Pro EZGC Library Update – Azo Dye Aryl Amines

Azo dyes are commonly used in many products, including food, textiles, and paints. In some azo dyes the azo functional group (R-N=N-R’) can be broken in reductive conditions to form carcinogenic aryl amines. The use of these dyes is regulated in the EU under the Registration, Evaluation, Authorization, and Restriction of Chemicals (REACH) regulation instated by (EC) 1907/2006. We’ve shown before that the Rxi-35Sil MS is an excellent solution for the analysis of 24 aryl amines with chromatograms (, and now we have updated our Pro EZGC libraries with the compounds found in the chromatogram plus some extras.


Name CAS
1,3-Benzenediamine, 4-methoxy- 615-05-4
1,4-Benzenediamine 106-50-3
2,4,5-Trichloroaniline 636-30-6
2,4,5-Trimethylaniline 137-17-7
2,4-Diaminotoluene 95-80-7
2,4-Xylidine 95-68-1
2,6-Xylidine 87-62-7
2-Amino-4-nitrotoluene 99-55-8
2-Aminobiphenyl 90-41-5
2-Naphthylamine 91-59-8
3,3′-Dichlorobenzidine 91-94-1
3,3′-Dimethoxybenzidine 119-90-4
3,3′-Dimethylbenzidine 119-93-7
3-Chloro-o-toluidine 87-60-5
4,4′-Diaminodiphenylmethane 101-77-9
4,4′-Methylenebis(2-chloroaniline) 101-14-4
4,4′-Oxydianiline 101-80-4
4-Aminobiphenyl 92-67-1
4-Chloroaniline 106-47-8
4-Chloro-o-toluidine 95-69-2
4-Methoxy-m-phenylendiamine 614-05-4
Aniline 62-53-3
Benzenamine, 4,4′-thiobis- 139-65-1
Benzidine 92-87-5
o-Aminoazotoluene 97-56-3
o-Anisidine 90-04-0
o-Toluidine 95-53-4
p-Aminoazobenzene 60-09-3
p-Cresidine 120-71-8


Peaks tR RS
1 Aniline 1.65 29.3
2 o-Toluidine 2.06 27.7
3 2,4-Xylidine 2.5 2.1
4 2,6-Xylidine 2.54 2.1
5 o-Anisidine 2.65 6.1
6 4-Chloroaniline 2.88 12.9
7 p-Cresidine 3.21 1.9
8 2,4,5-Trimethylaniline 3.24 1.9
9 3-Chloro-o-toluidine 3.42 1.9
10 1,4-Benzenediamine 3.46 1.1
11 4-Chloro-o-toluidine 3.48 1.1
12 2,4-Diaminotoluene 4.3 29.2
13 1,3-Benzenediamine, 4-methoxy- 4.92 7.9
14 2,4,5-Trichloroaniline 5.09 7.9
15 2-Naphthylamine 5.41 1.5
16 2-Aminobiphenyl 5.44 1.5
17 2-Amino-4-nitrotoluene 5.78 15.1
18 4-Aminobiphenyl 6.59 36
19 p-Aminoazobenzene 8.24 13.8
20 4,4′-Oxydianiline 8.56 2
21 4,4′-Diaminodiphenylmethane 8.61 2
22 Benzidine 8.66 2.1
23 o-Aminoazotoluene 9.07 15.1
24 4-Methoxy-m-phenylendiamine 9.53 6.1
25 3,3′-Dimethylbenzidine 9.75 6.1
26 Benzenamine, 4,4′-thiobis- 10.64 7.7
27 4,4′-Methylenebis(2-chloroaniline) 11.03 0.2
28 3,3′-Dichlorobenzidine 11.04 0.2
29 3,3′-Dimethoxybenzidine 11.12 1.5

As seen above, EZGC provides a <12min run on the Rxi-35Sil MS with one partial and one full coelution, both of which are mass resolved.


For more information on Azo dyes, see our previous work at


LC Preventative Maintenance Tips and Tricks – The UltraShield

We’re providing a creative trick for extending guard cartridges and column lifetimes.  This preventative maintenance consumable also acts as a handy troubleshooting tool for problematic back pressure readings.  Drum roll….The UltraShield!

Normally, the UltraShield is used as an affordable UHPLC pre-column filter and takes the place of a guard column.  These filters have minimal effect on efficiency with all UHPLC columns and systems and, when properly used, should not impact separations.  However, Paul Connolly at Restek had an idea to use the UltraShield as a preventative maintenance consumable for extending column and guard cartridge lifetimes and as a simple check point for troubleshooting back pressure issues which are common pain points for cannabis testing.  The UltraShield can be utilized for many new and refurbished LC systems that are frequently used for potency testing.  You can add the UltraShield in several locations in front of your guard cartridge and analytical column.  But ideally, adding it right after the injector will protect as much of your system and consumables as possible.   For this modification you’ll only need two 1/4” x 5/16” wrenches ( and an UltraShield as pictured below.

After acquiring these parts it’s time to get to work!  In the picture below, you can see a common 6 port valve set up.

To install the UltraShield, simply loosen and remove the sample loop fitting.  Then properly seat and tighten the UltraShield between the six port valve and the sample loop fitting.  In the picture below, you can see the completed installation of the UltraShield.

It’s that quick and simple!

Now your six port valve, lines, guard cartridge, and analytical column have an extra layer of protection from those dirty and complex cannabis samples!  In addition, when you notice elevated back pressure issues, you can easily change this filter out with another to see if the problem is corrected.  When you add the costs for replacing a guard cartridge and analytical column compared to replacing an UltraShield, it’s an affordable preventative maintenance consumable labs should be utilizing.

Helpful Tip:

I learned the hard way to keep UltraShields in a plastic container.  The ferrule for the UltraShield rolled off the lab bench during installation.  It took me a LONG time to find the PEEK ferrule.  It’s in the picture below….have fun spotting it (I had a joyful time with the task).

GC Inlet Liner Selection, Part IV: Liner Volume and Diameter

In parts I and II of this blog series I discussed the various liner configurations available for both split and splitless analyses.  One thing I did not mention was liner volume and inner diameter.  Most of the liners I compared are available in multiple inner diameters.  For the comparisons I did, I used a 4 mm inner diameter, which is relatively common for several instrument manufacturers, including Agilent.

Since many of Restek’s liner configurations are available in multiple inner diameters, what would make you choose one over the other?  One of the primary advantages of smaller inner diameters (which translates to smaller liner volumes), is the linear velocity of the carrier gas will be higher on the smaller I.D. liner, given the same flow setpoints.  This translates to faster column loading and less time for the analytes to spend in the inlet.  For volatile compounds this produces a narrower analyte band, which improves both resolution and sensitivity.  Figure 1 illustrates an example of peak shape differences for volatile compounds on a 2 mm liner vs a 4 mm liner.  Note that on the 4 mm liner the poorer focusing/peak shape significantly reduces resolution between these two peaks.  Another potential advantage of a smaller ID liner is that the active compounds will spend less time in the inlet, reducing the chances of adverse interactions, such as adsorption and reactivity.

Figure 1: Splitless injection of volatile compounds on a 2 mm ID single taper liner with wool vs. a 4 mm ID single taper liner with wool.

Ok great, so smaller ID’s lead to sharper peaks and less time for adverse interactions to occur in the inlet, so it’s better to always go smaller?  Well…not exactly.  When you inject a solvent into a hot inlet it is going to rapidly expand into a much larger vapor cloud.  The actual size of this vapor cloud will depend upon the solvent itself, as well as the temperature and pressure conditions found in the inlet.  Some solvents, such as water and methanol, will expand more than others (i.e. hexane, dichloromethane).  In fact, water can expand more than 1000 times in volume as it’s transferred to a vapor. Figure 2 gives an example of the difference in expansion volume when injecting 1 µL of hexane vs 1 µL of water under the same inlet conditions.

Figure 2: Comparison of solvent vapor expansion volumes for hexane and water, injected at 1 µL under the same inlet conditions.

If you do not have enough liner volume to accept the total size of the vapor cloud as it expands, the vapor can expand beyond the confines of the liner into carrier gas lines, purge vent lines, etc.  This phenomenon is known as backflash and can potentially cause issues.  For one, backflashing can lead to contamination of the system, as your solvent will possibly carry sample material into carrier gas lines and purge lines, which can later off-gas, leading to contaminant peaks.  In addition, if you are vaporizing outside of the confines of your liner, you are adding variability to your injections, making them less consistent and potentially leading to higher injection to injection %RSD’s.  Backflash can also produce poor peak shapes, including fronting, tailing, or split peaks.

There are some ways to mitigate backflash, even when injecting large volumes.  The use of wool can help to prevent backflash, since the wool provides a large surface area for solvent evaporation. This creates an evaporative cooling effect, trapping the analytes on the wool, keeping them within the liner even if some solvent may ultimately backflash.  Using an injection technique such as pulsed splitless can also help, as it uses a temporary increase in inlet pressure to load the analytes onto the column faster.  This pressure spike will reduce solvent expansion volume.  Care must be taken when using a pulsed injection mode, though, as sometimes poor chromatography of volatiles may result from the sudden pressure changes.

Beyond the issue of backflash, a smaller ID liner will also have less surface area and can become contaminated faster if injecting dirty samples.  This could result in more frequent liner changes compared to a larger volume liner.

So ultimately, liner volume can be a balancing act when injecting liquid samples.  Smaller volumes can help with peak shapes of volatiles, as well as reduce inlet residence time, but can also lead to backflash or quicker fouling of the liner from sample matrix.  Restek has a solvent expansion calculator, which can help you determine whether your solvent injection will stay within the confines of the liner: Solvent Expansion CalculatorFigure 3 gives some common liner configurations with the physical and effective volumes listed.  Effective volume takes the volume of carrier gas into consideration and is half of the physical volume of the liner.

Figure 3: Common liner physical and effective volumes.

A Word on Gas Analyses

The above discussion focused on liquid injections in split and splitless mode.  If you are introducing your sample as a gas, with techniques such as headspace, purge and trap, and SPME, you no longer have to worry about solvent and you are also working with more volatile analytes.  In this case, a small ID liner such as a 1 mm ID (or smaller) will lead to the quickest column transfer and therefore sharp, narrow peaks.  Also, since there is no need to evaporate the sample or protect the column from matrix material, the use of wool or another obstruction is not necessary when introducing gaseous samples.

Links to blogs in this series:

GC Inlet Liner Selection: An Introduction

GC Inlet Liner Selection, Part I: Splitless Liner Selection

GC Inlet Liner Selection, Part II: Split Liners

GC Inlet Liner Selection, Part IIB: Split Liners Continued

GC Inlet Liner Selection, Part III: Inertness

GC Inlet Liner Selection, Part IV: Liner Volume and Diameter

Not all 3.2 L air sampling canisters are 3 liters. Wait… what!?

The other day a customer contacted me to share the following discrepancy she has observed on several occasions: “when I collect ambient air samples with Restek 3 L canisters and collocated 3.2 L canisters from the competition, the canisters fill in the same amount of time.” Obviously this does not make sense, so I immediately go into trouble-shooting mode asking questions like: what flow controllers, what size critical orifices, what flow rates, are all the canisters evacuated the same, etc…

At no point in time are we able to find anything out of the order. The customer was running an apples-to-apples comparison and by all accounts the competitor’s 3.2 L canisters should have more vacuum left at the end of the sampling duration when compared to our 3 L canisters. At this point in time the only stone left un-turned (you already guessed it from the title of the blog) was the volume of the canisters.

We used the following two-pronged approach to determine the volume of our 3 L canisters and the competitors 3.2 L canisters:

  1. Weighed the canisters empty and weighed them filled with water on our verified shipping scales in the shipping department (yes, the same calibrated scales we used to show you our beefier 6 L canisters only weigh 8 oz. more than the competition). 3 weights were obtained from 3 different scales, like so:                                                                                                  
  2. Measured the volume of water, which came out of the canister (i.e., post weighing). We only had a 1,000 mL graduated cylinder, so we had to take 3 measurements to get the complete volume. This is a picture of the last reading from the competitor’s 3.2 L canister:

Here are the values we obtained from the aforementioned investigation:

As you may see in the above table, approaches 1 and 2 indicate the volume of the competitor’s 3.2 L canister is 2.849 and 2.860 L, respectively. These measurements were exactly the same as our 3 L numbers, which for the record are consistent with our internal specifications for min, nominal, and max 3 L canister volumes (i.e., 2821, 2853, and 2885 mL). Both approaches agree reasonably well with one another (99.6% agreement). We speculate the minor discrepancy is the graduated cylinder indicates an accuracy of ±6 mL. With three measurements used, we could have had upwards of an 18 mL swing, so 11 mL is well within limits.

Regardless, both measurements clearly indicate the competitors claimed 3.2 L volume is a far cry away from 3.2 L. Which all makes sense with why our customer saw the competitor canisters filling at the same rate as our 3 L canisters. Now I find myself wondering what other canister dimensions are not as claimed!? I also wonder why is there a 3.2 L canister on the market anyway? Maybe this is the competitor’s approach to excluding others from RFQs (request for quotation). We do not sell a 3.2 L, so we cannot compete on these quotes, but oddly enough neither does the competition.

A closer look at selectivity and tailing: Essential oils and Wax phases

The most common way to analyze essential oils is to use GC with either a FID or MS. The wax-based columns are the number one choice for many users and following up on an excellent blog post by Corby, I wanted to evaluate two wax in Restek inventory: Stabilwax and Rtx-Wax. The essential oil we chose was Citronella Java oil, which has high levels of citronellal (CAS 106-23-0), β-Citronellol (CAS 106-22-9) and Geraniol (CAS 106-24-1). Figure 1 shows a comparison between the two columns. While the chromatograms are very similar, a couple of things stand out.


Figure 1: Citronella Java oil (5% in acetone) on Stabilwax (#10673 , black trace) and Rtx-Wax (#12423 , blue trace). Method: 100 °C to 250 °C at 11 °C/min (hold 10 min), carrier gas: He at 1.31 mL/min, split: 100:1

Read the rest of this entry »

Want to learn about “Injection Techniques in GC” or “Practical Maintenance and Troubleshooting in GC”? Sign up to join our half day course during Pittcon, Chicago.

Please join one of our courses presented next Pittcon in Chicago:

“Injection Techniques in GC”:  Monday,  March 2,  08:30-12:00, Session: SC1230.

“Practical maintenance and troubleshooting in Gas Chromatography”: Tuesday, March 3, 08:30-12:00, Session: SC1231

For location, visit short course office at S100C.


Injection Techniques in Gas Chromatography

In Gas chromatography the most important process is to get the sample into the column. If sample transfer is not optimized, the results will not be reliable. Goal of this course is to understand the different injection techniques used and the process how to obtain a narrow injection band.In this half day course we will discuss the basics of the most popular injection techniques that are used in Gas chromatography. Techniques like split, splitless, direct, on-column and large volume injection will be discussed in detail. Also the selection of liners, retention gaps and columns will be addressed.All techniques will be explained using practical examples. At the end we will zoom into some typical “injection” troubleshooting examples


Practical maintenance and troubleshooting in Gas Chromatography

In Gas chromatography 90% of the troubles experienced, is happening in the injection system. In this training we will discuss the purpose and impact of the critical parts (consumables) present in split and splitless injection systems and how this impact in a maintenance schedule. At the end we will discuss a series of practical examples via troubleshooting exercises. In this half day course we will discuss the maintenance and optimization challenges for Split and Splitless injection techniques. We will zoom in carrier gas choice and purity, tubing, connections, septa, ferrules, seals, liners, column-coupling, installation and column maintenance. Also column operation/optimization and extending column life time will be discussed.


Chiral separation on a C18 column? Separation of d- and l-amphetamines, Part II

To continue my blog part 1 (Part 1: , where I have briefly discussed the importance of separating the d and l isomers to accurately identify the illicit isomer using an achiral method on a Raptor C18 column employing a pre-column derivatization technique. Today I’d like to discuss more about the matrix of interest, sample preparation and derivatization, chromatographic and mass spectrometric conditions.

Sample Preparation: 50 µL of calibration standard or QC sample prepared in analyte free pooled human urine (spiked in the range of 50-5000ng/mL) was aliquoted into a micro centrifuge tube. 10 µL of a working internal standard (20 µg/mL (±)-amphetamine-D11 and (±)-methamphetamine-D11 in water) and 20 µL of 1M NaHCO3 was added and vortexed at 3000 rpm for 10 seconds. After vortexing, 100 µL of 0.1% (w/v) Marfey’s reagent (1-fluoro-2-4-dinitrophenyl-5-L-alanine amide) prepared in acetone was added, vortexed, and heated at 45 °C for 1 hour in a water bath. Samples were allowed to cool to room temperature before the addition of 40 µL of 1M HCL in water. The d- and –l amphetamines are now converted to DNPA derivatives. The sample was then vortexed and evaporated to dryness under nitrogen at 45 °C. Samples were reconstituted in 1 mL (20x dilution) of 40:60 water: methanol (v/v) and filtered using Thomson SINGLE StEP standard filter vials (cat# 25893) and then injected for LC-MS/MS analysis.

Matrix…. What Matrix? Currently urine is the sample of choice for pain management or toxicology labs for routine drug testing, because of its ease of sample collection and clean up as well. So we bought 9 lots of drug free human urine lots from Bio IVT, the samples were derivatized and analyzed on LC-MS/MS to check for any presence of endogenous Amphetamines and other significant interference peaks. All the urine lots were free from interferences or presence of amphetamines. Pooled urine was prepared from the analyte free 9 urine lots and was used for method development and validation. Surine was evaluated as well and no matrix peaks were evident and amphetamines were well separated.

Urine dilution factor? How do I decide which dilution factor was right for this analysis. Well, I derivatized both 100uL (10x) and 50uL (20x) pooled urine: 20x dilution gave best results with good sensitivity and less matrix effects compared to 10x and increasing dilution rate more than 20x decreased the signal sensitivity.

Sensitivity issues? One of our customer complained about poor signal intensity of the analytes even after derivatization, a chemists nightmare. I blame the derivatizing reagent here, it is very light sensitive and should be prepared fresh in acetone and stored in dark at 40C. However, the derivatized samples were found to be stable for at least 48 hours in the auto sampler rack, no significant fluctuations in signal intensity was noticed.

Can Cleaning the MS source help? Definitely, especially in this particular case because you are just diluting the urine sample. I clean my MS source once a week with water, Methanol and Acetonitrile after running hundreds of urine samples or any other biological matrix like whole blood, oral fluid etc..

Chromatography: Now that our samples are derivatized and ready for the analysis, I’d like to discuss more about the LC conditions. A Raptor C-18 100 x 2.1 mm, 2.7 µm column was utilized for the separation of the DNPA derivatives of Amphetamines and their respective deuterated internal standards in 7 minutes (Figure 1 & 2). Using a simple derivatization and dilution procedure along with a Raptor C18 column— good baseline resolution of the target compounds was obtained, allowing easy peak identification and quantitation.

Fig 1: Separation of d- and l- Amphetamine and Methamphetamines enantiomers in fortified Human Urine at 500ng/mL (TIC)


Fig 2: Separation of d- and l- Amphetamine and Methamphetamines enantiomers in fortified Human Urine at 500ng/mL (XIC)

Instrument Conditions

Analysis of amphetamines by LC-MS/MS was performed on a Shimadzu Prominence HPLC equipped with a SCIEX API 4000 MS/MS. Instrument conditions were as follows and analyte transitions are provided in Table 1.

Analytical column: Raptor C18 2.7 µm, 100 mm x 2.1 mm (cat# 9304A12)

Guard column: Raptor C18 EXP guard column cartridge (cat# 9304A0252)

Mobile phase A: 0.1% Formic acid in water

Mobile phase B: 0.1% Formic acid in methanol

Flow rate:    0.5 mL/min

Injection volume: 10 µL

Column temp.:  35 °C

Ion mode: Negative ESI








Table 1: Analyte Transitions for the Analysis of Amphetamines by LC-MS/MS.










Hopefully this workflow and some of the tips discussed above on both sample prep and chromatography and can help get you some good baseline separation and resolution of these enantiomers to identify the illicit isomer.

In my next blog I will share some interesting data that mimics the real urine sample concentrations containing very high and low concentrations of the legal and illegal isomers present in urine and optimization of derivatizing procedure. Stay tuned till next one….


  1. Newmeyer, N. M, Concheiro, M and Huestis, A. M. J Chromatogr A. 2014; 1358: 68–74.
  2. Foster, S. B and Gilbert, D. D. J Analytical Toxicology.1998; 22:265-9.

Falling Victim to One of LC’s Classic Blunders: Mismatching Your Diluent and Mobile Phase

An early lesson most of us learn in liquid chromatography is this: Always match your diluent to your mobile phase.  Once the exams are done, if you learned this in a college course, or your manager has walked away, you start flexing this “requirement” a little to see how matched they really need to be.  I get it! You don’t want to have to blow off all your organic, because that takes time.  Or if you’re scouting various columns, you just want to use the same sample you used in HILIC and reversed-phase tests.

In some cases, you can get away with this possibly because your analyte has plenty of affinity for your stationary phase, and that extra bit of strong solvent isn’t enough to ruin your separation.  While in other cases your analytes elute well passed the dead volume of the column, so your peak shape is still good.  In these instances, it can be OK to bend the “rules” a little.  Sometimes though, you have an analyte so sensitive, that even 5% residual acetonitrile in your diluent is enough to ruin your chromatography altogether.  I present to you this case study on acrylamide.

Acrylamide is a small polar molecule (show image).  It is akin to Iocane powder.   It’s odorless, tasteless, dissolves instantly in liquid, and is [not] among the more deadly poisons known to man.  Though as a side note it is thought to be a probable carcinogen with repeated prolonged exposure.  In reality, acrylamide is soluble in water and methanol and is very soluble in acetonitrile. Additionally, it is difficult to retain on most LC columns.  Earlier this year, we developed a product specifically for acrylamide analyses and have developed applications in various matrices.

Early on in method development, we noticed that acrylamide was very sensitive to residual organic solvent leftover from the sample preparation procedure.  In the published EN 16618 method, the final extract is in a water/methanol eluate.  The step before analysis is a lengthy blow down; a step many, including myself, are tempted to short change.  In the QuEChERS method, the final sample diluent is acetonitrile (MeCN), which is even more detrimental to chromatographic analysis and requires complete solvent replacement.

To better understand the sensitivity of acrylamide to organic diluent, we deliberately performed an extraction of acrylamide from potato chips/crisps and added organic solvent in known volumes to see how this changed the chromatography from an ideal sample prep to one that might be more “lazy.”  In our method, we use a 100% aqueous mobile phase with 0.001% formic acid, meaning that our diluent should also be 100% aqueous.   As more organic was added to the diluent, retention time decreased and peak width broadened.  While 10% methanol in the diluent still gave decent retention and chromatography, 20% or more methanol resulted in poor peak shape and retention.  In the case where there is even 5% residual MeCN, massive peak broadening occurred and when more MeCN was present, the acrylamide peak split and then eventually eluted with the matrix contaminant peak.

While acrylamide qualifies as an extreme example, this data shows the necessity of ensuring your diluent matches your starting mobile phase.  We also noticed that having strong organic in your autosampler rinsing routine could result in a similar effect on your chromatography.  So in 2020, and beyond, next time you see poor chromatography with early eluting or poorly retained peaks, consider checking your diluent composition.  Your solution might be as simple as making sure your diluent better matches your mobile phase.

Happy New Year from all of us here at Restek!


What impurities are hiding in your monocyclic aromatic solvents?

As we approach the end of the year we are indulging in a bit of nostalgia and most experts agree that reminiscing about the past is a good thing1. It makes us more optimistic and generous. Researchers also noticed that in cold rooms we are more inclined to nostalgize than if we are in a warm environment. And, since my boss, Chris, is a fan of cold office spaces, I will spend a few minutes on what we did last year.

Past year, we devoted some time to validate the products for the ASTM Petrochemical Method Chromatography Product Guide. The ASTM guide contained mostly D02 GC methods – methods for the analysis of petroleum products. Now, we have expanded it to include D16 ASTM chromatographic methods. D16 Committee is responsible to standardize testing of the aromatic, industrial, specialty, and related Chemicals.

One of the more popular methods in the D16 section is ASTM D75042, a method for the analysis of trace impurities in monocyclic aromatic solvents by GC and the effective carbon number. This method uses one analytical procedure and column for an analysis of a variety of industrial solvents from different sources. This includes synthesized or refined solvents such as benzene, toluene, p,m,o-xylene, ethylbenzene, and styrene. Check out our ASTM guide for chromatograms of a few commercially available solvents.

The analysis is quite long, and if we simplify, the only resolution requirement in the method is to separate any impurity from the solvent being analyzed. The impurities in the solvent are reported as a sum of total non-polar hydrocarbons, a few individual mono-aromatic compounds, and later eluting aromatic compounds are summarized and reported as a group. Reported analytes can be easily separated on a long column which led to the use of shorter columns cited in the literature.

I evaluated the 20mm x 0.18mm x 0.36µm Stabilwax column and found it to have sufficiently separated the method resolution mixture in p-xylene.

Figure 1: Chromatogram of a method resolution mixture in p-xylene overlaid with a synthetic blend of compounds most commonly found as an impurity in monocyclic aromatic solvents on a 20m column. The analysis conditions were optimized using the Restek EZGC method development suite (method translator and chromatogram modeler)

Read the rest of this entry »

Two for the Price of One? Using CPP to Simultaneously Clean Pesticide and PCB Extracts

In a recently published blog, we introduced the advantages of cleaning EPA Method 8081 extracts using CarboPrep Plus (CPP) as a replacement.  The sample preparation for PCBs by EPA Method 8082 is commonly combined with the pesticides EPA Method 8081 for the extraction step then split into two portions for sample cleanup.  The logical question is: can I run the pesticide/PCB extract through CPP prior to splitting for separate analysis, thereby avoiding separate clean up procedures?

Before I answer, I should explain that the reason the single extract is sometimes split for post extraction processing is the sulfuric acid procedure cannot be applied to the pesticide portion, it will degrade the acid sensitive analytes.  Sulfuric acid cleanup has been used with PCB extracts since it can be effective at removing much of the colored material that Florisil leaves behind, so it is possible that CPP may be used instead of splitting the sample which results in additional extraction steps.

 Why clean the PCBs extracts at all?  If pesticide extracts are run on the same instrument regular maintenance would be required to meet the stringent inlet inertness requirements for endrin and DDT.  PCB methods typically do not have the same stringent requirements, however samples typically have high concentrations of non-volatile matrix compounds that build up in the inlet effecting chromatography, which effects quantitation.

An effective cleanup is also important since chromatographic interferences make it difficult to perform Aroclor pattern identifications.  Other challenges include weathering which can significantly change the Aroclor pattern combined with different Aroclor mixtures.

One characteristic to the CPP cartridge based upon experience with GCB, lies in its strong affinity for certain compounds.  In this case, the non-ortho PCB congeners have varying levels of interaction to GCB.  The logical question is how the pattern of various Aroclors will be affected by this preferential adsorption of some congeners relative to others when using CPP. Jason Thomas set off on a crusade to answer these questions. First, he started by running Aroclors through the cartridge using the standard elution method of 10 mL hexane/acetone 90:10.  Second, he ran the individual congeners through in a series of mixes, collecting fractions in 5 mL increments and identified congeners that required an additional amount of solvent for elution.  Finally, Jason compared our concentrations with the various Aroclor ratios of congeners reported in a paper by Frame & Cochran1.

First, all PCB congeners eluted in the first 5 mL elution volume except for the following PCBs in Table 1.  The congeners that exhibited retention beyond the first 5 mL of elution were those possessing no chlorine substitution at the otho positions (known as non-ortho PCB congeners) as illustrated in the chart below:

Figure 1: Recoveries of selected PCBs with different chlorine substitutions measured against amount of elution solvent necessary to remove the analyte from the carbon.

The non-ortho PCB congeners have a longer retention on the carbon, but for quantitation of Aroclors, these congeners may not be relevant as it is recognized that not all 209 are actually present in the Aroclor mixtures, which is where the Frame-Cochran paper comes in handy.  Ratios of these congeners in each specific Aroclor were reported.  Adding this information to the above table, it becomes apparent that only five of these non-ortho congeners (highlighted in orange) are expected to appear in any meaningful quantity as defined here with a cutoff of 0.3%.

As an example, let us compare the effect of CPP on the elution of Aroclor 1221 using the standard pesticide Florisil methodology, 10 mL 10% acetone/hexane elution solvent.  We will specifically focus on PCB-15 in Aroclor 1221 as it represents the congener with the highest % representation in a given Aroclor, (4.2% in 1221).

Chromatogram 1: Demonstration of the elution of Aroclor 1221 using standard pesticide Florisil methodology. Control standard is shown in blue and the extraction standard is shown in red.

There is a notable response missing from the CPP processed standard (red trace) versus the control standard (blue trace), roughly half, as expected with a 10 mL elution-based data in Table 1 above.  As indicated, PCB-15 is a minor component of the Aroclor 1221 pattern, the absence of which will not affect quantitation.

This does not constitute a significant change in the overall appearance of 1221.  In the event of quantitation however, choosing this as one of the quant peaks will provide some degree of error, but there are clearly several more prominent and representative peaks to choose from in order to establish the 3 to 5 peak calibration required by EPA 8082.

None of the heavily retained congeners are present in the heavier Aroclor mixtures, 1254 – 1268.  This may also be corroborated by comparing a 1254 standard processed with CPP and one without.  No discernable difference exists between the two suggesting that none of the components of this Aroclor is retained on the SPE cartridge.

Chromatogram 2: Demonstration of 1254 recoveries using CPP where blue is the standard and red is the extract. Little to no differences were observed.

In summary, CarboPrep Plus may be used to clean EPA 8082 extracts for Aroclor quantitation, provided caution is exercised in choosing the quant peaks, setting up the possibility of processing a combined pesticide/PCB extract without the sulfuric acid clean up step.

  1. Frame, G. M., Cochran, J. W., Bowadt, S. S. (1996), Taming Complete PCB Congener Distributions for 17 Aroclor Mixtures Determined by 3 HRGC Systems Optimized for Comprehensive, Quantitative, Congener-Specific Analysis. Journal of High Resolution Chromatography, 19: 657-668.