Welcome to the New Restek.com! Take a tour of the new site and feel free to send us your feedback.

Trans fatty acids analysis part 2: Let’s look at actual samples with incurred TFAs

  • Jana Rousova
  • #FAME
  • #Fatty Acids
  • #Product Selection
  • #Food
  • #Method Optimization
  • #Fused Silica Capillary Columns
  • #Blogs
  • #Food Safety
  • #GC
  • #Food & Agriculture
  • #GC Columns
  • #Columns
  • Share:

This blog a part of a series: part 1, part 3, and part 4

In my previous blog, I’ve tried and failed to acquire a sample that contained any trans fatty acids (TFAs). While this is a great news for everyone’s health, the scientist in me was somewhat disappointed. That’s why I first decided to cheat a little bit and look at different TFAs – ruminant TFAs. Ruminant TFAs are products of bacterial metabolism of polyunsaturated fatty acids in the rumen of cattle, sheep or goats, contributing up to 6% of total fat.1 They are present in both dairy products and meat. The major difference between artificial and ruminant TFAs is their distribution. Partial hydrogenation produces TFAs with almost Gaussian distribution, with highest abundance for trans-9 C18:1, while ruminant bacteria skews the distribution towards for trans-11 C18:1 (up to 42 wt%, Fig. 1). It’s also noteworthy that trans-C16:1 can contribute up to 20% of ruminant fats but it is not present in partially hydrogenated oils unless they originate in marine oil.1

Figure 1: Distribution of trans-C18:1 isomers. Adapted from Stender et al.1

The obvious choice for analysis was a dairy product with the most milk fat, i.e. butter. Butter contains up to 80% milk fat, therefore, theoretically, 5% of TFAs (about 0.5 g per serving). In many cases, TFA content is low enough that it doesn’t have to be declared on the label. Our analysis shows the trans-vaccenic acid (C18:1 trans-11), albeit in very small abundance (Fig. 2). While both columns perform well in terms of separating trans-vaccenic acid, the Rt-2560 provided better separation of C10:0 from the matrix. A different approach to sample preparation could result in better separation of C10:0 on Rtx-2330.

Figure 2: Separation of butter FAMEs using GC-FID

After analyzing butter, I decided to go back to a grocery store and take a good look at nutrition labels of more suspected products. Shelf-stable or frozen pastries were all in clear, but I had luck in the frosting aisle. I selected two types of chocolate frosting which declared 1g of TFAs per serving. The first frosting tested proved to be mislabeled since no TFAs were detected and the pattern of C12:0, C14:0 and C16:0 suggested replacement of partially hydrogenated oil by palm or palm kernel oil.

Fortunately for my analysis (not so fortunate for consumers), the second frosting showed a whole slew of TFAs. Figure 3 shows a close-up of the C18:1 region of chromatogram overlaid with cis/trans FAME standard mixture (#35079, in blue). Interestingly, Rtx-2330 was able to separate the standard mixture reasonably well but failed to separate the same isomers (namely C18:1 cis-6 and cis-9) in the real sample (Fig 3A). The second column (Rt-2560) separated those two peaks fairly well, moreover, it showed an additional peak directly following methyl oleate (C18:1 cis-9, Fig 3.B).

Figure 3: Separation of C18:1 isomers in frosting using GC-FID. The red trace is the frosting sample, blue trace is the cis/trans standard.


To conclude, while Rtx-2330 provides fairly good separation, it is crucial to use Rt-2560 to separate as much of cis/trans isomers as we can.

More about TFAs and Rt-2560 from Hansjoerg and rest of the blog series:

Part 1: Is the industry ready for a ban of partially hydrogenated oils?

Part 3: Running Rt-2560 with GC-MS

Part 4: Using hydrogen as a carrier gas for FAMEs analysis


  1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2596737/