Use the “Subscribe” link at right. It just takes a few seconds!
Is separation of benzo[b]fluoranthene and benzo[k]fluoranthene on 5% phenyl-type GC columns really that important for environmental analyses?
Some gas chromatographers might say so because of the following statement in US EPA Method 8270D (Semivolatile Organic Compounds by Gas Chromatography – Mass Spectrometry (GC-MS):
“22.214.171.124 Structural isomers that produce very similar mass spectra should be identified as individual isomers if they have sufficiently different GC retention times. Sufficient GC resolution is achieved if the height of the valley between two isomer peaks is less than 50% of the average of the two peak heights. Otherwise, structural isomers are identified as isomeric pairs. The resolution should be verified on the mid-point concentration of the initial calibration as well as the laboratory designated continuing calibration verification level if closely eluting isomers are to be reported (e.g., benzo[b]fluoranthene and benzo[k]fluoranthene).”
The language above ignores the fact that benzo[j]fluoranthene is a polycyclic aromatic hydrocarbon (PAH) isomer consistently found at significant levels in environmental (and food) samples, and more importantly, that it coelutes with either, or both, benzo[b]fluoranthene and benzo[k]fluoranthene on 5% phenyl-type GC columns. In this blog post, I will define the extent of the coelution on 30m x 0.25mm x 0.25µm Rxi-5Sil MS (5% phenyl as silphenylene, similar to DB-5MS) and 30m x 0.25mm x 0.25µm Rxi-5ms (5% phenyl as diphenyl, similar to DB-5).
PAHs were first analyzed using hydrogen efficiency-optimized flow (EOF), and optimal heating rate (OHR), with an Agilent 6890 GC-FID. Injections of PAH standards (prepared from SV Calibration Mix #5 / 610 PAH Mix and a custom benzo[j]fluoranthene standard) at a split ratio of 100:1 into a 4mm Precision split liner with wool served to minimize injection band widths to keep overall system efficiency high. In addition to OHR, chromatograms for each column were generated at 1.5 x OHR and 0.5 x OHR to force different elution temperatures for the benzofluoranthenes to see what effect that had on their separation.
As you can see in Figure 1, the Rxi-5Sil MS shows an intractable coelution for benzo[b]fluoranthene and benzo[j]fluoranthene under all experimental conditions, with elution temperatures around 278, 293, and 257°C for OHR, 1.5 x OHR, and 0.5 x OHR, respectively. The Rxi-5ms has a different coelution, that of benzo[j]fluoranthene and benzo[k]fluoranthene for OHR and 1.5 x OHR, but interestingly is able to partially resolve the benzofluoranthenes under 0.5 x OHR conditions (Figure 2). Elution temperatures in this case were approximately 280, 295, and 258°C (OHR, 1.5 x OHR, 0.5 x OHR).
What does this coelution situation mean for environmental analysts? First, all quantitative values reported for benzo[b]fluoranthene and benzo[k]fluoranthene are likely inaccurate for at least one of the isomers depending on the GC column choice if using a “5-type” stationary phase. Second, technically speaking, analysts should not be reporting the benzofluoranthenes as “individual isomers” according to the language of EPA Method 8270D when using a “5-type” GC column.
If an analyst needs to report benzofluoranthenes as “individual isomers”, a 30m x 0.25mm x 0.25µm Rxi-17Sil MS (or even a 15m version) will work well. If the challenge is to separate the benzofluoranthenes and, triphenylene and chrysene (228 m/z coelution on 5-type columns), consider the Rxi-PAH GC column.
Sample loading capacity for a GC column (also known as “column capacity” and “sample capacity”) is essentially the amount of non-active compound that can be put on a GC column and chromatographed at some set of conditions where the peak shape is symmetrical. Conversely, if a GC column is overloaded with a component amount, the peak will exhibit the classic “shark fin” shape. Most vendors estimate the sample loading capacity for a 0.25mm x 0.25µm GC column to be around 50-100 ng per analyte. But is this even close to accurate? Perhaps, but it depends on the compound of interest.
Environmental analysts already know that polycyclic aromatic hydrocarbons (PAHs) are prone to overload on GC columns, especially the typical “five” type columns used for semivolatile organic compound analysis in EPA methods, e.g., 8270 and CLP. Although the CLP semivolatiles method states that a GC column should be able to “accept up to 160 ng of each compound listed in Exhibit C (Semivolatiles), without becoming overloaded”, I could find no quantitative data indicating when overload occurs for Exhibit C PAHs on the commonly used 0.25mm x 0.25µm GC columns. That’s my lead-in to say, “I’ll test it myself!” With the follow-up… “And post about it on ChromaBLOGraphy”.
All work was done with a 30m x 0.25mm x 0.25µm Rxi-5ms utilizing hydrogen efficiency-optimized flow, and optimal heating rate, with an Agilent 6890 GC-FID. Injections of PAH standards (prepared from SV Calibration Mix #5 / 610 PAH Mix) at a split ratio of 10:1 into a 4mm Precision split liner with wool served to minimize injection band widths to keep any peak deformations associated with the column, and not the inlet. As you can see in Figure 1 by the “shark fin” shape of the peaks and the unacceptable resolution between benzo[b]fluoranthene and benzo[k]fluoranthene, and indeno[123-cd]pyrene and dibenzo[ah]anthracene, 200 ng of each of these PAHs greatly exceeds the sample loading capacity of this column. Figure 2, at 50 ng each compound, also shows gross overloading of PAHs, and although the separation starts to improve between critical pairs, the retention times are still shifted later based on peak deformation. Finally, at 12.5 ng (Figure 3), the chromatogram looks good in separations and retention times, comparing relatively nicely to Figure 4 at 3.13 ng, where we would not expect overload.
Even at 25 ng of later eluting PAHs (Figure 5), we have overload, which puts the maximum sample loading capacity of the 30m x 0.25mm x 0.25µm Rxi-5ms GC column under these operational conditions between 6.25 and 25 ng each PAH, an order-of-magnitude less than what is suggested to be possible by the CLP method. Can we tolerate overload and still get our work done? That depends on your data quality objectives, which might include:
Separation between critical, often isobaric, pairs.
Overall required peak capacity for the chromatogram (the 200 ng overload cut the peak capacity by almost 3).
Capacity of the detector to handle overload (MS systems are very sensitive these days).
I prefer the peaks look more like those seen in Figure 7, and given the sensitivity of today’s MS systems, we can inject less (maybe via “Shoot-and-Dilute”), see almost as much, and probably keep our systems up longer due to less involatile “dirt” put on the GC column.
Some of you have noticed that included with several of our baseplate traps are a package of two small O-rings. Technical Service has been informed that these are the O-rings which fit onto the check valves on your baseplate (see red arrows below). Does this mean that you need to replace these O-rings? Not necessarily. Carefully inspect the O-rings on the check valves, and if you do not see any signs of wear (cracks, deformation, etc…), you probably don’t need to replace them. However, if they do exhibit signs of wear, go ahead and replace them.
If you also need the larger O-rings that are located on the baseplate, at the base of the check valves (as identified by the green arrows), you can obtain those by purchasing Restek catalog number 22023.
“When you come to a fork in the road, take it.” Maybe that advice worked for Yogi Berra, but does not help much in decisions about your lab equipment.
If you are not certain whether you should use a Raptor™ column, please click on the title below to read my previous blog post before proceeding.
Once you have made the decision to take advantage of the benefits of SPP, your next question might be which particle size to use, the 2.7 or 5 µm Raptor™. The 2.7 µm column provides higher efficiencies, but the column must be matched with an instrument that is appropriate, in terms of system pressure and internal dwell volume. These considerations are described below:
Primary Consideration- System Pressure
As is the case with a 3 µm fully porous particle, the 2.7 µm Raptor™ column will operate at backpressures a little higher than the corresponding column with a 5 µm particle size. Pressure is a function of particle size, as well as column dimensions, flow rate and solvent properties, but is not a function of porosity. The following is a good example of how you can expect flow rate and particle size to affect the column back pressure. Please keep in mind that methanol and water both have higher surface tension, which results in higher backpressure. The data shown here was generating using acetonitrile/ water mobile phase.
Although the earlier blog post Building Up Pressure on HPLC? was written for fully porous particles, you can still get a pretty good idea of how pressures are affected by column dimensions, particle size, and solvent content by looking at the table shown in the post. The optimal flow rates will be a little different for SPP columns. Though the “optimal” flow rate for a SPP column is not much different versus a fully porous column, the SPP columns maintain their efficiency over a broader range of flow rates. Here is a good illustration, using a Van Deemter plot.
Secondary consideration-Dwell Volume
There is another factor to consider when transferring your method to a 2. 7 µm Raptor™, and that is the system dwell volume. In other words, dead volume in your LC system may negate the advantages of choosing a 2.7 µm SPP column over a 5 µm particle (either fully porous or SPP). This is of highest concern when using columns of smaller dimensions also. A symptom of excess dwell volume is be band broadening, and can be confirmed by running the same column on a system with less dwell volume. Although there are some system updates that can be made to reduce dead volume, the options are limited and it is simpler to start with a well-matched system and column from the beginning. Fortunately, LC systems with maximum pressures of about 600 bar (8700 psi) are usually also designed with a lower internal dwell volume relative to the lower pressure systems. This actually makes our choices a little easier. As a result of both considerations, we recommend using systems with a 600 bar maximum operating pressure for the 2.7 µm Raptor™ columns.
Here are examples of some LC Systems with a maximum pressure of at least 600 bar (8700 psi). Again, we recommend using a system with a max pressure of 600 bar, or higher, for use with a 2.7 µm Raptor™ column.
- Agilent 1200 Series- RRLC models only
- Agilent 1220 & 1260 Infinity
- Shimadzu Prominence UFLC-XR models only
- Thermo Ultimate 3000 -Basic Manual, Basic Automated, Binary Analytical, x2 Dual analytical, and Quaternary Analytical models
- Thermo Ultimate 3000 -RSLC nano, RSLC Binary, and RSLC X2 Dual (~700 Bar limit)
- Perkin Elmer Flexar -FX-10 models only
- ABSciex/Eksigent – microLC 200, nanoLC-Ultra and nanoLC 400 (~700 Bar limit)
You can also use a 2.7 µm Raptor™ column with a UHPLC system, which typically has a maximum operating pressure limit of 1200 bar/1700psi.
When to use the 5 µm Raptor™
Your next question might be when or why you should use a 5 µm Raptor™column. The 5 µm particle size for this SPP column allows you to obtain results that resemble that of a 3 µm particle fully porous particle column on an HPLC system that has a maximum pressure of 400 bar/ 5800 psi or less (which is usually too low to use a column with a 2.7 or 3 µm particles). There are also other advantages described in our article The Effects of LC Particle Choice on Column Performance 2.7 vs. 5 µm Diameter Superficially Porous particles (SPP), which I highly recommend reading, by the way. Here are just a few examples of some LC systems that could be optimized by using them with a 5 µm Raptor™column:
- Agilent 1100 & 1200 Series (except models listed above)
- Varian 920-LC & 940-LC
- Waters Alliance and Breeze models
- Jasco 2000 Series
- Perkin Elmer Series 200 & 275
The above list is definitely not all inclusive and there are many I have not shown. Generally, the older HPLC systems are more likely to have a lower pressure rating and many of them have 400 bar limits. The best way to find out is to consult your instrument manual or contact the manufacturer of the instrument.
I hope this has been helpful. Many thanks to our Innovations Chemists, Ty Kahler and Sharon Lupo for the data provided. Thank you for reading.
Persistent organic pollutants (POPs) like dioxins, furans, polychlorinated biphenyls (PCBs), and polybrominated diphenyl ether (PBDEs) are lipophilic and therefore may be found in fatty matrices like fish. In order to remove the high lipid coextractives, lengthy cleanup procedures that require multiple cleanup cartridges are often used. I recently presented work at the Dioxin meeting where we took a shortcut and analyzed PBDEs and other halogenated flame retardants (HFRs) in fish using a quick QuEChERS type extraction, and a PSA pass through cleanup (Get the poster here!). While the PSA pass through does a good job at removing a large amount of fat (we reduced the fat by 50-70%), there is still a considerable amount of nonvolatile residue remaining in the final extract. This is especially evident in the mackerel sample that we analyzed where 38 mg/mL of nonvolatile residue remains after the PSA cleanup.
In order to combat the negative effects of injecting so much nonvolatile material onto the column, such as decreased response of high MW compounds and increased inlet activity, we used a 10:1 split injection. A split injection (aka shoot-and-dilute GC) takes advantage of the sensitive detectors we have (ECD, MS/MS, HRMS) and keeps the system up at least twice as long when compared to a splitless injection. The flow through the liner is much faster when compared to a splitless injection and the majority of the sample is swept into the split vent and not onto the analytical column. Just remember to clean that split vent once in a while (it can get pretty gross)!
The chromatograms below highlight the advantage of using a split injection for fatty samples like fish. In all of the experiments I used a Sky Precision liner with wool and a 15m x 0.25mm x 0.10µm Rtx-1614 column. The sequence was the following: split injection of HFR standard, splitless injection of HFR standard, split injection of mackerel sample, split injection of HFR standard, splitless injection of HFR standard, 2 x split injection of mackerel sample, followed by the standards again. I repeated this sequence with increasing numbers of the mackerel sample injections in between standard injections until the system failed (I could no longer quantify my results). As you can see the response for the last two eluting peaks, syn and anti-Dechlorane Plus, are dramatically decreased after only three mackerel injections. The split injection standard did not fail until 15 sample injections.
The increased ruggedness of using shoot-and-dilute GC was also highlighted by Jack Cochran in this blog post where he made repeated injections of used motor oil. So if you have dirty samples and sensitivity to spare, try the shoot-and-dilute GC approach and think about all those extra things you could be doing instead of GC maintenance!
The answer seems fairly straightforward: Use a narrow internal diameter (ID) liner to keep peak shapes sharp by preventing band broadening; but this isn’t always the correct answer. There has to be enough internal volume in the liner to contain all of the sample. Let me explain.
Let’s first look at liners when using purge & trap (P&T). All instrumentation that I am familiar with uses a transfer line from the P&T unit to the GC injection port. In this case, you do want to use a narrow ID liner to minimize dead volume and prevent band broadening, especially for the compounds which will not condense and refocus within the capillary column.
Now let’s look at liners when using headspace (HS). Headspace units can transfer the sample to the injection port two ways:
(1) Using a transfer line (like those used on a P&T unit).
(2) Using a Gas-Tight syringe.
If the headspace unit uses a transfer line, then a small ID liner should be used. However, if a Gas-Tight syringe is used to inject the gas sample directly into an injection port liner, then a larger ID liner is commonly used. I briefly discussed liner selection for gas samples in the post below.
In this post, I stated:
The liner below is Restek 23302.1 Sky® 4.0mm ID Single Taper Inlet Liner , which is typically the best choice for 1-2µL injections of non-polar solvents (like hexane), 0.5-1µL injections of polar solvents (like methanol), and gas injections approximately > 250µL to 1mL.
The liner below is Restek 23315.1 Sky® 2.0mm ID Single Taper Inlet Liner , which is typically the best choice for 1µL or less injections of non-polar solvents (with low expansion volumes), or gas injections less than approximately 250µL.
Note: The liners above are typically used for splitless (or split/splitless) injections. If you are splitting your sample, choose an open-bottom liner (with no gooseneck or other restriction at the bottom), like 23301.1 or 23333.1 (shown below).
.Sky® 4.0 mm ID Straight Inlet Liner (23303.1)
If you are not using a syringe for sample introduction, but instead the sample is from a purge & trap unit or gas sampling valve, then the liner I usually recommend is 23333.1 (photo below). Sky® 1.0mm ID Straight Inlet Liner
Please note that these are representative liners for an Agilent GC with a split/splitless injection port; the catalog number of the liner you would select depends upon your specific instrument and injection port.
In summary, liner selection for gas samples will depend on your particular instrumentation and how the sample is introduced into the injection port liner. If your instrument uses a transfer line, then a small ID liner will likely provide you the best chromatography. If a Gas-Tight syringe is used to inject the gas sample, a larger internal diameter liner will usually provide you the best results.
 What do Chromatograms tell us? First Eluting Peaks are very Broad when using Direct Injection and Water as Matrix.
Chromatograms are like fingerprints. If you can “read” chromatograms well, you often can find a plausible cause. In this series, we will show a series of GC-chromatograms that are obtained from users and discuss some potential causes for the phenomena. Then we can move into some solutions for improvement.
Direct injection is often used if low levels of volatile components have to be analyzed in a higher boiling matrix. The main peak will elute later then the components, so we cannot use splitless injection. Direct injection is always a challenge as here a relative big injection-error is made. If the conditions are not optimized, one can get a chromatogram as shown in fig.1. Here the matrix is water which complicates the injection. Important is to minimize the band broadening during injection.
Conditions for a good direct injection:
- - Using smallest injection amount;
- - Inject fast;
- - Use wide bore columns with high retention;
- - Use a uniliner (see fig.2) for most ideal gas path.
- - Find the best oven temperature to start the analysis;
See also: http://blog.restek.com/?p=614
To get sufficient focusing of the early eluting compounds one need to find a compromise between initial oven temperature and retention of the stationary phase. Usually high retentive phases (thick films or PLOT’s) will have a good focusing effect, and result in sharp peaks.
The additional advantage is, that the starting temperature can be relative high, so there is reduced risk for the matrix to condense, which helps the chromatography.
Fig. 1 was using a 30m x 0.53mm Rt-Q BOND where oven temperature was 110C for 20 s, followed by a program ballistic to 130ºC, → 250ºC @ 10ºC/min
Under these conditions there is little focusing for volatiles happening which results in broadened peaks. The later eluting peaks are focused and sharp.
By reducing the oven temperature to 80 C and using the same program, the focusing improved significantly, resulting in good peaks for early eluting compounds, see fig. 3.
By using a divinyl benzene porous polymer, the hydrophobicity also helps to make the water peak move very fast, so it has minimal impact on the retained analytes.
Installation of a column with the uniliner , Fig 2, remains a challenge as this connection is made inside the injection port. See for practical details, here: Installation uniliner 05-04-002
For some systems, like Agilent, the uniliner must have a “hole” on the side of the liner, in order to make the EPC work. This “hole can be on top or on the bottom. For Direct injections, it is preferred to use the hole on the top
http://www.restek.com/catalog/view/9232 (4mm uni-liner) and
http://www.restek.com/catalog/view/9201 (1mm uni-liner)
More reading about Direct injections: http://www.restek.com/pdfs/59882B.pdf
Working in technical service has taught me many things, but one of the most important has been “don’t assume anything”. This is especially relevant when it comes to verifying the carrier gas flow through GC columns. To obtain reproducible results from column to column, this verification should always be done after installing a different packed/micropacked column into your GC oven.
Setting and verifying the carrier gas flow for packed/micropacked columns is commonly done using an electronic flow meter. This should absolutely be done with every column installation because each packed/micropacked column has a unique pressure drop. Remember that these columns contain packings which are not of a uniform particle size, but rather contain particles which fall within a specified range. In addition, particle size distribution within this specified range can vary. To read more about this topic, please see Understanding packed column mesh size ranges
As far as I know, there is no instrument software that can automatically set the desired flow rate. Even for those GC’s which have inlets controlled by a mass flow controller, it is still is a good idea to measure/verify the flow rate exiting the column.
- Cool all heated zones and then turn off all GC gases. Allow the current packed/micropacked column (if one is installed) to depressurize so that its removal from the GC (more specifically, the inlet) doesn’t cause a pressure surge which can expel packing from the column.
- Install the new column into the injection port. Do not connect to the detector at this time. Do not turn on any heated zones. Slowly increase the head-pressure just until carrier gas starts exiting from the column (holding a thin strip of tissue paper at the column outlet and observing when it moves is a good indicator).
- Attach an electronic flow meter (or soap-bubble flow meter) to the outlet of the column and once again begin slowly increasing the inlet’s head-pressure. When the desired column flow rate has been obtained, continue to monitor the exiting carrier gas flow rate for five minutes to make sure it is stable.
Very long packed columns, packings with irregular shaped particles, very small mesh sizes, and/or very small internal diameter micropacked columns, require high head pressures to obtain proper carrier gas flow. As a result, many of these columns are commonly used with valve (switching) systems because obtaining reproducible results using a syringe injection can be difficult (sample loss through a punctured injection port septum is common).
By verifying the column’s carrier gas flow rate when you install a different packed/micropacked column, you should obtain reproducible results every time. Thanks for reading.
Okay… so time to bring this story full circle. Recall in our first e-cigarette blog we demonstrated a quick proof of concept for analyzing the major constituents of e-juice on a GC-MS. Next in part II we demonstrated some faster capabilities on a GC-FID and even threw in a calibration curve. Then in our most recent post we showed you how our EZGC Method Translator and Flow Calculator could be used to easily switch from helium to hydrogen carrier gas for working on electronic cigarette solutions. NOW… time for us to pull back the curtain on what we can really see in in the e-juice.
So we took our GC-MS method from part I and slowed down the oven to near OHR @ 11 °C/min. No rush right now, time to carefully dissect the e-juice. We also dropped our split from 200:1 down to 10:1. Lastly, we decreased our dilution from 100:1 down to 2:1, just so we had enough methylene chloride in the sample to cut the viscosity. Here is what it looks like:
Not exactly anything to get too excited about, right!? Heck… we have seen the propylene glycol, glycerin, and nicotine peaks enough by now that you should be bored. Not to mention that the peak shape looks horrible due to the overloading situation created by the minimal dilution and split ratio. But wait just a minute… lets zoom in on the above chromatogram and really take a look:
BAM!!! Not just the propylene glycol, glycerin, and nicotine we’ve been talking about all along! Based on the above chromatogram, our e-juice has a lot more constituents than that. So what are all of these peaks? Well… the following chromatogram will hopefully help break down all the unknown compounds into several more digestible sections:
So as you may see in the above chromatogram, we have broken everything down into 5 different sections. You may also see in the table below, we have a list of the tentatively identified/unidentified compounds. Note: We only included compounds with a mass spectral quality of 80% or greater according to the NIST 2005 database. Compounds with 100% quality have been confirmed with standards.
38 [45 minus the 7 present in a solvent blank (see chromatogram below)] compounds in this e-juice. Some of which we were able to tentatively identify and some of which have even been confirmed with standards. So what does this all mean? Well…we will have to save that discussion for a later date, as this blog has now approached my maximum allowable length. In the meantime you may draw your own conclusions, but until then… stay tuned!
While this sounds like a very broad question, I’m sure it is crossing some of your minds. To start out, SPP stands for superficially porous particles. It has also been called “core-shell”. If you are a veteran in the world of HPLC, this term is old hat for you. If you are new to the game, some explanation might be helpful. Here are a couple of links to articles you could read to brush up on the topic:
SPP particles are silica based, but they have a solid core with an outside porous layer. In our case, for Raptor columns, we have columns with diameter of 2.7 µm, composed of a 1.7 µm non-porous core and a 0.5 µm porous outer layer. We also have 5 um diameter particles, composed of a 3.3 µm core and 0.6 µm outer shell.
The greatest benefit of using a column with SPP particles is increased efficiency without an increase in backpressure. Particularly for the 2.7 µm particles, you can obtain separation that resembles that of a UHPLC column (1.9 µm fully porous particle), at a pressure that is closer to that of traditional HPLC. This is because pressure is mostly a function of outer diameter, regardless of porosity. The consequence of this is that you don’t need an expensive UHPLC instrument to use our Raptor columns.
You might ask, “Why the increase in efficiency?”. Hopefully you can take time to read up on this if you’re interested, but in summary, the transfer time of molecules through the porous layer is greatly reduced. This, in turn, reduces the diffusion terms in the Van Deemter Equation, which automatically increases efficiency. More discussion of this phenomenon can be found in the FAQ section of our website. Other advantages of our Raptor columns are described in our product brochure, “Dissecting Raptor™ LC Columns”.
Please watch for the next post, where we will address, “Should I use a 2.7 or 5 µm Raptor™ column?”. Thank you for reading.