Half the Column, Same Chromatogram: Maintain Resolution of BDE 49 and BDE 71 with Proper Method Translation After Trimming an Rtx-1614 Column for Maintenance
Table of Contents:
Polybrominated diphenyl ethers (PBDEs) are additive flame retardants that are used in many household and office products, including furniture, electronics, and textiles. These lipophilic compounds have been found to be persistent and bioaccumulative in nature. Due to their toxicity and ubiquitous presence, PBDEs are included in many monitoring efforts across a wide array of biological and environmental sample matrices. The analysis of PBDEs is challenging due to structural isomers that need to be chromatographically separated and thermally-labile compounds of interest that may breakdown during gas chromatography. Furthermore, nonvolatile material may still persist even in cleaned final extracts, requiring GC column and inlet maintenance to be performed. PBDEs included in EPA Method 1614 are well resolved on a 15 m x 0.25 mm x 0.10 µm Rtx-1614 GC column that was specifically designed to meet method resolution requirements of less than 40% valley height between BDE 49 and BDE 71. The selectivity of this column, in combination with properly translating the original GC method, allowed the column to be trimmed significantly for maintenance (7.9 m) while maintaining the method resolution criterion. This allows chromatographers to extend GC column lifetime.
Flame retardants have been added to numerous products including polyurethane foams, electronics, plastics, and carpet padding to reduce the risk of burning. Technical mixtures of polybrominated diphenyl ethers (PBDEs) have been extensively used as additive flame retardants since the 1970s . However, the technical mixtures containing penta and octa congeners were voluntarily withdrawn in the United States in 2005 and the last remaining PBDE mixture, decaBDE, is being phased out. Unfortunately, the concentrations of PBDEs in the environment have not been declining and, due to their persistent and bioaccumulative nature, these compounds are still widely monitored. The health concerns of PBDEs are similar to that of polychlorinated biphenyls (PCBs) and at the most recent Stockholm Convention on Persistent Organic Pollutants, two technical mixtures (pentaBDE and octaBDE) were added to the priority pollutant list .
Monitoring PBDEs in biota and environmental matrices can be difficult due to the complexity of the sample, structural isomers that must be separated chromatographically, and thermally-labile compounds that can break down during gas chromatography (GC). A 15 m x 0.25 mm x 0.10 µm Rtx-1614 GC column, a 5% diphenyl, 95% dimethyl polysiloxane type stationary phase, was specifically designed by Restek to meet EPA Method 1614 resolution requirements for critical isobaric BDE congeners 49 and 71 . Using a short, thin-film column also allows the elution of decabromodiphenyl ether (BDE 209) without on-column thermal degradation that can compromise its qualitative and quantitative determination .
GC column and inlet maintenance is especially important for BDE analysis, because nonvolatile material persists in sample extracts and deposits onto the front of the column and liner. This can cause poor transfer of BDEs to the GC column, which compromises quantification and sensitivity. That same material at the head of the GC column also leads to poor peak shapes and reduced resolution between critical congeners, such as BDEs 49 and 71. Trimming a coil (0.5 m to 0.7 m) off the front of the column and changing the inlet liner restores performance and significantly increases the column lifetime. While fewer column changes saves time and money, care must be taken to properly translate the method for the shorter column length to ensure that the original separation is maintained. If the GC method is not translated after the column has been trimmed, the column flow would be very fast, which would decrease resolution and detectability. The flow may also exceed the pumping capacity of many mass spectrometers. Method translation software can help users adjust the GC oven program for the shorter column length and maintain the elution profile of the original chromatogram. Using a 15 m x 0.25 mm x 0.10 µm column, we investigated how much of the GC column could be clipped off for maintenance before the Method 1614 resolution requirement for BDE 49 and BDE 71 could no longer be met.
Original Instrument Conditions
A 15 m x 0.25 mm x 0.10 µm Rtx-1614 column (cat.# 10296) was used for the column trimming experiments. For all analyses, a 1 µL splitless injection (1 min purge time) of a native PBDE/BFR mix from Wellington Laboratories (Wellington cat.# BFR-PAR) was made into a split/splitless inlet set to 340 °C and fitted with a Restek Premium cyclo double taper inlet liner (cat.# 23310.5 ). The instrument was a 208 V Agilent 7890/5975 GC-MS equipped with an enhanced turbo molecular pump. Electron ionization (-70 eV) and selected ion monitoring (SIM) mode were used for MS.
The initial GC oven and flow conditions were optimized using Pro EZGC methods development software from Restek Corporation (cat.# 21487). This thermodynamic modeling software provided a separation which maximized resolution of target PBDEs with an analysis time of 25 min. The GC oven temperature program started at 75 °C (hold 1 min), then ramped at 18 °C/min to 210 °C, then ramped at 8 °C/min to 310 °C (hold 4 min). Helium was used as the carrier gas with a constant flow of 1.6 mL/min.
GC Method Translation
The original column length was accurately determined by measuring the retention time of an unretained air peak to determine the holdup time. Using the Agilent GC pressure/flow calculator, the column length parameter was increased until the calculated holdup time matched the experimental holdup time (Table I). The Rtx-1614 GC column that was used for this analysis had an initial length that was calculated to be 16 m. The determined column length was then used for method translation and instrument software inputs.
Table I: Column length was increased (or decreased) in the GC pressure/flow calculator until the calculated holdup time matched the measured holdup time.
|Measured Holdup Time = 0.420 min|
|Column Length (m)||Calculated Holdup Time (min)|
After the analysis of the BFR-PAR standard, column maintenance was simulated by immediately trimming approximately two coils (1.6 m) from the front of the column by removing the column from the inlet and pulling two full coils from the column cage and cutting the column where it lined up with the inlet piece again. The new column length was determined again by measuring the holdup time and using the GC pressure/flow calculator to find the column length. The new column length (14.4 m), along with the original GC conditions were input into the Agilent GC method translator software, operated in translate-only mode. The method translator calculated a new oven temperature program so that the PBDEs eluted at approximately the same temperatures as the original program and maintained resolution. The column flow of 1.6 mL/min was held constant. The BFR-PAR standard was reanalyzed to monitor method performance of critical resolution and peak shape. Approximately two coils were trimmed again from the front of the column and the GC method conditions were translated for the shorter column. This process of column trimming, method translation, and analysis of the BFR-PAR standard was repeated until the resolution between BDEs 49 and 71 was compromised or the instrument could no longer maintain the necessary oven programming rate or inlet pressure.
Results and Discussion
The Pro EZGC modeling software provided a GC oven temperature program and column flow conditions that maximized the resolution between the tetrabromodiphenyl ethers BDE 49 and BDE 71. These two analytes are isomers that must be chromatographically resolved in order to meet EPA Method 1614 resolution requirements of less than 40% valley height between BDE 49 and BDE 71. The optimized analysis conditions provided by the software yielded baseline resolution of BDE 49 and BDE 71 and a 25 min analysis time to reduce thermal breakdown of decabromodiphenyl ether (BDE 209). Resolution between isomers needs to be optimized, but speed of analysis is also an important consideration for the analysis of PBDEs. Easily debrominating during a fire to suppress the flames is a key characteristic that makes PBDEs ideal flame retardants, but it also makes them difficult to analyze in a hot GC oven. In previous work, we found that a slower analysis or a longer column (30 m) yielded thermal breakdown of BDE 209 . Using the 15 m, 0.25 mm, 0.10 µm Rtx-1614 column and a fast analysis reduces the thermal breakdown of the PBDEs while still providing baseline resolution of the structural isomers. Using an optimized injection port temperature of 340 °C and a cyclo double taper inlet liner was previously found to increase BDE 209 response in comparison to using an inlet temperature of 250 °C or 300 °C. The higher inlet temperature allowed better volatilization and transfer of the higher molecular weight flame retardants onto the column, without substantially increasing the degradation of BDE 209.
PBDEs are lipophilic compounds that will accumulate in the fatty tissue of animals. To determine the amount of PBDEs present in a sample, an extensive extraction and cleanup is required for soil, sediment, and tissue samples . After the extract cleanup there is often still a measureable amount of nonvolatile material that remains in the final sample. The nonvolatile residue collects in the inlet and at the beginning of the GC column and can cause reduced recoveries and peak tailing. Changing the inlet liner and trimming approximately one coil (0.5 m to 0.7 m) off of the front of the column as routine maintenance is often required to restore the desired system performance and continue analyzing samples; however, when this is done the method should be adjusted to accommodate the new shorter column length. Figure 1 shows the chromatographic separation achieved after two coils were trimmed off of the column. Due to the selectivity of the Rtx-1614 column for the target compounds and the use of properly translated method conditions, the critical separation of BDE 49 and BDE 71 was maintained on the shorter column.
GC Method Translation
The GC method translation software allows the user to input the original GC oven program conditions and scale the new program appropriately to the new GC column length so that the elution temperatures, and thus the retention order, of peaks remain the same. There are two possible ways to translate the method after column trimming. The first maintains the original retention times and oven program by reducing the column outlet flow. The benefit to this approach is that it is easier to maintain selected ion monitoring windows because the analytes elute at approximately the same retention times as when the column was longer. However, the utility of this approach is limited by the operating limits of the system. For example, starting with a 16 m column and an outlet pressure of 0 psi, due to the mass spectrometer, the minimum column length is 12.5 m (approx. 6 coils). As the column flow is reduced, the inlet pressure decreases. If another coil (0.7 m) is trimmed off, the column length would be approximately 11.8 m and the calculated inlet pressure would be an inoperable -0.4 psi.
The second way to translate the GC method after column trimming is to maintain the original column outlet flow while increasing the oven temperature program rate (Table II). Using this approach is beneficial because the column outlet flow into the mass spectrometer is constant. Decreasing or increasing the outlet flow into the MS will affect the vacuum pressure and change analyte response. In addition, as the column gets shorter the analysis becomes faster, which increases sample throughput. Maintaining the outlet flow and reducing the inlet pressure allows for a minimum column length of 7.9 m (~ 12 coils), which allows 6 more coils to be removed for maintenance compared to the method where retention times are held constant. At 7.9 m the inlet pressure was 0.85 psi. If another coil is trimmed off of the column, the length becomes approximately 7.2 m and when a vacuum-outlet detector (e.g., MS) is used the inlet pressure would be negative. In this case then, the inlet pressure was the limiting factor in determining whether to install a new column, not the failure to meet key congener separations, which were more than adequate (Figure 2).
Table II: As column length decreases with maintenance trimming (reduced by 0.5 m to 0.7 m for each coil removed), properly translated method conditions maintain the column flow rate and increase the oven temperature ramp rate so that analytes still elute at approximately the same temperature as with the original method.
|Original Column||~2 Coils Removed||~8 Coils Removed||~12 Coils Removed|
|Column Length (m)||16||14.4||10.4||7.9|
|Internal Diameter (mm)||0.25||0.25||0.25||0.25|
|Film Thickness (µm)||0.10||0.10||0.10||0.10|
|Inlet Pressure (psi)||7.4||6.3||3.1||0.85|
|Flow Rate (mL/min)||1.6||1.6||1.6||1.6|
|Average Velocity (cm/sec)||63||67||79||90|
|Oven Temperature Program||Ramp Rate (°C/min)||Temp (°C)||Hold Time (min)||Ramp Rate (°C/min)||Temp (°C)||Hold Time (min)||Ramp Rate (°C/min)||Temp (°C)||Hold Time (min)||Ramp Rate (°C/min)||Temp (°C)||Hold Time (min)|
Figure 2:   The minimum column length of 7.9 m achieved before the inlet pressure was no longer positive maintained adequate resolution between BDE 49 and BDE 71. Approximately twelve coils were trimmed from the column, and the analysis time was reduced by 16 min from the original 25 min method.
When adapting the method to the shorter column length, the ion monitoring windows must also be adjusted to accommodate the faster analysis time. The new windows can be determined using the speed gain factor calculated in the method translation software. For example, at 10.4 m the speed gain factor is 1.6 compared to the 14.4 m method. The retention times are all approximately shifted by a factor of 1.6; therefore, taking the previous method ion monitoring windows and dividing by the speed gain factor will give the expected new elution time (Table III). When column maintenance is later repeated, the new ion monitoring windows can again be accurately calculated from the actual measured retention times in the previous method (instead of from the measured retention times in the original method) using the speed gain factor that was calculated by the method translation software.
Table III: Retention times of new translated method can be predicted using the speed gain factor calculated by the method translation software.
|14.4 m Column||10.4 m Column|
|Actual tR (min)||Calculated tR (min)
= 14.4 m tR/ 1.6
|Actual tR (min)||Difference (min) =
Calculated tR - Actual tR
When properly translating the GC method, the column length of 7.9 m did not prevent the EPA Method 1614 resolution requirement for BDE 49 and BDE 71 from being met. The new method keeps the column flow the same but the oven programming rates are faster to accommodate the higher linear velocity through the column. Analysis times are decreased by a factor of 2.8 from the 16 m method to the 7.9 m method and peak widths are reduced when trimming the column and translating the method (Table IV). Narrower peak widths can increase the signal-to-noise ratio and detectability of the analytes if the data acquisition rate is adjusted to properly define the peak for quantitation.
Table IV: Full width half heights (FWHH) of the BDEs are reduced when the column length is reduced.
|14.4 m||10.4 m||7.9 m|
|FWHH (s)||FWHH (s)||FWHH (s)|
The Agilent 7890/5975 GC-MSD that we used for this work was a 208 V instrument with a fast ramping oven that allowed for an oven ramp rate of 52 °C/min to 210 °C and 23 °C/min to 310 °C for the shortest column length (7.9 m). Having the multi-ramp program with a slower ramp at higher temperatures also allowed for the faster programming of shorter columns. A single, fast oven programming rate may require additional systems so that the instrument can accurately provide a faster ramp rate at higher temperatures.
Trimming a coil (~0.5 m to 0.7 m) off of a GC column as part of routine maintenance extends column lifetime; however, in order to maintain resolution of the target analytes following maintenance, the method must be properly translated for the shorter column length. By translating the GC oven program method for the new column length, an Rtx-1614 column (0.25 mm x 0.10 µm) can be trimmed down to 7.9 m for maintenance and still exceed the EPA Method 1614 resolution requirement for BDEs 49 and 71 and produce a chromatogram very similar to the original 16 m column. Based on a “1 coil trim”, approximately 12 maintenance trims could be performed for the analysis of BDEs before the GC oven ramping speed and inlet pressure limitations did not allow further use of the column.