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
1 / 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.
2 / 24
Table 1. Gravimetric Preparation Mass Fraction and Homogeneity Evaluation of Target VOCs
Analyte benzene o-xylene m-xylene p-xylene
Mass Fraction, μg/g
Value 39.40
U95(Value) 0.39
15.77
0.28
20.87
0.20
28.25
0.28
ANOVA
Fc
Ft
1.83
2.39
1.39
2.39
1.16
2.39
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 BAM
CENAM
INTI KRISS
NIST NMIJ VNIIM VSL
Submitted by Tin Win, Rosemarie Philipp, Katja Kaminski Mauricio Maldonado Torres, Evangelina Camacho Frías, Melina Pérez Urquiza
Adriana Rosso
Dal Ho Kim Michele Schantz K. Ishikawa A.I. Krylov, Y.A. Kustikov Annarita Baldan
Country Germany
México
Argentina Korea USA Japan Russia Netherlands
Email Contact
tin.win@bam.de
meperez@cenam.mx
mariare@inti.gov.ar, cpuglisi@inti.gov.ar byungjoo@kriss.re.kr michele.schantz@nist.gov ishikawa-keiichiro@aist.go.jp lkonop@b10.vniim.ru abaldan@nmi.nl
3 / 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.
NMI
BAM
CENAM INTI
KRISS
NIST NMIJ VNIIM VSL
Table 3. Analytical Methods
Method GC/MS: benzene, o-xylene,
m&p-xylene (coelution) GC-FID: m/p-xylene ratio GC/MS GC-FID
GC-FID
GC-FID: xylenes GC/MS: benzene GC/MS GC/MS GC-FID
Phase DB-624
DB-WAX DB-WAX PEG (Supelcowax) PEG (5% phenyl) methyl
polysiloxane DB-WAX DB-VRX DB-WAX HP 5 MS CPWAX 52
Column L, m 30
60 60 60
30
15 60 60 30 60
ID, mm FT, µm
0.32
1.8
0.32
0.5
0.25
0.25
0.32
0.5
1
0.53
2.65
0.45
0.85
0.25
1.4
0.32
0.5
0.25
0.25
0.53
2
Table 4. Sample Mass of CCQM-K47 Solution used for Analysis
NMI BAM CENAM INTI KRISS NIST NMIJ VNIIM VSL
Mass solution used, g 0.05 to 0.09 0.3 0.5 0.2 0.4 0.4 0.5 0.7
NMI BAM
Table 5. Certified Reference Materials Used as Calibrants
Solvent Source Methanol KRISS
Certified Mass Fraction, μg/g benzene o-xylene m-xylene p-xylene 40.6 ±0.6 102.0±1.6 100.2±0.6 100.2±0.6
4 / 24
Table 6. Neat Materials Used to Prepare Calibrants
NMI CENAM INTI INTI KRISS NIST NMIJ NMIJ NMIJ NMIJ VNIIM VNIIM VNIIM VNIIM VSL VSL
Source Aldrich Merck AnalityCals Carlo Erba Sigma-Aldrich Sigma Aldrich NIM CRM 4002-a * NIM CRM 4011-a * NIM CRM 4012-a * NIM CRM 4013-a * SOP 0003-03 STC SOP 0020-03 STC SOP 0015-03 STC SOP 0016-03 STC Aldrich Fluka
benzene 0.9988
>0.99
0.9995 1.00 0.99996
Mass Fraction, mg/g o-xylene m-xylene
0.9915 0.9968
0.988 0.9916 0.9363
0.978 0.9977 1.00
0.9994
0.9981
0.9997
0.997
0.9966
0.999
0.995
0.995
* Deuterated Certified Reference material
p-xylene 0.9967
>0.99 0.9986 0.9924
0.9988
0.999 0.995
Table 7. Calibration Method used, Number of Calibration Levels, and Calibration Range
NMI BAM CENAM INTI KRISS NIST NMIJ VNIIM VSL
Use Internal Internal External External Internal IDMS Internal Internal
Internal standard was added at 2nd of 2 dilutions before injection
during sample preparation to sample (deuterated VOC) before subsample analysis to autosampler vials
benzene # mg/g 5 0.3 - 4.8 5 1 - 60 4 3.5 - 53 1 1
1 6 4 - 60
o-xylene # mg/g 5 0.7 - 12.0 5 1 - 60 3 3.5 - 53 1 1
m,p-xylene # mg/g 5 0.7 - 12.0 5 1 - 60 3 3.5 - 53 1 1
1
1
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.
5 / 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
o-Xylene, µg/g
16 17 18 19
Benzene, µg/g
32 34 36 38 40 42 VNIIM CENAM KRISS
INTI NMIJ NIST VSL BAM
<GravPrep>
NIST CENAM
BAM KRISS
INTI NMIJ VSL VNIIM
<GravPrep>
30 32
28
p-Xylene, µg/g
m-Xylene, µg/g
17 18 19 20 21 22 KRISS
INTI BAM CENAM NMIJ VSL NIST
26
<GravPrep>
24 KRISS
INTI CENAM
BAM NMIJ NIST VSL
<GravPrep>
Additional information on the methods and uncertainty calculations used by the participants is included in Appendix A. Representative chromatograms are provided in Appendix B.
6 / 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
Value U95 39.75 1.03
o-xylene
Value U95 17.06 0.42
m-xylene
Value U95 20.47 0.53
p-xylene
Value U95 28.71 0.72
toluene
Value U95 20.80 0.53
Figure 2. Dot-and-Bar Plots for the Four target VOCs with VSL Thermal Desorption Results
o-Xylene, µg/g
16 17 18 19
Benzene, µg/g
32 34 36 38 40 42 VNIIM CENAM KRISS
INTI NMIJ NIST VSL BAM
<GravPrep> <ThrmDsrp>
NIST CENAM
BAM KRISS
INTI NMIJ VSL VNIIM
<GravPrep> <ThrmDsrp>
30 32
28
p-Xylene, µg/g
m-Xylene, µg/g
17 18 19 20 21 22 KRISS
INTI BAM CENAM NMIJ VSL NIST
26
<GravPrep> <ThrmDsrp>
24 KRISS
INTI CENAM
BAM NMIJ NIST VSL
<GravPrep> <ThrmDsrp>
7 / 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:
uKCRV u 2 GravPrep s 2 Consensus . sConsensus MADe n
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:
U 95 KCRV 2 uKCRV .
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 benzene o-xylene m-xylene p-xylene
KCRV 39.4 15.8 20.9 28.3
u(KCRV) 0.4 0.2 0.3 0.3
U95(KCRV) 0.8 0.5 0.6 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:
d Value KCRV; u(d ) u 2 Value u 2 KCRV; U 95 d 2 u(d ); Rd d u(d )
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.
8 / 24
NMI BAM
CENAM INTI
KRISS NIST NMIJ
VNIIM VSL
Table 11. Degrees of Equivalence
benzene
d U95 Rd 1.0 1.0 2.0 -1.1 0.9 -2.5 -0.2 1.4 -0.3 -0.8 1.0 -1.7 0.0 1.1 0.0 -0.1 0.8 -0.3 -6.8 1.7 -8.0 0.4 1.6 0.5
o-xylene
d U95 Rd 0.3 0.5 1.1 0.2 0.5 0.7 0.7 0.8 1.6 0.6 0.5 2.1 -0.2 0.6 -0.6 0.7 0.5 2.7 3.3 1.0 6.8 1.0 0.8 2.5
m-xylene d U95 Rd -1.3 1.1 -2.5 -1.3 0.6 -4.0 -1.9 0.8 -4.9 -2.0 0.7 -6.0 -0.7 0.9 -1.6 -0.9 0.6 -3.0
-0.8 0.8 -1.8
p-xylene d U95 Rd -0.3 1.4 -0.4 -0.8 1.2 -1.4 -0.9 0.8 -2.5 -1.5 0.8 -4.0 0.0 1.0 0.0 -0.1 0.7 -0.4
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
o-Xylene, µg/g
16 17 18 19
Benzene, µg/g
32 34 36 38 40 42 BAM CENAM INTI KRISS NIST NMIJ VNIIM VSL
<GravPrep> <ThrmDsrp>
BAM CENAM
INTI KRISS
NIST NMIJ VNIIM VSL
<GravPrep> <ThrmDsrp>
30 32
28
p-Xylene, µg/g
m-Xylene, µg/g
17 18 19 20 21 22 BAM CENAM INTI KRISS NIST NMIJ <VNIIM> VSL
26
<GravPrep> <ThrmDsrp>
24 BAM CENAM INTI KRISS NIST NMIJ <VNIIM> VSL
<GravPrep> <ThrmDsrp>
9 / 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.
10 / 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
11 / 24
KRISS
Source of Uncertainty
Relative standard uncertainty (rel %) from each Source
Benzene
o-Xylene
m-Xylene
Standard Solution (rel%) Purity (rel%) Repeatability of gravimetric dilution of Stock solution (rel%)
0.55% 0.005%
0.52%
0.50% 0.005%
0.44%
0.56% 0.008%
0.18%
Repeatability of gravimetric preparation of Stock solution
0.04%
0.24%
0.53%
Measurement (rel%)
Repeatability of anaylsys(Within sampl
0.1-0.8%
0.1-0.8%
0.1-0.8%
Between sample Combined Uncertainty Expended Uncertainty(95%)
0.21-0.22% 0.79% 1.61%
0.03-0.15% 0.75% 1.52%
0.06-0.28% 0.71% 1.43%
p-Xylene 0.56% 0.005% 0.18%
0.53%
0.1-0.8% 0.09-0.32%
0.74% 1.49%
NIST
Measured Value (mean)
Uncertainty Components Measurement of Samples Measurement of Calib Stds purity of compounds used for prep of calibration solutions Conc of Calib Soln Combined Standard Unc. Effective degrees of freedom k (from t-distribution) Expanded Uncertainty Expanded Unc as %
benzene 39.38
o-xylene 15.60
m-xylene 20.19
p-xylene 28.26
degrees of freedom
0.177
0.070
0.214
0.208
7
0.002
0.001
0.001
0.006
5
0.002
0.003
0.001
0.001
2
0.394
0.156
0.202
0.283
inf
0.432
0.171
0.294
0.351
249
254
25
56
1.97
1.97
2.06
2.00
0.85029
0.34
0.61
0.70
2.16%
2.16%
3.00%
2.49%
NMIJ
Analyte CX (ug/g):
Rsample Rstd Mx Mz MYSmp MYStd Ccalib
CX
=
R sample R std
M
Z C calib M X M
M
YStd
YSmp
Analyte/D‑analyte ratio for sample blend Analyte/D‑analyte ratio for standard blend
Weight of sample solution (g) Weight of standard solution (g)
Weight of spike solution (g) added to sample solution Weight of spike solution (g) added to standard solution
Concentration of calibration solution (ug/g)
12 / 24
NMIJ: Benzene
Parameter Pmethod
Rsample Rstandard MZ (L) Ccalib MYsmp (L) MYstd (L) MX (L)
NMIJ: o-xylene
Parameter Pmethod
Rsample Rstandard MZ (L) Ccalib MYsmp (L) MYstd (L) MX (L)
Source of Uncertainty Precision for the entire method Measurement of Rsample Measurement of Rstd Balance linearity
Balance linearity Balance linearity Balance linearity
Source of Uncertainty Precision for the entire method Measurement of Rsample Measurement of Rstd Balance linearity
Balance linearity Balance linearity Balance linearity
Xi ug/g
39.26
1.15 1.05
g 0.42
ug/g 39.3
g 0.2
g 0.2
g 0.44
Xi ug/g
16.42
1.07
1.08 g
0.42 ug/g
16.6
g 0.2
g 0.2
g 0.44
U(xi)
ug/g 0.02
Covered Pmethod Covered Pmethod
within
within
g 0.0003
ug/g 0.09825
g 0.0003
g 0.0003
g 0.0003
U(xi)
ug/g 0.01
Covered Pmethod
Covered Pmethod
within
within
g 0.0003
ug/g 0.0415
g 0.0003
g 0.0003
g 0.0003
(∂f/∂xi)
(∂f/∂xi)u(xi)
Degrees of Freedom (i)
1
0.0160
39
-
93.477
-
0.028
Large
0.999
0.0982
Large
196.303
0.0589
Large
-196.303
-0.0589
Large
-89.228
-0.0268
Large
Type
A B B B B B
Source of
Data
Replicate
analysis of
the
4
ampoules
-
-
Balance calibration certificate Supplier’s specification
Balance calibration certificate Balance calibration certificate Balance calibration certificate
(∂f/∂xi)
(∂f/∂xi)u(xi)
Degrees of Freedom (i)
1
0.0128
39
-
39.095
-
0.012
Large
0.989
0.0410
Large
82.099
0.0246
Large
-82.099
-0.0246
Large
-37.318
-0.0112
Large
uc=
0.06
Type A -
Source of
Data
Replicate
analysis of
the
4
ampoules
-
-
-
Balance
B
calibration
certificate
Supplier’s
B
specification
Balance
B
calibration
certificate
Balance
B
calibration
certificate
Balance
B
calibration
certificate
13 / 24
NMIJ: m-xylene
Parameter Pmethod
Rsample Rstandard MZ (L) Ccalib MYsmp (L) MYstd (L) MX (L)
NMIJ: p-xylene
Parameter Pmethod
Rsample Rstandard MZ (L) Ccalib MYsmp (L) MYstd (L) MX (L)
Source of Uncertainty Precision for the entire method Measurement of Rsample Measurement of Rstd Balance linearity
Balance linearity Balance linearity Balance linearity
Source of Uncertainty Precision for the entire method Measurement of Rsample Measurement of Rstd Balance linearity
Balance linearity Balance linearity Balance linearity
Xi ug/g
19.92
0.92 0.94
g 0.42
ug/g 20.1
g 0.2
g 0.2
g 0.44
Xi ug/g
28.12
0.97
1 g
0.42 ug/g
28.4
g 0.2
g 0.2
g 0.44
U(xi)
ug/g 0.01
Covered Pmethod Covered Pmethod
within
within
g 0.0003
ug/g 0.05025
g 0.0003
g 0.0003
g 0.0003
U(xi)
ug/g 0.02
Covered Pmethod Covered Pmethod
within
within
g 0.0003
ug/g 0.071
g 0.0003
g 0.0003
g 0.0003
(∂f/∂xi)
(∂f/∂xi)u(xi)
Degrees of Freedom (i)
1
0.0144
39
-
47.417
-
0.014
Large
0.991
0.0498
Large
99.575
0.0299
Large
-99.575
-0.0299
Large
-45.262
-0.0136
Large
uc=
0.07
Type
A B B B B B
Source of
Data
Replicate
analysis of
the
4
ampoules
-
-
Balance calibration certificate Supplier’s specification
Balance calibration certificate Balance calibration certificate Balance calibration certificate
(∂f/∂xi)
(∂f/∂xi)u(xi)
Degrees of Freedom (i)
1
0.0192
39
-
66.959
-
0.020
Large
0.990
0.0703
Large
140.614
0.0422
Large
-140.614
-0.0422
Large
-63.915
-0.0192
Large
uc=
0.10
Type
A B B B B B
Source of
Data
Replicate
analysis of
the
4
ampoules
-
-
Balance calibration certificate Supplier’s specification
Balance calibration certificate Balance calibration certificate Balance calibration certificate
14 / 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.
15 / 24
APPENDIX B: Chromatograms
BAM
GC-MS, K47 sample
16 / 24
GC-FID K47 sample
FID2 B, (2006\_S\\KAMINSKI\\008F0801.D) pA
45
p-Xylene
40
m-Xylene
35
o-Xylene
30
25
20
20
20.25
20.5
20.75
21
21.25
21.5
21.75
22
min
CENAM
17 / 24
18 / 24
19 / 24
INTI
28 standard 35 ug/g
26
24
22
20
18
16
14
12
10
8
6
4
2
0
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
28 sample CCQM-K47
26
24
22
20
18
16
14
12
10
8
6
4
2
0
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
20 / 24
KRISS 21 / 24
NIST
GC-FID for Xylenes
22 / 24
NMIJ
Abundance
GC/MS for Benzene – Calibration Solution
Ion 78.00 (77.70 to 78.70): MK47007.D
200000 150000 100000
50000 0
Time--> Abundance
16.00
18.00
20.00
22.00
24.00
26.00
28.00
30.00
32.00
34.00
Ion 116.00 (115.70 to 116.70): MK47007.D
60000 50000 40000 30000 20000 10000
0
Time-->
16.00
18.00
20.00
22.00
24.00
26.00
28.00
30.00
32.00
34.00
GC/MS for Benzene – CCQM-K47 Solution
Abundance
Ion 78.00 (77.70 to 78.70): MK47008.D
7000 6000 5000 4000 3000 2000 1000
0
Time--> Abundance
16.00
18.00
20.00
22.00
24.00
26.00
28.00
30.00
32.00
34.00
35000
30000
25000
20000
15000
10000
5000
Time-->
0 16.00
18.00
Ion 116.00 (115.70 to 116.70): MK47008.D 20.00 22.00 24.00 26.00 28.00 30.00
32.00
34.00
23 / 24
VSL 24 / 24
Ver+/-