FTÅ200 Measurement Capabilities

by Roger P. Woodward, Ph.D.

First Ten Ångstroms, 465 Dinwiddie Street, Portsmouth, VA 23704
Tel: 757.393.1584 Toll-free: 800.949.4110 Fax: 757.393.3708 http://www.firsttenangstroms.com

The FTÅ200 is an instrument for measuring phenom- memory. A single image may be captured (a snap-
ena such as surface tension, surface energy, and shot) or a sequence may be captured (a movie).
absorption. It is built around rapid video capture of Normally, the FTÅ200 captures images and analyzes
images and automatic image analysis. them after they have been stored. However, the “real-
time” mode analyzes images on-the-fly. This option is
Measurement Processes somewhat less flexible because the images are not
preserved, although the results are stored in a disk
Measurements are made by observing the drop shape
file. The image analysis algorithms employed by the
of a fluid which reveals information about the fluid
software seldom take longer than one second per
itself or about the surrounding media. In some experi-
image, so images can be obtained roughly once per
ments, the fluid will be the known quantity and the
second in real time mode.
media the unknown being explored, and in other cases
these roles will be reversed. Figure 1 shows a typical situation in which contact
angle is being determined for a fluid on an unknown
Dispensing. In general, drop volume is quite small, in
substrate. Air surrounds the fluid and substrate.
the range 1–15µl. Sometimes small volumes are only
a matter of convenience, but in other experiments Image Analysis. The FTÅ200 follows one strategy for
small volumes permit better spatial resolution and all image processing: a set of analytic curves are
sometimes they are used to avoid distortions due to formed which describe the drop, then these expres-
gravity. The FTÅ200 uses a highly accurate syringe sions are solved for the desired data. This is the heart
pump driven by a stepper motor to dispense the test of drop-shape analysis implemented by the FTA200
drops. The pump may also be run in reverse to software: reduce a gray-scale image to a set of
aspirate, or pick up, a drop. It may be used with a equations describing the drop’s edges.
variety of commercially available syringes (glass or
Figure 2 illustrates the edge finder algorithm that
plastic) and dispensing needles (stainless steel or
accurately locates an edge from a gray-scale image. A
Teflon-coated).
large number of X,Y points are generated in this
Image Acquisition. Once there is a drop to observe, manner and then a least-squares fit is used to derive
images are captured by the computer and stored in the curve equation. An advantage of the least-squares

T
Tangent “T” to drop’s profile at point “P”

Θ = Contact

,,,,,
angle Drop profile
B
P

Baseline “B” tangent Fluid drop

to specimen surface P

Specimen
Figure 1. Contact angle measuresment.
 we may solve them in various fashions. The software
does all of the following automatically.
For contact angle, it obtains the intersections of the
Y baseline with the drop profile by solving the equations
simultaneously. At each intersection point it obtains
the slopes by differentiating the equations. Next the
arctangents of the slopes are taken to obtain angles.
X The difference between the baseline angle and the
Plot Z (gray level) for the
middle scan line in the
drop profile angle is the contact angle.
magnified pixel For other analyses, the software can solve the equa-
image above.
Z tions for distance and area, or it might solve the
Laplace-Young equation for surface tension. All solu-
tions are obtained automatically by the software.

X The FTÅ200 software also offers unique hybrid image

“Center of mass” analysis where the user points and clicks on certain
features of the image to define some of the lines. The
software will take what the user has provided, deter-
∆Z mine the remainder automatically, and then proceed
with solving the equations. This facilitates automatic
analysis of complicated or unclear images. In particu-
Edge location
lar, a long movie of complex images can often be
X
analyzed automatically with the operator intervening
Figure 2. Sub-pixel resolution of the edge finder. only on the first frame. Subsequent frames will be
analyzed as perturbations of the first. In all of these
curve fit is that it effectively averages many data cases, the operator’s input is limited to locating the
points into one equation, thus smoothing the inevi- edges in the image. The software then fits these edges
table noise in the video image. to analytic expressions and proceeds with the same
At the top of Figure 2, we see a magnified portion of a equation solution process in all cases.
gray-scale image edge. Each individual pixel is visible
Data Reporting. After the images have been analyzed
as a block. Only five Z levels of gray are shown; the
the software provides a variety of editing, smoothing,
real images have roughly 250. The graph in the mid-
graphing, and statistical reporting capabilities. Data
dle shows the gray values as a function of X for a
can be exported to other Windows™ programs or to
single selected scan line. The bottom graph shows the
industry-standard spreadsheets and databases.
first differences (or rate of change) of the Z values
plotted against X for the same scan line. The “edge” is Movie Capture. The FTÅ200 is a dynamic system in
defined as the centroid of this slope plot. In general, that it can capture motion and analyze non-static
this center of mass will occur at an arbitrary location situations. It has extremely fast image capture—60
within a pixel; i.e., the location will be determined independent frames per second, but it can also run
with a resolution less than the size of one pixel. This slower. It can capture long high speed movies, typi-
“sub-pixel” resolution allows for great accuracy in the cally 150 to 300 frames. These movies are stored in
resultant analytic curve. The FTÅ200 measures edges the main memory of the computer, which has the
both horizontally, as pictured here, and vertically; advantage of being easily expanded at any time by the
then it combines them in an optimal fashion. user. Slower movies, with frame intervals of a few
The second part of image analysis is to solve the seconds or more, can be stored directly on the hard
analytic equations for the desired result. In a sense, the disk, in which case there is no practical limit on the
edge finder does all the work, because it searches the number of images in a movie.
image for edges and converts them to the coefficients The FTÅ200 also has flexible “triggering” capabili-
of the desired equations. Once we have the equations, ties. The trigger sets the reference frame from which
2
other frames are measured. For example, the user may Unfortunately, the Laplace-Young equation can not be
request five frames before the trigger and 25 after. solved analytically in the general case. Therefore, all
The system keeps a rotating buffer of images; when solution methods rely on numerical techniques and
triggered, it marks the requested number of prior interpolations. The classic solution was provided by
frames for permanent storage and proceeds to acquire Bashforth and Adams. When done manually, as with
the remaining images after the trigger. A trigger may a ruler on a photograph, it suffers from the difficulty
be generated in many ways: manually by the operator, in finding SE , the equatorial diameter, because many
by a fiber optic detector observing the drop, by the points have to be measured to determine the true
syringe pump after a specified volume has been maximum width. When done automatically, it suffers
pumped, by a user-set timer, or by external user- from errors introduced by noise at the specific mea-
provided sensors. Finally, the system has provisions surement points. The FTÅ200 extends this technique
for non-uniform frame periods, so images may be by optimally fitting curves to the regions of interest in
taken rapidly at first and then slowly later on to better the pendant drop in order to minimize noise effects.
match the rapidity of drop shape changes. The various distances are then determined by solving
the equations analytically. An alternative technique is
Experimental Techniques to determine a drop shape parameter by interpolating
the drop shape against a set of reference shapes. After
Many different parameters can be determined from
the shape parameter is known, the solution follows the
drop shape analysis. The more commonly used ex-
classical Bashforth-Adams technique. Although called
periments are described below.
Bashforth-Adams, most workers use the more modern
Pendant Drop Surface Tension. The Laplace-Young recalculated tables of Padday.2
equation1 describes the shape of a fluid drop under
Pendant Drop Interfacial Tension. Interfacial tension
equilibrium conditions. A hanging pendant drop can
between two immiscible fluids, at least one of which
be analyzed more reliably than can a sitting sessile
is clear enough to transmit light, can be determined by
drop, since one can safely assume axial symmetry for
exactly the same technique as simple surface tension.
the pendant drop but not for the sessile drop. Figure 3
One must have an appropriate container to hold the
illustrates a pendant drop and the Bashforth-Adams
surrounding fluid and the operator must supply the
technique for solving the Laplace-Young equation.
specific gravity of both fluids. First Ten Ångstroms
can supply suitable interfacial tension chambers.
Dynamic Surface or Interfacial Tension. Temporal
variations in surface tension may be caused by the ad-
sorption of fluid components at the drop surface. For
variations on the order a few hundred milliseconds or
longer, the pendant drop can be dispensed and the
SW
movie taken after the pump stops. Figure 4 on the
next page illustrates this type of experiment; time
resolution is 0.5 seconds. Surface tension decreases in
SE
time as the surfactant adsorbs on the drop's surface
(i.e., it moves from being uniformly dispersed in the
fluid to being preferentially found at the surface).
H = SE

An alternative method is to first form a pendant drop

and let it come to equilibrium. Then the pump adds a
small amount of fluid in a short burst. This causes a
Surface tension is derived from drop shape. stress in the surface of the drop (new drop area is
Given fluid density, the value of SW / SE is an entry
into a lookup table which yields surface tension. 1. Also known as the equation of capillarity. See A.W. Adamson,
Physical Chemistry of Surfaces, Wiley, ISBN 0-471-61019-4.
Figure 3. Pendant drop surface tension 2. J. F. Padday, in Surface and Colloid Science, Vol. 1, (E.
measured by the Bashforth-Adams technique. Matijevic, ed), Wiley, New York, 1969.

3
 0.4% Tween Solution Surface Tension (dy/cm) “reflected” image. The FTÅ200 software lets the user
45
optionally specify a baseline type.
44 Another option allows the operator to choose between
a spherical drop profile fit and a non-spherical fit.
43
The non-spherical mode fits the drop profile only in
42
the regions adjacent to the contact angle point. This is
illustrated in Figure 6. The spherical mode fits over
41 the top portion of the drop and down as far as possible
towards the contact angle points. Note that it is not
40 necessary for the curve fits to actually reach the
0 5 10 15 20
Time (s) After Pendant Drop Formation contact angle point because the algorithm extrap-
olates the curves during the process of solving the
Figure 4. Dynamic surface tension.
equations analytically for the intersections. The non-
formed). If only one motor step is requested, the spherical fit is useful when the dispensing needle is
change is very rapid and is effectively an impulse. left in contact with the drop (for reasons described in
The change in surface tension can be measured with a following section) and when the drop is large, and

,,,,







16ms resolution by capturing a movie at the highest so is distorted by gravity. The spherical fit may be
speed. Time constants in the 50ms range will have used with smaller drops, and has the advantage that it
three data points and can be determined with is more robust against noise in the image.
reasonable accuracy using least squares fits. Longer
time constants will, naturally, have more data points Side-on, no
and higher accuracy. camera tilt








,,,,



Contact Angle. Contact angle is probably the most fre- Baseline
quently run experiment, as it is used to measure
wettability of solid surfaces and can also be used for Specimen
absorption experiments. The FTÅ200 provides several
Camera
options for measuring contact angles.




tilted down
In general, it is easy for the software to find the drop
Baseline
profile, but hard for it to find the baseline. This is
because many image defects can obscure the baseline,
plus the specimen may be irregular. Conversely, it is
easy for an operator to locate a baseline but fitting a








,,,,
tangent to the drop profile is difficult. The FTÅ200 Specimen
software lets you “mix and match” between operator Figure 5. Two possible viewing angles.
baselines and software determined drop profiles.
Spherical curve fit
Two kinds of baselines are found in practice. The type
of image illustrated in Figure 1 is an idealized one

,,,,
where the camera is looking exactly side-on, with a
viewing line parallel to the specimen surface. In many
cases, this is not desirable. Often one desires to look
down at a slight angle, e.g., two or three degrees. Fig-
Non-spherical curve fitting regions
ure 5 illustrates the two cases for the same drop.
If the drop is not close to the specimen edge, then the
downward tilt method will yield a clearer baseline,
since the specimen’s front edge cannot be in focus.
The baseline is located at the inflection point of the
profile between the upper drop image and the lower Figure 6. Spherical and non-spherical fitting regions.

4






,,,,

Finally, the specimen surface has been assumed to be Receding Contact Angle and Hysteresis. Figure 9 il-
flat so far. The FTÅ200 has the capability to correct lustrates the case in which the pump is reversed and
for curved specimen surfaces. The operator measures fluid is removed from the drop. The receding mode
the specimen curvature, using the FTÅ200 tools, then contact angle will normally be significantly lower
this curvature is subtracted from the subsequently than the advancing contact angle.
measured contact angles.
Contact angles obtained from a combination of mea-
suring advancing and receding angles are sometimes
plotted together as shown in Figure 10. The difference
between the advancing and receding contact angles is
Θ = Contact angle known as the “hysteresis.” The FTÅ200 will average
(same value the two nominally linear regions and compute the
as Figure 1) hysteresis. Various explanations are offered for hys-
Air
teresis based on microscopic compositional hetero-
Tangent line geneities or surface roughness of the specimen.3 Low
hysteresis is considered the sign of a “quality” surface
in some applications.
Fluid

Dispensing needle

Figure 7. Inverted mode.



Profile at time: t3 > t2 > t1
Inverted Experiments. Inverted experiments are those
in which the surrounding media is the fluid and the t
t3 t2 1
“drop” is air, or some other gas. The software will
function the same as with normal sessile drops, except Fluid
the user must specify the inverted mode. Figure 7
shows the geometry in this case. The contact angle is
the complement of what would have been the angle in Figure 8. Advancing contact angle.
the normal case. The situation depicted in Figure 7
represents the same fluid and substrate as that shown
in Figure 1. Note the shape of the “drop” is entirely

,,,,
different.
Dispensing needle
Advancing Contact Angle. If we keep the needle in
contact with the drop and continue dispensing while
capturing images, we obtain an “advancing” contact
Profile at time: t6 > t5 > t4
angle. In the laboratory, the contact angle of a prop-
t
erly dispensed static drop, as depicted in previous t4 t5 6
figures, will be the same as an advancing angle.
However, the advancing experiment does have the
advantage of covering new territory as the drop
expands, so contact angle versus drop width or posi-
tion data is obtained. This results in averaged data for Figure 9. Receding contact angle
(continuation of Figure 8).
the overall surface, which is useful since most speci-
mens exhibit variations from point to point. Figure 8
shows the measurement of advancing contact angle. 3. R. E. Johnson and R. H. Dettre, in Wettability, Surfactant
Non-spherical analysis is used to avoid the top of the Science Series, Vol. 49, (J. C. Berg, ed.), Dekker, ISBN 0-8247-
drop which is distorted by the dispensing needle. 9046-4.

5
 its shape reflects its surface tension. In other words,
Advancing surface energy remains hidden by the lack of deform-
contact angle ability in the solid. However, we can estimate surface
energy from the contact angles made by various
fluids. Loosely speaking, the more different fluids we
Θ Hysteresis use, the better the estimate. Choosing the best fluid(s)
is beyond the scope of this discussion (see reference
in footnote 3, page 5).
Receding The FTÅ200 provides four models, or equations, for
contact angle
relating contact angles to surface energy. These are
t6
solved automatically once the operator has specified
t1 t2 t3 t4 t5 Time (t)
the fluids and obtained the contact angles. The sim-
Figure 10. Contact angle hysteresis. plest, the Girifalco model, is often omitted in other
software, yet it is useful in that it only requires one
Absorption. Absorbent substrates can be studied with fluid and one contact angle measurement. It is accu-
the FTÅ200. A movie is taken of the drop as it rate at low contact angles but less accurate at the high
absorbs into the specimen. The initial contact angle is contact angles which occur at low surface energies. It
often of interest, as is volume of fluid in the drop can be improved by interpolating, or forcing the
which is not yet absorbed. A standing volume plot is answer to be correct for, say, Teflon (data for which
shown in Figure 11. is well known). With this improvement, the Girifalco
model is useful and convenient in many practical
12
situations.
For more precise work, the software offers the
10
geometric mean and harmonic mean models, both of
8
which require two fluids and two contact angle mea-
µl surements, and the acid/base model which uses three
6 fluids and three measurements. In all cases, the soft-
ware reports the components of surface energy appro-
4 priate to the chosen model.
2 Surface Tension from Contact Angle. When it is de-
0.0 0.5 1.0 1.5
sired to estimate surface tension for a sessile drop on
Time (s) After Drop Application a known surface, the Girifalco model can run in
Figure 11. Volume of drop not yet absorbed. “reverse,” inputting surface energy and contact angle
and calculating surface tension. Along similar lines,
the software will predict contact angles given fluid
This particular plot shows two regimes, one up to 0.5 surface tension and substrate surface energy.
seconds and the other after 0.5 seconds. Most likely
Zisman Critical Wetting Tension. The Zisman tech-
sizing in the paper was being dissolved during the
nique (see reference in footnote 1, page 3) is an
first half-second, after which absorption could take
alternative to surface energy models. It estimates the
place more rapidly. The software can also plot base
fluid surface tension which would just completely wet
width as a function of time to show spreading, and
(i.e., have a zero contact angle) the solid. The esti-
can compute flow rate into the specimen. These are all
matation is performed by extrapolating contact angle
derived from the geometry of the drop profile.
data from fluids which do not completely wet. Figure
Surface Energy Estimates. Whereas the surface ten- 12 shows hypothetical data in a Zisman plot. The
sion of a fluid can be determined with accuracy, the vertical axis is the cosine of the contact angle. Noting
same is not true of the corresponding quantity for a that the cosine of zero is one, Zisman extrapolated a
solid, surface energy. A solid surface will support a best fit line through the data and called its tension
shear stress, whereas a liquid will not support one, so when the line hit one the critical wetting tension. The

6
 Intersection of extrapolated best Robotic Motion. Stepper motor stages are available to
cos Θ
fit line with cos Θ = 1 (Θ = 0˚) position specimens under the dispensing needle.
1.0 These are available in any combination of X, Y, Z,
and rotational stages. These can be used to precisely
0.8 Best fit line position samples and to achieve “step and repeat”
patterns for analyzing specimen surfaces on a grid.
0.6 Tilting stages are available for those who prefer this
method of obtaining advancing and receding data.
0.4
The table is tilted until the drop begins to slide
0.2
“downhill.” The downhill contact angle will be the
advancing angle and the uphill, or trailing angle, will
0.0 be the receding contact angle.
10 20 30 40 50 60 γL
Dilution Sequences. An alternative form of the instru-
Critical Wetting Tension
ment is the Robotic Dilutions System. This config-
Figure 12. Zisman plot. uration uses a valved syringe pump, which is different
in that it is placed apart from the dispensing tip and is
FTÅ200 can make this plot and compute the critical connected by Teflon tubing. The valve allows the
wetting tension; the operator measures contact angle syringe chamber to be connected to either the
with different fluids of appropriate surface tension. dispensing tip or a fluid reservoir. The fluid being dis-
Critical Micelle Concentration. In a fashion similar to pensed need not be the fluid in the syringe chamber;
the making of a Zisman plot, critical micelle concen- instead an air gap may separate the two in the
tration can be determined from surface tension data at connecting tubing. The tip is then moved about and
various concentrations of the surfactant. Surface ten- positioned by a 3-axis robot while the syringe pump
sion plotted against log concentration will resemble remains fixed.
Figure 13. The operator makes up solutions of known This instrument can pick up test fluids from a matrix
concentrations and uses the FTÅ200 to measure sur- of locations (typically a 96-well plate) and bring them
face tension. This data is plotted by the software and into the measurement position in front of the camera.
the curve fits obtained. The critical micelle concentra- Surface tension can be measured and a drop can be
tion is inferred from the shape of the curve of surface placed on a test surface for contact angle measure-
tension plotted against concentration. ments. The utility of this instrument is that it can be
easily programmed to carry out sequences of tip
γL movements, syringe pump operations, and video
Increasing surfactant concentration
measurements automatically, all without operator
intervention during the sequence. Literally hours of
machine operation can be setup in advance. This
Intersection point
of line segments
capability lends itself to the following situations:
1. Dilution Sequences. A dilution sequence involves
a large number of fluid samples, progressively more
diluted. Sequences can be prepared with as little as
log C
100ul of initial concentrate. Surface tension can be
Critical concentration
measured as a function of time for each dilution,
allowing the calculation of diffusion constants and
Figure 13. Critical Micelle Concentration. critical micelle concentrations. Typically dilution
sequences are prepared in 96-well plates. The instru-
Hardware Accessories and Additional Techniques ment can be programmed to wash the dispensing tip
between each sample.
Hardware accessories can be added to the basic sys-
tem to extend its capabilities and, in some cases, allow 2. Large Numbers of Samples. Instead of preparing
specialized measurements. dilution sequence samples, the instrument can simply

7
measure independent samples placed in a 96-well Interfacial Tension Chamber. These special chambers
plate or in vials placed on a grid. Again, measure- are leak and pressure proof and are fluid-loop
ments include surface tension as a function of time heated/cooled by an external constant temperature
and, possibly, contact angle measurements. bath. A built-in RTD temperature detector measures
the actual chamber temperature which can be used to
3. Multi-Fluid Surface Energy Analysis. With a
control a pump on the external bath. As above, the
single-valved syringe pump and the built-in washing
specimen mount has a degree of freedom mechan-
protocols, the instrument can place multiple fluids on
ically for multiple measurements, and temperature
adjacent locations of a test solid for surface energy
may be ramped to obtain experimental data as a
determination. This greatly simplifies acid/base sur-
function of temperature.
face energy analysis.
Viscosity Measurement. A simple viscosity determi-
3-D Measurements. Another variation of the instru-
nation is possible by measuring the pressure drop
ment is the Robotic Mapping System. This system has
across the dispensing needle at a known flow rate.
robotics both for positioning the sample and moving
Although not as accurate as conventional viscosity
multiple dispense tips. Each dispense tip is connected
meters, it has the advantage of making the mea-
to a valved syringe pump of the same type used in the
surement on the same fluid that is used in the surface
Robotic Dilutions System.
tension or contact angle work, and it only requires a
These pump/tip combinations can be assigned the very small volume of fluid.
tasks of dispensing different fluids (useful in acid/base
determinations), or they can be divided between dis- FTÅ200 Software Support
pensing and picking up drops. This latter is needed for
First Ten Ångstroms has been shipping Windows™
2-D mapping work, where the first dispense tip
software since 1994. The software runs equally well
deposits drops on specimens for contact angle deter-
on Windows 3.1 or Windows 95. It makes full use of
mination. The second tip then comes along and picks
built-in Windows capabilities for printing, networks,
up the drop, leaving the specimen surface relatively
and transferring data via the Clipboard to other
free of fluid. This is necessary so previous drops will
programs. The software is licensed to instrument cus-
not interfere with automatic image analysis as new
tomers on a multi-computer basis which means that
drops are placed on the specimen. The 2-D grid of
customers may make copies of the software for other
data is displayed as a 3-D map of surface energy or
computers they own without charge. This is useful for
contact angle on the computer screen.
educational institutions and for users who may wish
Environmental Chamber. Environmental chambers to analyze data on a separate computer from the one
provide controlled temperature and atmosphere for the connected to the instrument.
specimen and fluids; often this is useful to maintain a
First Ten Ångstroms has a policy of continuously up-
constant relative humidity. These chambers can be
grading the software and customers are encouraged to
electrically heated to 200˚C, and to 300˚C with an
submit suggestions for improvements or extensions.
optional insulating shroud. They also can be heated
Upgrades are furnished to customers without charge,
and cooled through a fluid loop with an external
so there is never a need to requisition more funds to
constant-temperature bath. Temperature control reso-
have the latest software.
lution is 0.1˚C and accuracy is a few tenths of a
degree. First Ten Ångstroms’ chambers are unique in
that a specimen positioning slide allows multiple
measurements to be taken on a single chamber tem-
perature cycle. While the obvious mode is to bring the
chamber to temperature and then make measure-
ments, an alternative mode is to make measurements
continuously as temperature is ramped up or down.
This obtains surface tension or contact angle as a
function of temperature.