis, OR 97331, USA.
14 of 14