The following is a slide show of the presentation I gave at the
4th Research Forum on Recycling in Quebec, QC, Canada in Oct,1997:
The central tower to the Chateau Frontenac in the old city of
Quebec where the Forum was held! Pretty nice!
Click on title or image for full page scale slides.
Quebec Slide 1: Fluid Interfacial Separations
for Secondary Fiber Recovery as Probed with Atomic Force
Microscopy
I'm Eric Aston from the Department of Chemical Engineering at the
University of Washington working for Professor John Berg. The behavior of
ink particles at water-fluid interfaces is fundamental to all separation
processes for secondary fiber recovery. We've begun research in this
field by studying a system that we are already familiar with, that is,
oil-assisted agglomeration for de-inking toner-printed papers.
Specifically, we've focused on a direct, single-particle analysis using
atomic force microscopy (a.k.a. AFM) set up for colloidal probing. This
colloidal force microscopy is not a conventional AFM mode and has only
come into practice for fluid/fluid interfacial studies in the last two
years. Our very first results were published in Langmuir this last
February.
Quebec Slide 2: Toner-Ink
Oil-Assisted Agglomeration
Industry has recently moved away from oil-assisted agglomeration for
recycling toner-printed office waste in favor of densification processes.
This change was spurred on by the much larger volume of toner usage and by
ever increasing restrictions on volatile organic carbon (VOC) emissions.
However, oil-assisted agglomeration is still useful in other respects.
The critical separation step in toner de-inking by oil-assisted
agglomeration is the coalescence of dispersed toner and oil into large
aggregates that can be easily removed from the repulped paper slurries
leaving clean cellulose fibers. Furthermore, the key is to create
aggregates of adequate strength to remain stable against the high shear
forces experienced in the repulpers. It has been shown that cationic
polymers commonly present in most papers inhibit oil-toner coalescence by
forming electrosteric adlayers. Agglomerating agents of simple oils with
surfactants are able to overcome this steric stabilization [to wetting the
toner]. The addition of a surfactant into the oil phase allows
coalescence to occur even with starch-covered particles.
Quebec Slide 3: Cationic Starch
Adlayers
In recent tests, we used a commercial brand, modified potato starch
(STA-LOK 400--derivatized with quaternary ammonium groups) as an
adsorbing polymer similar to cationic starches present in many papers
known to inhibit toner aggregation (separation from paper slurries by oil
agglomeration processes). The cationic groups on the starch chain
(electrosterically) associate with the negative surface charges on the
toner particle and the oil-aqueous interface, forming adlayers found to be
49 nm and 33 nm in thickness from dynamic light scattering. Significant
amounts of starch were also adsorbed by the oil/surfactant blend, but with
no detectable change in droplet size. Along with trends in zeta potential
(electrokinetic), this reveals the quite remarkable situation of
hydrophilic starch engulfment into the oil drop leaving no adlayer on the
aqueous side. Our oil/surfactant blend was a commercial agglomerating
agent from Betz Paperchem, Inc. It consists of roughly 60% simple
aliphatic oils and 40% largely hydrophobic (HLB--low hydrophile-lipophile
balance) surfactants which reside almost completely in the oil phase. The
presence of the oil surfactants is enough to allow starch penetration into
the oil. Even though the Betz product requires elevated temperatures for
maximum performance, the AFM is only suited for making room temperature
comparisons.
Quebec Slide 4:
Objectives
Our main objective was to use the atomic force microscope (AFM) to examine
forces of interaction between an oil droplet under water and a hydrophobic
toner particle. The three parameters we've chosen to focus on with AFM
are the attractive "snap-in" force pulling the toner and oil together, the
applied force required to induce particle attachment or film rupture, and
the induction time, which we estimate from the time the toner first
encounters the interface until it enters the oil (initial repulsion until
final snap-in). The resulting force-distance curves are compared with
advancing oil contact angles on toner flats under water. This
single-particle AFM analysis is also the first step toward a direct
comparison to bulk agglomeration studies that gauge the final degree of
toner dispersion (aggregation).
Quebec Slide 5: AFM Cell For
Liquid-Liquid Interfaces
This schematic shows an unfused toner particle roughly 10 microns in
diameter glued at the end of a flexible, (commercial) AFM cantilever.
This assembly is then lowered into the liquid cell for probing. A Teflon
tube (ID = 2.25 mm) glued to the bottom of the cell anchors the
n-hexadecane, or oil/surfactant, under water. The cell sits on the AFM
scanner that uses piezoelectric transducers to move, specifically, the
oil-water interface toward and away from the toner. When the toner is
near the interface, it will experience attractive or repulsive forces
deflecting the cantilever down or up, respectively. The interface is also
free to deform as a result of long-range interactions and physical contact
with the toner. A laser reflecting from the back of the cantilever is
directed to a photodetector (PSPD) measuring the shifts in position. The
stiffness of the cantilever is known in terms of a spring constant, so the
deflection corresponds to a quantifiable force typically in the nanoNewton
range (from pico- to micro-Newtons).
Quebec Slide 6: AFM Force vs.
Distance
We've observed three common interactions in our oil-aqueous-toner systems.
Force profile "A" illustrates the main features for contacting oil and
toner in pure water, and curve "B" shows the decidedly different behavior
when cationic starch is added to the water. The horizontal line is the
undeflected cantilever signal, or zero interaction force, as the oil
approaches the toner. There is a (1) rapid increase in force at or near
contact with the oil interface. Hydrodynamic forces initially repel the
two hydrophobic materials until the long-range attraction precipitates
snap-in and contact. An intervening aqueous film may also resist the
attraction until it drains and ruptures. Further advancing continues with
the cantilever deflecting in constant compliance with the oil interface.
Reversing the direction, (2) the toner drags the interface along and must
be pulled back more than 50 microns to detach it from the oil. This large
pull-off force can only correspond to a capillary force resulting from a
three-phase contact line with radius approaching the toner size. When
cationic starch is present to form steric adlayers, (3) the approach curve
is monotonically repulsive. No attractive snap-in is observed, and little
or no pull-off force is required to detach the toner. In this case, the
oil and toner cannot coalesce and agglomeration will fail.
Quebec Slide 7: Toner Particle
Approach to Oil-in-Water
These AFM experiments used a toner particle approaching the interface of
interest at approximately 30 um/s. This simulates the high-shear
environment experienced in repulpers. This approach curve for the pure
water case shows the toner being engulfed by the oil after some initial
viscous repulsion. Optical images of a larger toner particle (~100 um)
interacting with the oil under water were acquired to confirm the events
as interpreted from the AFM data. While watching the live video, the
snap-in was quite obvious, though it is difficult to appreciate here. The
interface was rapidly pulled up and formed a neck with the toner particle.
Quebec Slide 8: Toner and Oil
with Cationic Starch Adlayers
The difference with the 0.1 wt% aqueous starch dispersion was anticipated
from bulk agglomeration studies but is still quite remarkable. The
approach curve, given on the same scale as the previous plot, is
monotonically repulsive; that is, no attractive snap-in was detected.
Again, optical images show the extreme to which a toner particle can
deform the oil drop without wetting, and then be pulled back cleanly, when
both are covered with starch adlayers.
Quebec Slide 8B: Toner and
Oil/Surfactant Blend
This slide show images and force profiles never used in the talk although
the outcome of these data are summed up in the tables of comparative
results following. The Betz oil/surfactant always wet the toner surfaces
whether in the presence of cationic starch or not. The plots show not
great differences in character for either case.
Quebec Slide 9: Review of
Results
Now, we can make some gross, quantitative comparisons with the snap-in
forces, or distances, the induction times, and the repulsive loads before
wetting contact. For oil in pure water, the induction time for the toner
was 52 ms at an average applied force of 5.6 nN before the very aggressive
snap-in which was off scale and greater than 16.6 nN. The snap-in
distance of >332 nm was calculated directly from the force and includes a
significant degree of toner entry into the oil; with further analysis
we'll get a better idea for the true interfacial separation just prior to
snap-in which will be much smaller than this value. The addition of
cationic starch at least delayed interfacial rupture beyond 191 ms, we
suspect indefinitely, and above 77 nN applied load. These pure oil
results are in agreement with the starch-induced stability against
coalescence previously observed.
When the pure oil is replaced with the Betz oil/surfactant blend, we
once again record a measurable induction time, about 133 ms (± 26), at
loads around 0.7 nN (± 0). The snap-in is much weaker than recorded with
pure oil (3.5 ± 0.57 nN). This time the addition of starch has a much
less dramatic effect on coalescence (151 ± 47 ms, 1.4 ± 0.16 nN, 2.5 ±
0.18 nN) agreeing with the past observations of the oil/surfactant's
ability to imbibe the cationic starch adsorbed to both toner and oil. The
induction time and load are slightly elevated with a smaller snap-in,
suggesting that the starch may still be working against complete wetting
of the toner.
Quebec Slide 10: Final
Comparisons
These single-particle measurements in oil-aqueous-toner systems further
illustrate the roles of cationic starch and oil surfactants in the
coalescence of toner and agree with bulk agglomeration studies. Steric
stability against toner attachment to oil due to the adsorbed starch is
overcome by the addition of a largely hydrophobic surfactant to the oil
phase. A qualitative comparison can be made between the snap-in forces as
from the AFM and the advancing oil contact angles. Only the case of pure
oil in the presence of cationic starch showed interfacial stability (no
interfacial rupture) and a nonwetting contact angle (95°).
Direct, quantitative comparisons between snap-in force and contact
angle for the oil and the Betz product are not easily made because of the
effects that interfacial stiffness, viscosity and particle entry have on
the AFM analysis. But wetting vs. nonwetting comparisons are still
applicable. Both pure water cases show that toner is wet by the oil
phase, but toner covered with starch more easily attaches to the
oil/surfactant than the pure oil. Looking the Betz experiments with
contact angles suggests that the starch has little effect on toner
wettability with the oil/surfactant additive.
Quebec Slide 11: Snap-in
Interpretation
Our most recent work with recording interactions at liquid interfaces
shows great promise for finally generating a true force-separation curve
which can be directly compared to theoretical plots of strictly
intermolecular forces. Previously, we've been restricted to snap-in
values from the AFM with the knowledge that they included not only the
actual long-range jump into contact but also a significant degree of
particle entry into the oil. A new technique uses an oscilloscope to
collect AFM data much faster than is normally considered. Through a
simple dynamic force analysis, we are able to subtract most, if not all,
of the external, competing forces, like mechanical, capillary, and
hydrodynamic effects, to reveal the intermolecular forces.
The toner and oil interface seem to interact in a three-step sequence
during snap-in. A long-range attraction first pulls the interface toward
the toner particle. When the force field gradient exceeds the stiffness
of either the interface or the cantilever to which the particle is glued,
the dynamic snap-in begins. The separation rapidly decreases until the
intervening water phase forms a thin film that resists the motion via
viscous film drainage. The toner particle is finally wet by the oil only
after the thin film drains and ruptures against the applied pressure.
The snap-in distance is commonly used to compare the range of
attractive interactions between systems. But it is experimentally
underestimated because the AFM measurement does NOT include the film
thickness. Ideally, the snap-in distance would be the maximum separation
at which a free particle is attracted into direct, physical contact with
the interface. Because our particle is not free but is attached to a
cantilever, the AFM again underestimates this value.
Quebec Slide 12: Dynamic
Force Analysis
After the AFM data are modified by this dynamic force analysis, the
already significant differences in the behavior of oil and oil/surfactant
(with toner) are magnified. The new force curve shown in red emphasizes
the incredibly strong attraction for hydrophobic toner particles to the
pure n-hexadecane. The toner particle jumps about 70-100 nm closer to the
oil droplet where it meets with what we've interpreted as viscous film
drainage. After film rupture, the oil wets the particle and pulls it in
(by capillary forces). The oil/surfactant blend provides a very much
weaker attraction relying mostly on its enhanced wetting ability for
coalescence. The amphiphilic surfactant molecules at the oil interface
reduce the hydrophobicity and no long-range attraction is evident from the
dynamic analysis. A snap-in of several nanometers or less would not have
been detected with this design. Increased magnification of the force
profile shows slow wetting taking place, and optical images confirm
engulfment.
We're really one step away from a complete analysis that can
extract the intermolecular force profile from this type of AFM experiment.
Hopefully, the last phase will be to model the viscous film drainage to
confirm our thin film hypothesis and then subtract its effects from the
existing curve. The result will be an even greater well depth in the
force profile. In addition to generating new fundamental information, we
appear to be on our way to developing a practical tool for testing the
affect of surface and solution chemistry on such processes as
air-flotation and oil-assisted agglomeration.