Satas D., Tracton A.A. (ed.). Coatings Technology Handbook
Подождите немного. Документ загружается.
This Page Intentionally Left Blank
10
Color Measurement for the Coatings
Industry
Harold Van Aken
Gretcq&frrcbeth,
New
Windsor,
New
York
Color is the most important appearance
of
coatings for their formulation, application,
or inspection. Color is also the most subjective parameter to characterize visually, and
characterization is often attempted under uncontrolled conditions that result in poor color
judgement. Proper viewing conditions require controlled lighting in a viewing booth where
the different types
of
light, such as simulated daylight, tungsten, and fluorescent light
sources, can be used for evaluation. Visual evaluation always requires a physical standard
for comparison, since the “color memory”
of
the brain is quite poor without one, but
very good when two samples are compared beside one another. Even when proper viewing
conditions are used, it is often difficult to determine the direction and intensity of color
difference between two samples. This process requires a trained colorist to make the
evaluation.
A
more accurate and consistent approach
to
evaluate color difference is the use
of
a color measurement instrument. The two types of instruments that can be used for this
purpose are colorimeters and spectrophotometers.
A
colorimeter uses optical filters to
simulate the color response
of
the eye, and a spectrophotometer breaks the visible spectrum
into intervals that mathematically simulate the color response of the eye. The advantage
of
using spectrophotometers to determine color difference is in their accuracy, stability,
and ability to simulate various light sources. Spectrophotometer cost and complexity
of
operation are greatly reduced on new versions
of
the instruments.
There are three different technologies that are used in modern industrial spectropho-
tometers: interference filters, gratings, and
LEDs.
Interference filters require a filter for
each wavelength measured and usually have
16
or
31
filters depending on the resolution
required. Grating-based instruments have diode arrays of
20
to
256
elements to provide
higher resolution for applications that require it. The advantage of interference filters
is
in their simplicity of operation and mechanical ruggedness. However, they are difficult
93
94
VAN AKEN
to make consistent and deteriorate over time. High-performance instruments usually have
gratings that give more resolution and better consistency, but they are usually more expen-
sive and complex to build and calibrate.
A
new market entrant for spectrophotometers is
based
on
LEDs
of different illumination colors. Up to nine separate color
LEDs
are now
available to cover most of the visible spectrum. The instruments operate by illuminating
one
LED
at
a
time while measuring the reflected light. The advantage is that they can be
made very small and cost less
to
manufacture. The disadvantages are reduced accuracy
and stability, but the technology is improving with the advent of newer
LEDs
with better
methods for compensation.
There are several different measurement geometries: sphere,
4.570,
and multiangle.
A
sphere instrument illuminates a sample from all directions and views the sample at near
normal or perpendicular. The
45/0
illuminates the sample at
45
degrees from all directions
and views the sample normal. It is also possible to illuminate at
0
and view at
45.
The
multiangle approach illuminates at multiple angles and views at
a
fixed angle. It is also
possible to illuminate at a fixed angle and view at multiple angles.
The use of proper geometry is important for color formulation or color inspection.
Color formulation with sphere geometry eliminates the need
to
characterize the gloss and
mathematically removes the gloss that
is
independent from the color formulation. Color
inspection usually requires the instrument to have agreement with visual methods.
A
451
0
instrument will give better correlation to visual assessment since it better approximates
the conditions in a viewing booth.
A
sphere instrument with
a
specular exclusion port can
eliminate sample high gloss to give good visual correlation but has difficulties with semig-
loss
samples. The assessment can yield misleading information. This is very important
when trying to match a coating to a plastic molded part.
Effect pigments such as metallic, pearlescent, and interference materials require
multiple angles
of
illumination and viewing to characterize color at different angles. Mul-
tiangle instruments or goniospectrophotometers are available
to
measure three
to
five
separate angles.
A
minimum of three angles are usually required
to
characterize effect
pigments:
(l
)
the near specular at
15
to
25
degrees from gloss.
(2)
45/0,
and
(3)
far from
gloss
of
75
to
1
10
degrees.
Consideration of the sample type to be measured should determine the variety of
spectrophotometer to use. If samples are large and cannot be brought to the instrument,
the instrument needs to be a portable. There are high-performance portable instruments
for each geometry, but consideration should be given to the correlation to laboratory
instruments, since the communication of the measurements to a color lab is often required.
If
very small samples such as paint chips or small color bars are measured, the measurement
aperture needs to be small. If the sample is nonuniform, the aperture should be
as
large
as
possible. Many instruments have changeable apertures that can be used for both samples.
Fluorescent coatings require a spectrophotometer with a carefully controlled illumination
source that is usually specified
as
daylight. Tungsten does not have the necessary
UV,
and cannot be used as a good daylight simulator. However, pulsed xenon is
a
very good
daylight simulator and can be adjusted in some instruments to match the spectrum of
natural daylight exactly.
The required tolerances for color measurement are one
of
the most important consid-
erations when selecting
a
color measurement instrument.
If
the comparison is always to
a
physical standard, and the formula is the same, a colorimeter can be sufficient. When
high accuracy is needed for producing coatings in different locations throughout the world
COLOR
MEASUREMENT
FOR
THE COATINGS INDUSTRY 95
using numerical standards, only the very high-end instruments will be capable
of
results
within acceptable visual agreement.
BIBLIOGRAPHY
1.
GretagMacbeth,
Fundatnentals
of
Color and Appearance.
2.
ASTM Standards
on
Color and Appearance Measurement,
4th ed. Philadelphia, PA: ASTM,
3.
Color
in
Business, Science, and Industry,
2d ed. New
York:
John Wiley, 1987.
4.
Practical Color Measurement,
A
Printer,for the Beginner, A Reminder for the Expert.
Anni
1994.
Berger Schunn, New York:
John
Wiley, 1994.
This Page Intentionally Left Blank
l1
The
Use
of X-ray Fluorescence for
Coat Weight Determinations
Wayne
E.
Mozer
Oxford Analytical,
Inc.,
Anclover, Mrr.s.sacllusetts
1
.O
INTRODUCTION
The technique of elemental analysis by x-ray fluorescence (XRF) has been applied to the
quality control of coating weights at the plant level. Measurements by nonlaboratory
personnel provide precise and rapid analytical data on the amount and uniformity
of
the
applied coating. XRF has proved to be an effective means of determining silicone coating
weights on paper and film, titanium dioxide loading in paper, and silver on film.
2.0
TECHNIQUE
X-ray fluorescence is a rapid, nondestructive, and comparative technique for the quantita-
tive determination of elements in a variety of matrices. XRF units come in a variety of
packages; however, the type of unit most prevalent in the coating industry is described
in this chapter.
The XRF benchtop analyzer makes use of a low level radioisotope placed in close
proximity to the sample. The primary x-rays emitted from the excitation source strike the
sample, and fluorescence
of
secondary x-rays occur. These secondary x-rays have specific
energies, which are characteristic of the elements in the sample and are independent of
chemical or physical state. These x-rays are detected in
a
gas-filled counter that outputs
a
series of pulses, the amplitudes of which are proportional to the energy of the incident
radiation. The number of pulses from silicon x-rays, for example is proportional to the
silicone coat weight of the sample. Since the technique is nondestructive, the sample is
reusable for further analysis at any time.
To ensure optimum excitation, alternate radioisotopes may be necessary for different
applications. For silicone coatings and titanium dioxide in paper, an iron-55 (Fe-55) source
97
98
MOZER
is used. Fe-55 x-rays are soft (low energy) and do not penetrate far into
a
sample.
For
silver on film.
a
more energetic americanum-241 source has been used.
Placing samples just
a
few millimeters from the excitation source enables high
sensitivities to be obtained. Irradiation of the sample is more efficient, the closer the sample
is brought to the source. In the case of low energy x-rays, such
as
silicon, which are
easily absorbed by the atmosphere, a helium purge should be used. With optimum sample
irradiation and helium purging when necessary, measurements within
2
minutes of count-
ing time are typical for most samples.
Calibration curves for different elements and materials can be stored directly in the
instrument and are available for recall. Curves are established by measuring
a
set of
known samples
or
standards
of
the same material. Since
XRF
is a comparative technique,
subsequent analyses are only
as
good
as
the quality of the calibration standards supplied.
The total accumulated intensity is actually a combination of signals from the analyte
element and from the matrix in the form of background. Background intensities may vary
depending on the thickness or the basis weight of the material. These differences may be
determined by measuring uncoated
or
blank samples, which are automatically incorporated
into the calibration.
3.0
METHOD
Three procedural steps must precede analysis. First, spectrum scan helps
to
set up various
calibration parameters. Next a calibration is established. Almost
all
the time the instrument
is in use, it is dedicated to analysis
of
production samples. Typically, spectrum scans are
performed on representative standards to identify the occurrence of elemental peaks of
interest in the spectrum,
as
illustrated in Figure
1.
The scan is a qualitative tool and will
show the presence of any potential interfacing element. For example, an Fe-55 source will
excite the following range of elements in the periodic scale: aluminum to vanadium and
zirconium
to
cerium. With the proper combination of sources, it is theoretically possible
to measure from aluminum
to
uranium. Once any interference has been recognized by a
spectrum scan,
a
calibration may be established. A sample is prepared by cutting a disk
and placing it onto a nickel-coated paper holder. The holder ensures that the sample will
be held
as
flat
as
possible, since an uneven surface will introduce errors in reproducibility,
caused by scattering and vacancies at the incident surface. Table
1
gives a typical calibra-
tion for silicon coat weights.
Once
a
calibration has been established, two samples are selected which bracket the
operational range of the samples measured. These samples will then be used to correct
for changes in instrumental performance with time. The measurement of these two stan-
dards and of an uncoated sample for background correction purposes is
a
push-button
procedure. Termed restandardization, is performed once per shift. Analyses may then be
performed: an operator simply cuts
a
sample to be analyzed, inserts it into the instrument,
and initiates the measurement. Within 23 minutes the instrument will determine the concen-
tration of the element of interest, both printing and displaying the results.
4.0
ACCURACY
To
ensure optimum accuracy, it is important
to
be aware
of
three sources of errors: suppliers
of coating material, background changes, and interferences. It has been recognized,
al-
though not explained, that silicones of different suppliers give different sensitivities. When
THE
USE
OF
X-RAY
FLUORESCENCE
FOR
COAT
WEIGHT
DETERMINATIONS
99
I
L
.
I
100
150
200
Channel
Figure
1
Spectrum scan
of
silicone coatings on super calendar craft paper.
different silicones are analyzed on the same basis weight paper, lines with different slopes
are obtained (Fig.
2).
Thus, to avoid this source of error, separate calibrations are estab-
lished for each supplier
of
silicone. Second, correction
for
changes in background will
help reduce error.
Figure
3
illustrates two sets of standards
of
the same silicone on supercalendar kraft
prepared on different dates. Quite evidently, the samples give the
same
slopes, they are
Table
1
Typical Results for
a
Silicone-on-paper Calibration"
Given concentration X-ray concentration
Sample
( g/m'
)
(dm?) Error
0.33
0.41
0.54
0.62
0.70
1
SO
0.34
0.4
1
0.52
0.62
0.72
l
SO
0.01
0.00
-
0.02
0.00
0.02
0.00
.'
Standard
error
=
0.013
g/m2
100
MOZER
150
-
X-ray
CPS
.25
.50
.l5
.l00
Concentration
g/m
2
Figure
2
Differences in sensitivity in products of different suppliers of silicone.
.25
.50
.75
1.00
Concentration
g/m
2
Figure
3
Differences In paper backings.
THE
USE
OF
X-RAY FLUORESCENCE
FOR
COAT WEIGHT DETERMINATIONS
101
offset. The analyzer will correct for this background change by having the operator insert
an uncoated sample from the lot of paper to be coated. Thus, all subsequent analyses will
be corrected for the apparent background change.
The final source of error is the presence
of
another material in high concentrations.
Examples include silicone coatings on polyvinyl chloride
(PVC)
or titanium dioxide filled
films.
As
the levels in these films change, the effect on the background for the silicon
region of the spectrum will change. Thus, there will be raising or lowering of the apparent
silicone coat weight for that film sample. This type
of
interference is easily corrected for
by software that can be built into the analyzer. The interferences are recognized by using
the spectrum scan, and automatically compensated for during calibration.
5.0
REPEATABILITY AND REPRODUCIBILITY
Repeatability in terms
of
precision of measurement is
a
statistical function determining
the variability of repeat measurements on the accumulated intensity and x-ray counts.
Since an iron-55 source excites titanium x-rays, more efficiently than silicon x-rays, it
follows that sensitivities for titania coatings are higher, and therefore, measurements more
precise. The precision may always be improved by increasing the coating time, but this
does not always result in substantial improvements in accuracy. Typically, for silicone
coatings, reported standard deviations correspond to
It
0.01
g
per square meter
of
silicone.
6.0
CONCLUSION
On-site XRF determinations are rapid enough and precise enough for effective quality
control programs for elemental determinations on a variety of substrates. The use of
XRF
is not limited to measuring coatings; it is flexible enough to measure related products,
such
as
tin, platinum, and rhodium catalysts, and other solutions.