Field of the Invention
The present invention relates to methods and apparatus for the characterisation
of solid/liquid interfaces and their interaction with particles such as cells.
Determination of the adhesion of particles to surfaces finds many
applications. For example, a need exists in the evaluation of biopolymers, where
the minimization of adhesion is an objective of specific interest, see "Biological
evaluation of medical devices" - Part 4. ISO 10993-4 (1992), "A micro mechanical
technique for monitoring cell-substrate adhesiveness: Measurements of the strength
of red blood cell adhesion to glass and polymer test surfaces" V.M. Bowers, L.R.
Fisher, G.W. Francis and K.L. Williams. J. Biomed. Mat. Res. 23, 1453 (1989), "Catheter
sepsis due to staphylococcus epidermis during parenteral nutrition" A. Sitges-Serra,
P. Puig, E. Janrietta, M.D Garau, A. Alastrue and A. Sitges-Creus, Surg. Gynecol.
Obsstet. 151, pp. 481-483 (1980), "In vitro quantitative adherence of bacteria
to intravascular catheters", N.K. Shet, H.D. Rose, T.R. Franson, F.L.A. Buckmire
and P.G. Sohnie J. Surg. Res. 34, pp. 213-218 (1983).
The behaviour of particles in a flowing liquid has been investigated
theoretically - Taylor G. "Dispersion of soluble matter in solvent flowing slowly
through a tube", Proc Roy Soc London A 1953; 219:187-203; Golay M. J. E, Atwood
J.G, "Early phases of the dispersion of a sample injected in Poiseuille flow",
J. of Chromatogr. 1979;186:353-70.
The dispersion of a sample injected in convective cell or particle
transport in the conditions of Poiseuille flow is subject to shear, gravitational
settling and hydrodynamic lift, all of which appear interrelated. Constrictions
of the flow path in the sampling valve do not affect the laminar flow character
provided the Reynolds number is kept within certain ranges as shown by Ahmed SA,
Giddens DP, in "Velocity measurements in steady flow through axisymmetric stenoses
at moderate Reynolds numbers", J Biomechanics 1983;16(7):505-16.
In the article by Cherukat P and McLaughlin B entitled "The inertial
lift on a rigid sphere in a linear shear flow field near a flat wall", J. Fluid
Mech 1994;263:1-18, the magnitude of lift in the linear shear zone was calculated,
as occurs close to the capillary walls. On the basis of their findings, a physical
model was formulated that describes the effects of lift and settling forces. The
model predicts transition from fully developed Segré-Silberberg transport at high
flow rates of the carrier liquid to transport conditions where wall contacts must
be expected. The model teaches that at any given flow speed only cells or particles
of well determined size/density ratios are able to translate near the wall, in
the absence of resultant forces. Larger and/or more dense ones are expected to
settle and contact the wall while smaller and/or less dense cells/particles will
be pulled away from it, to end up in regions of much higher flow speeds. The paper
does not deal with the problems of adhesion.
Basic problems, in fact, are the slow equilibration of adhesion processes,
- see "Mechanics of particle adhesion", Rimai DS, DeMejo LP, Mittal KL, eds.,
Fundamentals of Adhesives and Interfaces. Utrecht:VSP;1995. pp. 1-34, and their
susceptibility to surface roughness - see Shaeffer DM, Carpenter M, Gady B, Reifenberger
R, DeMejo LP, Rimai DS. "Surface roughness and its influence on particle adhesion,
using atomic force techniques." in Rimai DS, DeMejo LP, Mittal KL, eds. Fundamentals
of Adhesives and Interfaces. Utrecht:VSP;1995. pp. 35-48.
Although not directly related to the present invention, gas and liquid
chromotography are usful techniques for comparison purposes. Both chromotographies
relate to the determination of varied species in mixtures of gases or liquids or
solids dissolved in liquids whereas the present invention relates to particles,
cells, or other types of solid matter or discrete liquid matter which are dispersed
in liquids. A suitable reference for the discussion of liquid and gas chromotaography
is "A practical guide to Instrumental Analysis", Ernö Pungor, CRC Press, 1995.
Various industrial processes are known for separation of solids from
liquids. Examples are flotation and sedimentation. In these processes gravimetric
effects are used to separate the particles from the liquids. Generally, in these
processes flow is a disadvantage.
There remains a requirement to provide apparatus and methods for the
rapid and accurate assessment of interactions between particles such as cells and
a solid/liquid interface, especially adhesive interactions.
There is also a requirement to provide apparatus and methods for the
rapid and accurate assessment of solid/liquid interfaces, especially adhesive interactions
with particles such as cells.
Summary of the Invention
The present invention provides a qualitative or quantitative method
of evaluating an adhesion characteristic between at least one set of particles
in a liquid and a stationary phase, the stationary phase being provided on an inner
surface of a chamber, comprising: passing the particles in the liquid through the
chamber at a flow rate so that adhesion between the stationary phase and the particles
retards elution, and measuring a value relating to the retardation of elution of
the at least one set of particles as a measure of the adhesion characteristic.
The stationary phase provides a solid/liquid interface in the chamber. The stationary
phase is provided on an inner surface of the chamber, for instance, on an inner
wall or on the outer surface of beads in the chamber. The flow in the chamber is
preferably convective. The flow rate of passing liquid through the chamber is preferably
lower than that required for Segré-Silberberg transport of the set of particles.
On the other hand, the flow rate of passing liquid through the chamber is preferably
higher than required to allow partial sedimentation in the chamber and rolling
of the particles along the solid to liquid interface.
The present invention also provides a method of sorting one set of
particles from another set of particles in a liquid using a stationary phase, the
stationary phase being provided on an inner surface of a chamber, comprising: passing
the particles in the liquid through the chamber at a flow rate so that adhesion
between the stationary phase and the particles retards elution for at least one
of the sets of particles, and selecting a sample of the particles depending on
their degree of retardation. The flow rate of passing liquid through the chamber
is preferably lower than that required for Segré-Silberberg transport of at least
one of the sets of particles. The flow rate of passing liquid through the chamber
is preferably higher than required to allow partial sedimentation in the chamber
and rolling of at least one of the sets of particles along the solid to liquid
interface. The stationary phase is provided on an inner surface of the chamber,
for instance on an inner wall or on the outer surface of beads in the chamber.
The present invention also provides an apparatus for the qualitative
or quantitative evaluation of an adhesion characteristic between at least one set
of particles in a liquid and a stationary phase, comprising: a chamber with an
inner surface having the stationary phase, a delivery device for supplying the
particles in the liquid to the chamber at a flow rate so that adhesion between
the stationary phase and the particles of the at least one set retards elution,
and a measuring device for measuring a value relating to the retardation of elution
as a measure of the adhesion characteristic.
The present invention also provides an apparatus for sorting one set
of particles from another set of particles in a liquid using a stationary phase:
a chamber with an inner surface having the stationary phase, passing the particles
in the liquid through the chamber at a flow rate so that adhesion between the stationary
phase and the particles of at least one of the sets retards elution, and means
for selecting a sample of the particles depending on their degree of retardation.
The present invention will now be described with reference to the
Brief Description of the drawings
Detailed Description of the illustrative embodiments
- Figure 1 shows an experimental set-up for the registration of the transport
retardation of cells/particles in capillaries in accordance with an embodiment
of the present invention.
- Figure 2 shows an AC-operated gating suitable for use with the present invention.
Variations of the gate impedance upon passage of any cells/particles generate burst
LF-signals measurable across a sensing circuit such as a bridge.
- Figure 3 shows signal detection schematically based on logarithmic demodulation/amplification.
AD8307 : Log demodulator. LTC1560-1: active low-pass filter. VM1, AM1: bridge equilibration.
- Figure 4 shows elution diagrams of 10 µm latex particles at different flow speeds
(Um) of the carrier liquid as carried out in accordance with an embodiment
of the present invention. Capillary : 530 µm bore, methyl-coated and 5m in length.
- Figure 5 shows travelled distance of the leading flank of elution peaks as a
function of time and at various flow rates of the carrier liquid as carried out
in accordance with an embodiment of the present invention. Sample: 10 µm latex
particles. The column: methyl coated, 530 µm bore. Q: volume flow rate. R2:
- Figure 6 shows cumulative elution at different flow speeds (Um) of
the carrier liquid as carried out in accordance with an embodiment of the present
invention. The column: methyl coated, 530 µm bore and 5 m in length. Sample material:
10 µm latex particles.
- Figure 7 shows elution of 10 µm latex particles from capillaries with different
wall coatings, measured at various flow speeds of the carrier liquid as carried
out in accordance with the present invention. Capillaries: 530 µm bore, 5 m in
- Figure 8 shows the effect of the integral cell/particle number on transport
retardation as carried out in accordance with an embodiment of the present invention.
Sample: 10 µm latex particles. Column : glycol coated, 530 µm bore.
- Figure 9 shows forces implicated in the dynamic equilibrium in accordance with
certain theoretical considerations which may be relevant to the present invention:
Hydrodynamic lift (FL), gravitational settling (Fg), Archimedes
force (FA) and shear (Fs).
- Figure 10 shows the magnitude of lift as a function of slip velocity (Us),
according to Cox and Cherukat.
The present invention will be described with reference to certain
embodiments and drawings but these are provided as examples of the invention only
and the skilled persons will appreciate that many other implementations of the
present invention are included within the scope of the attached claims.
The present invention may provide in one aspect an apparatus and an
instrumental method for quantitative and/or qualitative evaluation of cell or particle
adhesion to interfaces especially to solid-liquid interfaces. The present invention
may also provide in another aspect an apparatus and an instrumental method for
quantitative or qualitative evaluation of solid-liquid interfaces especially by
their adhesive interaction with cells or particles. In one aspect of the present
invention, an apparatus and a method are described which provide a measure of adhesion
determined from a characteristic of the retardation of convective cell or particle
transport as measured in a chamber. The chamber may be a generally elongate hollow
structure such as a capillary tube or a column. Generally, the column will have
a liquid inlet and a liquid outlet. The chamber may have a uniform cross-section
along its length and this is generally preferred. The chamber may also be filled
with objects, e.g. beads or chunks of solid material. The characteristic may be
directly or indirectly related to the retardation. Within the chamber there is
an internal surface portion which is coated with the material under investigation.
The portion which is coated preferably on the walls of the chamber but can also
be a coated portion on structures such as beads which are located within the chamber.
The retardation may be measured with respect to a reference material or arrangement.
The cells or particles are preferably in the form of a dispersion, e.g. a suspension
or an emulsion, in a carrier liquid and are passed through the chamber, e.g. through
a capillary, in conditions of controlled flow. The controlled flow is preferably
convective flow. An emulsion may be defined as a dispersion of non-soluble liquid
in another. A suspension may be defined as a dispersion of solid particles in a
liquid. Further, the dispersion is preferably not subject to significant sedimentation
or floatation during the time of one test, i.e. gravimetric separation preferably
does not occur.
It is generally preferred if the concentration of particles in the
liquid remains constant during a test, that is that the flow is isocratic (= equal
strength) during the test. The particles may be any form of solid or liquid matter
which has a clearly defined separation from the liquid and has some form of distinguishable
and more than transient existence in the carrier liquid. The particles do not have
to be rigid - they may be deformable and when cell-liquid/solid interfaces are
being examined they will normally be deformable. Cells comprise large quantities
of water and the particles need not be solid but can comprise liquids. The cells
may be any type of cells, e.g. eukaryotic or prokaryotic cells. In particular,
cells from the human or animal body may be suitable such as red or white blood
corpuscles. Also the particles may be other structures such as micelles or liposomes
which form discretely identifiable bodies in a liquid even if these bodies are
liquid or mainly liquid. The present invention does not excludes the testing of
diffuse solid bodies such as obtained by flocculation.
In accordance with an aspect of the present invention delayed elution
at specific flow rates of the carrier liquid has been shown to reflect adhesive
interaction of the cells or particles with a coated surface, e.g. with the coated
chamber walls or with coated objects within the chamber. This delayed elution may
be used for qualitative or quantitative comparison purposes, e.g. as compared with
a reference substance. Various parameters may be used to characterise particle
interactions with solid/liquid interfaces. First of all the delay in eluting particles
may be taken as a measure of the adhesion between the particles and the solid/liquid
interface. The delay may be an absolute delay or may be relative, e.g. relative
to a standard. In order to compare with a standard one or more calibration graphs
may be prepared, e.g. a standard particle/solid/liquid combination may be used
to prepare calibration graphs of retardation time versus flow rate, retardation
time versus particle concentration, etc. Multi-dimensional calibration graphs may
also be prepared, e.g. a calibration graph relating flow rate:particle concentration:retardation
time or a calibration graph relating flow rate:particle concentration:retardation
More complex parameters may be used as the characteristic of particle
- liquid/solid interface adhesive interactions. For instance, based on a suitable
theory (see below) retardation values may be converted into an adhesive coefficient
or other similar parameter. Other parameters may be derived from simple time retardation
measurements, e.g. the number of particles retained in the chamber by the adhesive
effects during the time of retardation; the volume of liquid eluted within the
time of retardation, etc.
Fig. 1 shows a measurement arrangement 1 in accordance with an embodiment
of the present invention. Arrangement 1 comprises a delivery device 11 for delivering
cells or particles in a suitable fluid. For instance, the delivery device 11 may
comprise a loading syringe 2 for introducing, e.g. by gravity feeding or injecting,
a liquid containing particles or cells in suspension. A storage device 5 may be
provided to allow storage of a certain quantity of liquid containing particles,
e.g. sufficient for an experiment. A further inlet 3 may be provided for introducing,
e.g. gravity feeding or injecting, a sample or further sample to be tested (e.g.
for suction by the syringe 2 out of a vial for storing in the storage loop 5).
One of the syringe 2 and the inlet 3 may be used to introduce a standard particle/liquid
mixture for comparison or calibration. A second syringe may be attached to inlet
3. Further, a supply 6 of carrier liquid is provided. Optionally, a sample valve
4 is provided which is a convenient way of connecting a measurement chamber 7 such
as a capillary column to a variety of inputs, e.g. the syringe 2, the storage loop
5, the inlet 3 or the carrier liquid supply 6.
The transport and separation processes in the chamber 7 can be highly
dependent on the viscosity and density of the carrier liquid. Carrier liquids suitable
for cell transport and having a wide range of densities and viscosities can be
obtained by dissolving small amounts of povidone (polyvinylpyrrolidone, Sigma Pharm.),
methylcellulose or percoll (Amersham Bioscience, Gibco, Invitrogen) in isotonic
saline (0.9 % NaCl in H2O). Isotonic carrier liquids are preferred for
non-destructive measurements of cell transport.
Neither the viscosity of the carrier nor the length or bore of a capillary
column create any problems with respect to the control of carrier flow. Commercial
HPLC pumps (high pressure liquid chromatography) are suitable for pumping the carrier
Temperature effects on the viscosity of carrier liquids may have a
marked impact on transport retardation and adhesion and is preferably controlled.
An increased cell or particle retardation may be associated with increasing temperature.
Chamber 7 is hollow and may be generally elongate, for example a capillary
tube, and enclose a volume (rather than being an open tray), e.g. it can be cylindrical.
The elongate chamber should have a length sufficient to retard the particles or
cells so that different elluent peaks can be distinguished in time as is discussed
below with reference to Figs. 4b and 5.
A suitable cell or particle detector such as an electronic gate detector
8 is located at the outlet of the measurement chamber 7 and monitors the elution
of particles from the chamber 7 by measuring the passage of individual cells or
particles from the chamber 7. Two basic methods are known to count particles such
as cells - aperture- impedance counters and optical counters. The first type of
counter includes devices such as those supplied by Coulter, Hialeah, FL, Baxter
Diagnostics, Waukegan, IL (Sysmex) and Abbott Diagnostics, Santa Clara, CA (Cell-Dyn).
Cells are counted by measuring the change in electrical impedance as a cell passes
through an orifice. As the cells are in the orifice there is a momentary drop in
conductance and this is registered as an electrical pulse. The second type of counter
depends on light scattering properties of the particles. Examples of such devices
are supplied by Bayer Diagnostics, Kent, WA (the Technicon series) and Abbott Diagnostics,
Santa Clara, CA (Cell-Dyn). In these devices the particles pass through a narrowly
focussed beam of light, usually a laser light. The particles interrupt or alter
the light beam thereby generating an electrical pulse from suitably arranged photovoltaic
sensors. Any such detectors as mentioned above may find advantageous use with the
Generally, the detector 8 transposes the passage of individual cells
or particles into recognisable signals, e.g. electrical pulsed signals. The output
of the detector 8 may be fed to an amplifier and base line restorer 9. The output
of 9 may be supplied to a counter and pulse height analyser 10 for measuring the
pulse rate and height. The pulse rate is a measure of the cell or particle elution
rate. It is preferred if the detector 8 does not affect the flow pattern in the
column 7 significantly. AC-gate detectors of axial symmetry and operating at working
frequencies in the (low frequency) LF-domain as described in the article by De
Bisschop FR, Vandewege J, Li W, De Mets M., "Low frequency electronic gate technique
for the counting and sizing of cells, bacteria and colloidal particles in liquids",
Proc. 18th IEEE IMTC Conf Budapest 2001:1309-13, may be found suitable for use
as detector 8. However, other cell or particle detectors may be used. For example,
optical detectors can be used where appropriate for the particular particles or
For example, an AC-gating detector may be implemented as follows.
A diluted sample, consisting of a conductive carrier liquid and cells or particles
to be measured, is introduced into a device containing a gate, and flows towards
the gate. Electrodes are provided on either side of the gate. Preferably, these
electrodes are platinum electrodes. These electrodes are electrically interconnected
to either side of a low frequency AC signal generator (best shown in Fig. 3), having
a frequency below 1 MHz, whereby DC is excluded. The frequency is preferably in
the range 100 kHz to 1MHz, more preferably in the range 100 to 600 kHz for measurements
on cells, most preferably in the range 200 to 600 kHz. In the following the words
"low frequency" represent a voltage in any of these ranges. This current source
sends current through the gate via the conductive carrier liquid. Two signal sensing
electrodes are applied, before and after the gate (with before and after seen in
function of the direction of flow of the carrier liquid), which measure the current
going through the gate. One of these signal sensing electrodes is grounded, the
other one is connected with a bridge, the signal sensing electrodes forming the
nodes A and D of the bridge respectively (see Fig. 2). In order not to have any
current leaking to the earth through the carrier liquid, and thus in order to eliminate
external influences, guard electrodes may be introduced. Both guard electrodes
are kept on the same potential as the second signal sensing electrode by means
of a voltage follower. Signal recovery in ac-operated gating may be based on phase-sensitive
AM-demodulation as shown schematically in Fig. 2, the bridge actuating signal serving
as a reference. The kind of modulation however that takes place in the bridge does
not exactly respond to a multiplication of signals required for proper AM-modulation.
The dynamic range of the modulation/demodulation, suffering from such limitations,
seldom exceeds 28 dB.
The circuit of Fig. 2 may be described as an active cell with a gate
Cg) and a device matched in impedance to the active
cell (Cv and Rv), the AC signal (E0(t)) from the
signal generator being applied to both the cell and the matched device in parallel
to generate a non-resonant electrical signal at the output of the active cell and
the matched impedance device. The detector of Fig. 2 comprises an AC bridge which
measures each particle transition through the gate as a variation of resistive
and reactive impedance of the gate. Voltage division, by the bridge, converts such
variations into amplitude- and phase modulation of the LF carrier signal applied.
The gate forming part of the AC-bridge between nodes A and D, appears to function
as a capacitor bypassed signal source, with equivalent source resistance Re
= (Rg.R2)(Rg + R2) and equivalent load capacitance Ce
= Cg + Cw + Ci, where Cg represents
the intrinsic gate capacitance, Cw represents the wiring capacitance
and CI represents the preamplifier input capacitance.
The elements (Cv and Rv) of the bridge are adjusted
such that, when only carrier liquid is passing through the gate, the bridge is
in equilibrium, and no current is going from node B to node D. More particularly,
R1, R2, and Rv, are preferably equal to the gate
resistance Rg, in order to have optimum sensitivity. For example Rv
might be a 10 kΩ potentiometer and Cv a 25 pF variable capacitor,
in order to adjust the bridge to equilibrium. Cg is the gate capacitance.
When a cell or particle passes through the gate, the bridge is out
of equilibrium, and a voltage ΔE(t) is measured between node B (voltage EL(t))
and node D (voltage Er(t)). When the bridge is balanced, the signals
at nodes B and D are equal. Both signals are led to a differential amplifier (AD847).
With EL(t) and Er(t) being in phase, the bridge trimmed to
equilibrium operates with almost disappearance of the LF carrier signal. Demodulation,
however, is preferably not done at this stage since a true difference signal, at
the occurring frequencies, is not readily available. An embodiment of the present
invention includes the use of a differential pre-amplifier comprising signal followers
(e.g. operational amplifiers) to drive a broadband torroidal transformer T so that
the difference signal only is passed to a demodulator (multiplier demodulator).
For example two operational amplifiers AD847 supplied by Analog Devices may be
used. The transformer T acts as a common mode rejection device. The output signal
from each operational amplifier is fed back to its inverting input. The voltage
gain is unitary but the power and current gain is appreciable which is necessary
to drive the torroidal transformer T. The followers have a high input impedance
which reduces the load in the bridge.
Diodes d1 and d2 (such as IN914 or IN4148) protect
the demodulator input in the common cases of clogging. Demodulation in the demodulator
may be performed, preferably, by synchronous multiplication with the bridge input
signal Eo(t) supplied from node C as a reference, followed by band-pass
filtering and final amplification.
The dynamic range problem of the circuit of Fig. 2 can be eliminated
by the use of a logarithmic demodulator amplifier specifically developed for the
handling of pulsed RF signals. Demodulators of this type detect the envelope of
burst RF signals which is of particular interest for the present purposes. The
dynamic range of log demodulators such as the AD606, AD640 or AD8307, in the order
of 60 dB or better, in practice remains limited by the noise level of the gate.
The latter responds to a 1/f frequency dependence. The effect of cell "transparency"
at working frequencies much higher than 800 kHz degrades the detection sensitivity.
Optimal working frequencies for such reasons remain situated in the LF-domain for
cells, e.g. from 200 to 600 kHz. As in the detector of Fig. 2, the gate (GATE)
is incorporated in an AC bridge in the circuit of Fig. 3. The detector detects
particle transition through the gate as a variation of the gate's resistive and
reactive impedance. The elements (R3 C1) of the bridge are
adjusted such that, when only carrier liquid is passing through the gate, the bridge
is in equilibrium.
Both signals from the bridge are fed to a differential preamplifier,
e.g. comprising operational amplifiers AD827 supplied by Analog Devices. The differential
preamplifier drives a broadband torroidal transformer TR1 so that only the difference
signal is passed to a logarithmic demodulator (AD8307). The transformer TR1 acts
as a common mode rejection device. The output signal from each operational amplifier
is fed back to its inverting input, for example over a resistance R5, resp. R6
of e.g. 1 kΩ.
The output signal of the torroidal transformer TR1 is demodulated
by the logarithmic demodulator such as an AD8307 supplied by Analog Devices. The
use of such a logarithmic demodulator shows improved functional stability, simplicity
of operation and reduced costs over the use of a phase sensitive demodulator as
described in the first embodiment of the present invention.
Logarithmic demodulation, which is phase-insensitive, needs neither
a reference signal (from the RF signal generator) nor compensation for the phase-lag
introduced by the differential preamplifier, so that the circuit is greatly simplified.
The circuit, furthermore, demands no tuning. Logarithmic signal detection, contrary
to phase-sensitive demodulation, is essentially broad-band. The frequency response
of the preamplifier and demodulator is substantially flat within the range from
200 kHz to 20 MHz. Logarithmic detection, as such, offers access to the range of
working frequencies. This demodulator allows determination of the frequency dependence
of cell impedance, within that range. This is a point of actual interest, specifically
in the field of biotechnology. It also allows optimum setting of the frequency
dependent upon the type of cells or particles used.
Coupling of the differential preamplifier to the logarithmic demodulator
requires matching of input/output impedances and blocking of dc offset voltages
that might be present in the preamplifiers' output signal. This is carried out
by capacitors C3, C4. The capacitance of C3, C4
should be high enough to avoid oscillation. The output of the demodulator is fed
to a low-pass filter (LTC1560) required for the elimination of any residual ripple
that might be present on the demodulated gate signal. The output from the filter
is supplied to a main amplifier and from there, for example, to a multichannel
analyser and desk top computer.
The equilibration of the measuring bridge is greatly simplified by
direct reading of the magnitude of the difference signal ΔE(t) across the
bridge. For example a phase comparator (see Fig. 3) may be used to bring the measurement
signals across the bridge in phase, e.g. while adjusting the capacitance C1.
The phase difference is visualised by a meter VM1. The amplitude of the bridge
output signals (after proper phasing) is equalised by potentiometer P1. This is
visualised by the meter AM1 which takes on an extreme value when equality is reached.
The use of panel meters such as AM1 and VM1 is more preferable than using an oscilloscope.
The panel meters also indicate any clogging of the gate, even partial, or a changing
conductivity of the carrier liquid as a consequence e.g. of a changing room temperature,
because any of these events will put the bridge out of equilibrium.
The circuits of Fig. 2 and Fig. 3 are described in detail in European
Patent application EP 1162449 and International Patent Application, WO01/94914,
respectively, both of which are incorporated herewith in by reference in their
entirety. In particular it will be appreciated that the circuits of Fig. 2 and
3 provide a detection method for detecting objects in a liquid such as colloidal
particles, cells, or the like, comprising the following steps: applying an AC signal
in parallel to an active cell with a gate and a device matched in impedance to
the active cell, this AC signal generating a non-resonant electrical signal at
an output of the active cell as well as at an output of the matched impedance device,
sending at least one object to be measured through the gate of the active cell,
this causing an electrical measurement pulse across the output of the active cell,
modulating the non-resonant AC electrical signal at the output of the active cell,
with the electrical measurement pulse, feeding the outputs from the cell and the
matched impedance device to an AC common mode rejection device, thus generating
a difference signal modulated with the non-resonant AC signal, demodulating the
difference signal from the common mode rejection device, and filtering the demodulated
signal to remove components of the AC signal and its harmonics to retrieve a signal
relating to the volume of an object passing through the gate of the active cell.
The signal generator may provide a signal with a frequency in the range 100 kHz
to 1MHz, more particularly in the range 100 to 600 kHz. The common mode rejection
device may be a Wheatstone bridge. Also a detection device for detecting an object
in a liquid such as colloidal particles or cells is provided, comprising: an AC
signal generator for generating an AC signal; an active cell with a gate and a
device matched in impedance to the active cell, the AC signal from the signal generator
being applied to both cells in parallel to generate a non-resonant electrical signal
at the output of the active cell and the matched impedance device; a modulator
for modulating the signal applied to the active cell with a measurement pulse generated
when an object passes through the gate of the active cell, an AC common mode rejection
device, to which the outputs from the active cell and the matched impedance device
are fed, generating a difference signal; a demodulator for demodulating the difference
signal; and a filter for filtering the demodulated signal to remove components
of the AC signal and its harmonics to retrieve a signal relating to the volume
of an object passing through the gate of the active cell. The AC signal generator
may generate a frequency between 100 kHz and 1 MHz, more preferably in the range
100 to 600 kHz, especially when detecting cells. The matched impedance device may
comprise resistors and/or capacitors. The AC common mode rejection device may be
a Wheatstone bridge. The common mode rejection device may be a differential amplifier.
The demodulator is preferably a logarithmic demodulator.
Log demodulators, full-wave rectifiers, do not suffer from feed-through
of the carrier signal. Feed-through is problematic in phase-sensitive demodulation
due to the proximity of the carrier frequency to the signal pass-band. Log. Demodulation
has the advantage that it does not require very specific filters. Any residual
RF ripple present in the demodulated signal is at the double of the carrier frequency
or higher. The increased frequency distance between noise and the wanted signal
makes filtering less critical.
Log demodulation/amplification also increases the counting accuracy
which is limited by so-called dead time in A/D conversion, e.g. in the counter
and pulse-height analyser 10, specifically at high elution rates. Dead time is
markedly reduced by the logarithmic scaling of signals, because of the reduction
in converted address numbers to which dead-time is proportional. Cell/particle
sizing is done on the basis of pulse-height measurement, e.g. by a pulse height
analyser - see Fig. 2 and 3. The size axis however, linear in the logarithm of
the cell/particle sizes, is related to a much larger range of values.
For adequate functioning of such detectors, carrier liquids of a well
determined and stable electrical conductivity are preferred. The gate resistance
should be kept as low as possible for reasons of signal-to-noise ratio and detection
sensitivity. The resistivity of carrier liquids and diluents appears very susceptible
to PH. The use of PH
buffering is advised therefore, but not
strictly necessary. The buffering capacity doesn't need to be high.
If conductive liquids cannot be used, other types of detectors can
be used. Alternative particle or cell detection methods include radioactive labelling
of the cells or particles and then detection of the radiation, the use of fluorescing
particle counters, the use of colorimetric detectors. All these methods are suited
for the monitoring of cell or particle elution. However, the detection sensitivity
of the above detection methods is lower, commonly, than that of electronic gating.
An exception is a photo-multiplier based system applied in cell sorters, the sensitivity
of which can be almost the same as that of electronic gating, see for example,
"The art of Electronics, Horowitz and Hill, pages 599-600, Cambridge Press, London,
Most of the above detection systems are of the integrating type, that
is to say: the detectors do not signal the passage of single cells or particles.
The relationship furthermore between the cell or particle size and the amplitude
of the detection signal is not well defined. The above techniques may be used therefore
in counting experiments. The present invention is not limited to integrating detectors
or those capable of detecting individual particles.
When the apparatus is to be used for cell sorting, a means should
be provided (not shown) for selecting a specific fraction of the cells in dependence
upon their retardation.
When a charge of particles or cells is injected into the chamber 7
along with a carrier liquid, the cell or particle elution appears retarded with
respect to that calculated on the basis of physical models for the transport of
solutes. This retardation is not dependent upon the adhesion of the cells or particles
to the chamber wall and marks the absence of transport near the capillary axis,
a phenomenon known as the Segré-Silberberg effect. In Segré-Silberberg flow transport
takes place predominantly at a lateral distance about halfway between the axis
and the capillary wall - see Segré G, Silberberg A., "Behaviour of macroscopic
rigid spheres in Poiseuille flow" (Parts 1 and 2), J. of Fluid Mech 1962; 14:115-35
and 136-56; Feng J, HU HH, Joseph DD, "Direct simulation of initial value problems
for the motion of solid bodies in Newtonian fluid, Part 2, Couette and Poiseuille
flows.", J. Fluid Mech 1994; 227:271-301; and Schönberg JA, Hinch EJ., "Inertial
migration of a sphere in Poiseuille flow", J. Fluid Mech 1989; 203:517-24.
The occurrence of Segré-Silberberg transport explains why elution
diagrams show a sharp and delayed leading flank as shown in Fig. 4a, and why tailing
is absent, at least at relatively high flow rates of the carrier liquid. The phenomenon,
plainly developed in vertical flow tubes, appears disturbed in horizontal ones
by effects of settling and adhesion. On decreasing the carrier flow speed elution
peaks become more and more distorted and eventually split up into two (or more)
as shown in Fig. 4b.
The doublet of Fig. 4b may be attributed (without being limited to
theory) to sample fractions that behave in different ways: a first one that moves
at a radial distance about halfway the axis and the wall (that is with Segré-Silberberg
transport) and a second one that moves close to the wall, prone to (and delayed
by) adhesion. Fig. 4b may be referred to as a "chromatogram". In accordance with
an aspect of the present invention, a flow rate is used in the measurement chamber
which is lower than the rate for Segré-Silberberg transport of at least one set
of particles in the liquid/particle mixture in that chamber. However, the flow
rate should be higher than that required to allow gravitational separation of the
same set of particles, e.g. higher than the rate for partial sedimentation and
rolling of the particles along the solid/liquid interface.
It has been found that the retardation of the second fraction shown
in Fig. 4b effectively depends on cell or particle sizes and densities and, most
importantly, on the chemical nature of the coatings on the wall of the chamber.
The present invention makes use of this retardation of the second peak to sort
cells or particles or to determine a characteristic relating to the interaction
between the cells or particles and a solid/liquid interface within the chamber
Analysis of results or optimisation of practicable particle sizes,
viscosities/densities of carrier liquids and Reynolds numbers, may be based on
an approximate transport model that describes the motion of cells or particles
on column, in translation near the wall. This model is presented for capillary
tubes for simplicity's sake.
According to Cherukat and McLaughlin the magnitude of lift in linear
shear near the wall corresponds to (see Fig. 9)
FL = aµVReI
where a : the cell/particle radius
The quantity I, according to Cherukat and McLaughlin and to Cox (see "The lateral
migration of solid particles in a laminar flow near a plane", R.G. Cox and S.K.
Hsu, Intl. J. Multiphase Flow 3, 201-222 (1977)), should be close to 1.8, at least
in circumstances where cell or particle slip is absent. This obviously is no longer
the case when cell or particle motion is retarded due to adhesion and the model
has to be amended. The quantity I in such cases increases markedly. The slip (velocity)
Us of the cell/particle with respect to the surrounding liquid, is defined
Us = Vcell - GL
where L stands for the distance of the cell/particle to the wall, and G for the
velocity gradient. The latter, in the present case of Poiseuille flow can be represented
G = 2νor / (r02)
where vo: axial speed, ro: capillary radius, r : distance
to the axis. Near the wall, i.e. at a distance r = ro - a from the axis,
the gradient G is seen to take on an almost constant value, close to 2vo/ro.
The characteristic velocity V of the carrier, in the present case of linear shear
near the wall corresponds to
V = G a
The Reynolds number Re is defined here as
Re = Ga2 / (v)
where v : cinematic viscosity, i.e. v = µρL,
ρL standing for the density of the carrier liquid.
- µ: dynamic viscosity
- V: a characteristic velocity of the carrier liquid
- Re: Reynolds number
- I: dimensionless number, numerical outcome of an integral
As a result, the magnitude of lift in the absence of slip is formulated
FL = 4a4v04 / (r02)ρLI
Lift, acting in a radial direction, is seen to add-up with the gravitational force
above the axis while countering it below, at least for cells or particles that
settle in the carrier at rest. Conditions of cancelling lift, settling and Archimedes
forces are expected to occur when:
4a4v04 / (r02)ρLI
≈ 4 / (3)πa3(ρP - ρL)
where ρp represents the particle/cell density.
It should be noticed that eq. 7 is to be considered as an approximation,
since Archimedes' law doesn't strictly hold in non-static fluids. Nevertheless,
the r.h.s of eq. 7 is found to produce the right order of magnitude, in most cases.
Taking into consideration the fact that a « ro in many cases, it is
seen that eq. 7 can be simplified and rearranged to read
a ≈ 10,27 r02 / (v02)
(ρP / (ρL)-1)
v0 = [10,27 r02 / (aI)
(ρP / (ρL)-1)]1/2
At any flow speed vo of the carrier, only cells or particles of a well
determined density/size ratio will satisfy eq. 9 and translate near the capillary
wall in the absence of resultant forces.
When adhesion comes into play, cell or particle translation and rotation
slow down, giving rise to negative slip velocities with respect to the surrounding
carrier liquid. The Cherukat-McLaughlin model in such circumstances predicts increasing
lift. Viscous drag is animated also. Cells or particles in such conditions presumably
can be involved in repeated adhesion-and-release processes. The times needed for
the build-up of lift and drag, sufficient for release, are expected to be in proportion
with the strength of the adhesive interaction. More dense cells or particles will
settle and contact the wall, unless the capillary radius ro is decreased
or a carrier liquid of higher density is adopted, so that the ratio ρp/ρL
is kept constant. Less dense cells/particles that do not satisfy eq. 9 will be
pulled away from the wall to end up in regions of much higher flow speed. On decreasing
the carrier flow speed, adhesion comes into play so that the cell/particle translation
and rotation slow down. Slip obviously isn't zero any longer and magnitude of lift
increases - see Fig. 10.
Increasing lift and drag very well may cause the release of adhering
cells or particles. The latter can be involved, as such, in repeated adhesion-and-release
processes or in a hopping-like motion. Cells or particles that almost exactly meet
equations 8 and 9 are expected to be involved in such processes as a consequence
of the short-range adhesion force, superimposed on settling force.
The validity of equations. 8 and 9 has been confirmed by experiments
with latex particles, having a known particle size (10 µm) and density (1038 kg/m3).
The capillaries were of 530 µm and 220 µm bore. Physiologic saline (0.9 % NaCl
in water) was used as carrier liquid. Its density at 24 °C was 1024 kg/m3. The
carrier speeds, in full agreement with eq. 9, where close to 3 cm/s.
For the sake of completeness, it should be mentioned that rolling
of particles (over the capillary wall) was never observed as a representative type
of motion: the elution would be much more retarded in that case, than experimentally
Cell or particle and flow parameters can be optimised, amended or
changed to meet the equations 7, 8 and 9, and to obtain the balance of forces that
is at the basis of the present transport model.
It should be mentioned that apparent noise in elution diagrams like
that shown in Fig. 4 is not necessarily of electrical origin: it can reflect statistical
variations in the count rate due to the limited number of cells or particles present
in the sample.
An application of the present technique is the relative measurement
of cell or particle adhesion. The proposed technique provides accurate and highly
reproducible information about the extent of adhesion in real-world situations,
so as encountered in medical devices (artificial organs, catheters, membranes).
The present technique aside elution diagrams also allows cell or particle size
distributions to be determined, with high accuracy. Cell or particle size distributions
of column effluents measured at regular time intervals confirm the separation of
sample fractions, on the column, on the basis of cell or particle sizes and densities,
as predicted by the transport model.
One embodiment of the present invention relates to quantitative evaluation
of cell or particle adhesion upon convective transport in a chamber such as a capillary
tube. The objective is to measure the transport retardation due to adhesion, and
its relation to the chemical nature of coatings in the chamber, e.g. of capillary
wall coatings. Cells or particles in suspension in a carrier liquid (e.g. physiological
saline solution) are passed through the chamber 7 such as a capillary several meters
in length. On the inside of the chamber 7 an internal coating is applied whose
adhesive properties are to be measured. Such a coating may be made of a chemical
compound having methyl, methyl-phenyl or glycol groups, for example. For instance,
the superficial -OH groups of a fused silica capillary column are modified to form
active groups, e.g. methyl-, phenyl, or glycol- groups. The layer thickness is
negligible with respect to the capillary bore. Other coatings may be obtained by
any suitable deposition technique, e.g. evaporation from solution, vapour deposition
from a gas. The thickness of such coatings may not be negligible and should be
taken into account, e.g. when calculating flow speed (see eq. 9) and when comparing
different materials. The adherence of these coatings is expected to be differ from
each other. A co-axial gate detector 8 inserted at the chamber outlet monitors
the elution and translates the passage of individual cells or particles into electrical
pulsed signals. The pulse rate measured reflects the elution rate. Calibration
tests for reproducibility reasons may be done with synthetic particles such as
latex particles, having a known and uniform density and size distribution. Gates
used in these experiments for reasons of detection sensitivity may have an aperture
of about 50 µm, for example. If a capillary tube is used for the measurement chamber
7, this can have a bore of 530 µm, similar to a commercial type used in chromatography.
Such a combination allows a perfect match of gate transition times and capillary
flow rates. The inside of the capillary is coated with the material to be tested.
Gate transition times in the order of 10 to 20 µs ensure optimal signal-to-noise
ratios of the detection system. Flow rates in the range of 1 to 7 mm3/sec
allow the transition to be observed from fully developed Segré-Silberberg transport,
where adhesion effects are absent, to transport regimes in which adhesion induced
retardation is present. The flow path inside the sampling valve for the present
purposes should preferably be metal free.
Of importance for the applicability of the technique is the rate of
establishment of Segré-Silberberg transport as soon as possible after injection
of the sample. The sample is presumed to move as a plug, at that moment of injection.
Samples in the present experiments of horizontal flow usually contain thousands
of cells or particles, or even more. The rate of transition from a distribution
like that to Segré-Silberberg transport remains largely unknown. Relevant information
can be obtained empirically by measuring transition times of capillaries cut to
decreasing lengths, and at various flow speeds of the carrier liquid as shown schematically
in Fig. 5. To produce this curve a capillary of a certain length was taken, e.g.
2,5 m and the times required for transition of that length at each of the different
flow rates indicated was determined. Next, the capillary was shortened by 0.5 m
and the measurements at each flow rate repeated. This procedure was continued until
a capillary length of about 30 cm was reached where the measurements became irreproducible.
Tables of travelled distances (by elution peaks) and the associated (transition)
times were produced from this data. By transition time, is meant the time the leading
flank of an elution peak takes to travel a capillary of given length. That leading
flank is known to travel with the characteristic speed of Segré-Silberberg transport.
The slope of the plot that shows the travelled distance (capillary length) versus
time reflects that speed, at any flow rate of the carrier liquid. The Y axis is
the distance travelled by the elution (= capillary length) and the X axis is time.
The slope of these curves, a measure of the speed of Segré-Silberberg
transport, remains constant down to capillary lengths of about 30 cm, irrespective
of the chemical nature of the capillary wall coatings. Experiments with capillaries
shorter than that show transient flow regimes and some dependency on sample storage
times (see fig. 1: storage column). The constancy of these speeds allows the position
of the leading flank of the first peak in elution diagrams (see fig. 4) to be considered
as a reference in measurements of the retardation of the second peak, as explained
The apparent noise in elution diagrams is caused by statistical variations
of the count rate. The interpretation of elution diagrams is greatly facilitated
therefore by considering the cumulative elution (Figs. 6-8), rather than the differential
one (see Fig. 4). Steep flanks of cumulative diagrams followed by a sharp knee
indicate fast elution, and absence of adhesion. A broad knee region, on the contrary,
reflects retarded elution as an effect of adhesion. This becomes more evident on
comparing elution diagrams obtained at different flow speeds of the carrier liquid.
As shown in Figs. 6 and 7, the knee is broader at the slower flow rates.
It should be noticed that elution at relatively high flow speeds of
the carrier (Fig. 7, volume flow rates of 6.8 mm3/sec.) does not show
any retardation: the diagrams coincide, irrespective the chemical nature of the
wall coatings. It is assumed that the reason is the absence of wall contacts. The
Segré-Silberberg effect is fully developed and transport takes place at a radial
distance about 45 % from the capillary axis. This is no longer the case at lower
carrier speeds (Fig. 7, volume flow rates of 3.6 mm3/sec.). Segré-Silberberg
transport is disturbed by gravitation. Wall contacts are enabled. The degree of
disturbance actually depends on particle or cell sizes and densities, which should
be almost uniform in the present case. The retardation effectuated by the methyl-phenyl
coating (Fig. 7) is clearly greater than by the glycol coating. In the elution
curves of a methyl-coated capillary adhesion is almost absent as can be concluded
on the basis of the initial steepness of the curves and the sharp knee. The diagrams
for methyl may serve as a reference in comparative tests with other wall coatings.
The results indicate that the area enclosed between the reference
curve (e.g. methyl-coating) and the ones obtained with other coatings may serve
as a quantitative or qualitative measure of transport retardation and of adhesive
interaction. The determination of the area enclosed between the reference curve
and any other is greatly facilitated by the use of the multi-channel analyser 10
that allows storage and simultaneous display of multi-scaling data. The wanted
area is readily obtained by a subtraction of integrals, operation done with digital
The reproducibility of these experiments, measured by the relative
variation of the enclosed areas obtained upon repeated tests, is of the order of
2 % or better. An important parameter in that context appears to be the integral
number of cells or particles involved in the experiments. The transport retardation
markedly increases with decreasing cell or particle numbers as shown in Fig. 8.
As an explanation, it should be noticed that the cross sectional area of the capillary
tube where transport actively takes place remains very low. As a consequence of
the Segré-Silberberg effect only a small proportion of the cross-section is involved
in transport. Cell or particle number densities can be quite high. Mutual hindrance
is most probably involved in the observed effect: cells or particles translating
at a certain distance from the wall might push partners closer to the wall so as
to decrease their retardation. Another cause might be a saturation of the accessible
surface. It should be mentioned in this context that small sample fractions have
been found involved in what may be called irreversible adhesion to the wall, an
effect that causes progressively decreasing adhesion upon repeated experiments.
Results like those shown in Fig. 8 indicate the importance of standardisation of
the number of cells or particles used in each test. In particular the results of
Fig. 8 may be interpreted as a chamber 7 having an "overload value", and that for
repeatable tests the flow rates and particle concentrations should preferably be
below the overload value. Suitable numbers in the case of 530 µm capillaries, may
be in the order of 5000. On the other hand much lower numbers may degrade the reproducibility.
Tests with several different coatings other than the above ones have
been carried out. Among these, polyurethanes, intended for use in artificial organs
have been tested. The polymer, introduced in a dissolved state, is deposited at
the inner surface of the capillary as a layer, about 2 µm thick, by evaporation
of the solvent. Measurements with latex particles as a test material show a retardation
that is almost the same as with methyl-phenyl coatings. Experiments have been performed
with silica capillaries of 220 µm bore, methyl coated. The objective is to investigate
the transport process in artificial kidneys. The capillaries are of the same bore
as that of the semi-permeable membranes. Results show the occurrence of Segré-Silberberg
transport, similar to the one in larger bore capillaries.
The lower limit of particle sizes in experiments that respond to the
transport mechanism described is about 0,9 µm. Much smaller cells or particles
appear increasingly subject to diffusion. The result is a less marked Segré-Silberberg
transport. This size limit also corresponds to the detection limit of electronic
gating, with gates of 30 µm aperture. Experiments with gates like that require
precautions regarding working conditions, e.g. to prevent or monitor clogging of
the gate caused by dust. Gates of smaller aperture are not commercially available,
however this is not a limit on the present invention.
The largest commercial electronic gates are of 300 µm aperture, although
this is not a limit on the present invention. Experiments with such gates and particles
of tens of micrometer in diameter are suitable. The signal-to-noise ratio is excellent,
due to low gate impedance. The detection limit (minimal diameter of spherical particles
detected) as a rule remains about 3 % of the gate aperture.
It should be noticed that the adhesion/transport behaviour of different
cells or particles can be investigated as well, in capillaries of given bore and
wall coating. Various carrier liquids can be applied also.
The chamber 7 may be constructed from one or more tubes or capillaries
or may comprise a column packed with coated material, e.g. beads. However, a capillary
chamber is preferred as it is easier to analyse the results theoretically. Capillaries
of the length 0,2 m up to 10 m or more can be successful. The upper limit to tube
length is probably only fixed by pressure demands imposed to the pump system that
must control the carrier flow.
With tubes shorter than 30 cm, poor reproducibility of elution diagrams
may occur. This may find its origin in the lack of a transition from a plug-like
distribution of sample materials, immediately after injection, to the one associated
with Segre'-Silberberg transport, i.e. the transport on a preferential pathway
half-way the axis and the capillary wall.
The injection of samples is preferably done by the use of a sampling
valve 4 with storage column 5. With measurement chamber lengths in the order of
30 cm, the dimensions of the storage column 5 and settling that takes place inside
mean that samples don't really behave as a plug. Storage times become important
and are preferably kept as short as possible.
Fused silica columns with internal diameters of 100 µm, 115 µm, 220
µm... 530 µm and with various (internal) coatings are commercially available from
SGE Inc.; Varian, Altech. Polymer coatings (e.g. polyurethanes) are obtained by
evaporation deposition from solutes.
The skilled person will appreciate that the methods of the present
invention offer qualitative and quantitative evaluation of cell or particle adhesion
at solid-liquid interfaces. The technique as a measurement method is sensitive,
in the way that it reveals the slightest difference in adherence of the particles
to the capillary wall coatings applied. The reproducibility is high, comparable
to that of liquid chromatography. The technique, in addition to its application
in adhesion measurements, also offers a cell or particle separation process. Discriminating
parameters in that case may be at least one of size, density and adherence. Any
applications that require a characterization/separation of cells or particles on
the basis of cell/particle density, size or adherence are also included within
the scope of the present invention. For example, the present invention may be used
to evaluate the adherence of biomaterials as an alternative to the methods proposed
in ISO 10993-4 norm (addendum 2). Further, any separation of cell populations into
fractions of depending on at least one of different cell adherence, cell size and
density is included within the scope of the present invention, e.g. in the fields
of haematology, histology, cancerology, artificial organs or similar. For example,
adhesion in 200 µm bore capillaries has been investigated, successfully. Capillaries
of that bore are representative for the hollow fibres used in renal dialysers (artificial
kidneys). Adhesion, related to the chemical nature of these fibres, is a point
of actual interest. The present invention may also be used in the quality control
of industrial processes, e.g. in the detection and discrimination of mastitis cells
from somatic cells (not related to the occurrence of mastitis) in raw milk, greatly
facilitated by the logarithmic scaling of cell size distributions. The present
invention may also be used in comparative tests on adherence of dyes to various
substrates. For example, the present invention may be used to investigate the colloidal
stability of suspensions/emulsions in contact with solids. The following is a non-exhaustive
list of materials which can be tested:
- biopolymers (poly-urethanes);
- membranes used in filtration processes;
- semi-permeable (hollow) fibres used in renal dialysers; milk and its derivatives;
- liquid foods for parenteral administration;
- pharmaceuticals (in emulsion/suspension).
While the invention has been shown and described with reference to
preferred embodiments, it will be understood by those skilled in the art that various
changes or modifications in form and detail may be made without departing from
the scope and spirit of this invention. For example, the chamber 7 may be subject
to electric or magnetic fields to modify the flow and adhesion of the particles.
Similarly, the particles carry electric charge and the solid/liquid interface may
be adapted to attract or repel these particles. For example, certain unwanted particles
may be severely retarded by such means so that they do not contribute to the relevant
part of the ellute.