Quick & Painless Email Subscription

Use the “Subscribe” link at right. It just takes a few seconds!

Why is my LC Retention Time Shifting?

If you have experienced retention time shifts with LC, you know it can be complicated and can cause problems with quantitation if not resolved. Hopefully this post will help you navigate through troubleshooting this situation. For the purposes of this discussion, we will be discussing changes in retention on one individual column, not differences from one column to another. Also we are assuming that the analyst is following an established method that worked for a significant period of time prior to seeing changes in retention.

It is always useful to define the specific problem to help narrow down the possible causes.  An important factor is whether all of your peaks have shifted or whether it is only certain ones.

If all of the peaks are shifted by about the same time interval, the variations are likely caused by variation in flow rate in the instrument. This can be a shift in either direction, but is often toward longer retention times/slower flow rate. This can be verified by collecting eluent from the column into a graduated cylinder and using a stopwatch to measure the time taken to fill the container to the mark. Flow rate variations can be caused by any of the following:

  • Leaks in the flow path- Check to make sure all fittings are tight and in good condition.
  • Worn or faulty pump seals- Replacing the seals is often a good precaution to prevent this.
  • Faulty check valves- Get help from an instrument service person if needed to verify performance of check valves and other parts in the flow path.
  • Bubbles in the pump or tubing- Make sure your system has a degasser built into the system that degasses solvent prior to pulling it into the solvent pump. If a degasser is not available, mobile phase can be degassed manually before using for LC. Please note that any mobile phases prepared to contain buffer salts or any modifiers in solid form should be filtered prior to use for LC. We have a membrane microfiltration system available for this if needed. Filtration systems such as this also serve to degas mobile phases because the solvent is pulled through by vacuum.
  • Changes in instrument settings- This may seem obvious, but one should always confirm that program parameters have not changed from time to time. When multiple users are performing the same analysis on shared equipment, this can often be an issue.
  • Changes in column temperature- this can affect the column in a way similar to flow rate and can move the retention in either direction. Make sure you have the column in a thermostatic column compartment to avoid this.

If early eluting peaks are shifted, but late eluting peaks are relatively unaffected, the issue is usually related to differences in the solvent ratio with sample and/or mobile phase. This could be caused by any of the following:

  • Solvent composition of sample is not compatible with the mobile phase or gradient- This is usually accompanied by poor peak shape. Try making your sample in your mobile phase solvents at the same ratio as the starting conditions of your gradient (or same ratio as your mobile phases if using isocratic conditions).
  • Sample volume too high, overloaded- Peaks usually are fronting under these conditions, but sometimes can tail. Try injecting a smaller volume.
  • Error in preparation of mobile phases- It is always helpful to carefully remake the mobile phases to rule this out.
  • Changes in instrument settings- see discussion above for this.
  • Guard cartridge is saturated with matrix- This should be replaced any time a change in peak shape or retention is noticed. Preferably it should be replaced as a regular preventative maintenance procedure before adverse effects are noticed.
  • Column may be damaged or contaminated with matrix at the inlet- If changing the guard doesn’t help (or if you are not using a guard), try flushing the column to clean. Instructions are located here: https://www.restek.com/Technical-Resources/Technical-Library/General-Interest/general_MIS_0322

 

If only some of the peaks are shifted, seemingly random, or if all peaks are shifted in varying degrees, several different things could be happening:

  • Sample matrix is interfering with elution of some compounds- Try making injections of those compounds in mobile phase with no matrix to rule this out. Additional cleanup of the sample extract may be needed.
  • Sample pH is different from the mobile phase- Sometimes this explains why some peaks are affected, but others are not. Try matching the pH with that of the mobile phase and check it regularly.
  • Column has a buildup of sample matrix- Often this is accompanied by an increase in pressure, noisy baseline, and/or ghost peaks. Try flushing the column to clean (see link above).
  • Column has reached the end of its lifetime- If flushing does not correct the problem, the column needs to be replaced.

 

Restek also has a video you may be interested in on this topic:

https://www.restek.com/Technical-Resources/Technical-Library/Video-Library/LC-Troubleshooting-Retention-Time-Shift

More information can also be found at the following locations on our website:

Shifting Analyte Retention Times or Matrix Interferences?

Technical Service “Red Flags” -LC

LC FAQ #3 -What happens if my sample solvent is stronger than my mobile phase?

https://blog.restek.com/?p=14352

 

I hope you find this post helpful. Thank you for reading.

SPME Arrow Configurations

If you have been looking at SPME Arrows, because we convinced you of the Arrows’ mechanical robustness, sensitivity, speed, etc., you could find yourself in a decision dilemma. Beyond phase selection, which is the subject of another blog, the various SPME Arrow configurations may leave you stymied. So, it is the intent of the current blog to help remove any of the mystique surrounding the following 3 “major” SPME Arrow configurations:

  1. 1.1 mm Arrow
    1. 100 μm Polydimethylsiloxane (PDMS) (cat.# 27485)
    2. 100 μm Polyacrylate (cat.# 27488)
    3. 120 μm Carbon Wide Range (WR)/PDMS (cat.# 27487)
    4. 120 μm Divinylbenzene (DVB)/PDMS (cat.# 27486)
    5. 120 μm DVB/Carbon WR/PDMS (cat.# 27875)
  2. 1.5 mm Wide Sleeve Arrow
    1. 100 μm Polydimethylsiloxane (PDMS) (cat.# 27877)
    2. 120 μm Carbon Wide Range (WR)/PDMS (cat.# 27879)
    3. 120 μm Divinylbenzene (DVB)/PDMS (cat.# 27878)
    4. 120 μm DVB/Carbon WR/PDMS (cat.# 27876)
  3. 1.5 mm Arrow
    1. 250 μm Polydimethylsiloxane (PDMS) (cat.# 27484)

As alluded to earlier, the current blog will focus on the “major” (i.e., 1-3) options. We will address the “minor” (i.e., sub-bullets with phases) options in another blog. First, we will make the choice easier by ignoring option 3, which is a configuration I consider an anomaly to be discussed later. Now that we are down to 2 choices the decision between option 1 (1.1. mm Arrows) and 2 (1.5 mm Wide Sleeve Arrows) becomes quite simple: are you going to conduct direct immersion sampling? If no [i.e., headspace (HS) extractions only], then you should stick with the 1.1 mm Arrows under option 1. If yes [i.e., HS and/or direct immersion (DI) extractions], then you should run with the 1.5 mm Wide Sleeve Arrows under option 2. The 1.5 mm Wide Sleeve Arrows (option 2) were specifically developed for DI sampling. In fact, this version of the Arrow is really just a 1.1 mm Arrow housed in a 1.5 mm septum piercing needle, as shown in the Figure and Table below. To elaborate, option 1a and 2a (above) both have 100 μm of PDMS, because both Arrows have the same phase support tubing diameter and phase diameter, rather the septum piercing needle is different.

So, how do these Wide Sleeve Arrows (option 2) aid in DI sampling? When the fiber phase swells from DI sampling (I said when, not if, because this happens often), the larger 1.5 mm septum piercing needle ensures the phase is not sloughed off and/or damaged when the SPME Arrow support tubing is retracted inside the septum piercing needle. So, why not make the choice even simpler and always use the Wide Sleeve Arrows (option 2). I mean, they are good for both HS and DI sampling, so it is the best of both worlds right? The answer is simple: cost and septum wear. The 1.5 mm Wide Sleeve Arrows cost more than 1.1 mm Arrows, and I would speculate they will put more wear and tear on GC inlet septa and vial septa.

Now what about those 1.5 mm Arrows (option 3) I told you to ignore earlier? Truth is, this 1.5 mm option is only available in PDMS; hence the “anomaly.” They are nice in that there is 2.5x the phase volume of the comparable 1.1 mm PDMS and therefore a potential 2.5x increase in response. However, we have found that the increased phase thickness can have negative impacts on the chromatography, as discussed in our recent manuscript.

Solid phase microextraction (SPME) Arrows; (1) 1.1 mm SPME Arrow; (2) 1.5 mm Wide Sleeve SPME Arrow; and (3) 1.5 mm SPME Arrow. All SPME Arrows composed of the following parts: a color coded screw hub (A), a sealing septum (B), a septum piercing needle (C), a fiber attachment needle (D), and a coated metal fiber (E). d1: Support tubing; d2: septum piercing needle; d3: phase diameter; d4: phase support tubing diameter; l3: phase length; a3: phase area; and v3: phase volume.

One last talking point: We told you there is no free lunch when incorporating the SPME Arrow into your laboratory, because you will have to install a GC Inlet Conversion Kit. It is important to stress that the inlet conversion kits work with all 3 types of SPME Arrow AND you may use the converted inlet with all your standard injection techniques (e.g., liquid syringe, headspace syringe, etc.). So, there is no need to swap back and forth between inlets.

Over tightening your fittings has consequences.

If you use the correct fittings and seals in the gas chromatography system, and follow some simple steps, there is absolutely no need to over tighten any fitting to achieve a leak free seal.

For almost all fittings that use seals in a GC system, whether they are ferrules, septa or o-rings, the tightening process should be the same:

  1. Make sure the place where you are applying the fitting is cooled down and the gas flow is turned off
  2. Select the correct size fitting and seal for the application
  3. Correctly fit the fitting and seal together according to the manufacturer’s guidelines
  4. Finger-tighten the fitting on to the receiving part
  5. Using the appropriate tool (wrench or fingers) tighten the fitting and extra ¼ turn.
  6. Establish a gas flow and pressurize the section of the GC system.
  7. Use a Restek electronic leak detector (28500) to check for leaks around the fitting
  8. If you do detect a leak further tighten the fitting in 1/8 turn increments until the leak ceases.

Note: If you continue to notice a leak and you have done one complete revolution of the fitting STOP.  There is something wrong.  Rather than damaging the fitting, or the receiving part, remove the fitting, get a new one and a new seal and start again.

Common fittings and seal, and the consequences of over tightening

Septa

This is the easiest seal to get leak free.  You should not need a wrench; minimal finger tightening is usually sufficient.  If you over tighten you compress the septum.  This puts the septum under stress, and so when the septum is penetrated by a needle it is more easily ripped apart leading to coring and leaks.  Coring can result in particles of septum in the liner which can adversely affect chromatography results.

You should leak check with and electronic leak detector around the septum fitting every day.  It is one of the hardest working seals, and the easiest and cheapest to change.

Reducing Nut

This is the fitting at the bottom of the inlet where the column nut attaches.  Over tightening this nut can damage the threads on both the nut and the inlet fitting.  We have also seen the inlet bent by over tightening to such an extent that liners could not fit down inside.

For GCs like Agilent, Thermo Trace 1300/1310 and Perkin Elmer 590/690 the simple way to avoid over tightening, or the use of excessive force, is to use a dual vespel ring inlet seal.  The dual vespel ring allows for a soft but effective seal, and enables you to follow the tightening procedure above.

Dual Vespel Inlet Seals come in several different surface finishes.

It should be noted that when moderately high inlet temperatures are employed the dual vespel ring can shrink slightly, which can cause a leak, but that is easily remedied by a slight additional tightening after a couple of heat cycles.  With good, frequent leak monitoring with a Restek electronic leak detector this would not cause any issues.

Column Nuts

Over tightening column nuts has many consequences.  First you could crush the column, whether that is a fused silica or MXT column.  This in turn can lead to leaks and/or flow restrictions, and in some cases the column may break off in the fitting.  Second, the ferrule may crack which can lead to leaks.  Lastly the ferrule may deform so much that it may extrude into the fittings where it can block the flow or lead to leaks, and getting stuck.

Standard connections

These are like Swagelok or Parker fitting that are brass of stainless steel, and sometimes use a graphite ferrule.  Over tightening these fittings can easily lead to stripping the threads, especially in brass fittings.  We have seen where over tightening has led to the crushing of copper tubing, and this restricting flow, and in extreme cases cracking the tubing.  If you use graphite ferrules in these fittings, as before the graphite can be extruded into the fitting which can cause leaks.

Never mix material.  If you have a steel fitting, don’t use a brass nut, and vice versa.  This can lead to easier thread stripping.  Also as a rule of thumb use brass fitting with copper piping and stainless steel fittings with stainless steel tubing.

Screw top vials

Don’t over tighten screw top vials.  This can pinch the septum resulting in a poor fit and potential for loss of sample.  We have seen it where an over tightened cap will pinch the septum so much that when a syringe tries to penetrate the septum it pushes it right into the vial.  It is almost impossible to leak check this seal.  So you have two choices: a gentle touch, or switch to crimp top vials.

Using a liquid leak detection solution has consequences

We have been warning about the consequences of using liquid leak detection products for years (see Blog).  These products, like Snoop, contain a whole host of compounds that can easily contaminate your GC system if you have a leak.  The solution can enter the leak site by capillary action, and is very hard to remove, leaving a lasting background signature.  Furthermore, squirting a liquid onto an electronic device or instrument sounds like a really dumb idea.  We have also seen that some customers have used an electronic leak detector after using a liquid leak detection solution and sucked the solution up into the device and damaged it!!

Recently we found a new reason not to use liquid leak detection solution.  In order to evaporate the solution a little quicker these products use volatile solvents as well as water.  These solvents can degrade plastic components of a GC system.  In the following example you can see before and after images of what can happen to the plastic housing of a gas filter cartridge.

Gas filter cartridge before

Gas filter cartridge after

 

The only way to detect gas leaks in a GC system that is quick, easy, cheap, effective, safe and accurate is to use an electronic leak detector like Restek’s latest version of our leak detector 28500

Restek’s electronic leak detector, 28500.

Discovering Analyte Breakdown On-Column

I completed my winter internship at Restek and my main task was to expand the Pro-EZGC libraries.
So I created a library for Nitrosamines on the Rxi-1301Sil MS and while acquiring data we noticed a foot on a couple of my peaks.

Figure 1: On-column breakdown can be observed for the late eluting nitrosamine compounds as indicated by the foot at the leading edge of the peak.

I attempted to translate the simulation to a column with the same phase, but with a thicker film. Since the Rxi-624Sil MS and Rxi-1301Sil MS are the same stationary phase, I modeled a run with the Rxi-1301Sil MS Library using the column dimensions of an Rxi-624 Sil MS. As I ran the conditions provided by EZGC we noticed a strange rise of the baseline.

Figure 2: Rise in the baseline is observed which is most likely caused by analyte breakdown on-column. Breakdown components remain in the stationary phase and are observed as carryover.

One of my colleagues immediately suggested that the nitrosamines could break down as a function of elution temperature. In order to test this theory, I prepared three different run conditions with different elution temperatures for the late eluting nitrosamines on the Rxi-624 Sil MS:

Figure 4: Adjusting the final hold changes the elution temperature of the last three compounds. If the compound response drops as elution temperature increases it will support our theory of on-column breakdown.

These are the compounds I used (sorted by elution order):

Figure 5: List of nitrosamines found in the chromatogram where the last three compounds are labeled as 1, 2 and 3.

Figure 5: Three chromatograms illustrating the three different programs with increasing elution temperature of the last three nitrosamines. Notice the significant increase in breakdown.

I labeled the last three peaks; the ones that break down appear to have the most activity on-column. Thanks to the MS I was still able to determine the retention times and it became evident that the breakdown increases with higher elution temperature. Was this a general problem with these compounds on this phase, or was this specific to the Restek column?

 

Figure 6: Chromatograms with three different oven conditions. The last three nitrosamines have increasing elution temperatures from top to bottom. These chromatograms were performed on a competitor column to verify the on-column breakdown is most likely related to the cyano stationary phase and not specific to a manufacturer.

We only had a 60m competitor column available, so it’s not a direct comparison, however, we see similar breakdown. We can tell that the nitrosamines break down on-column on this phase in general. But if you use a film that is thin enough and run the compounds with a low elution temperature it will minimize on-column breakdown. Using an Rxi-1301Sil MS is still faster than using a “5-type” phase, but you will have to enter your nitrosamines into ProEZGC and find out for yourself.

The eVol – Investigating a Liquid Handling Tool to Improve Consistency and Reliability for Calibrations

We’re taking you to Austin, Texas to meet the amazing team at Santé Laboratories for this investigation! This project, led by the analytical team at Sante Laboratories, put the eVol in use and compared it to standard pipetting techniques in the generation of analytical calibration curves.

Wait….what?

You might be thinking to yourself….What’s an eVol?

An eVol is a digitally controlled dispensing system that’s used for liquid handling.  The eVol covers a liquid volume range from 200 nL to 1000 uL.  You can use this system for a variety of applications including serial dilutions, addition of standards, standard preparation, and delivery of derivatization agents. While the eVol might be a new technology to the cannabis and hemp market, the eVol has been use for a long time by analytical laboratories providing testing services in other industries.  You can find out more information on Trajan’s website.

Now back to the purpose of this study.

There are many factors that can contribute to errors in liquid handling like variability in glassware, volatility of compounds and evaporation, and incomplete fluid transfer.  These errors start to have a larger impact on uncertainties as transfer liquid volumes decrease.  The team at Santé Laboratories also emphasized that, “additional effort to improve precision during preparation of calibration and quality control samples better allows the high throughput analytical laboratory to improve accuracy and repeatability in unknown concentration determinations. Producing a more robust and accurate curve and quality control samples also allows the laboratory to monitor the lifetime and health of their calibration curves, minimizing recalibrations and reducing usage of expensive reference materials (CRM).”

To evaluate the hypothesis that improved liquid handling produces a more reliable and repeatable calibration, Santé Laboratories prepared two sets of external standards; one using calibrated low retention air-cushion pipettes and one using the eVol.  The curve screened Restek’s 11 cannabinoid standards with a serial dilution procedure.  The table below compares the R2 values, multiple minus adjusted R2 values, and variance coefficients.

 

Table 1. R2 Values (Multiple and Adjusted) and Variance Coefficients

Table 2. Multiple Minus Adjusted R2

 

The Santé team commented, “Immediately, comparing the multiple and adjusted R2 values and variance coefficients, we observed lower variance coefficients (between 34 and 66%, 56% on average) and adjusted R2 values closer to unity using the eVol analytical syringe versus air cushion pipettes with low retention tips. Furthermore, when looking at the multiple vs adjusted R2 for the pipette, we see that the adjusted R2 decreases across all cannabinoids, as the adjusted R2 penalizes variance in the individual replicates. Comparing the differences with the eVol, we see a much smaller change.”

But their assessment did not stop there—to really assess the calibration procedure, the team reviewed the residual plots—the differences between the measured and predicted response values in a regression curve. For linear regression, residuals ideally distribute normally across the x-axis. Observing the differences in the residual plots for CBD, CBDa, d9-THC, and THCa, we observed significantly more linear residuals with the eVol than standard pipetting, an important feature as nonlinear residuals lead to repeatable biases in various sections of the calibration curve. In the case of this analysis, the residuals indicate that samples which fall between calibrators 4 and 7 with the pipette would be positively biased from the calibration curve, before taking into account any other random or systematic errors in the analysis.

Figure 1. Residual Plots for CBD, CBDa, THC, and THCa.

Santé Laboratories had the following conclusion, “Precision liquid handling improves calibration procedures, particularly when utilizing small volumes of concentrated materials. This is extremely important for cannabinoids analyses, as commercially available CRMs are concentration and quantity limited due to the controlled nature of the molecules. This requires laboratories in many cases to utilize serial dilutions and low volume inserts or ultra-recovery vials to prepare sufficiently concentrated calibration curves. The eVol analytical syringe, when used for serial dilutions, shows modest improvements in several markers of calibration health. It also enables the user to move from a serial dilution to a straight dilution procedure, allowing for the use of fixed concentration internal standards.

 

The eVol has the additional benefit of being able to pierce septa, allowing use with closed autosampler vials to minimize losses of volatile materials during preparation. This is a particularly useful feature in gas chromatography, especially for laboratories still utilizing full evaporation technique, where analytes are extremely volatile and sample sizes are extremely small—work on this is under way as well.”

 

A special thanks to Tyler West, Andrea Clemente, and Brian R. Sloat at Santé Laboratories (www.santelabs.com).

Poor press-tight connections have consequences

My colleagues have written plenty of articles over the years with how to cut columns and get the perfect press-tight connection:

Jaap – How to make a good fused silica seal using a Press Tight type connections

Jaap – Column cutting, for making the optimal coupling

Chas – How (not) to cut your capillary column

But we haven’t really focused on what are the consequences of poor cuts and poor installation.

As Chas pointed out in his article you can make a bad cut.  This can be achieved by doing things like pressing too hard on the column on the column with the wafer when scoring or using the serrated edge of the wafer.  The consequences are a jagged edge that would be hard to seal or even creating fissures in the column that compromise the strength of the column.  These in turn can lead to poor connections or a crushed column.

Poor cut and uneven seal

We have seen poor seals like this disconnect during the heating cycle of the GC program, and thus lead to a leak, and probably a damaged analytical column.

Fractures in column because of a poor cut leading to a poor seal

Another problem we have seen is people using too much force to insert the column into the press-tight to get a seal.  This has consequences too.  In the following examples we that it can lead to both column and press-tight damage

Crushed end with poor seal after column strength compromised

Press-tights shattered and cracked after columns that were inserted to firmly expanded during the GC program and fractured the glass.

Shattered

 

 

 

Cracked

Analyzing avocado: How to deal with lack of water and keep the fats out

Avocados are a popular produce item. In fact, the consumption of avocados rose by 450% since 2000! The good news – avocados are ranked number 1 on the clean 15 list, meaning they have the least amount of pesticide residues. With so many avocados on the market, the testing is as important as ever, which may not be that easy. Avocados are a commodity that consists of approximately 70% water and 15 % fat. For GC-MS/MS analysis, most of the fat needs to be removed because excess fat can lead to poor analyte recoveries and an increase in instrument maintenance.

Image source: https://daily.jstor.org/the-illustrious-history-of-the-avocado/

Read the rest of this entry »

Are Wax columns always used for essential oils?

My last blog examined the analysis of essential oils using GC columns with wax phases. While these columns are commonly used for natural oils, they are not the only option. Another choice that falls on the other end of the selectivity spectrum, in this case, a non or low polar; dimethyl diphenyl polysiloxane stationary phase. The two non-polar columns I’ve compared were Rxi-5MS and Rxi-5Sil MS (both 30×0.25×0.25). The essential oil used was the same as in the previous blog; citronella java oil (Figure 1).

Figure 1: Citronella Java oil (5% in acetone) on Rxi-5Sil MS(#13623, black trace) and Rxi-5MS (#13423, blue trace). Method: 100 °C to 300 °C at 11 °C/min (hold 10 min), carrier gas: He at 1.31 mL/min, split: 100:1. The Rxi-5Sil MS results in a slightly faster overall runtime.

From the two chromatograms we can see that analysis on Rxi-5Sil MS is faster with comparable resolution. We’ve decided to take it a step further and speed up the analysis by using a shorter Rxi-5Sil MS column with a narrower internal diameter, i.e. 20×0.18×0.18 and 10×0.15×0.15 (Fig 2). The methods were translated using the EZGC method translator.

Figure 2: Citronella Java oil (5% in acetone) on Rxi-5Sil MS 20×0.18×0.18 (#43602, black trace) and 10×0.15×0.15 (#43815, blue trace). Method 20x018x0.18: 100 °C to 300 °C at 17.5 °C/min (hold 10 min), carrier gas: He at 1.01 mL/min, split: 100:1. Method 10×0.15×0.15: 100 °C to 300 °C at 45 °C/min (hold 10 min), carrier gas: He at 1.01 mL/min, split: 100:1. Translated conditions with smaller bore columns can significantly decrease runtimes.

There is very little resolution lost between the long and the shorter columns, making it a great choice for fast screening.

Here is a link to the wax column selection blog post and blog post about citral analysis.

Pro EZGC Update: Comprehensive 209 compound Library of Brominated Diphenyl Ethers and a New Column Format

While attending Dioxin 2018 in Krakow, I noticed that several academic researchers were studying  the toxicity of specific PBDE congeners not on the standard target compound list for EPA method 1614 (or the EU equivalent). Though PBDE mixtures have been phased out of production and use, the concentrations in the environment have not been declining and are currently still widely monitored. Researchers at Environment Canada demonstrated that although decabromodiphenyl ether (BDE 209) is the primary PBDE in the flame retardant decaBDE, it can be metabolically debrominated by fish (and possibly other animals), forming a variety of penta- and hexa- brominated diphenyl ethers (Stapleton, Alaee et al. 2004).

The standard column used for the analysis of brominated diphenyl ethers by EPA Method 1614 (Figure 1) is the Rtx-1614, a 15 m x 0.25 mm ID x 0.10 µf column with a 5% diphenyl type phase modified for elevated thermal stability. The short column length is critical for EPA method 1614 because BDE 209 is thermally labile; its response relates directly with elution temperature, and because BDE 209 is the primary ingredient in decaBDE, accurate quantitation is critical for determining the scope of contamination.

Figure 1 – Wellington Laboratories BFR-PAR calibration standard collected on a 15 m Rtx-1614. Decabromodiphenyl ether is eluting just after 20 minutes, and the resolution of BDEs 49 and 71 is close, but meets method selectivity criteria.

It is difficult to identify individual PBDE congeners in large homologue groups because their mass spectra are virtually identical (isobaric) and their retention times are similar on the 15 meter column (Figure 2).

Figure 2 – (Top) The elution profile of the 42 hexabromodiphenyl ether homologues collected on a 15m Rtx-1614 column installed in an Agilent GC-MS equipped with an HES. (Bottom) The elution profile of the 42 hexabromodiphenyl ether homologues modeled in ProEZGC

In an attempt to help improve the speed and accuracy of congener identification, especially in the tri- through hepta- brominated homologue groups, we are offering a new high efficiency Rtx-1614 column format (60 m x 0.25 mm x 0.1 μm) as a custom column (PN CC1915) and expanded our ProEZGC PBDE library include all 209 congeners (plus 16 other significant BFRs). A quick literature search leads me to believe I’ve generated the first comprehensive PBDE retention index library. With better separation over a larger time scale, it will be easier to identify individual congeners by using the relative retention time calculated from the EZGC model and select carbon 13 labeled isomers as internal standards.

Figure 3 – (Top) Elution profile of the 42 hexabromodiphenyl ether homologues collected on a 60m Rtx-1614 installed in a Thermo TSQ 9000. The filter used to collect the native hexa-brominated species (black trace) was 641.5 to 483.7. The C13 labeled relative retention time compounds (blue trace) used a mass filter of 653.6 to 493.8. (Bottom) Elution profile of the 42 hexabromodiphenyl ether homologues modeled on a 60m Rtx-1614 using online ProEZGC. The red boxes show an area where there was some disagreement between the real world and modeled chromatogram – the purple box shows what appears to be a gap in the MRM data and possibly missing peaks

It should be noted that model accuracy is highly dependent on an accurate column length and flow. Due to the internal diameter variation inherent to fused silica capillary columns, it is essential that the effective column length be calculated using the column holdup time and head pressure. I translated the 15m run conditions to the 60m column, but my retention times did not match with the model as well as the 15m runs did. These changes in elution temperatures could explain the minor elution profile differences seen between the TSQ 9000 run and the online ProEZGC model (Figure 3) in the region enclosed by the red box. It is also possible that some of the isomers have a much reduced response because I’m only looking at one transition, and different substitution patterns can yield different fragmentation patterns (missing peaks in the purple box). The SIM chromatogram shown for the 15m column data is a sum of multiple ions, but primarily m/z = 483.6.

Finally, the 60m Rtx-1614 is not appropriate for quantitative analysis of the octa-, nona-, or decabrominated diphenyl ethers because they can experience extensive thermal degradation at elevated elution temperatures and extended time on column.

References

Stapleton, H. M., et al. (2004). “Debromination of the Flame Retardant Decabromodiphenyl Ether by Juvenile Carp (Cyprinus carpio) following Dietary Exposure.” Environmental Science & Technology 38(1): 112-119.