From: Food Quality & Safety magazine, August/September 2013

Vegetable Oil Analysis

by Melissa N. Dunkle, Frank David, Pat Sandra, and Martin Vollmer

Vegetable Oil Analysis

Quality control and quality assurance testing are increasingly important at every point in the food supply chain, from manufacturing and packaging to distribution and retail sale. The focus on efficient and reliable food analysis has become more acute over the past few years, as high-profile cases of food fraud and adulteration have come to light.

Food fraud is a decades-old problem. According to researchers at the U.S. Pharmacopeial Convention (USP), an independent scientific non-profit organization, vegetable oils—especially olive oil—have a high vulnerability to adulteration and represent the most documented cases of food fraud, with dilution being the most common cause of problems. Over the past 30 years, more than 270 studies and articles have been published on the adulteration of olive oil alone.

In recent years, food analysis has improved dramatically and many types of adulterated food are now unlikely to escape detection. Extensive research has been done in the field of vegetable oil analysis to test for authenticity and chemical properties. For example, gas chromatography (GC) and high-performance liquid chromatography (HPLC) have frequently been used to evaluate triglyceride content in vegetable oil samples. The physical and chemical properties of vegetable oils are closely related to the type and relative amount of each constituent triglyceride in the sample.

A New Technique

Supercritical fluid chromatography (SFC) in combination with evaporative light scattering detection (ELSD) is a valuable technique for the determination of triglyceride composition of vegetable oils. Compared to GC, SFC separates triglycerides at much lower temperatures; compared to HPLC, SFC permits greater selectivity with shorter analysis times.

The potential of SFC for the separation of triglycerides has been demonstrated for many years. Using a reversed stationary phase, the separation is similar to that obtained in reversed phase HPLC. Separation is based on carbon number (total number of carbons in fatty acids) and on the total number of double bonds. Using a silver-loaded column, separation is primarily based on the degree of unsaturation (total number of double bonds). These two separation mechanisms are complementary.

This technical article demonstrates the SFC separation of triglycerides in three vegetable oil samples.

Experimental Methods

Sunflower seed oil, peanut oil, soybean oil reference oils, tripalmitin (PPP), triolein (OOO), and trilinolein (LLL) standards were purchased from Sigma-Aldrich (Bornem, Belgium). The oils were dissolved in chloroform at the 5 percent (50 mg/mL) level.

Analyses were performed on an Agilent 1260 Infinity Analytical SFC System combined with an Agilent 1260 Infinity ELSD Evaporative Light Scattering Detector. The ELSD was coupled to the SFC module using a procedure similar to the one used for SFC-MS 4.

The addition of a make-up flow before the backpressure regulator, together with additional heating at the entrance of the ELSD, was found necessary to obtain good sensitivity, reproducibility, and avoid solute deposition in the transfer capillary. Experiments show that switching off make-flow or heating immediately results in low sensitivity and unstable baseline in ELSD detection.

Analyses were performed on two different stationary phases: ZORBAX SB-C18 and Chromspher 5 Lipids silver loaded column. For the reversed phase separation, three ZORBAX SB-C18 columns were coupled in series.

Reversed Phase Separation

Figure 1 shows the UV and ELSD chromatograms for the separation of triglycerides in sunflower seed oil. As seen, the ELSD detector is more sensitive than UV detection, giving S/N ratios approximately five times better than in UV detection. In addition, the baseline is more stable than in the UV signal at this low wavelength. Moreover, the response in ELSD is more universal and less dependent on the number of double bonds in the lipid molecule.

In SFC, using a reversed phase C18 column, triglycerides are separated according to the carbon number and the total number of double bonds. By approximation, the elution order is set according to:

  • PN = CN – NDB


  • PN = partition number
  • CN = carbon number (sum of carbons in fatty acid chains)
  • NDB = sum of number of double bonds

Therefore, the PN for OOO is (18+18+18)–(1+1+1) = 51, and this compound elutes later than OLO with a PN = (18+18+18)-(1+2+1) = 50. Within a group of triglycerides with an equal PN number, additional separation can be obtained. For example, LLL and PLL (PN = 48), OLL and PLO (PN = 49), and OLO and POO (PN = 50) are separated.

Three different vegetable oil samples were analyzed in another experiment (see Figure 2). In all cases, distinct pro-files and ideal separation were obtain-ed for all oil types when using SFC with ELSD detection.

Separation of sunflower seed oil
Figure 1: Separation of sunflower seed oil triglycerides with UV detection at 210 nm (left) and ELSD detection (right).
3x C18 column vegetable oils
Figure 2: 3x C18 column vegetable oils 50mg/mL (5 percent) (CHCl3).

Separation Conditions: Column = 3X Zorbax SB-C18 (4.6 x 250mm, 5µm), Injection = 5µL, Flow Rate = 2.5 mL/min, Outlet P = 150 bar, SF = CO2, Mod = 9:1 ACN/MeOH, Gradient = 0 – 90min: 2-10%, Column T = 25°C, Make-up = IPA at 0.6mL/min, Caloratherm = 60°C, UV = 210/4nm REF 360/100nm, ELSD: Evap = Neb = 30°C, 1.60 SLM, Gain = 1, Smoothing = 5s, 10Hz.

Separation on a Silver-Loaded Stationary Phase

The separation of the vegetable oils on the silver loaded column is shown in Figure 3. On this column, separation is mainly based on the number of double bonds, resulting in a group type separation of lipids. Within a group of triglycerides with the same number of double bonds, some partial separation could be observed but to a lower degree as compared with separation on C18 (for example, PLL/OLO).

Retention time and peak area repeatability was tested on both columns for a test mixture containing PPP, OOO, and LLL. The RSDs percent on retention times were below 0.2 percent on ODS and around 1 percent on the silver loaded column. Peak area repeatability was around 2 percent on ODS and around 4 percent on the ChromSpher lipid column.

Separation of triglycerides of sunflower seed oil
Figure 3: Separation of triglycerides of sunflower seed oil (right), peanut oil (center), and soybean oil (left) on silver loaded ­column.

In Closing

This experiment demonstrated the separation of triglycerides in vegetable oil samples using the Agilent 1260 Infinity Analytical SFC System coupled to ELSD. The ELSD results were reproducible and provided enhanced sensitivity compared to UV detection. The separations obtained on octadecyl silicagel (reversed phase) and on a silver-loaded stationary phase (ChromSpher Lipid) were complementary. Analyzing the three vegetable oils on both column types demonstrated that the combination of both SFC methods creates an ideal quality-control tool for vegetable oil samples.

International standards for the analysis of vegetable oils are evolving. The methods described open the door for continued advances in the assurance of food quality and the fight against food fraud, particularly in regards to vegetable oils like olive oil. In addition, some of the methods, such as SFC separation on a silver-loaded column, could easily be applied to quality-control protocols for other types of food oils, including fish oils.


Dunkle, David, and Sandra work for the Research Institute for Chromatography in Belgium. They can be reached at Vollmer is employed at Agilent Technologies, Inc. in Waldbronn, Germany. Reach him at


References Furnished Upon Request



Current Issue

Current Issue

February/March 2015

Site Search

Site Navigation