LOD on HPLC vs UPLC

[tommymouse123]

I recently compared a UPLC column (1.6 uM) with HPLC (5 uM) column on acidic condition from same brand. The LOD/LOQ on UPLC is not better than HPLC on most metabolites (mildly polar like HILIC mode). Theoretically the UPLC's LOD or LOQ should be better than HPLC because the peak shapes are sharper and more narrower on UPLC's (whereas, HPLC has reasonably good peak shapes too). For example, Glycine has good peak shape, narrower, sharper on UPLC, but peak area or S/N is 3 times lower on UPLC than that of HPLC. It DOES NOT make sense.The background of UPLC has no bleeding from column, but mainly are ACN or formic acid adducts like 2ACN or water+ACN etc, which are similar to HPLC column on using. Additionally, ESI settings has also been considered and tuned on UPLC of 0.3 ml/min flow or HPLC of 0.2 min/min.

[per_oxid]

I recently compared a UPLC column (1.6 uM) with HPLC (5 uM) column on acidic condition from same brand. The LOD/LOQ on UPLC is not better than HPLC on most metabolites (mildly polar like HILIC mode). Theoretically the UPLC's LOD or LOQ should be better than HPLC because the peak shapes are sharper and more narrower on UPLC's (whereas, HPLC has reasonably good peak shapes too). For example, Glycine has good peak shape, narrower, sharper on UPLC, but peak area or S/N is 3 times lower on UPLC than that of HPLC. It DOES NOT make sense.The background of UPLC has no bleeding from column, but mainly are ACN or formic acid adducts like 2ACN or water+ACN etc, which are similar to HPLC column on using. Additionally, ESI settings has also been considered and tuned on UPLC of 0.3 ml/min flow or HPLC of 0.2 min/min.

I would also expect impoved sensativity for a given injection volume if lc system dispesion volume is low enough.
For the following reason
Injected sample volume is less diluted on column
Flowrate is usually lower with improved ESI ionisation
ESI spray can be tuned for higher sampling to ms orifice

I tried to move from 2.1 to 1.0 mm column but did not get the improvments. This was likely due to the system dispersion volume being to high. I could probably get the improvements if i lowered system volume but I wish to keep column manager and diverter valves for practical reasons.

Your flows does not make sense for transfering a method, the narrower column would have much lower flow for the same solvent velocity use a calculator tool for hplc to uhplc migration?
Is your system volume low enough to use 2.1 ID columns?
Have you tuned the ESI spray position for optimal s/n?

[tommymouse123]

Thermo U3000- QE plus system is in use. Injection volume=3uL. It seem that no problem for compatibility. UPLC should be higher flow rate for practice. Everything is considered but no improvement. 0.2 or 0.3 ml/min is not significantly different. so maybe ESI tune is not a big role. However, if the peak is sharper and narrower, S/N or peak area should be better, it is very puzzled.

[per_oxid]

Sorry I was thinking of hplc as larger diameter column than uhplc. For same column dimension I also would expect some improvements in sensativity with smaller particles and narrower peaks but mainly better peak capacity. Please post if you find the explanation.

[Multidimensional]

"tommymouse123": Your question is irrational. "UPLC", which is a trademark or Waters, not a technique is HPLC. HPLC is the technique being used so it never makes any sense to compare a trademark to a technique. HPLC is the correct name. Now some vendors do indeed market their products and even columns as "UPLC" or "UHPLC", but this is pure marketing/sales language of no scientific value. The sales people love the idea that they can fool new and inexperienced users into thinking that they have a new product to sell by giving it a silly name. SO please be cautious of this and use the correct terminology.

BTW: My boss tells me that columns of all dimensions (320m to 10mm ID) with all reasonable particle sizes (1.8 to 10u) were used back in the 1980's for analytical separations and the technique was still called HPLC (just as it is now). No one ever changed the name of the technique because they used a column that had a narrow bore or was packed with 2 micron particles. Changes to the column packing or dimensions are normal variables used to optimize an HPLC method. They often used 1mm and 2.1mm ID columns packed with tiny particles for LC and LC-MS methods at very low flow rates (same as we do know and it was called HPLC).

It sounds like in your case you certainly did not make a fair comparison between a column labeled 'UPLC' and HPLC. If you take a standard HPLC method and re-scale it for use with a narrower column OR a column with even smaller particle size (MUST BE THE SAME EXACT COLUMN SUPPORT inside, just a different particle size, or ID. Otherwise, you are not making an actual comparison), then IF THE METHOD's are optimized AND the instrument's flow path is ALSO optimized (never assume as flow cell, injector, tubing ID etc must be evaluated), then you may see better response rates or LODs with the smaller particles, BUT as noted, there are so many variables to consider. A lot if training is needed and familiarity with the specific instrumentation is needed to do it correctly. Look in the journals to find plenty of examples where untrained users developed methods with ultra small particles in tiny columns and achieved worse results than if they had used std sized columns and 5 or 10u particles.

In other words, if you do not correctly optimize the flow rate, mobile phase comp, flow cell, tubing ID and gradient parameters for each method (esp. when scaling down), then you will never see improvements to the chromatography (these are all advanced level tasks).

[Multidimensional]

Here are some links to authoritative articles that my boss provides on the web for free. These articles come from his training classes and consulting advice.

"Terminology. Which is it? "UPLC" (TM) , UHPLC or HPLC? The correct name is still HPLC"; https://hplctips.blogspot.com/2015/08/t ... lc-or.html

"HPLC to UHPLC Conversion Notes (Column Dimensions, Flow Rate, Injection Volume & System Dispersion)"; https://hplctips.blogspot.com/2011/07/h ... notes.html

"Speed Up HPLC Analysis Time Using Higher than "Normal" Flow Rates with SMALLER Particles "; https://hplctips.blogspot.com/2021/01/s ... using.html

[JI2002]

Assuming all other conditions (including the column dimension, injection volume, flow rate, and the MS parameters) except partial size are the same, the noise level should remain the same for sub 2 um particle size column and 5 um particle size column. Smaller particle size column should generate narrower peaks (higher signal) and therefore higher S/N.

The LCMS sensitivity is flow rate dependent.

[Multidimensional]

"JI2002 wrote: Assuming all other conditions (including the column dimension, injection volume, flow rate, and the MS parameters) except partial size are the same, the noise level should remain the same for sub 2 um particle size column and 5 um particle size column. Smaller particle size column should generate narrower peaks (higher signal) and therefore higher S/N"

- That is not true. The flow path optimization of the specific instrument is the number one factor that will determine the output data and actual limits of detection. Even if the column dimensions are the same, the reduced particle size may not show improved resolution or better peak shape unless steps are taken to optimize the flow path. Use of narrow bore columns and/or sub 2 micron supports (which are often packed into very low volume columns) requires that the HPLC system's flow path also be optimized for the method. A std analytical system with a 1 mL dead volume will literally wash away and dilute many of the possible advantages of using the smaller particles or narrow bore column. When you use a typical 4.6 x 250mm column with 3 to 5 u particles inside, the column volume will be around 2.8 mLs. This is a large enough volume that small differences in extra volume do not degrade the results. However, when you replace that "std" column with a short column or narrow bore column, then the column volume drops to a very low value increasing the chances of dilution and diffusion, messing up the results). In order to have any chance of improving the results (seeing narrower peaks, better resolution etc), the overall connection tubing should be as short as possible and as narrow as allowed to reduce diffusion lose. The flow cell's volume may need to be reduced too. Also, the injector's flow path must be reduced too for the same reasons. We see so many "new" methods (journal articles, white papers, application notes) where the novice users are unaware of these basics and develop methods using column with very small particles and show no improvement, or worse results than observed with std 5u particles.

SO true LOD is determined by METHOD and FLOW PATH optimization, not just adding in a higher efficiency column.

[Multidimensional]

"Determination of HPLC Column Void Volume / Dead Volume, Dead Time (T zero)"; https://hplctips.blogspot.com/2011/05/d ... -time.html

[Multidimensional]

This article, again, is really worth reading and understanding to get to best results of any column which uses very low volumes.

"HPLC to UHPLC Conversion Notes (Column Dimensions, Flow Rate, Injection Volume & System Dispersion)"; https://hplctips.blogspot.com/2011/07/h ... notes.html

[JI2002]

I thought there is a new HPLC theory contradicting what I said. All you said is basic HPLC knowledge. When EVERYTHING else is the same other than particle size, the column with smaller particle size always generates better efficiency although the extra column volume needs to be minimized to take FULL advantage of the smaller particle size.

[DR]

The smaller things get (especially column ID, particle size), the harder it is to get appreciable marginal returns in S/N, LOD etc. as what used to be insignificant extracolumn effects are more and more significant as diameters shrink. It's all covered by the math presented in any thorough HPLC review of theory. The math does not change with the addition of a U to the name of the technique. Shorter, straighter tubing runs, perfect unions and perfect packing on the part of the column vendor, coupled with a high quality determination of ideal linear velocity and its corresponding flow rate are all requirements of an improved LOD via sharper peaks.