ASTM D5917 Test Method for Trace Impurities in Monocyclic Aromatic Hydrocarbons
ASTM D5917 Standard Test Method for Trace Impurities in Monocyclic Aromatic Hydrocarbons by Gas Chromatography and External Calibration
12. Calibration
12.1 Prepare a synthetic mixture of high purity p-xylene containing impurities at concentrations representative of those expected in the samples to be analyzed. The volume of each hydrocarbon impurity must be measured to the nearest 1 µL and all liquid reference compounds must be brought to the same temperature before mixing. Refer to Table 2 for an example of a calibration blend. n-Nonane will represent the nonaromatic fraction, o-xylene the o-xylene fraction, m-xylene the m-xylene fraction. Cumene will represent the aromatic hydrocarbons containing nine carbon atoms or greater, with exception of PDEB. If PDEB is included in the calibration, PDEB will represent PDEB.
12.1.1 Prior to preparing the calibration standard, all reference compounds and any samples to be analyzed must be brought to the same temperature, preferably 25°C.

12.2 Using the exact volumes and densities in Table 2, calculate the weight % concentration for each impurity in the calibration blend as follows:
Ci = 100 DiVi/(VtDp)
where:
Di = density of impurity i from Table 2,
Vi = volume of impurity i, mL,
Dp = density of p-xylene from Table 2,
Vt = total volume of standard blend, mL, and
Ci = concentration of impurity i, weight %.

12.2.1 Alternatively, calibration standards may be used that have been gravimetrically prepared in accordance with Practice D4307.

12.3 Inject the resulting solution from 12.1 into the chromatograph, collect and process the data. A typical chromatogram is illustrated in Fig. 1.

12.4 Determine the response factor for each impurity in the calibration mixture as follows:
RFi = Ci/Ai
where:
RFi = response factor for impurity i,
Ai = peak area of impurity i, and
Ci = concentration of impurity i, as calculated in 12.2, weight %.

12.5 Analyze the calibration solution(s) a minimum of three times and calculate an average RF.

12.6 Determine the sample standard deviation for RF of each impurity using a scientific calculator or spreadsheet program. Determine the coefficient of variation for each RFas follows:
CVi = 100SDi/Avgi
where:
CVi = coefficient of variation for RFi,
SDi = standard deviation for RFi, and
Avgi = average RF of impurity i.

12.7 The coefficient of variation for the response factor of any impurity, as calculated from a minimum of three successive analyses of the standard, shall not exceed 10 %.