ASTM D6591 Standard Test Method for Determination of Aromatic Hydrocarbon Types in Middle Distillates - High Performance Liquid Chromatography Method with Refractive Index Detection
8. Sampling
8.1 Unless otherwise specified in the commodity specification, samples are taken by following Practice D4057 or D4177, or a similar standard. In certain situations, sampling is done in accordance with the requirements of national standards or regulations for the sampling of the product under test.

9. Apparatus Preparation
9.1 Set up the chromatograph, injection system, column, backflush valve, column oven, refractive index detector, and computing integrator in accordance with the appropriate equipment manuals. Install the HPLC column and backflush valve in the column oven. Insert the backflush valve so that the detector is always connected independently of the direction of flow through the column (see Fig. 1). Maintain the sample injection valve at the same temperature as the sample solution; in most cases this will be at room temperature.

NOTE 11 - The column oven is optional if alternative arrangements are made to maintain a constant temperature environment, for example, a temperature-controlled laboratory (see 6.5). It is recommended to install the backflush valve in the column oven and to install the apparatus away from drafts (that is, not near air-conditioning unit or fume cupboard). Pipework and/or valving which is not temperature controlled should be insulated.

NOTE 12 - Regular maintenance of the liquid chromatograph and its components is important to ensure consistent performance. Leakages and partial blockage of filters, frits, injector needles and valve rotors can produce flow rate inconsistencies and poor injector performance.

9.2 Adjust the flow rate of the mobile phase to a constant 1.0 +/- 0.2 mL/min, and ensure the reference cell of the refractive index detector is full of mobile phase. Allow the temperature of the column oven (and refractive index detector, if equipped with temperature control) to stabilize.
9.2.1 To minimize drift, it is essential to make sure the reference cell is full of solvent. The best way to accomplish this is either (1) to flush the mobile phase through the reference cell (then isolate the reference cell to prevent evaporation of the solvent) immediately prior to analysis, or (2) to continuously make up for solvent evaporation by supplying a steady flow through the reference cell. The make-up flow is optimized so that reference and analytical cell miss-match due to drying-out, temperature, or pressure gradients are minimized. Typically, this can be accomplished with a make-up flow set at one tenth of the analytical flow.

NOTE 13 - The flow rate may be adjusted (typically within the range from 0.8 to 1.2 mL/min) to an optimum value in order to meet the resolution requirements specified in 9.4.3.

9.3 Prepare a system performance standard (SPS) by weighing cyclohexane (1.0 +/- 0.1 g), o-xylene (0.5 +/- 0.05 g), dibenzothiophene (0.05 +/- 0.005 g) and 9-methylanthracene (0.05 +/- 0.005 g) into a 100 mL volumetric flask and making up to the mark with heptane. Ensure that the dibenzothiophene and 9-methylanthracene are dissolved in the o-xylene-cyclohexane mixture (for example, by using an ultrasonic bath) before adding heptane.

NOTE 14 - The SPS may be kept for up to one year if stored in a tightly stoppered bottle in a dark place between 5 and 25°C.

9.4 When operating conditions are steady, as indicated by a stable horizontal baseline, inject 10 µL of the SPS (see 9.3) and record the chromatogram, using the data system. Ensure the baseline drift over the period of the HPLC analysis run is less than 0.5 % of the peak height for cyclohexane.

NOTE 15 - A baseline drift greater than this indicates problems with the temperature control of the column/refractive index detector or polar material eluting from the column, or both. A period of up to 1 h may be required before the liquid chromatograph reaches steady state conditions.

9.4.1 Ensure that baseline separation is obtained between all four components of the SPS (see Fig. 2).

9.4.2 Ensure that the data system can accurately measure the peak areas of dibenzothiophene and 9-methylanthracene.

NOTE 16 - The S/N (signal to noise) ratio for dibenzothiophene and 9-methylanthracene should be 3:1 or greater.

9.4.3 Ensure that the resolution between cyclohexane and o-xylene is not less than 5.0.
9.4.3.1 Column Resolution - Calculate the resolution between cyclohexane and o-xylene as follows:
Resolution = 2 x (t2 - t1) / 1.699 x (y1 + y2)
where:
t1 = retention time of cyclohexane peak in seconds,
t2 = retention time of o-xylene peak in seconds,
y1 = half-height peak width of cyclohexane in seconds, and
y2 = half-height peak width of o-xylene in seconds.

If the resolution is less than 5.0, check to see that all system components are functioning correctly and that the chromatographic dead volume has been minimized. Adjust the flow rate to see if this improves the resolution, and make sure the mobile phase is of sufficiently high quality. Finally, regenerate or replace the column.

9.5 Measure the retention times of the dibenzothiophene and 9-methylanthracene peaks, using the data system.

9.6 Calculate the backflush time, B, in seconds, using the following equation:
B = tA + 0.4 (tB - tA)
where:
tA = retention time of dibenzothiophene in seconds, and
t B = retention time of 9-methylanthracene in seconds.

NOTE 17 - The backflush time is the time after injection at which the backflush valve will be actuated in order to elute T+AHs as a single sharp peak.

9.7 When operating conditions are steady, as indicated by a stable horizontal baseline, inject 10mL of the SPS (see 9.3) and record the chromatogram, using the data system. Actuate the backflush valve at the predetermined time (see 9.6) to elute the T+AHs as a single sharp peak (see Fig. 3). When the analysis is finished, reverse the flow direction of the middle phase (that is, return to forward flush) and allow the baseline to stabilize before the next injection.

9.8 Repeat 9.7, and ensure that the repeatabilities for peak area measurements of o-xylene, dibenzothiophene, and 9-methylanthracene are within the precision of this test method.

NOTE 18 - If peak area repeatabilities are poor, check to see that the injection system is working optimally and that the baseline is stable (minimal drift) and noise-free.