ASTM D3606: Capillary options

The American Society for Testing and Materials (ASTM) provides standardized methods for the testing of a wide variety of materials. The method for determining benzene and toluene in fuel, designated method D3606, utilizes multidimensional gas chromatography (GC) in two different ways: capillary columns (procedure A) or packed columns (procedure B). We have some work from Barry Burger, using our procedure B column set, so now, we are presenting some tips for procedure A.

Setup: Setting up an instrument for the capillary option of method D3606 can be daunting. But once the method is set up, it is robust. The method utilizes a mid-run backflush, creating the backflush in only one of the two columns. The idea is that flow will move continuously through the first and second column until toluene elutes from the first column. Everything remaining in the first column after toluene elutes is backflushed, while additional flow carries toluene and earlier eluting compounds through the second column. This means that a second source of flow is necessary. And why do we bother with all of this? The second column is a long, highly polar ‘wax’ column that is prone to bleed at high temperature, so without the backflushing, run times would be dramatically longer, and column longevity may be compromised.

There are a few devices available that can be used to provide additional flow for the second column, with popular options being a flow-splitter or a Dean’s switch. However, any device that can maintain flow though the second column while the first column is backflushed, will work. For this blog, I am using a three-way microfluidic flow-splitter which connects to an auxiliary EPC to provide additional helium flow.  My configuration is diagrammed below, and follows ASTM D3606 procedure A, configuration A (Figure 1). I am using the Rxi-1ms, 30m x 0.25mm x 0.5umm (cat# 13338), as the pre-column, and the Stabilwax, 60m x 0.32mm x 1um (cat# 10657), as the analytical column, with a 0.52m piece of IP deactivated guard column as the restrictor.

Figure 1: Basic setup for ASTM D3606, procedure A, configuration A.

Method D3606 offers a variety of configurations, where configuration A allows for continuous monitoring of effluent from the first column, specifically, toluene. Normal wear and tear of the first column will shift your elution times, and we want to keep track of that toluene to make sure we aren’t backflushing too early, which makes the restrictor useful. But, not everyone has a dual-FID instrument, so alternatives are described in the method. It will probably take a while before you see toluene shift out of its backflushing window, so you should know the signs: peaks may get smaller and wider, or you may start seeing additional noise along the baseline.

Creating a backflush mid-run: The concept of backflushing is fairly simple; pressure in the inlet drops, and flow moves backwards and out the split vent. Under normal conditions (not backflushing), pressure at the inlet is higher than pressure at the outlet (the auxiliary EPC connected to the flow splitter) so flow moves toward the auxiliary EPC and through the splitter. The backflush is going to occur between the inlet and auxiliary EPC, so pressure at the inlet is going to have to drop below the pressure at the auxiliary EPC (Figure 2). This allows the gases still in the column to move backwards and out the split vent. Meanwhile, the flow that is continuously produced the by auxiliary EPC allows everything that has already eluted from the Rxi-1ms to continue though the Stabilwax and restrictor. So, does the flow from the EPC get pulled backwards when backflushing starts? It shouldn’t. The microfluidic splitter is designed with flow paths that minimize that phenomenon so long as your pre-column is installed in the correct position (check manufacturer guides).

Figure 2: Flow paths before and after back-flushing, with instrument parameters. Red = flow from the inlet, blue = flow from the auxiliary EPC, purple = combined flow from the inlet and EPC, dashed red = reversed flow.

Getting the software to cooperate: Understanding backflushing is one thing, but getting your instrument software to backflush is another. I used an Agilent 7890A equipped with ChemStation for my setup. After making sure my auxiliary EPC was properly configured on the instrument front, I did the same in the software. See Table 1 for all of my settings. When you run this for the first time, watch your GC front for indications that the system is backflushing! When I performed this method there were no indications from the software that the backflush would occur, but my GC front said ‘backflushing’ at the proper time and inlet pressure dropped to about 5.5psi as expected. Your settings may have to be adjusted based on your setup, especially if you have a different restrictor length. ASTM D3606 provides a variety of strategies for optimizing your GC settings.

Table 1: GC settings and column parameters.

Oven
Equilibration time (min)	1
Max temperate (degC)	220
Oven program	75 °C for 8 min
	then 5 °C/min to 85 °C for 3 min
	then 40 °C/min to 140 °C for 5 min
Run time min)	19.375
Inlet
Carrier gas	Helium
Mode	Split
Heater (degC)	200
Pressure (psi)	35.029
Gas saver	On 20 mL/min after 2 min
Split ratio	400:1
Split flow (mL/min)	800
Column #1	Cat#13338 - Rxi-1ms: 30 m x 0.25 mm x 0.5 μm
In	Front SS Inlet He
Out	Aux EPC
Pressure (psi)	35.029
Flow (mL/min)	2
Average velocity (cm/s)	27.038
Holdup time (min)	1.8492
Flow program	2 mL/min for [backflush time]
	then 100 mL/min per min to -1.6 mL/min for 0 min
Column #2	Cat#10657 - Stabilwax: 60 m x 0.32 mm x 1 μm
In	Aux EPC
Out	FID A
Pressure (psi)	21.792
Flow (mL/min)	2.6
Average velocity (cm/s)	34.5
Holdup time (min)	2.8985
Flow program (mL/min)	2.6 mL/min
Column #3	IP deactivated guard: 0.52m x 0.1mm x 0 μm
In	Aux EPC
Out	FID B
Pressure (psi)	controlled by column #2
Flow (mL/min)
Average velocity (cm/s)
Holdup time (min)
Flow program (mL/min)
FID's
Heater (degC)	200
H2 flow (mL/min)	40
Air fow (mL/min)	400
Makeup flow (mL/min)	45
Aux EPC
Carrier gas	Helium
Pressure (psi)	13

The restrictor length plays an important role in your flow settings. You are going to be relying on physics to split your flow between the restrictor and analytical column. The pressure required to move the gaseous sample through the short and narrow restrictor will be higher than the pressure needed to move the sample through the long and wide Stabilwax column, but the flows should be equivalent.  Many splitters come with calculators that can help you determine what restrictor length you would need for your splitter. The split may not be 100% even, but you should be able to reach the detection limits outlined in ASTM D3606 on your analytical column.

Troubleshooting your backflushing system: ASTM D3606 has instructions on calculating backflush time, so I won’t be spending time instructing you on how to do that, but I wanted to highlight some troubleshooting strategies I learned from my colleague Jason Hoisington.

If the difference in pressure between your auxiliary EPC and inlet is not great enough, you may only get a ‘pull-back’ of toluene instead of a backflush. This means, instead of moving backwards, the toluene is just delayed. While calculating your backflush time, if your peak height just isn’t decreasing but your retention time is shifting, this may be your issue. Run a blank after your sample to check if toluene is still in your restrictor. Increasing pressure at the auxiliary EPC can correct this, but it will increase flow through your analytical column. You can also try decreasing your inlet flow further at the backflush time, but you don’t want your inlet flow to drop too low, depending on your instrument.

Another way to check if you are backflushing is to change your split to 2:1 and measure flow out of the split vent. At the backflush time, the flow will increase rapidly then settle at [split flow]+[-backflush flow]. For example, if I have a split of 2:1, and my backflush is set to be -1.6mL/min at 4 minutes, then my 2mL/min flow out of the split vent will jump up to 3.6mL/min at 4 minutes. If you see this jump, then congratulations, your backflush is working!

Results with the Rxi-1ms and Stabilwax

Toluene was found to elute from the Rxi-1ms around 4.36min, and the total run-time was just under 20min. We ran ASTM D3606 standard 8.5.2 and a sample of gasoline using our ASTM D3606 setup, and saw excellent separation of compounds of interest (Figure 3). As expected, the shorter hydrocarbons that elute before toluene clutter the baseline of the chromatogram from the restrictor. The restrictor is showing everything eluting from the Rxi-1ms, which is not yet suitable for any quantitative work. After moving through the Stabilwax, those same hydrocarbons elute early in the chromatogram providing a cleaner baseline for determining ethanol, toluene, benzene, and butanols in the gasoline sample.

Figure 3: Overlaid chromatograms of gasoline (blue) and ASTM D3606 standard 8.5.2 (red) from the restrictor (top) and Stabilwax (bottom).

Let us know if you want to see more work with this column set for ASTM D3606!