If I use a 30 meters capyllar DB-column is acceptable to use linear velocity at 50 cm/sec?

Are there any reference abut limits for linear velocity in GC (capillar and non-capillar) analysis?

sure, use any rate from about 20cm/sec to as high as you find usable.

Most likely, the (average) linear velocity of 50 cm/s is somewhat too high for most of the 30 m columns. Even more so if the carrier gas is not hydrogen. Generally, the optimal linear velocity depends on the type of a carrier gas, the column diameter, and on the column length. Unfortunately, there is no simple formula for the optimal linear velocity for all useful column dimensions, and, in many cases, the linear velocity that is recommended in the literature (including the recommendations provided by the column and the GC instrument manufacturers) is far from the optimal values. This can cause a loss in a column separation power and/or in the speed of analysis.

However, there is a simple solution to the problem of setting the pneumatic conditions of a column: Ask NOT what the gas linear velocity should be. Ask what the gas flow rate should be.

And the simple answer is: The default flow rate that would most likely lead to the best separation-speed tradeoff can be found for a capillary column as

(default flow rate in mL/min) = c*(column internal diameter in mm)

where
c = 10 for hydrogen,
c = 8 for helium,
c = 2.5 for nitrogen or argon

Example. For helium in a 0.32 mm column (of any length), the default flow rate is: 8*0.32 = 2.6 [mL/min].

Another way is to use the following program and aim for 1.5 times the optimum linear velocity for your carrier gas - this allows for the decrease in flow rate that occurs during a temp program

the optimum linear velocity for each gas from the van Deemter curves is

Nitrogen 12cm/s
Helium 20cm/s
Hydrogen 40cm/s

http://www.chem.agilent.com/cag/servsup/usersoft/main.html#flowcalc205

Regards,

Ralph

Dear Ralph,

I agree that, in many cases, especially for typical megabore and median-bore columns, our recommendations might lead to roughly the same column performance representing a compromise of separation efficiency and speed of analysis. Unfortunately, it is not so for all column dimensions, and, in many cases, our recommendations lead to substantially different column performance. Following are several examples. In all examples, carrier gas is helium at 100C. The flow rate is measured at 25C. (You recommend to use Agilent’s Flow Calculator for conversion of velocity to flow rate and vice versa. Here I used Agilent’s Method Translation. Either way, the results must be the same.)

For helium, you recommend (average) velocity of 20cm/s.

Example 1. For a 3m x 0.1mm column, 20cm/s gas velocity corresponds to flow rate of 0.09mL/min that is 9 times lower than 0.8mL/min that I recommend. The gas velocity corresponding to 0.8mL/min flow rate is 93cm/s. This means that my recommendations lead to 4.6 times shorter analysis time (something like 0.5min vs. 2.3min). Granted that my recommendations correspond to a speed-optimizing flow rate (SOF) which is about 40% higher than conventional efficiency-optimizing flow rate (EOF), still the flow rate that comes from your recommendation is 6.4 times lower than the flow rate corresponding to the highest column efficiency. This means that your recommendations lead to about 3 times lower plate number than my recommendations. If my recommendations are correct then the summary of the differences between our recommendations is this: Your recommendations lead to about 3 times lower plate number and 4.6 times longer analysis time compared to my recommendations. Of course, you can question my recommendations.

Example 2. For a 10m x 0.1mm column, the difference between our recommendations is less dramatic, but it is still significant. For this column, the flow rate corresponding to 20cm/s gas velocity (as you recommend) is 0.14mL/min – nearly 6 times lower than 0.8mL/min that I recommend. The gas velocity corresponding to 0.8mL/min flow rate is 55cm/s.

Example 3. Even for a 30m x 0.53mm column, there is still a noticeable difference between our recommendations. The flow rate corresponding to 20cm/s gas velocity is 2.3mL/min compared to 4.2mL/min that comes from my recommendations. The gas velocity corresponding to 4.2mL/min flow rate is 35cm/s.

One can conclude from these examples that our recommendations are substantially different, and, at least in some cases, the difference could lead to substantially different column performance.

You suggest that your recommendations come from Van Deemter curves. Unfortunately, I am not aware of any curves that lead to these recommendations, and, therefore, I can not verify them. As for my recommendations, they can be found in two publications [1,2] and in manuals for 6890 GC (since 1994). Needless to say that the recommendations were tested in numerous experiments.

Regards

Leon


References

1. L.M. Blumberg. "Theory of Fast Capillary Gas Chromatography. Part 3: Column Performance vs. Gas Flow Rate", J. High Resolut. Chromatogr. 22: 403-13 (1999).

2. M.S. Klee and L.M. Blumberg. "Theoretical and Practical Aspects of Fast Gas Chromatography and Method Translation", J. Chromatogr. Sci. 40: 234-47 (2002).