Quick & Painless Email Subscription

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

The Chromatography Picture-Show

If you’re caught up on all your favorite shows, might I suggest the Restek Video Library for your next binge. Most of the episodes only run about two minutes, so there are few extraneous subplots. (And, so far, no car chases. Not a single one in over 50 videos! But new episodes are added all the time, so who knows?)

screen shot showing the library and its menu location

The cast is superb, and delivers performances packed with practical wisdom. Their costumes probably won’t dominate the awards, but the concise and helpful info they’re providing outshines such considerations. Also of note is the use of exotic filming locations such as laboratories and even the abstract realm of cyberspace. You feel like you’re right there. The cinematographer’s boldness is most apparent in the unflinchingly informative depictions of the guts of LC and GC instrumentation.

Diverse plotlines offer something for nearly everyone. From the heartbreak of tailing peaks to the triumph of a clean, square cut on a fused silica column, from the thrill of proper LC column storage to the drama of detector lamp replacement, it’s all there. There’s even a very special episode about finding the reference standards that match your requirements.

Although these gems have yet to show up on any of the major streaming services, they’re easy to track down. You can even use your phone and take them into the lab with you. Here’s the hookup:



Have there been any new applications developed for Diquat and Paraquat?

I am happy to answer “yes” to this question. Those of you who have been doing this analysis might have been using our Ultra Quat column that I talked about in an earlier blog post here, “I have an Ultra Quat column and I know how to use it…”


This updated method uses the Raptor HILIC-Si, which is a newer product we are excited to have. Here is what the application looks like:



We are pleased with this application, particularly since it does not require any ion pairing agents or chaotropic salts. The mobile phase modifiers are very common ones that most labs will have on hand and extended equilibration times are not needed (as compared to NaPF6 or Heptafluorobutyric acid). The run time is short and sweet.  :-)

Here is a link to the product information for the column we used:



I hope you find this post helpful and thanks for reading.

Which Headspace Vial Cap Is Right For Me?

Today, I’d like to cover a question that I have been receiving from cannabis labs who are doing residual solvents analysis. This blog may not only help cannabis labs, but also help labs analyzing other matrices looking for residual solvents or very volatile compounds. The question is, “Which type of vial and vial cap combination should we use to do our analysis?”

This is referring to the screw-thread caps (18 mm) and crimp caps (20 mm) for 20 mL headspace (HS) vials. So instead of just recommending one or the other based on “feel” or price, we decided to run a little experiment to compare the performance of the two styles. The experiment was designed to analyze a few very volatile compounds and gather data on how well these two types of caps prevent the compounds from escaping the HS vial over the course of a specific period of time. To familiarize yourself with the products that we were working with, please reference Figures 1 & 2.


Figure 1. Screw Cap HS Vial: PN#23082 and Screw Cap: PN#23092


Figure 2. Crimp Cap HS Vial: PN#21162 and Crimp Cap: PN#21761


Sample preparation included adding propane, butane, and isobutane standards in N,N-Dimethylacetamide (DMA) along with α,α,α-Trifluorotoluene as the ISTD to a 20 mL HS vial. To model the vial cap’s ability to keep these volatile compounds from escaping the vial, we designed a time study. In this study, 14 samples were prepared for each vial type and the samples were split into two sets of 7. Set one (T0) was run immediately after the samples were prepared and set two (T6) was run 6 hours later. It should also be noted that the crimp cap pressure was set to model the tightness of a screw cap. This meaning that the amount of torque needed to remove a screw cap was similar to that of the torque needed to spin a crimped, crimp cap.

The average compound area count from T0 to T6 were compared by calculating the relative percent loss over the 6 hour period of time. The relative percent difference for each HS vial type can be seen in Figure 3.


Figure 3. HS Vial Cap Comparison


From the data shown in Figure 3, we can see that the screw caps (blue) did not seal as well from T0 to T6 compared to the crimp caps (red). It is also worth noting that we see very little change in our ISTD over the 6 hour time period for both the screw cap and the crimp cap.

After looking at the data and taking everything in, I think it is suffice to say that if you are working in a lab doing HS analysis of very volatile compounds (gases), you should be using crimp caps. However, even if crimp caps are being used, one still needs to be cautious of the fact that the lighter, more volatile compounds, like the compounds shown here, are still capable of escaping the HS vial. Now, with all this being said, I don’t want to give the impression that screw cap HS vials are worthless. They can definitely be used in HS analysis, especially if the compounds being analyzed are on the less volatile side or your time from sample preparation to analysis is kept to a minimum. As an example take a look at the minimal differences of α,α,α-Trifluorotoluene, so for many customers analyzing the standard list of solvents and volatiles a screw cap may be appropriate.

Tailing solvent peak during split injection? Check your column installation distance!

When performing splitless injections, it is not uncommon to witness a tailing solvent peak, given the large amount of solvent introduced onto the column over a relatively long period of time.  For split injections, however, only a small portion of the solvent loads onto the column and it occurs very quickly. This results in a relatively sharp solvent peak.

Recently, I set up a method with a 20:1 split and upon my first injection I noticed a badly tailing solvent peak, which I wasn’t expecting (Figure 1 red trace).  My gut feeling told me something was occurring within the inlet and the solvent wasn’t split properly.  I removed my column from the inlet to find that the installation distance was too low. The recommended installation distance for an Agilent split/splitless inlet is 4-6mm above the ferrule and my column was only 1-2mm above it.  This would lead to the column being below the inlet seal surface, creating an area of dead volume, where residual solvent would not get swept away by the higher split flow.  After correcting the installation distance, my problem disappeared (Figure 1 blue trace)!

Figure 1: Improper installation distance of column can create excess dead volume, leading to tailing.

Figure 2: Capillary installation gauge


Generally, I pre-swage my ferrule to the column at the proper distance using a capillary installation gauge, but I was in a rush and didn’t do that this time. I probably bumped the column while installing, inadvertently pushing it down.  I highly recommend using a capillary installation gauge, which allows you to pre-swage your ferrule to the column at the exact specified distance for installation (Figure 2).  It only takes a few seconds and greatly improves consistency of installation.  The table below shows the capillary installation gauges that Restek carries for various instrument makes.


Are My Shut-off Valves Contaminated?

Figure 1: Experimental set-up for testing valve contamination (plug valve shown). Gas flow was set at 20mL/min for 1 hour (1200mL total extraction volume) with oven held at 35⁰C and the split vent closed. After a one hour hold, oven was ramped to 330⁰C @ 20⁰C/min. An Agilent 7890A was used with an Rxi-5MS (30m x 0.53mm x 0.25µm) (cat. #: 13425) to allow for high flows. Inlet liner was a Topaz 4mm single taper liner with wool (cat #: 23303).

Shut-off valves provide a convenient means to control gas flows and are often used in conjunction with gas chromatography systems.  There are several types of shut-off valves including ball valves, plug valves, and diaphragm-sealed valves.  Some of these valves, ball and plug valves included, use a silicone-based lubricant to make opening and closing the valve easier and to extend lifetime by reducing wear on moving parts.  Unfortunately, this lubricant can off-gas into the flow path and lead to system contamination, causing baseline interferences.

Use a gas filter containing a hydrocarbon trap between lubricated valves and the GC to prevent any off-gassed components from entering your system.

I performed an experiment where I attached ball valves (cat. #: 23144) and plug valves (cat. #: 23146) directly behind the GC EPC and set a high flow of carrier gas for one hour (1200mL of Helium total volume), while leaving the split vent closed (Figure 1).  The GC oven was left at 35⁰C to trap off-gassed components on the head of the column.  Following this, I ran a series of “no-injection” blanks, with a temperature program, to qualitatively see if anything was off-gassing from the valves.  This experiment was also performed with a Super Clean Triple Gas filter (cat. #: 22019) installed between each valve and the instrument.

I found that the plug valves introduced a significant amount of contamination to the system (Figure 2).  Ball valves produced some contamination, as well, but at a much lower level than the plug valves (Figure 3).  This is likely due to the different opening and closing mechanisms of each valve and the much larger surface area of the lubricated plug.  Based on this, I would never recommend using a plug valve without a filter for carrier gas lines and would encourage the use of filters even with a ball valve.  Testing of multiple valves of each type showed similar results.

When these experiments were repeated using a filter in between the valve and the instrument, this contamination was effectively captured, as can be seen in Figure 2 and Figure 3, for each respective valve.  Note that I used a triple filter (traps hydrocarbons, moisture, and oxygen); however, a hydrocarbon filter is most critical for trapping this specific type of contamination.  Nonetheless, gas lines should have moisture and oxygen traps prior to the instrument, as well, hence my choice of the triple filter.

Figure 2: Contamination introduced from plug valve (red trace), measured with a GC-FID. By contrast, using a filter between the valve and the instrument (blue trace) produced a clean baseline.

Figure 3: Contamination introduced from ball valve (red trace), measured with a GC-FID. Note that this is at a much lower level than the plug valve, but could still interfere with trace analyses. Using a filter between the valve and the instrument (blue trace) produced a clean baseline.


I was surprised at the amount of contamination witnessed, especially from the plug valves.  Many GC gas plumbing guides show shut-off valves downstream from the filters, placed directly before the instrument.  Keep in mind that with the large volume of gas passed over the filter with the split vent closed, this was designed to be somewhat of a worst-case scenario; however, similar results could be witnessed simply from letting an instrument sit idle for a period of time.

For these experiments I used an oven temperature of 35⁰C to trap off-gassed components on the column.  There are undoubtedly additional volatile components that are not effectively captured using this method.  Off-gassing of these volatile components could potentially create major contamination issues with sampling techniques where valves are placed before analytical traps for volatiles, such as thermal desorption units, purge and trap, etc.


The usage of Reference Standards in MOSH/MOAH Analysis


The unintended migration of mineral oil compounds into food, e.g from packing materials or from production processes are a raising topic in industry and scientific discussions. For a first screening, an LC-GC/FID method was established in the last years. In collaboration with Axel Semrau, Restek has developed ready to use standards for quantification and and identification of the different fractions to be controlled. Saturated Hydrocarbons and Aromatic Hydrocarbons have to be cut of from LC fraction before transferred into the GC and the different areas which have to be taken into consideration have to be identified properly. The Blog entry gives an overview about how to use the different available standards correctly.


Food may be contaminated with mineral oil compounds by unintended migration from different sources. One of the most common sources for such an unintended contamination is coming out of recycled paperboards, but also during production processes such a failure may occure.

Due to their different toxicological capability, mineral oil compounds are subdivided into two fractions:

  • MOSH: Mineral Oil Saturated Hydrocarbons
    -saturated hydrocarbons
    -paraffines & naphthenes (cyclic)
    -highly alkylated
  • MOAH: Mineral Oil Aromatic Hydrocarbons
    -highly alkylated mono- and polyaromatic compounds
    -also saturated ring linked to aromatic rings
    consistence often unknown

The toxicity of both groups are described as follows:

  • MOSH:
    –accumulated in Human Tissue
    –toxicological relevance unknown
  • MOAH
    -most substances unknown and toxicological nonvaluated
    -carcinogenic potential can not be excluded
    -unaccepted risk
    -unaccepted in food

The different toxicological risks have found its expressions in different legal restrictions, as to be found in the EU Commission  Recommendation (EU) 2017/84 of 16 January 2017 or in the German Mineral Oil Regulation, wherein the concentration in food contact material (recycling paper & cardboard) is limited for MOAH to 6 mg/kg or, if limits in the material does not comply, it has to be verified that limits in the food comply to MOAH C16-C35  less than  0,5 mg/kg.

For the MOSH fraction the ALARA Principle has to be applied.

  • ALARA Principle
    –In Europe the ALARA principle is regulated by the European Food Safety Agency EFSA and translates to “As Low As Reasonably Achievable”

Analytical Aspects

In both fractions, MOSH and MOAH, a tremendous amount of different compounds have to be covered with an analytical method. Therefore both fractions can only be described as a sum parameter method.

Due to EFSA “Currently, the most efficient methods for analysis of MOSH and MOAH in food and feed comprise extraction followed by pre-separation by high performance liquid chromatography (HPLC) on-line coupled to GC with flame ionisation detection (FID).” – EFSA Journal, 2012, 10(6):2704, is the best choice for this analysis.

This method was first described by K. Grob and M. Biedermann from the Kantonales Labor Zurich. (BIEDERMANN M., GROB K. On-line coupled high performance liquid chromatography-gas chromatography (HPLC-GC) for the analysis of mineral oil: Part 1: method of analysis. J. Chromatogr. A. 2012, 1255 pp. 56–75)  and has found its way into a European norm via EN 16995:2017 (Foodstuffs – Vegetable oils and foodstuff on basis of vegetable oils – Determination of mineral oil saturated hydrocarbons (MOSH) and mineral oil aromatic hydrocarbons (MOAH) with on­line HPLC­GC­FID analysis).

Restek is accompanying the development of an automated system for the analysis of MOSH and MOAH of the German company Axel Semrau GmbH since years, adding our knowledge in separation science as well as in LC and GC but also by preparing the needed reference materials.

The method uses a “Normal Phase” HPLC with a Silica column to separate both fractions, MOSH and MOAH, from sample matrix, than transfers either the MOSH or the MOAH fraction as a large volume injection onto a 7 to 10 m retention gap, which uses solvent focusing and a solvent vapor vent to get rid of most of the used solvent. This approach delivers both fractions, MOSH and MOAH as a chromatographic hump, as shown in the chromatograms below. Both chromatograms were provided by Axel Semrau GmbH


Sunflower Oil MOSH Fraction


Sunflower Oil MOAH Fraction

Reference Standards

For quantification, to prevent losses of low boiling compounds and to make sure that the complete fraction was transferred to the GC, an internal standard (IS) is used. (Restek cat. No. #31070). The 9 compounds included are used as shown in the chromatograms below.

Sunflower Oil MOSH fraction, spiked with IS, courtesy of Axel Semrau GmbH

CyCy: Cyclohexyl Cyclohexane, standard for Quantification, not present in mineral oil products
C13: Tridecane, proves no coelution with CyCy. Peak Identification. Has to be a peak pair with CyCy.
C11: Undecane, detection of losses of volatile compounds
Cho: Cholestane, end of MOSH Fraction

Sunflower Oil MOAH Fraction, spiked with IS, courtesy of Axel Semrau GmbH

2-MN, 1-MN: 1-/2-Methylnaphtalene: standard for Quantification, peak pair for better identification
5B: n-Pentylbezene: detection of losses of volatile compounds
TBB: Tri-tert.-butylbenzene: start of MOAH Fraction
PER: Perylene: end of MOAH fraction. Can be determined also with UV detection in automated approaches

Different regulations are looking for different integration limits for the detection of the two fractions. Restek therefore has developed a Retention Time Standard (Restek cat. No. #31076), This standard has to be run before and after every Sequenz to avoid retention time shifts and therefore wrong results by using the wrong integration limits.

The compounds are used as follows:

  • C10 = lowest boiling compound to be detected
    C11 = Marker out of Restek-Mix #31070 – retention time stability
    C13 = Internal Standard out of Restek-Mix #31070 – ret. Time stability
    C16 = requested boiling point limit for fraction quantification
    C20 = requested boiling point limit for fraction quantification
    C24 = nowadays requested boiling point limit for fraction
    C25 = New supposed requested boiling point limit for fraction
    quantification. If C24 or C25 will be used is not yet fully
    determined (both included gives future proof)
    C35 = requested boiling point limit for fraction quantification
    C40 = todays detection limit
    C50 = desired new detection limit


MOSH/MOAH Analysis helps to improve our Food Quality


Annual Replacement of Cartridge Gas Filters

A significant number of instrument and column complaints are carrier gas related, which could be caused by breakthrough from filters that are not replaced in time.

The possible consequences of a filter breakthrough are:

  • Gas distribution system behind the filter will be contaminated (fast cleaning nearly impossible, bleeds for months)
  • Instrument gets contaminated, expensive maintenance required
  • Column lifetime reduced, bad analytical results, high cost of ownership, unnecessary changing of column brands
  • MS source gets contaminated, expensive maintenance required, long system shutdown.

This article explains why it is important to replace your gas filters annually instead of waiting for the indicators to change color.

Filter Media Types

A typical GC-MS laboratory gas filter system contains three types of filtering media:

  • Oxygen Catalyst for absorbing traces of oxygen
  • Activated Carbon for adsorbing hydrocarbons
  • Molecular Sieve for adsorbing moisture

These filter media types can be divided into two groups:

  • Chemically absorbent media

Like a Venus fly trap, when it comes in contact with a contaminate it absorbs it and does not let it go. Adsorption differs from absorption, which also removes things, but the result is swelling of the media. The media size increase equals the amount of material removed







  • Surface adsorbent media

Like a floor mop, where the contaminant is trapped onto the surface of the media and retained. The material doing the adsorption does not change in size.






Filter Media Breakthrough Indicators

The visual indicators are mainly for urgent situations such as a leak or high amount of impurities breaking through.

Indicators are typically placed behind the filter media bed. When they change color shortly after installation of the filter it usually means there is a leak or that the gas distribution system including the manifold to which the filter is connected was not flushed properly prior to – or during – installation of the filter.

When the indicator changes color during normal operation of the filter it indicates that the filtration media has reached its capacity, and the filter should be replaced immediately to avoid contaminant breakthrough.

Other components such as branched hydrocarbons which are also trapped by the molecular sieve are not shown by the moisture indicator.

Also note that most filters do not provide an indicator for hydrocarbons, so you would never know whether the activated carbon has reached its adsorption capacity.
























Filter Media Breakthrough

Both adsorption and absorption media have fixed capacities, meaning they hold just so much, since they are storing the material removed from the gas, not destroying it.

Normal situation

Hydrocarbon breakthrough

Over time, the pores of the activated carbon fill up. The molecules that are adsorbed with higher energy (larger mass) can displace the lower-energy molecules that are less tightly held. This phenomenon, called displacement, may knock the smaller particles off the media, straight to the column and into the instrument.

Moisture breakthrough

The molecular sieve adsorbs moisture until it cannot adsorb anymore. If the humidity level of the gas is lower than in the molecular sieve, it will de-adsorb its moisture until it is in “balance” with the lower humidity level of the gas, which means the filter could increase the amount of moisture in your gas.

Oxygen breakthrough

The media size increase equals the amount of material removed. When the media has reached its absorption capacity it will not release the already trapped contaminants but it will also not absorb any new ones.

Replace your filter before the indicators start changing color to prevent breakthrough and to avoid high maintenance and repair cost of your instrument

As explained previously, filters which are not maintained on a regular interval can cause the outgoing gas to become more contaminated than the original source gas.

The color indicators used in gas filters are so called ‘last minute’ indicators and require quick action.

You can compare it to the engine oil indicator in your car. When the engine oil indicator is blinking on your dashboard, the car should not be driven and ignition should be switched off unless topped with engine oil – In event of taking risk to drive – high probability of engine getting seized causing high expenditure to repair or replace the engine.

The same is valid for gas filters. When one of the indicators starts to change color, the filter should not be used and instrument should be switched off unless the filter is immediately being replaced with a new one – In event of taking risk to continue – there will be a high probability of contaminant breakthrough causing high expenditure on instrument maintenance as compared to planned annual preventive maintenance on your instrument and gas filters.

So what you can do as an end-user is to always buy an additional set of filters to keep on standby in case the indicator starts changing color or to use a good preventive maintenance plan or tool (such as the electronic indicator) to replace your filters at least once a year.

Find out more about cartridge filter and details on the product offering by following the following weblink: http://www.restek.com/Supplies-Accessories/GC-Accessories/Gas-Purification?s=type:baseplate


VMS for when your TO-15 air lab is hazy, hot, and humid!

When it comes to analyzing volatile organic compounds (VOCs) in air canisters, it is no secret that my go-to-column is the VMS. I have raved about the VMS in the following pieces:

  1. The NJDEP-SRP Low Level TO-15 blog series
  2. My Favorite Column blog
  3. A white paper
  4. And some other articles I am forgetting…

You get the point! However, do you remember why I love my VMS so much? Well, we better just recap with the following reasons the VMS rocks:

  1. The polars (e.g., ethanol, IPA, etc…) look symmetrical (unlike all the other phases) with no hint of tailing.
  2. Butane and 1,3-butadiene are separated, which is a common coelution most air labs are completely unaware of.
  3. There is no clumping of compounds (i.e., everything is spread out across the entire GC run).

So why then today’s blog? What more could I say about the VMS (Quick Reminder: VMS = Volatile Mass Spec). Well, as you are sitting there in your hot pants (no shorts for this lab rat, because of EH&S attire rules), on this hot and muggy summer day, reading this blog; you may remember that all of my previous work on the VMS had GC oven starting temperatures from 32 – 35 °C. I know you do not want to wait the extra time as your GC oven struggles to cool from 40 down to 32. So, as the mercury rises today we show you the following:

  1. The complete resolution of 79 VOCs (75 targets and 4 internal standards) in 16 minutes; with all of the aforementioned awesomeness, but now with a 40 °C start temperature.
  2. The complete resolution of chloromethane and butane (a coelution discussed in the NJDEP-SRP Low Level TO-15 blog series).
  3. The complete resolution of hexane, MTBE, and TBA.
    1. IDK why, but Hexane always finds a way to try and coelute on the other phases. For example:
      1. Ethyl acetate and Hexane (1-type)
      2. MEK and Hexane (5-type)

Without further ado, Table I gives you all the pertinent Preconcentrator-GC-MS parameters; and Figures 1 – 3 and Table II shows the end result.

Table I – Preconcentrator-GC-MS Parameters

Figure 1 – 75 TO-15 VOCs and 4 ISTDs in 16 min

Table II – Compounds and Retention Times (RTs) * Internal Standard t Tuning standard


Figure 2 – Try to get your polars looking like that on a 1- or 5-type.

Figure 3 – You can drive a bus in between these peaks! And look at that TBA symmetry!

There you have it! Everything I normally brag about when using the Rtx-VMS for VOCs in air, but now with a 40 °C starting temperature. So, whether your scorching in Shanghai or roasting in Hotlanta, you no-longer have to wait forever to run your next canister sample as the GC oven cools.

How to Determine the Size of 1/4” and 1/8” National Pipe Thread tapered (NPT) Fittings

Restek’s Technical Service team gets quite a few questions from customers about fittings. There is a bewildering assortment of fitting types and sizes as you can see here on the Restek web site. A blog post titled, “I need a fitting, but which one?” is where we often direct customers for help with these questions. One of the things discussed in that blog post is National Pipe Thread tapered (NPT) fittings as one of the main types used, but many people have difficulty identifying the correct size NPT fitting for their needs. The confusion comes from the fact that the outside diameter (OD) of an NPT fitting does not match the “name” of the fitting. For an example let’s look at Restek catalog # 23187 shown below, which is a 1/4″ to 1/8″ NPT Male Connector.


The 1/4″ designation in the name refers to the compression fitting side of the fitting (on the left in the picture above) which has a nut and ferrules to accept and connect to 1/4″ OD tubing. Makes sense, right? However, the threads on the right side of the fitting in the picture above are called 1/8″ Male NPT, but if you measure the OD of the 1/8” NPT side you will find it is about 0.4″ in diameter…certainly not very close to 1/8″ (0.125”). As a general “rule of thumb” an NPT thread is approximately 1/4″ (0.25”) larger than its “name.” For a 1/4″ NPT fitting the “nominal” OD is 0.533”.


NPT fittings are slightly tapered so the “nominal” diameter is the diameter in the middle of the threaded portion, as measured by the top (crest) of the threads. This is a bit confusing, but NPT threads are made to the ANSI B1.20.1, SAE AS71051 standard and anything complying with a standard with a name like that is bound to be bewildering. Hopefully the image below will help.


The charts below are from the Swagelok Thread and End Connection Identification Guide with the first chart (from page 12) showing dimensions for male NPT threads and the second chart (from page 13) containing the dimensions for female NPT threads.


A few other helpful blog posts related to NPT fittings are:

Don’t forget the end fittings when you purchase an inline gas regulator”

How to connect 1/8 inch tubing to a Restek gas regulator

Swagelok® and Parker® Tube Fitting Manuals


Finally, you should always use PTFE tape when making a connection with an NPT fitting.


Thanks for reading!

The main source of artificial trans fatty acids is banned on June 18, 2018. Are we ready?

One of the leading causes of deaths in the world is heart disease, accounting for approximately 24% of deaths in the USA in 2015.1 One of the factors linked to cardiovascular diseases is consumption of artificial trans fatty acids (TFAs). Artificial TFAs are a by-product of partially hydrogenated vegetable oils, which were introduced as a replacement for saturated fatty acids found in butter, after concerns about adverse health effects. The TFAs have a benefit of a higher boiling point than their cis counterparts allowing for denser packing leading to a more solid form of hydrogenated oils.2 However, this is not the only way that cis and trans fatty acids differ. While cis fatty acids help maintain a healthy balance between low-density and high-density lipoprotein cholesterols (LDL and HDL, respectively), the artificial TFAs increase LDL levels and decrease the HDL levels. In addition, TFAs increase total levels of cholesterol, effectively increase the risk of cardiovascular disease.2

There is a push to eliminate the partially hydrogenated oils (the main source of artificial TFAs) globally. In Europe Denmark was the first to execute legislative implementation in 2004, followed by Austria, Sweden, Hungary and Latvia, while other European countries implemented guidelines for voluntary self-regulation.3 In the United States, the FDA declared partially hydrogenated oils as not “generally recognized as safe” in June of 2015 and food manufacturers have been given three years to phase them out (unless otherwise approved).4 On May 14, 2018, World Health Organization (WHO) together with the non-profit Resolve to Save Lives released a step-by-step guide for the elimination of industrially produced TFAs with the goal of completely phasing out partially hydrogenated oil and artificial TFAs by 2023.5

Less than a month away from the US ban of artificial TFAs, we have decided to test products that traditionally contain TFAs– margarine stick, shortening and butter flavored popcorn. The margarine and shortening were purchased at a local grocery store, while the popcorn was bought at a dollar store. The fatty acids were trans-esterified using sodium methoxide, following a method by Ichihara et al6 and analyzed on multiple Restek columns, namely Rtx-2330, Rt-2560, and FAMEWAX. All these columns can be used for fatty acid methyl esters (FAMEs) analysis, where Rtx-2330 is a general use polar columns with cyano-polymer stationary phase, while FAMEWAX is formulated to deliver comparable elution order to other Carbowax columns with better FAMEs resolution. RT-2560 is specifically formulated to ensure accurate quantification of critical cis/trans FAMEs.

Let’s start with the individual products. Margarines previously contained 10-20 % of saturated, 30-40 % monounsaturated, 20-30% polyunsaturated fatty acids and approximately 15% TFAs.7 We tested the new margarine and the composition was approximately 40% saturated (peaks 1-6, Fig. 1), 23% monounsaturated (peaks 7a- 8, Fig. 1), 37% polyunsaturated fatty acids (peaks 9-10, Fig. 1) and no TFAs (peak 7b, Fig.1). This shift is caused mostly by addition of palm and palm kernel oils, which contain more saturated fatty acids (approximately 50% and 80%, respectively).8-9 In Figure 1 we can see chromatograms of the trans-esterified margarine stick without any alterations (red), and a sample spiked with methyl elaidate (methyl trans-9-octadecenoate), the most common artificial TFA added.

Figure 1: GC Analysis of trans-esterified margarine stick. Red – original sample, blue – sample spiked with methyl elaidate (peak 7b).

Figure 1: GC Analysis of trans-esterified margarine stick. Red – original sample, blue – sample spiked with methyl elaidate (peak 7b).

The differences between the individual columns are apparent. Both cyano polymer-based columns (Rtx-2330 and Rt-2560, Fig. 1A and B, respectively) elute trans isomers before the cis. This is particularly advantageous because even when using GC-MS we cannot distinguish the isomers based on their EI spectra. This helps to unambiguously determine the identity of peak 8 as methyl cis-11-octadecenoate. When it comes to Carbowax polymer-based columns, trans isomers elute right after its cis counterparts. Nevertheless, FAMEWAX column is capable of separating the cis/trans isomers (we can see two distinct peaks, Fig. 1C peak 7a and 7b).

The next product we analyzed was shortening. Traditionally shortening can contain as much as 40% of industrial TFAs.10 As with the margarine, the shortening we bought had no significant TFAs present (Fig. 2). In contrast to the margarine, the polyunsaturated fatty acids in shortening accounted for almost 50% of FAMEs, which was also in direct contradiction with the product’s label of 20.8% polyunsaturated fatty acids.

Figure 2: GC Analysis of trans-esterified shortening

Figure 2: GC Analysis of trans-esterified shortening

It is worth noting that the relative content saturated of fatty acids’ in shortening is lower than in margarine. This is most likely due to the absence of palm kernel oil, however, the palm oil was still present. Our third sample, butter-flavored popcorn (Fig. 3), contained no palm or palm kernel oil, resulting in the least amount of saturated fatty acids, approximately 16%. In addition, the popcorn contained the most monounsaturated fatty acids (mainly oleic acid) – approximately 67% of total FAMEs. This can be attributed to the oil source – while fats in both margarine and shortening are comprised of soybean oil, palm oil and palm kernel oil (margarine only), popcorn’s source of fat is canola oil and fully hydrogenated cottonseed oil.

Figure 3: GC Analysis of trans-esterified popcorn’s butter flavoring

Figure 3: GC Analysis of trans-esterified popcorn’s butter flavoring

Overall, our evaluation confirmed that US food producers ditched the artificial TFAs. On the other hand, the introduction of palm and palm kernel oil means increases in the saturated fatty acids’ content. When it comes to column selection, FAMEWAX provides great separation of the majority of FAMEs with shorter analysis times (especially for longer chains), however, if you need to quantitatively determine the TFAs, Rt-2560 is the best choice.


  1. https://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm
  2. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3955571/
  3. https://ec.europa.eu/food/sites/food/files/safety/docs/fs_labelling-nutrition_trans-fats-report_en.pdf
  4. https://www.federalregister.gov/documents/2015/06/17/2015-14883/final-determination-regarding-partially-hydrogenated-oils
  5. http://www.who.int/news-room/detail/14-05-2018-who-plan-to-eliminate-industrially-produced-trans-fatty-acids-from-global-food-supply
  6. https://www.sciencedirect.com/science/article/pii/S0003269715005357
  7. https://ndb.nal.usda.gov/ndb/foods/show/04629
  8. https://ndb.nal.usda.gov/ndb/foods/show/299928
  9. https://ndb.nal.usda.gov/ndb/foods/show/299949
  10. https://ndb.nal.usda.gov/ndb/foods/show/04667

Next time we’ll look into how does margarine compare to other fats, such as butter, coconut oil, palm oil and olive oil.