11-1

11
The Ise ol X-ray
!Iuorescence lor Coaf
WeIghf IefermInafIons
11.1 Intioduction ......................................................................11-1
11.2 Technique...........................................................................11-1
11.3 Method...............................................................................11-2
11.4 Accuiacy.............................................................................11-3
11.5 Repeatability and Repioducibility....................................11-3
11.6 Conclusion.........................................................................11-5
11.1 Intruductiun
The technique of elemental analysis by x-iay ßuoiescence (XRF) has been applied to the quality contiol
of coating weights at the plant level. Measuiements by nonlaboiatoiy peisonnel piovide piecise and iapid
analytical data on the amount and unifoimity of the applied coating. XRF has pioved to be an effective
means of deteimining silicone coating weights on papei and flm, titanium dioxide loading in papei, and
silvei on flm.
11.2 Technique
XRF is a iapid, nondestiuctive, and compaiative technique foi the quantitative deteimination of elements
in a vaiiety of matiices. XRF units come in a vaiiety of packages; howevei, the type of unit most pievalent
in the coating industiy is desciibed in this chaptei.
The XRF benchtop analyzei makes use of a low level iadioisotope placed in close pioximity to the
sample. The piimaiy x-iays emitted fiom the excitation souice stiike the sample, and ßuoiescence of
secondaiy x-iays occuis. These secondaiy x-iays have specifc eneigies that aie chaiacteiistic of the
elements in the sample and aie independent of chemical oi physical state. These x-iays aie detected in
a gas-flled countei that outputs a seiies of pulses, the amplitudes of which aie piopoitional to the eneigy
of the incident iadiation. The numbei of pulses fiom silicon x-iays, foi example is piopoitional to the
silicone coat weight of the sample. Because the technique is nondestiuctive, the sample is ieusable foi
fuithei analysis at any time.
To ensuie optimum excitation, alteinate iadioisotopes may be necessaiy foi diffeient applications. Foi
silicone coatings and titanium dioxide in papei, an iion-55 (Fe-55) souice is used. Fe-55 x-iays aie soft
(low eneigy) and do not penetiate fai into a sample. Foi silvei on flm, a moie eneigetic ameiicanum-
241 souice has been used.
Wayne £. Nozer
Oxford Ano|yríco|, Inc.
© 2006 by Taylor & Erancis Group, LLC
11-2 Cooríng· Tec|no|ogy Hondboo|, T|írd ídíríon
Placing samples just a few millimeteis fiom the excitation souice enables high sensitivities to be
obtained. Iiiadiation of the sample is moie effcient the closei the sample is biought to the souice. In
the case of low eneigy x-iays, such as silicon, which aie easily absoibed by the atmospheie, a helium
puige should be used. With optimum sample iiiadiation and helium puiging when necessaiy, measuie-
ments within 2 minutes of counting time aie typical foi most samples.
Calibiation cuives foi diffeient elements and mateiials can be stoied diiectly in the instiument and
aie available foi iecall. Cuives aie established by measuiing a set of known samples oi standaids of the
same mateiial. Since XRF is a compaiative technique, subsequent analyses aie only as good as the quality
of the calibiation standaids supplied.
The total accumulated intensity is actually a combination of signals fiom the analyte element and
fiom the matiix in the foim of backgiound. Backgiound intensities may vaiy depending on the thickness
oi the basis weight of the mateiial. These diffeiences may be deteimined by measuiing uncoated oi blank
samples that aie automatically incoipoiated into the calibiation.
11.3 Methud
Thiee pioceduial steps must piecede analysis. Fiist, spectium scan helps to set up vaiious calibiation
paiameteis. Next a calibiation is established. Almost all the time the instiument is in use, it is dedicated
to analysis of pioduction samples. Typically, spectium scans aie peifoimed on iepiesentative standaids
to identify the occuiience of elemental peaks of inteiest in the spectium, as illustiated in Figuie 11.1.
The scan is a qualitative tool and will show the piesence of any potential inteifacing element. Foi example,
an Fe-55 souice will excite the following iange of elements in the peiiodic scale: aluminum to vanadium
and ziiconium to ceiium. With the piopei combination of souices, it is theoietically possible to measuie
fiom aluminum to uianium. Once any inteifeience has been iecognized by a spectium scan, a calibiation
FIGURE 11.1 Spectia scan of silicone coatings on supei calendai ciaft papei.
X
-
R
a
y

C
o
u
n
t
s
/
S
e
c
o
n
d
100
75
50
25
100 150 200
Channel
C1
Ca
Si
© 2006 by Taylor & Erancis Group, LLC
T|e I·e of Xroy í|uore·cence for Coor Weíg|r Derermínoríon· 11-3
may be established. A sample is piepaied by cutting a disk and placing it onto a nickel-coated papei
holdei. The holdei ensuies that the sample will be held as ßat as possible because an uneven suiface will
intioduce eiiois in iepioducibility, caused by scatteiing and vacancies at the incident suiface. Table 11.1
gives a typical calibiation foi silicon coat weights.
Once a calibiation has been established, two samples aie selected that biacket the opeiational iange
of the samples measuied. These samples will then be used to coiiect foi changes in instiumental
peifoimance with time. The measuiement of these two standaids and of an uncoated sample foi back-
giound coiiection puiposes is a push-button pioceduie. Teimed iestandaidization is peifoimed once
pei shift. Analyses may then be peifoimed: an opeiatoi simply cuts a sample to be analyzed, inseits it
into the instiument, and initiates the measuiement. Within 23 minutes, the instiument will deteimine
the concentiation of the element of inteiest, both piinting and displaying the iesults.
11.4 Accuracy
To ensuie optimum accuiacy, it is impoitant to be awaie of thiee souices of eiiois: supplieis of coating
mateiial, backgiound changes, and inteifeiences. It has been iecognized, although not explained, that
silicones of diffeient supplieis give diffeient sensitivities. When diffeient silicones aie analyzed on the
of eiioi, sepaiate calibiations aie established foi each suppliei of silicone. Second, coiiection foi changes
in backgiound will help ieduce eiioi.
diffeient dates. Quite evidently, the samples give the same slopes, they aie offset. The analyzei will coiiect
foi this backgiound change by having the opeiatoi inseit an uncoated sample fiom the lot of papei to
be coated. Thus, all subsequent analyses will be coiiected foi the appaient backgiound change.
The fnal souice of eiioi is the piesence of anothei mateiial in high concentiations. Examples include
silicone coatings on polyvinyl chloiide (PVC) oi titanium dioxide flled flms. As the levels in these flms
change, the effect on the backgiound foi the silicon iegion of the spectium will change. Thus, theie will
be a iaising oi a loweiing of the appaient silicone coat weight foi that flm sample. This type of inteifeience
is easily coiiected foi by softwaie that can be built into the analyzei. The inteifeiences aie iecognized by
using the spectium scan and aie automatically compensated foi duiing calibiation.
11.5 Repeatabi!ity and Repruducibi!ity
Repeatability in teims of piecision of measuiement is a statistical function deteimining the vaiiability
of iepeat measuiements on the accumulated intensity and x-iay counts. Since an iion-55 souice excites
titanium x-iays moie effciently than silicon x-iays, it follows that sensitivities foi titania coatings aie
highei, and theiefoie, measuiements aie moie piecise. The piecision may always be impioved by incieas-
ing the coating time, but this does not always iesult in substantial impiovements in accuiacy. Typically,
foi silicone coatings, iepoited standaid deviations coiiespond to 0.01 g pei squaie metei of silicone.
TABLE 11.1 Typical Results foi a Silicone-on-Papei Calibiation
a
Sample
Given Concentiation
(g/m
2
)
X-iay Concentiation
(g/m
2
) Eiioi
1 0.33 0.34 0.01
2 0.41 0.41 0.00
3 0.54 0.52 0.02
4 0.62 0.62 0.00
5 0.70 0.72 0.02
6 1.50 1.50 0.00
a
Standaid eiioi · 0.013 g/m
2
.
© 2006 by Taylor & Erancis Group, LLC
same basis weight papei, lines with diffeient slopes aie obtained (Figuie 11.2). Thus, to avoid this souice
Figuie 11.3 illustiates two sets of standaids of the same silicone on supeicalendai kiaft piepaied on
11-4 Cooríng· Tec|no|ogy Hondboo|, T|írd ídíríon
FIGURE 11.2 Diffeiences in sensitivities in pioducts fiom diffeient supplieis of silicone.
FIGURE 11.3 Diffeiences in papei backings.
X
-
R
a
y

C
P
S
200
150
100
50
.25 .50 .75 1.00
Concentration g/m
2
Vendor A
Vendor B
Vendor C
X
-
R
a
y

C
P
S
200
150
100
50
.25 .50 .75 1.00
Concentration g/m
2
© 2006 by Taylor & Erancis Group, LLC
T|e I·e of Xroy í|uore·cence for Coor Weíg|r Derermínoríon· 11-5
11.6 Cunc!usiun
On-site XRF deteiminations aie iapid enough and piecise enough foi effective quality contiol piogiams
foi elemental deteiminations on a vaiiety of substiates. The use of XRF is not limited to measuiing
coatings; it is ßexible enough to measuie ielated pioducts, such as tin, platinum, and ihodium catalysts,
and othei solutions.
© 2006 by Taylor & Erancis Group, LLC