Part 4: Putting Our Interpretation to the Test

Guest Author: Dr. Jeff Layne

Welcome back to our final installation of this blog series on the effect of column inner diameter on chromatographic performance, focused in our case on LC-MS applications.  In the last article, we examined the data generated when using columns with three different inner diameter values (4.6 mm, 3.0 mm, and 2.1 mm) under a range of flow rates, and focused on the resulting effects on peak height response, system pressure, and run time.  What we had found on our system (Agilent 1200 series LC and an API 4000™ MS (SCIEX)), was an increase in peak height intensity as we decrease column ID. This is fantastic if we want to maximize LOD/LOQ values, but the greatest benefit is going from 4.6 mm to 3.0 mm, and then moving from 3.0 mm to 2.1 mm, which generates only a modest increase in peak height response, but at the cost of longer run time and pressure.  But will the interpretation of the data, based upon a single analyte (17-alpha-OH-Progesterone) translate to a more complex, realistic sample?

To answer this, we ran a panel of about 40 pesticides that typically are used in cannabis cultivation.  The pesticides have a wide range of polarity values and differ greatly in the chemical nature (a mixture of bases and neutrals primarily, with a few acids as well).  Since sensitivity was the primary concern, we ran the method using three of the top column ID / flow rate combinations (based upon peak height intensity) from our previous evaluation – the 50 x 4.6mm at 0.5 mL/min, the 50 x 3.0 mm at 0.5 mL/min, and the 50 x 2.1 mm column at 0.2 mL/min.  Representative overlaid XIC are in Figures 1-3, below.  For this screen, we used a simple water/acetonitrile/0.1% formic acid mobile phase, with a gradient of 3 – 100% B over 5 min, with a 2 min hold from 5 – 7 min.  The gradient profile was held constant for the three different columns, but the flow rate was adjusted as specified.

The first thing that was noticed was the shift in retention times as we went between the three different column/flow rate formats.  Since the 50 x 3.0 mm and 50 x 4.6 mm columns were run at the same flow rate of 0.5 mL/min, retention time decreases using the 3.0 mm column, which is expected from greater linear velocity.  Then, when the 2.1 mm column is moved at 0.2 mL/min, run time increases again as expected.

We then pulled out the XIC (extracting ion current) for three analytes – dimethoate, thiamethoxam, and carbaryl – and calculated the S:N (Signal to Noise) ratios under the three different column and flow rate combinations (XICs in Figures 4-6).  Tabulated results for S:N ratio, retention time, and back pressure are in Tables 1-3.  However, how do these results compare to our test probe results from our previous work?

Overall, the data shown in Tables 1 -3 seem to confirm our previous findings.  In each case, the 50 x 2.1 mm column operated at 0.2 mL/min gave the greatest peak height intensity, although the increase in peak height over the 3.0 mm column for dimethoate and thiamethoxam was very marginal.  However, the S:N ratios, which are critical in determining LOD and LOQ values, were significant better for the 2.1 mm conditions, despite an only marginal increase in peak height.  The exception being dimethoate, for which the 50 x 3.0 mm gave the best S:N ratio according the software S:N calculation.  Looking at the respective chromatograms (Figures 4b and c), it’s hard to see a true difference in the two XIC, and perhaps that S:N calculation was skewed by a subjective choice of regions to use for the S:N calculation.  But for carbaryl and thiamethoxam you definitely can see a visual improvement S:N ratio that supports the calculated value.

It’s safe to say that this data largely supports our initial findings.  On our LC-MS system, the best sensitivity was achieved using a 2.1 mm column operated at a low flow rate of 0.2 mL/min.  But, the trade-off was that it also gave the longest retention times.  The 50 x 3.0 mm operated at 0.5 mL/min offered a good compromise, providing sensitivity close to the 2.1 mm in most cases, but was a shorter analysis time.  In the end, each analyst is going to have to decide which factor is most important to them, but I hope that this exercise at least helps to point those testing in the correct direction for column and flow rate selection.

Figure 1. Pesticide panel using the 50 x 4.6 mm @ 0.5 mL/min

Figure 2. Pesticide panel using the 50 x 3.0 mm @ 0.5 mL/min

Figure 3. Pesticide panel using the 50 x 2.1 mm @ 0.2 mL/min

Figure 4. XIC for Dimethoate using the three different column/flow rate combinations

a. 50 x 4.6 mm @ 0.5 mL/min

b. 50 x 3.0 mm @ 0.5 mL/min

c. 50 x 2.1 mm @ 0.2 mL/min

Figure 5. XIC for Thiamethoxam using the three different column/flow rate combinations

a. 50 x 4.6 mm @ 0.5 mL/min

b. 50 x 3.0 mm @ 0.5 mL/min

c. 50 x 2.1 mm @ 0.2 mL/min

Figure 6. XIC for Carbaryl using the three different column/flow rate combinations

a. 50 x 4.6 mm @ 0.5 mL/min

b. 50 x 3.0 mm @ 0.5 mL/min

c. 50 x 2.1 mm @ 0.2 mL/min