I'm developing a HPLC method to analyse triglycerides with C8 to C18 length alkyl chains, and my question concerns the detection of these compounds.
UV detection is impossible (especially for saturated triglycerides).
I have a method developed with the refractive index detection. However, i think it's not the best solution (obligatory to work in isocratic mode, low sensibility).
So I think it would be a better idea to work with an ELSD. The parameters are a nebulization temperature of 40°C, and a gas pressure of 3.5 bar. I checked that trioctanoate glyceride is well detected (and not evaporated) : I conclude that the longer triglycerides will also be detected.
Consequently, I think ELSD is a good solution for the detection of triglycerides.
However, I was told that RI should be preferred due to the instability of the triglycerides in the ELSD without further explanation. But I do not unterstand : the nebulization gaz is N2, so no risk of oxidation, and the ELSD temperature is rather low (40°C).
My question : does anyone have an explanation why ELSD should be avoided ? Do you agree RI should be preferred ?
Hi Gaëtan,
if you have access to an ELSD then why not use it. Do not expect the response to be highly linear, the light scattering from small particles is fundamentally not really proportional to concentration or mass as size distrubution of particles will be dependent on analyte concentration entering the nebulizer. If you run a gradient, reponse may also depend on where in the gradient your peak elute, but this is not a major problem if you calibrate your target analytes individually. Good luck!
yes i know that the response is not linear if you trace peak area as a function of concentration. But, il you work with ln(area) versus ln(cc°), it will be linear.
indeed, why not use a ELSD if i have one. But i try to know why i was told that RI detection should be preferred to ELSD. But maybe it is a mistake ....
Personally I can't see the reason why not to use ELSD. As you pointed out, there's inert nebulising gas, low temperature and analytes reside inside detector's drift tube only for few seconds. For me RI is a no-go option.
As you can see in my paper "Triacylglycerols from viper bugloss (Echium vulgare L.) seed bio-oil" using ELSD and DAD detectors is very suitable for TAGs.
Using the RI isn't a good choice because of it's low sensitivity and baseline drift when works with gradient.
In agreement with the previous comments, in my opinion the best choice for analysing TG is actually ELSD. UV is not convenient because most solvents absorb in the same wavelengths as TG. IR is useless if you are using a gradient because of a high drift. I have used ELSD for many years with very good results. The only problem is that the response is not linear. You can overcome this using logarithms but you can also adjust the points to polynomial curves. There is also a chance to have linear responses in your concentration range, but this is quite unlikely.
Also you can use ELSD to quantify total TG in a lipid sample or to separate the molecular species.
You can check these papers: Perona et al. J Chromatogr B Biomed Sci Appl. 1998;706(2):173-9. Perona JS and Ruiz-Gutierrez V. J
Chromatogr B Analyt Technol Biomed Life Sci. 2003;785(1):89-99. Perona and. Ruiz-Gutierrez, J Liquid Chromatogr. Rel Technol, 22(11), 1999, pp. 1699-1714
Why analyse triglyceries by HPLC? Existing GC-FID methods are quite effective. Look up ISO 11053 as a starting point.
For TG analysis, besides the GC-FID is the ELDS a good choice, especially if there is a need in a gradient elution profile.
A constant nitrogen flow is very important but also the remarks of Mr. Javier Perona concerns the linearity of the response.
Have used ELDS for sugars analysis. The non linearity of the calibration curve will be dependant upon how wide the dynamic range of instrument response used for the calibration curve. Ensure that the calibration is well defined for a known response range, then ensure that the sample response is within this range.
We used ELSD years ago and continue to use it for TAG as well as DAG. It is a very sensitive method and the linearity issues are not that great if you are in the right part of the curve. Like anything else, there are pluses and minuses, but in the end this is a robust detector with a relatively broad linearity over a wide-dynamic range. Give it a go.
I think as compared to RI, this is the way to go when using an HPLC. The real question is what is it you are trying to analyze? Total TAG or TAG molecular species? If it is total TAG, this is a great method. For others, GLC or LC-MS is a better route.
@Eric Murphy : i want to analyse the repartition in each TAG
Otherwise, i was answered for DAD, but concerning the ELSD, are the TAG response factors similar ? So we could assimilate their area pourcentage as their proportion in concentration
The use of ELSD for TAG analysis is well recognized. I still remember the 20 years old article by W.S. Letter (Journal of Liquid Chromatography, 16, 1993, 225) who used ELSD for qualitative analysis of TAG. Regarding quantitative analysis and detector linearity, these are issues inherent to any analytical technique that could be addressed by selecting the appropriate linear range and regression model. So, carry on developing your TAG method with your ELSD detector.
Hello Gaëtan: CAD and ELSD types of units can both can be used for triglyceride analysis by HPLC. Both types of detectors are not "universal" detectors (no such thing), but have shown to be useful with these types of compounds. *CAD units are much more expensive than ELSD units, without any difference in sensitivity (sensitivity is more of a function of the exact sample used, the mobile phase chosen and the detector settings employed).
RID can of course be used, but the sensitivity will be poor with many compounds AND you will be limited to isocratic analysis only. RID operation will be much easier than ELSD and requires very little training or expertise. ELSD operation and optimization requires a great deal more technical skill, plus regular cleaning and maintenance of the detector.
All of these detectors work and you do not need to derivatize the sample (a big plus in my book). Use whichever one is available and provides a reproducible result.
Thank you Pedro for the citation to one of my old papers !
"Evaporative HPLC Detectors; CAD (Charged Aerosol Detector) and ELSD. (Evaporative Light-Scattering Detector)"; https://hplctips.blogspot.com/2017/12/evaporative-hplc-detectors-cad-charged.html
"ELSD Optimization Procedure";
http://www.chiralizer.com/uploads/8/9/7/7/89770565/elsd_optimization_procedure_evaporative_light_scattering_detector_hplc.pdf
- Badges
- Science method