Selection of cylinder type for preparation
As VOCs may adsorb, desorb or react with each other in the active site of the cylinder inner wall, it is necessary to select the suitable cylinder before preparation. Imported treated cylinders, domestic treated cylinders and domestic non-treated cylinders were included in this research. The cylinder containing 30VOCs CRMs was prepared and stored as the mother cylinder while the cleaned and evacuated cylinder was used as sub-cylinder. The 30VOCs CRMs was transferred from the mother cylinder to the sub-cylinder by pre-treated pipeline as short as possible until the pressure balanced between the two cylinders. After waiting for 7 days, the 30VOCs CRMs in the mother cylinder and sub-cylinder were determined by GC-FID and the relative deviations(E) of the response values were calculated to evaluate the applicability of sub-cylinder. Consider the high cost and long transportation time of the imported treated cylinders, only domestic treated and non-treated cylinders were selected as sub-cylinders.
Relative deviations(E) of domestic treated cylinders (A-type) and non-treated cylinders (B-type) were listed in Table 3, only one set of test data was listed considering that others have shown the same trend. The relative deviations of A-type were less than 1.2% for all components which were closed to the repeatability of analytical method. The relative deviations of B-type showed that the components with heavier molecular mass, such as n-propylbenzene, 1,3,5-trimethylene and 1,2,4-trimethylene were more than 2% which indicated adsorption reaction may occur in the inner wall. Therefore, only A-type can be used to prepare 30 VOCs CRMs.
Table 3
The relative deviations between A and B at 48 hours and 30 days
Component | E/% | Component | E/% |
A | B | A | B |
ethylene | -0.3 | -0.2 | n-hexane | -0.5 | -0.9 |
propane | -0.6 | -0.2 | 3-methyhexane | -1.0 | -0.7 |
propylene | -0.3 | -0.2 | methylcyclohexane | -0.9 | -0.8 |
isobutane | -0.4 | -0.4 | n-heptane | -1.1 | -0.8 |
n-butane | -0.5 | -0.5 | benzene | -0.8 | -0.8 |
1-butene | -0.3 | -0.6 | n-octane | -0.9 | -1.1 |
tran-2-butene | -0.5 | -0.4 | methylbenzene | -0.8 | -1.0 |
1,3-butadiene | -0.4 | -0.4 | n-nonane | -0.7 | -1.5 |
cis-2-butene | -0.5 | -0.4 | ethylbenzene | -0.9 | -1.4 |
2-methybutane | -0.5 | -0.4 | styrene | -0.5 | -1.8 |
n-pentane | -0.6 | -0.4 | p-xylene | -1.1 | -1.7 |
1-pentene | -0.6 | -0.3 | o-xylene | -0.8 | -1.9 |
cis-2-pentene | -0.6 | -0.6 | n-propylbenzene | -1.2 | -2.3 |
2-methypentane | -0.7 | -0.3 | 1,2,4-trimethylbenzene | -1.1 | -2.6 |
isoprene | -0.4 | -0.8 | 1,3,5-trimethylbenzene | -0.8 | -2.9 |
Within-cylinder homogeneity test
In the view of that VOCs with high molecular mass may be stratified in the cylinder which may cause the decrease or increase of their concentration under different pressure, within-cylinder homogeneity was carried out to study the stable of concentration while the pressure decreased. 30VOCs CRMs were prepared with a pressure of 10 MPa and determined three times at the pressure of 10, 8, 6, 4, 2 and 1 MPa.
F-test was applied to evaluate the within-cylinder homogeneity of all the components and the uncertainty of within-cylinder homogeneity\({(u}_{bb})\) was computed according to Eq. 1–4. The results were displayed in Table 4.
$$\stackrel{̿}{X}=\frac{\sum _{j=1}^{m}{\stackrel{-}{X}}_{j}}{m}$$
1
$${MS}_{among}=\frac{n\sum _{j=1}^{m}{\left({\stackrel{-}{X}}_{j}-\stackrel{̿}{X}\right)}^{2}}{(m-1)}$$
2
$${MS}_{within}=\frac{\sum _{j=1}^{m}\sum _{i=1}^{n}{\left({X}_{ij}-{\stackrel{-}{X}}_{j}\right)}^{2}}{(N-m)}$$
3
$${u}_{bb}=\sqrt{\frac{{MS}_{among}-{MS}_{within}}{n}}$$
4
Where m is the number of pressure tested setting, n is the determine time under the same tested pressure, \({\stackrel{-}{X}}_{j}\) is the concentration at each tested pressure, and N is the number of all determinations.
The concentrations of most components were stable in the within-cylinder homogeneity test, expect that styrene, p-xylene, o-xylene, n-propylbenzene, 1,3,5-trimethylene and 1,2,4-trimethylene were increased significantly when the pressure decreased to 1 MPa. Therefore, the concentrations of 30 VOCs CRMs can remain stable above 2 MPa and the uncertainty caused by with-in homogeneity was range from 0.23–1.13%.
Table 4
With-in homogeneity of 30 VOCs CRMs
Component | \({\text{u}}_{\text{bb,rel}}\)/% |
30903093# | 11051007# | 11051015# |
1 MPa | 2 MPa | 1 MPa | 2 MPa | 1 MPa | 2 MPa |
ethylene | 0.14 | 0.15 | 0.26 | 0.18 | 0.31 | 0.23 |
propane | 0.00 | 0.17 | 0.26 | 0.21 | 0.38 | 0.44 |
propylene | 0.33 | 0.39 | 0.32 | 0.29 | 0.45 | 0.43 |
isobutane | 0.32 | 0.33 | 0.39 | 0.35 | 0.54 | 0.47 |
n-butane | 0.43 | 0.37 | 0.27 | 0.21 | 0.57 | 0.44 |
1-butene | 0.00 | 0.00 | 0.37 | 0.36 | 0.51 | 0.48 |
tran-2-butene | 0.30 | 0.26 | 0.49 | 0.48 | 0.49 | 0.46 |
1,3-butadiene | 0.19 | 0.21 | 0.41 | 0.38 | 0.52 | 0.48 |
cis-2-butene | 0.21 | 0.23 | 0.34 | 0.34 | 0.43 | 0.43 |
2-methybutane | 0.43 | 0.46 | 0.61 | 0.55 | 0.47 | 0.34 |
n-pentane | 0.10 | 0.17 | 0.37 | 0.31 | 0.44 | 0.41 |
1-pentene | 0.25 | 0.30 | 0.60 | 0.57 | 0.48 | 0.45 |
cis-2-pentene | 0.17 | 0.19 | 0.17 | 0.04 | 0.37 | 0.36 |
2-methypentane | 0.39 | 0.14 | 0.34 | 0.31 | 0.74 | 0.68 |
isoprene | 0.00 | 0.00 | 0.40 | 0.37 | 0.89 | 0.75 |
n-hexane | 0.28 | 0.22 | 0.47 | 0.44 | 0.51 | 0.41 |
3-methyhexane | 0.00 | 0.00 | 0.00 | 0.00 | 0.49 | 0.42 |
methylcyclohexane | 0.34 | 0.23 | 0.41 | 0.20 | 0.54 | 0.46 |
n-heptane | 0.36 | 0.05 | 0.40 | 0.30 | 0.40 | 0.28 |
benzene | 0.49 | 0.31 | 0.46 | 0.28 | 0.51 | 0.43 |
n-octane | 0.63 | 0.26 | 0.75 | 0.37 | 0.65 | 0.49 |
methylbenzene | 0.73 | 0.27 | 0.80 | 0.44 | 0.72 | 0.55 |
n-nonane | 1.30 | 0.40 | 1.35 | 0.62 | 1.01 | 0.63 |
ethylbenzene | 1.22 | 0.44 | 1.16 | 0.55 | 1.01 | 0.65 |
styrene | 2.37 | 0.71 | 1.86 | 0.78 | 1.54 | 0.89 |
p-xylene | 1.54 | 0.56 | 1.43 | 0.54 | 1.14 | 0.73 |
o-xylene | 1.49 | 0.51 | 1.37 | 0.43 | 1.14 | 0.68 |
n-propylbenzene | 2.11 | 0.61 | 2.12 | 0.66 | 1.42 | 0.81 |
1,2,4-trimethylbenzene | 2.65 | 0.50 | 2.36 | 0.57 | 1.73 | 0.91 |
1,3,5-trimethylbenzene | 3.38 | 0.78 | 2.97 | 0.91 | 2.02 | 1.13 |
Long-term Stability test
Long-term stability is an important factor to monitor the change of concentration with storage time. According to ISO Guide 35, 30VOCs CRMs were measured by GC-FID at different monitor time (0, 3, 6, 9, 12, 16 months).
T-test was used to determine whether there is a significant difference in concentration over monitor time. And then, a linear curve of monitor time and concentration was fitted as Eq. 5, where X denotes time and Y denotes the concentration. The uncertainty caused by long-term stability\({(u}_{lts})\) were calculated by Equations 6–9 and shown in Table 5.
$$\text{Y=}{\text{b}}_{\text{1}}\text{∙X+}{\text{b}}_{\text{0}}$$
5
$${\text{b}}_{\text{1}}\text{=}\frac{\sum _{\text{i=1}}^{\text{n}}\left({\text{X}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{X}}\right)\left(\text{Y-}\stackrel{\text{-}}{\text{Y}}\right)}{\sum _{\text{i=1}}^{\text{n}}{\left({\text{X}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{X}}\right)}^{\text{2}}}$$
6
$$\text{s=}\sqrt{\frac{\sum _{\text{i=1}}^{\text{n}}{\left({\text{Y}}_{\text{i}}\text{-}{\text{b}}_{\text{0}}\text{-}{\text{b}}_{\text{i}}{\text{X}}_{\text{i}}\right)}^{\text{2}}}{\text{n-2}}}$$
7
$$\text{s}\left({\text{b}}_{\text{1}}\right)\text{=}\frac{\text{s}}{\sqrt{\sum _{\text{i=1}}^{\text{n}}{\left({\text{X}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{X}}\right)}^{\text{2}}}}$$
8
\({\text{u}}_{\text{lts}}\text{=s}\left({\text{b}}_{\text{1}}\right)\text{*t}\) , where t is the shelf life (9)
Table 5
Long-term stability result of 30 VOCs CRMs
Component | b1 | b0 | s(b1) | \({t}_{0.95,n-2}\)*s(b1) | ults |
ethylene | 3.86E-04 | 1.12 | 1.07E-03 | 2.97E-03 | 1.3 |
propane | -2.25E-04 | 1.13 | 7.84E-04 | 2.18E-03 | 1.0 |
propylene | 4.79E-04 | 1.12 | 1.01E-03 | 2.82E-03 | 1.3 |
isobutane | -8.77E-04 | 1.10 | 1.47E-03 | 4.09E-03 | 1.9 |
n-butane | 5.21E-04 | 1.14 | 1.02E-03 | 2.84E-03 | 1.3 |
1-butene | 1.31E-03 | 1.07 | 1.54E-03 | 4.27E-03 | 2.0 |
tran-2-butene | 3.01E-04 | 1.10 | 9.93E-04 | 2.76E-03 | 1.3 |
1,3-butadiene | 8.90E-05 | 1.08 | 1.03E-03 | 2.87E-03 | 1.3 |
cis-2-butene | 1.65E-04 | 1.15 | 9.54E-04 | 2.65E-03 | 1.2 |
2-methybutane | -5.85E-04 | 1.12 | 8.53E-04 | 2.37E-03 | 1.1 |
n-pentane | -5.59E-04 | 1.15 | 8.74E-04 | 2.43E-03 | 1.1 |
1-pentene | 1.48E-04 | 1.16 | 1.28E-03 | 3.56E-03 | 1.5 |
cis-2-pentene | 4.87E-04 | 1.08 | 1.06E-03 | 2.96E-03 | 1.4 |
2-methypentane | 2.01E-03 | 1.56 | 2.16E-03 | 6.00E-03 | 1.9 |
isoprene | 4.07E-04 | 1.09 | 7.28E-04 | 2.02E-03 | 0.9 |
n-hexane | -9.87E-04 | 1.41 | 2.05E-03 | 5.71E-03 | 2.1 |
3-methyhexane | -1.14E-04 | 1.33 | 1.45E-03 | 4.04E-03 | 1.5 |
methylcyclohexane | -1.02E-03 | 1.40 | 1.35E-03 | 3.75E-03 | 1.4 |
n-heptane | -1.59E-03 | 1.41 | 1.33E-03 | 3.70E-03 | 1.3 |
benzene | 2.22E-03 | 1.37 | 1.97E-03 | 5.48E-03 | 2.0 |
n-octane | -1.13E-03 | 1.34 | 9.66E-04 | 2.69E-03 | 1.0 |
methylbenzene | -4.96E-04 | 1.29 | 1.28E-03 | 3.56E-03 | 1.4 |
n-nonane | -4.45E-04 | 1.42 | 1.22E-03 | 3.38E-03 | 1.2 |
ethylbenzene | -1.78E-03 | 1.30 | 1.52E-03 | 4.24E-03 | 1.7 |
styrene | -2.53E-03 | 1.39 | 1.62E-03 | 4.49E-03 | 1.7 |
p-xylene | -2.39E-03 | 1.42 | 1.94E-03 | 5.38E-03 | 1.9 |
o-xylene | -1.79E-03 | 1.32 | 1.46E-03 | 4.05E-03 | 1.6 |
n-propylbenzene | -1.47E-03 | 1.14 | 1.35E-03 | 3.75E-03 | 1.7 |
1,2,4-trimethylbenzene | -1.50E-03 | 1.28 | 1.63E-03 | 4.54E-03 | 1.8 |
1,3,5-trimethylbenzene | -2.31E-03 | 1.40 | 1.34E-03 | 3.72E-03 | 1.4 |
As a consequence, the concentrations of 30VOCs CRMs have no significant trend within 16 months which indicate it can main stable for at least 16 months and the uncertainty caused by long-term stability\({(u}_{lts,rel})\) ranged from 1.0 to 2.1%.
Short-term Stability test
The short-term stability test was carried out to determine whether the extreme low temperature during transportation affected the concentrations of 30 VOCs CRMs. The freezing operation procedure of 30VOCs CRMs was: frozen to -20℃ for 24 hours and then returned to room temperature. The 30VOCs was determined by GC-FID before and after freezing operation procedure and the relative deviation (E) was calculated to estimate the short-term stability. The results were shown in Table 6 and the E is close to \({u}_{lts}\), which indicates that the low temperature during transportation has no effect on 30VOCs CRMs.
Table 6
The result of low temperature freezing test
Component | C/(µmol/mol) | E/% | Component | C/(µmol/mol) | E/% |
Before freezing | After freezing | Before freezing | After freezing |
ethylene | 1.192 | 1.188 | 0.3% | n-hexane | 1.503 | 1.521 | -1.2% |
propane | 1.199 | 1.200 | -0.1% | 3-methyhexane | 1.426 | 1.416 | 0.7% |
propylene | 1.187 | 1.207 | -1.7% | methylcyclohexane | 1.499 | 1.483 | 1.1% |
isobutane | 1.155 | 1.158 | -0.3% | n-heptane | 1.499 | 1.479 | 1.4% |
n-butane | 1.208 | 1.213 | -0.4% | benzene | 1.475 | 1.481 | -0.4% |
1-butene | 1.140 | 1.118 | 2.0% | n-octane | 1.405 | 1.416 | -0.8% |
tran-2-butene | 1.168 | 1.160 | 0.7% | methylbenzene | 1.355 | 1.367 | -0.9% |
1,3-butadiene | 1.153 | 1.134 | 1.7% | n-nonane | 1.496 | 1.508 | -0.8% |
cis-2-butene | 1.222 | 1.213 | 0.7% | ethylbenzene | 1.352 | 1.348 | 0.3% |
2-methybutane | 1.195 | 1.201 | -0.5% | styrene | 1.440 | 1.449 | -0.6% |
n-pentane | 1.392 | 1.386 | 0.4% | p-xylene | 1.475 | 1.501 | -1.7% |
1-pentene | 1.123 | 1.133 | -0.9% | o-xylene | 1.381 | 1.398 | -1.2% |
cis-2-pentene | 1.229 | 1.227 | 0.2% | n-propylbenzene | 1.178 | 1.186 | -0.7% |
2-methypentane | 1.745 | 1.718 | 1.6% | 1,2,4-trimethylbenzene | 1.325 | 1.336 | -0.8% |
isoprene | 1.226 | 1.233 | -0.6% | 1,3,5-trimethylbenzene | 1.446 | 1.474 | -1.9% |
Uncertainty evaluation of 30VOCs CRMs
The certified value of 30VOCs CRMs was calculated through gravimetric method which was recommended by ISO 35 and its uncertainty mainly comes from: preparation, within-cylinder homogeneity and long-term stability. The preparation uncertainty (\({\text{u}}_{\text{prep}}\)) was evaluated by identifying, evaluating and synthesizing all the uncertainty factors related to preparation which was shown in Fig. 3
The uncertainty of 30VOCs CRMs (\({u}_{CRM,grav}\)) was certified by Eq. 10 and the relative expanded uncertainty Ugrav (where k = 2) was 2.2%-4.1%. The results were all listed in Table 7.
$${u}_{CRM,grav}=\sqrt{{u}_{prep}^{2}+{u}_{lts}^{2}+{u}_{bb}^{2}}$$
10
Table 7
Uncertainty of 30 VOCs CRMs
Component | \({\text{u}}_{\text{prep,rel}}\)/% | \({\text{u}}_{\text{bb}}\)/% | \({\text{u}}_{\text{lts}}\)/% | \({\text{U}}_{\text{grav}}\)/% |
ethylene | 0.2 | 0.2 | 1.3 | 2.7 |
propane | 0.2 | 0.4 | 1.0 | 2.2 |
propylene | 0.2 | 0.4 | 1.3 | 2.7 |
isobutane | 0.2 | 0.5 | 1.9 | 3.9 |
n-butane | 0.2 | 0.4 | 1.3 | 2.7 |
1-butene | 0.7 | 0.5 | 2.0 | 4.3 |
tran-2-butene | 0.7 | 0.5 | 1.3 | 3.0 |
1,3-butadiene | 0.6 | 0.5 | 1.3 | 3.1 |
cis-2-butene | 0.6 | 0.4 | 1.2 | 2.8 |
2-methybutane | 0.3 | 0.3 | 1.1 | 2.3 |
n-pentane | 0.3 | 0.4 | 1.1 | 2.4 |
1-pentene | 0.4 | 0.5 | 1.5 | 3.3 |
cis-2-pentene | 0.4 | 0.4 | 1.4 | 2.9 |
2-methypentane | 0.3 | 0.7 | 1.9 | 4.1 |
isoprene | 0.4 | 0.8 | 0.9 | 2.5 |
n-hexane | 0.4 | 0.4 | 2.1 | 4.3 |
3-methyhexane | 0.4 | 0.4 | 1.5 | 3.3 |
methylcyclohexane | 0.4 | 0.5 | 1.4 | 3.0 |
n-heptane | 0.4 | 0.3 | 1.3 | 2.9 |
benzene | 0.4 | 0.4 | 2.0 | 4.2 |
n-octane | 0.4 | 0.5 | 1.0 | 2.4 |
methylbenzene | 0.4 | 0.6 | 1.4 | 3.1 |
n-nonane | 0.4 | 0.6 | 1.2 | 2.8 |
ethylbenzene | 0.4 | 0.7 | 1.7 | 3.7 |
styrene | 0.4 | 0.9 | 1.7 | 3.8 |
p-xylene | 0.5 | 0.7 | 1.9 | 4.3 |
o-xylene | 0.4 | 0.7 | 1.6 | 3.5 |
n-propylbenzene | 0.4 | 0.8 | 1.7 | 3.8 |
1,2,4-trimethylbenzene | 0.4 | 0.9 | 1.8 | 4.1 |
1,3,5-trimethylbenzene | 0.6 | 1.1 | 1.4 | 3.7 |
Comparison of certified value
The 30VOCs CRMs were compared with other related gases CRMs by Linde Spectra Environment Gases and Air Liquid Company to verify its accuracy of certified value. The determination data which can be confirmed the consistency of certified value was evaluated by Formula 11.
$$\left|{\text{x}}_{\text{CRM}}\text{-}{\text{x}}_{\text{meas}}\right|\text{≤}\text{2}\sqrt{{\text{u}}_{\text{CRM}}^{\text{2}}\text{+}{\text{u}}_{\text{meas}}^{\text{2}}}$$
11
Where xCRM is the certified value of 30VOCs CRMs, xmeas is the determined concentration of 30VOCs CRMs, uCRMs is the uncertainty of 30VOCs CRMs and umeas is the uncertainty caused by determination.
The comparison results were shown in Table 8 and the differences between certified value and determined value were less than their uncertainty which illustrated the consistency of 30VOCs CRMs.
Table 8
The results of cross-check
Component | \({\text{x}}_{\text{CRM}}\) | \({\text{u}}_{\text{CRM}}\) | Using Linde standard gas | Using Air Liquid standard gas |
\(\left|{\text{x}}_{\text{CRM}}\text{-}{\text{x}}_{\text{meas}}\right|\) | \(\text{2}\sqrt{{\text{u}}_{\text{CRM}}^{\text{2}}\text{+}{\text{u}}_{\text{meas}}^{\text{2}}}\) | \(\left|{\text{x}}_{\text{CRM}}\text{-}{\text{x}}_{\text{meas}}\right|\) | \(\text{2}\sqrt{{\text{u}}_{\text{CRM}}^{\text{2}}\text{+}{\text{u}}_{\text{meas}}^{\text{2}}}\) |
ethylene | 1.21 | 0.021 | 0.05 | 0.078 | 0.00 | 0.130 |
propane | 1.22 | 0.022 | 0.04 | 0.080 | 0.02 | 0.133 |
propylene | 1.22 | 0.024 | 0.03 | 0.081 | 0.00 | 0.134 |
isobutane | 1.19 | 0.029 | 0.02 | 0.087 | 0.01 | 0.137 |
n-butane | 1.24 | 0.024 | 0.03 | 0.081 | 0.00 | 0.134 |
tran-2-butene | 1.20 | 0.023 | 0.03 | 0.079 | 0.01 | 0.133 |
1-butene | 1.06 | 0.027 | 0.02 | 0.081 | 0.12 | 0.135 |
cis-2-butene | 1.25 | 0.025 | 0.00 | 0.085 | 0.00 | 0.134 |
2-methybutane | 1.21 | 0.021 | 0.01 | 0.079 | 0.01 | 0.133 |
n-pentane | 1.24 | 0.022 | 0.00 | 0.079 | 0.00 | 0.134 |
1-pentene | 1.26 | 0.026 | 0.02 | 0.086 | 0.02 | 0.139 |
cis-2-pentene | 1.17 | 0.023 | 0.00 | 0.083 | 0.02 | 0.136 |
2-methypentane | 1.41 | 0.035 | 0.08 | 0.097 | 0.08 | 0.146 |
n-hexane | 1.27 | 0.032 | 0.01 | 0.093 | 0.00 | 0.143 |
isoprene | 1.18 | 0.020 | 0.00 | 0.079 | 0.01 | 0.134 |
methylcyclohexane | 1.25 | 0.023 | 0.06 | 0.079 | 0.04 | 0.131 |
n-heptane | 1.26 | 0.023 | 0.00 | 0.087 | 0.01 | 0.132 |
benzene | 1.24 | 0.030 | 0.00 | 0.089 | 0.01 | 0.142 |
n-octane | 1.20 | 0.017 | 0.03 | 0.070 | 0.03 | 0.127 |
methylbenzene | 1.16 | 0.021 | 0.05 | 0.081 | 0.01 | 0.131 |
n-nonane | 1.27 | 0.018 | 0.06 | 0.071 | 0.05 | 0.128 |
ethylbenzene | 1.16 | 0.023 | 0.04 | 0.079 | 0.01 | 0.132 |
styrene | 1.24 | 0.025 | 0.06 | 0.080 | 0.00 | 0.130 |
p-xylene | 1.27 | 0.028 | 0.04 | 0.083 | 0.02 | 0.137 |
o-xylene | 1.18 | 0.023 | 0.05 | 0.078 | 0.06 | 0.132 |
n-propylbenzene | 1.02 | 0.020 | 0.01 | 0.070 | 0.02 | 0.129 |
1,2,4-trimethylbenzene | 1.14 | 0.025 | 0.02 | 0.079 | 0.06 | 0.133 |
1,3,5-trimethylbenzene | 1.25 | 0.027 | 0.05 | 0.087 | 0.05 | 0.136 |