1 / 24
CCQM Key Comparison – Organic Solutions
CCQM-K47 Volatile Organic Compounds in Methanol
Final Report
April 8, 2009
Melina Pérez Urquiza, Mauricio Maldonado Torres and Yoshito Mitani
Centro Nacional de Metrologia (CENAM)
Querétaro, Mexico
Michele M. Schantz, David L. Duewer, Wille E. May, Reenie M. Parris and Stephen A. Wise
National Institute of Standards and Technology (NIST)
Gaithersburg, MD USA
Katja Kaminski, Rosemarie Philipp and Tin Win
Federal Institute for Materials Research and Testing (BAM)
Berlin, Germany
Adriana Rosso
Instituto Nacional de Tecnología Industrial (INTI)
Buenos Aires, Argentina
Dal Ho Kim
Korea Research Institute of Standards and Science (KRISS)
Taejon, Korea
Keiichiro Ishikawa
National Metrology Institute of Japan (NMIJ)
Ibaraki, Japan
A.I. Krylov and Y.A. Kustikov
D.I. Mendeleyev Institute for Metrology (VNIIM)
St. Petersburg, Russia
Annarita Baldan
Van Swinden Laboratorium (VSL)
Delft, the Netherlands

2 / 24
Introduction
At the October 2005 CCQM Organic Analysis Working Group Meeting (IRMM, Belgium), the
decision was made to proceed with a Key Comparison study (CCQM-K47) and a concurrent
subsequent pilot study (CCQM-P61.1) addressing the calibration function for the determination
of volatile organic compounds (VOCs) used for water quality monitoring. Both studies were
coordinated by CENAM and NIST. Benzene, o-xylene, m-xylene, and p-xylene were chosen as
representative VOCs. The solvent of choice was methanol. This report summarizes the results
for CCQM-K47.

Summary of Pilot Study CCQM-P61
At the October 2004 CCQM Organic Analysis Working Group Meeting (NRCCRM, China), the
decision was made to proceed with a pilot study addressing the calibration function for the
determination of volatile organic compounds (VOCs) used for water quality monitoring.
Benzene, toluene, ethylbenzene, and three xylene isomers were chosen as representative VOCs.
The solvent of choice was methanol. The pilot study CCQM-P61 Volatile Organic Compounds
in Methanol was coordinated by CENAM and NIST.
Each CCQM-P61 participant received five ampoules of the a gravimetrically prepared solution.
Each 2 mL ampoule contained approximately 1.2 mL of a methanol solution including benzene,
toluene, ethylbenzene, o-xylene, m-xylene, and p-xylene at concentrations between 13.37 µg/g
(o-xylene) to 49.55 µg/g (toluene). The instructions requested the analysis of duplicate
subsamples from each of four ampoules using the laboratory’s analytical procedure for
determination of the mass fraction concentrations of the target analytes.
Nineteen laboratories received samples for CCQM-P61; eighteen laboratories reported
measurement results to the coordinators. Two laboratories used IDMS, six laboratories used
internal standards, six laboratories used external standards, and the other four laboratories did not
report the method of quantification used. Most of the laboratories used gas chromatography with
mass spectrometry (GC/MS) and/or GC with flame ionization detection (GC-FID).
The majority of reported results agreed with the gravimetric preparation values to within ± 5 %,
with a minority of values ranging up to ± 20 %.

Design, Conduct, and Results from CCQM-K47
Study Material
The material used in CCQM-K47 was similar to that used in the CCQM-P61 pilot study: a
gravimetrically prepared methanol solution of the four target VOCs (benzene and the three
xylene isomers) plus toluene and ethylbenzene. The solution was prepared from neat materials
procured from commercial sources. The purities of the starting materials were determined by
GC-FID. Table 1 lists the gravimetric preparation impurity-adjusted mass fraction of the target
VOCs in the study solution.

3 / 24
Table 1. Gravimetric Preparation Mass Fraction and Homogeneity Evaluation of Target VOCs

Mass Fraction, μg/g ANOVA
Analyte Value U95(Value) Fc Ft
benzene 39.40 0.39 1.83 2.39
o-xylene 15.77 0.28 1.39 2.39
m-xylene 20.87 0.20 1.16 2.39
p-xylene 28.25 0.28 1.04 2.39
Homogeneity Evaluation
Table 1 also summarizes the results of an Analysis of Variance (ANOVA) homogeneity
evaluation for the target VOCs. Ten ampoules (A043, A073, A173, A203, A373, A381, A486,
A502, A579, A581) were measured in triplicate using GC-FID. The null hypothesis was that the
measurement means in all of the ampoules are identical; the alternate hypothesis was that the
mean in at least one of the ampoules differed from the others. For all four of the target VOCs,
the null hypothesis of homogeneity is accepted. The relative standard deviation of the
measurements ranged from 0.2 % to 0.5 %.
Study Design
Each participant in CCQM-K47 received five ampoules of the study solution. Each 2-mL
ampoule contained approximately 1.2 mL of solution. Participants were informed that the mass
fractions of all target VOCs ranged between 10 µg/g and 50 µg/g. Participants were requested to
analyze duplicate aliquots from each of four ampoules using an analytical procedure validated in
CCQM-P61.
Participants
The eight laboratories listed in Table 2 received samples for CCQM-K47. All eight laboratories
reported measurement results to the coordinators.

Table 2. CCQM-K47 Participants

NMI Submitted by Country Email Contact
BAM Tin Win, Rosemarie Philipp,Katja Kaminski Germany tin.win@bam.de
CENAM
Mauricio Maldonado Torres,
Evangelina Camacho Frías,
Melina Pérez Urquiza
México meperez@cenam.mx
INTI Adriana Rosso Argentina mariare@inti.gov.ar, cpuglisi@inti.gov.ar
KRISS Dal Ho Kim Korea byungjoo@kriss.re.kr
NIST Michele Schantz USA michele.schantz@nist.gov
NMIJ K. Ishikawa Japan ishikawa-keiichiro@aist.go.jp
VNIIM A.I. Krylov, Y.A. Kustikov Russia lkonop@b10.vniim.ru
VSL Annarita Baldan Netherlands abaldan@nmi.nl

4 / 24
Analytical Methods
Table 3 summarizes the analytical methods used by the eight participants. Table 4 summarizes
the amount of sample used in each analysis. Tables 5 and 6 summarize the materials used to
prepare calibration solutions. Table 7 summarizes how the calibration solutions were used.


Table 3. Analytical Methods

Column
NMI Method Phase L, m ID, mm FT, µm
BAM
GC/MS: benzene, o-xylene,
m&p-xylene (coelution)
GC-FID: m/p-xylene ratio
DB-624

DB-WAX
30

60
0.32

0.32
1.8

0.5
CENAM GC/MS DB-WAX 60 0.25 0.25
INTI GC-FID PEG (Supelcowax) 60 0.32 0.5
KRISS GC-FID
PEG
(5% phenyl) methyl
polysiloxane
30 0.53
1

2.65
NIST GC-FID: xylenes GC/MS: benzene
DB-WAX
DB-VRX
15
60
0.45
0.25
0.85
1.4
NMIJ GC/MS DB-WAX 60 0.32 0.5
VNIIM GC/MS HP 5 MS 30 0.25 0.25
VSL GC-FID CPWAX 52 60 0.53 2


Table 4. Sample Mass of CCQM-K47 Solution used for Analysis

NMI Mass solution used, g
BAM 0.05 to 0.09
CENAM 0.3
INTI 0.5
KRISS 0.2
NIST 0.4
NMIJ 0.4
VNIIM 0.5
VSL 0.7


Table 5. Certified Reference Materials Used as Calibrants

Certified Mass Fraction, μg/g
NMI Solvent Source benzene o-xylene m-xylene p-xylene
BAM Methanol KRISS 40.6 ±0.6 102.0±1.6 100.2±0.6 100.2±0.6

5 / 24
Table 6. Neat Materials Used to Prepare Calibrants

Mass Fraction, mg/g
NMI Source benzene o-xylene m-xylene p-xylene
CENAM Aldrich 0.9988 0.9915 0.9968 0.9967
INTI Merck >0.99
INTI AnalityCals Carlo Erba 0.988 0.978 >0.99
KRISS Sigma-Aldrich 0.9995 0.9916 0.9977 0.9986
NIST Sigma Aldrich 1.00 0.9363 1.00 0.9924
NMIJ NIM CRM 4002-a * 0.99996
NMIJ NIM CRM 4011-a * 0.9994
NMIJ NIM CRM 4012-a * 0.9981
NMIJ NIM CRM 4013-a * 0.9988
VNIIM SOP 0003-03 STC 0.9997
VNIIM SOP 0020-03 STC 0.997
VNIIM SOP 0015-03 STC 0.9966
VNIIM SOP 0016-03 STC 0.999
VSL Aldrich 0.999
VSL Fluka 0.995 0.995 0.995

* Deuterated Certified Reference material

Table 7. Calibration Method used, Number of Calibration Levels, and Calibration Range

benzene o-xylene m,p-xylene
NMI Use Internal standard was added # mg/g # mg/g # mg/g
BAM Internal at 2nd of 2 dilutions 5 0.3 - 4.8 5 0.7 - 12.0 5 0.7 - 12.0
CENAM Internal before injection 5 1 - 60 5 1 - 60 5 1 - 60
INTI External 4 3.5 - 53 3 3.5 - 53 3 3.5 - 53
KRISS External 1 1 1
NIST Internal during sample preparation 1 1 1
NMIJ IDMS to sample (deuterated VOC)
VNIIM Internal before subsample analysis 1 1 1
VSL Internal to autosampler vials 6 4 - 60 6 1 - 35 6 1 - 35

Results
The values, the combined uncertainties on the values, u, and the 95 % expanded uncertainties,
U95, as submitted by the participants are summarized in Table 8 and displayed in Figure 1. Each
panel of Figure 1 displays the results, their consensus summary, and the gravimetric preparation
values for one of the target VOCs. Each “dot” denotes a value and each vertical “bar” denotes
the 95% confidence interval on the value. The solid black horizontal line represents the
consensus median, the red horizontal lines represent the robust 95 % confidence interval about
the median, and the dotted black line goes through the gravimetric value. The black curve to the
right of each panel is the additive probability density function for the reported results; the light
blue curve is the Gaussian defined by the consensus summary values. In all four panels, the
mass-fraction axis spans a 30 % interval about the consensus value.

6 / 24
Table 8. Submitted Values, µg/g

benzene o-xylene m-xylene p-xylene
NMI Value u U95 Value u U95 Value u U95 Value u U95
BAM 40.38 0.31 0.62 16.07 0.14 0.28 19.56 0.43 0.86 27.95 0.61 1.22
CENAM 38.26 0.27 0.53 15.93 0.05 0.10 19.58 0.08 0.16 27.42 0.48 0.97
INTI 39.17 0.49 1.21 16.42 0.24 0.68 18.95 0.16 0.46 27.32 0.16 0.38
KRISS 38.58 0.31 0.62 16.33 0.12 0.25 18.83 0.13 0.27 26.71 0.20 0.40
NIST 39.38 0.43 0.85 15.60 0.17 0.34 20.19 0.29 0.61 28.26 0.35 0.70
NMIJ 39.26 0.14 0.27 16.42 0.06 0.12 19.92 0.07 0.14 28.12 0.10 0.20
VNIIM 32.64 0.76 1.52 19.04 0.42 0.84 Reported as sum of m- & p-xylene
VSL 39.83 0.69 1.37 16.77 0.32 0.64 20.11 0.27 0.54 28.55 0.36 0.72
Consensus* 39.22 0.33 0.76 16.38 0.18 0.42 19.58 0.30 0.70 27.95 0.30 0.70

* Robust consensus values calculated as: Value = median, u = median absolute deviation from the median
(MADe), and U95 = u * t0.025,n-1 where t0.025,n-1 is the two-sided Student’s t for 95 % confidence and n measurements.

Figure 1. Dot-and-Bar Plots for the Four Target VOCs


<G
rav
Pr
ep
>
BA
M
VS
L
NI
ST
NM
IJ
IN
TI
KR
ISS
CE
NA
M
VN
IIM
32
34
36
38
40
42
Be
nz
en
e,
µg
/g
<G
rav
Pr
ep
>
VN
IIMVS
L
NM
IJ
IN
TI
KR
ISSBA
M
CE
NA
M
NI
ST
16
17
18
19
o-X
yle
ne
, µ
g/g
<G
rav
Pr
ep
>
NI
STVS
L
NM
IJ
CE
NA
M
BA
M
IN
TI
KR
ISS
17
18
19
20
21
22
m-
Xy
len
e,
µg
/g
<G
rav
Pr
ep
>
VS
L
NI
ST
NM
IJ
BA
M
CE
NA
M
IN
TI
KR
ISS
24
26
28
30
32
p-X
yle
ne
, µ
g/g

Additional information on the methods and uncertainty calculations used by the participants is
included in Appendix A. Representative chromatograms are provided in Appendix B.

7 / 24
Key Comparison Reference Values

As displayed in Figure 1, the consensus summary values agree well with the gravimetric
preparation values for benzene and p-xylene but not for o-xylene nor m-xylene. Review of the
solution preparation records failed to identify any experimental or data analysis oversight.
Inspection of the chromatograms in Appendix B suggests that discrepancy for m-xylene is not
caused by insufficient chromatographic separation of m- and p-xylene.
At the April 2006 CCQM Meeting in Paris, VSL was asked to analyze the CCQM-47 solution
with thermal desorption GC-FID. Dynamic dilution was used to prepare six calibration levels of
gas mixtures following the procedure of ISO 6145-8. These reference gas mixtures were loaded
onto tenax TA sorbent tubes. The CCQM-K47 solution was loaded onto tenax TA sorbent tubes
using a calibrated 10 μL syringe, three tubes each for five ampoules. Table 9 summarizes the
results of this evaluation; Figure 2 displays the results relative to the measurement and
gravimetric preparation values
Table 9. Results from VSL Thermal Desorption Analysis, μg/g

benzene o-xylene m-xylene p-xylene toluene
Value U95 Value U95 Value U95 Value U95 Value U95
39.75 1.03 17.06 0.42 20.47 0.53 28.71 0.72 20.80 0.53

Figure 2. Dot-and-Bar Plots for the Four target VOCs with VSL Thermal Desorption Results

<T
hrm
Ds
rp>
<G
rav
Pre
p>
BA
M
VS
L
NI
ST
NM
IJ
INT
I
KR
ISS
CE
NA
M
VN
IIM
32
34
36
38
40
42
Be
nz
en
e,
µg
/g
<Th
rm
Ds
rp>
<G
rav
Pre
p>
VN
IIMVS
L
NM
IJ
INT
I
KR
ISSBA
M
CE
NA
M
NI
ST
16
17
18
19
o-X
yle
ne
, µ
g/g

<T
hrm
Ds
rp>
<G
rav
Pr
ep
>
NIS
T
VS
L
NM
IJ
CE
NA
M
BA
M
INT
I
KR
ISS
17
18
19
20
21
22
m-
Xy
len
e,
µg
/g
<Th
rm
Ds
rp>
<G
rav
Pr
ep
>
VS
L
NIS
T
NM
IJ
BA
M
CE
NA
M
INT
I
KR
ISS
24
26
28
30
32
p-X
yle
ne
, µ
g/g

8 / 24
The thermal desorption results agree well with the gravimetric preparation values for benzene,
m-xylene, and p-xylene. The result for o-xylene does not agree well with either the gravimetric
or consensus value.
At the April 2008 CCQM Meeting in Paris, consideration was given to defining the Key
Comparison Reference Value (KCRV) from the consensus, gravimetric preparation, thermal
desorption, or some combination of the three values. Consideration was also given to voiding
the study completely and to dropping o-xylene as a target measurand. On the basis of the
agreement between the gravimetric preparation and thermal desorption values for m-xylene and
the absence of any evidence of an analytical blunder that would be specific to o-xylene, the
decision was made to accept the complete study as valid and use the gravimetric preparation
values as the KCRV for all for four target VOCs.

Key Comparison Reference Value Uncertainties
Recognizing that the uncertainty estimated from the gravimetric preparation process may not
fully represent the variability of the VOC levels in the solution levels as delivered to the
participants, the combined uncertainty for the KCRV is estimated as:
       ns
suu
MADeConsensus
ConsensusGravPrepKCRV 22

 .
where MADe is a robust estimate for the standard deviation of the n reported values. The 95 %
level of confidence expanded uncertainty is estimated with the standard metrological factor, k=2:
   KCRV2KCRV95 uU  .
Table 10 lists the KCRV, u(KCRV), and U95(KCRV) values for the four target VOCs.
Table 10. Key Comparison Reference Values and Their Uncertainties, μg/g

VOC KCRV u(KCRV) U95(KCRV)
benzene 39.4 0.4 0.8
o-xylene 15.8 0.2 0.5
m-xylene 20.9 0.3 0.6
p-xylene 28.3 0.3 0.7

Degrees of Equivalence
Since participant results are not used in the estimation of the KCRV, the degree of equivalence
for a given participant value for a given VOC, d ±U95(d), is estimated as:
      )();(2;KCRVValue)( KCRV;Value 9522 dudRddudUuudud 
where Value and u(Value) are the participant’s reported value and combined uncertainty. Table
11 lists the degrees of equivalence for the four target VOCs.

9 / 24
Table 11. Degrees of Equivalence

benzene o-xylene m-xylene p-xylene
NMI d U95 Rd d U95 Rd d U95 Rd d U95 Rd
BAM 1.0 1.0 2.0 0.3 0.5 1.1 -1.3 1.1 -2.5 -0.3 1.4 -0.4
CENAM -1.1 0.9 -2.5 0.2 0.5 0.7 -1.3 0.6 -4.0 -0.8 1.2 -1.4
INTI -0.2 1.4 -0.3 0.7 0.8 1.6 -1.9 0.8 -4.9 -0.9 0.8 -2.5
KRISS -0.8 1.0 -1.7 0.6 0.5 2.1 -2.0 0.7 -6.0 -1.5 0.8 -4.0
NIST 0.0 1.1 0.0 -0.2 0.6 -0.6 -0.7 0.9 -1.6 0.0 1.0 0.0
NMIJ -0.1 0.8 -0.3 0.7 0.5 2.7 -0.9 0.6 -3.0 -0.1 0.7 -0.4
VNIIM -6.8 1.7 -8.0 3.3 1.0 6.8
VSL 0.4 1.6 0.5 1.0 0.8 2.5 -0.8 0.8 -1.8 0.3 1.0 0.6

Values of Rd less than -2 or greater than +2 indicate results that deviate from the KCRV by more
than is likely given the estimated uncertainties.
Figure 3 provides an approximate graphical display of the degrees of equivalence. Note that in
the panels of this Figure the participant results are listed in alphabetical order, the solid black line
represents the KCRV, and the solid red lines bound the symmetric interval KCRV ±U95(KCRV).

Figure 3. Approximate Degrees of Equivalence

<T
hrm
Ds
rp>
<G
rav
Pre
p>VS
L
VN
IIMNM
IJ
NI
ST
KR
ISSIN
TI
CE
NA
M
BA
M
32
34
36
38
40
42
Be
nz
en
e,
µg
/g
<Th
rm
Ds
rp>
<G
rav
Pre
p>VS
L
VN
IIMNM
IJ
NI
ST
KR
ISSIN
TI
CE
NA
M
BA
M
16
17
18
19
o-X
yle
ne
, µ
g/g

<T
hrm
Ds
rp>
<G
rav
Pre
p>VS
L
<V
NI
IM
>
NM
IJ
NI
ST
KR
ISSIN
TI
CE
NA
M
BA
M
17
18
19
20
21
22
m-
Xy
len
e,
µg
/g
<Th
rm
Ds
rp>
<G
rav
Pre
p>VS
L
<V
NI
IM
>
NM
IJ
NI
ST
KR
ISSIN
TI
CE
NA
M
BA
M
24
26
28
30
32
p-X
yle
ne
, µ
g/g

10 / 24
How Far Does the Light Shine?
Key Comparison CCQM-K47 demonstrated the capabilities of participating NMIs to identify
and measure the four target VOCs benzene, o-xylene, m-xylene, and p-xylene in a calibration
solution using GC-based methods. These measurands were selected to be representative of
VOCs monitored in water quality assessments. The measurement challenges in CCQM-K47,
such as avoiding volatility loss, achieving adequate chromatographic resolution, and isolating
potential interferences, are typical of those required for value-assigning volatile reference
materials. Participants achieving comparable measurements for all four VOCs in this Key
Comparison should be capable of providing reference materials and measurements for VOCs in
solutions when present at concentration levels greater than 10 µg/g.

11 / 24
Appendix A: NMI Reports

BAM


CENAM
Uncertainty was calculated by using ISO-GUM guide 1993 "Quantifying Uncertainty in
Analytical Measurement" standard deviation between ampoules, same ampoule standard
deviation, calibration curve error and uncertainty by solvent purity was combined and the
combined uncertainty multiplied by k=2.


INTI
Uncertainty estimated from the calibration curve (sample variability agrees with the variability of
the calibration curve

12 / 24
KRISS
Source of Uncertainty Relative standard uncertainty (rel %) from each Source
Benzene o-Xylene m-Xylene p-Xylene
Standard Solution (rel%) 0.55% 0.50% 0.56% 0.56%
Purity (rel%) 0.005% 0.005% 0.008% 0.005%
Repeatability of gravimetric dilution
of Stock solution (rel%) 0.52% 0.44% 0.18% 0.18%
Repeatability of gravimetric
preparation of Stock solution 0.04% 0.24% 0.53% 0.53%
Measurement (rel%)
Repeatability of anaylsys(Within sampl 0.1-0.8% 0.1-0.8% 0.1-0.8% 0.1-0.8%
Between sample 0.21-0.22% 0.03-0.15% 0.06-0.28% 0.09-0.32%
Combined Uncertainty 0.79% 0.75% 0.71% 0.74%
Expended Uncertainty(95%) 1.61% 1.52% 1.43% 1.49%

NIST
benzene o-xylene m-xylene p-xylene degrees of freedom
Measured Value (mean) 39.38 15.60 20.19 28.26
Uncertainty Components
Measurement of Samples 0.177 0.070 0.214 0.208 7
Measurement of Calib Stds 0.002 0.001 0.001 0.006 5
purity of compounds used for prep of calibration solutions 0.002 0.003 0.001 0.001 2
Conc of Calib Soln 0.394 0.156 0.202 0.283 inf
Combined Standard Unc. 0.432 0.171 0.294 0.351
Effective degrees of freedom 249 254 25 56
k (from t-distribution) 1.97 1.97 2.06 2.00
Expanded Uncertainty 0.85029 0.34 0.61 0.70
Expanded Unc as % 2.16% 2.16% 3.00% 2.49%

NMIJ
Rsample
Rstd
Mx Weight of sample solution (g)
Mz Weight of standard solution (g)
MYSmp Weight of spike solution (g) added to sample solution
MYStd Weight of spike solution (g) added to standard solution
Ccalib Concentration of calibration solution (ug/g)
Analyte CX (ug/g):
Analyte/D‑analyte ratio for sample blend
Analyte/D‑analyte ratio for standard blend






YStdXstd
YSmpcalibZsample
X MMR
MCMRC =

13 / 24
NMIJ: Benzene
Parameter Source of Uncertainty (∂f/∂xi) (∂f/∂xi)u(xi) Degrees of
Freedom (i)
Type Source of
Data
Pmethod Precision for the entire method
39.26
ug/g
0.02
ug/g
1 0.0160 39 A
Replicate
analysis of
the 4
ampoules
Rsample Measurement of Rsample 1.15 Covered withinPmethod - - - - -
Rstandard Measurement of Rstd 1.05 Covered withinPmethod - - - - -
MZ (L) Balance linearity
0.42
g
0.0003
g
93.477 0.028 Large B
Balance
calibration
certificate
Ccalib
39.3
ug/g
0.09825
ug/g
0.999 0.0982 Large B
Supplier’s
specification
MYsmp (L) Balance linearity
0.2
g
0.0003
g
196.303 0.0589 Large B
Balance
calibration
certificate
MYstd (L) Balance linearity
0.2
g
0.0003
g
-196.303 -0.0589 Large B
Balance
calibration
certificate
MX (L) Balance linearity
0.44
g
0.0003
g
-89.228 -0.0268 Large B
Balance
calibration
certificate
Xi U(xi)

NMIJ: o-xylene
Parameter Source of Uncertainty (∂f/∂xi) (∂f/∂xi)u(xi) Degrees of
Freedom (i)
Type Source of
Data
Pmethod Precision for the entire method
16.42
ug/g
0.01
ug/g
1 0.0128 39 A
Replicate
analysis of
the 4
ampoules
Rsample Measurement of Rsample 1.07 Covered withinPmethod - - - - -
Rstandard Measurement of Rstd 1.08 Covered withinPmethod - - - - -
MZ (L) Balance linearity
0.42
g
0.0003
g
39.095 0.012 Large B
Balance
calibration
certificate
Ccalib
16.6
ug/g
0.0415
ug/g
0.989 0.0410 Large B
Supplier’s
specification
MYsmp (L) Balance linearity
0.2
g
0.0003
g
82.099 0.0246 Large B
Balance
calibration
certificate
MYstd (L) Balance linearity
0.2
g
0.0003
g
-82.099 -0.0246 Large B
Balance
calibration
certificate
MX (L) Balance linearity
0.44
g
0.0003
g
-37.318 -0.0112 Large B
Balance
calibration
certificate
uc= 0.06
Xi U(xi)

14 / 24
NMIJ: m-xylene
Parameter Source of Uncertainty (∂f/∂xi) (∂f/∂xi)u(xi) Degrees of
Freedom (i)
Type Source of
Data
Pmethod Precision for the entire method
19.92
ug/g
0.01
ug/g
1 0.0144 39 A
Replicate
analysis of
the 4
ampoules
Rsample Measurement of Rsample 0.92 Covered withinPmethod - - - - -
Rstandard Measurement of Rstd 0.94 Covered withinPmethod - - - - -
MZ (L) Balance linearity
0.42
g
0.0003
g
47.417 0.014 Large B
Balance
calibration
certificate
Ccalib
20.1
ug/g
0.05025
ug/g
0.991 0.0498 Large B
Supplier’s
specification
MYsmp (L) Balance linearity
0.2
g
0.0003
g
99.575 0.0299 Large B
Balance
calibration
certificate
MYstd (L) Balance linearity
0.2
g
0.0003
g
-99.575 -0.0299 Large B
Balance
calibration
certificate
MX (L) Balance linearity
0.44
g
0.0003
g
-45.262 -0.0136 Large B
Balance
calibration
certificate
uc= 0.07
Xi U(xi)
NMIJ: p-xylene
Parameter Source of Uncertainty (∂f/∂xi) (∂f/∂xi)u(xi) Degrees of
Freedom (i)
Type Source of
Data
Pmethod Precision for the entire method
28.12
ug/g
0.02
ug/g
1 0.0192 39 A
Replicate
analysis of
the 4
ampoules
Rsample Measurement of Rsample 0.97 Covered withinPmethod - - - - -
Rstandard Measurement of Rstd 1 Covered withinPmethod - - - - -
MZ (L) Balance linearity
0.42
g
0.0003
g
66.959 0.020 Large B
Balance
calibration
certificate
Ccalib
28.4
ug/g
0.071
ug/g
0.990 0.0703 Large B
Supplier’s
specification
MYsmp (L) Balance linearity
0.2
g
0.0003
g
140.614 0.0422 Large B
Balance
calibration
certificate
MYstd (L) Balance linearity
0.2
g
0.0003
g
-140.614 -0.0422 Large B
Balance
calibration
certificate
MX (L) Balance linearity
0.44
g
0.0003
g
-63.915 -0.0192 Large B
Balance
calibration
certificate
uc= 0.10
Xi U(xi)

15 / 24
VSL
The content of each ampoule is divided in two aliquotes and placed in an autosampler vial and added of internal standard.
2 series of measurements are carried out (samples + gravimetrically prepared liquid calibrants (BTEX in MeOH))
A weighted line of regression is obtained for each component in each series.
The component average result per series (4 samples analysed twice) and corresponding uncertainty are further calculated
The contributions to the standard uncertainty of each series (U (k=2) between 1-5%) are mainly:
a- calibrants preparation (weighing, purity and potential losses due to evaporation)
b- lack of fit and regression coefficient uncertainties (main uncertainty source)
c- repeatability of the sample measurement
The final value is the mean value of the two series of measurements.
The uncertainty in the final result is the pooled estimate of uncertainty of the series of measurements added of the mean standard deviation of the two series results.
The reported expanded standard uncertainty has a coverage factor of two.

16 / 24
APPENDIX B: Chromatograms

BAM
GC-MS, K47 sample

17 / 24
GC-FID K47 sample

min20 20.25 20.5 20.75 21 21.25 21.5 21.75 22
pA
20
25
30
35
40
45
FID2 B, (2006_S\KAMINSKI\008F0801.D)
p-Xylene
m-Xylene
o-Xylene



CENAM

18 / 24

19 / 24

20 / 24
INTI

0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
standard 35 ug/g

0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
sample CCQM-K47

21 / 24
KRISS

22 / 24
NIST
GC-FID for Xylenes

23 / 24
NMIJ

GC/MS for Benzene – Calibration Solution
16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.000
50000
100000
150000
200000
Time-->
Abundance
Ion 78.00 (77.70 to 78.70): MK47007.D
16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.000
10000
20000
30000
40000
50000
60000
Time-->
Abundance
Ion 116.00 (115.70 to 116.70): MK47007.D


GC/MS for Benzene – CCQM-K47 Solution

16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.000
1000
2000
3000
4000
5000
6000
7000
Time-->
Abundance
Ion 78.00 (77.70 to 78.70): MK47008.D
16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00
0
5000
10000
15000
20000
25000
30000
35000
Time-->
Abundance
Ion 116.00 (115.70 to 116.70): MK47008.D

24 / 24
VSL