Optimium flows in capillary GC.

[stanlee]

Does anyone have a link to a site where I can obtain optimum flow/pressure for various capillary dimensions ?

TIA

Stanlee

[Bernie]

Hi Stanlee,

you may find this link helpful

http://www.chem.agilent.com/cag/main.html

You can download some useful little helpers.

greets
Bernie

[WK]

Stanlee,
Try this link from Restek:

http://www.restek.com/info_calcs.asp?oldcookie1=

Click on: "Typical GC column characteristics"

WK

[lmb]

The following flow rates, measured in mL/min at 1 atm and room temperature regardless of actual column temperature, are recommended in Agilent 6890 GC Operating Manual, Table 8 (www.chem.agilent.com/scripts/Library.asp):

Column ID (in mm): 0.05; 0.1; 0.2; 0.25; 0.32; 0.53
Hydrogen flow: 0.5; 1; 2; 2.5; 3.2; 5.3
Helium flow: 0.4; 0.8; 1.6; 2; 2.6; 4.2

Notice that the flow rates are proportional to column ID (internal diameter).

By the way, be careful with other recommendations. For example, in 3m-0.1mm column, helium linear velocity of 20 cm/sec (recommended in some sources) corresponds to about 0.1 mL/min. This is far below the optimal flow rate of helium for this column.

lmb

[JI2002]

I downloaded the software from Agilent website and played with it, I can manually calculate the inlet pressure, average linear vilocity, tm (hold time) based on the carrier gas flow rate, the dimention of the column and oven temp. They all match the results from the software. But I have trouble to calculate the inlet flow rate at the given conditions. Any input is appreciated.

[Bernd]

Hello Stanlee,

You can get from me a data sheet with pressure/ flows for WCOT and PLOT columns.
Go to my profile and send me an email, You will get back the right data sheet.
BTW, for PLOT columns the flows are different from WCOT, this is not calculated in the FlowCalc software!
Regards
Bernd

[lmb]

To JI2002.
All gas that enters a column, leaves it. Therefore, inlet flow rate is the same as outelet flow rate. It is just called flow rate.

[JI2002]

lmb.

Thanks for your reply. This is how I figure the flow rate: There are different kinds of flow rates in GC, the most common one is the flow rate at room temp (25C) and 1 atm (14.7 psia), I believe it's the Outlet flow rate in HP FlowCalc software. Another flow rate is average flow rate at oven temp and Average Pressure across the column, it can be calculated from average linear velocity of the carrier gas. What I don't know is how to calculate Inlet flow rate. Here is an example from HP FlowCalc:

Carrier gas: He
Outlet pressure: o psia (GC/MS)

Column length: 30 m
Column i.d.: 0.250 mm
Oven Temp: 180C

If set up Outlet flow rate at 1.00 ML/min, then

Inlet pressure: 15.1 psig
Average flow rate: 38.1 cm/s
Inlet flow rate 0.809 ml/min @ 175C

I just don't know how to come up with 0.809 ml/min. Thanks for any input.

[lmb]

To JI2002

I just downloaded the HP FlowCalc 2.0 from the Agilent site (www.chem.agilent.com/cag/main.html) and played with it. I also called Agilent for clarifications.
The terms "Outlet Flow (mL/min)" and "Inlet Flow (mL/min)" in the FlowCalc mean the following:

"Outlet Flow (mL/min)" is a column volumetric flow rate measured at atmospheric pressure, AtmPress, and at some reference temperature, RefTemp, which is the same for all measurements regardless of actual column temperature. What exactly RefTemp is, is not clear, but it looks like RefTemp = 0C (more about it later).

"Inlet Flow (mL/min)" is the column volumetric flow rate measured at inlet pressure and temperature. (This measure of the column flow is necessary for calculation of time that it takes for the column flow to sweep an inlet liner of a known volume.)

If Outlet Flow, Fout, inlet gauge pressure, pg, and inlet temperature, Tin (in degree Celsius), are known then Inlet Fow, Fin, can be found as:

Fin = ((273 + Tin)/(273 + RefTemp)).(AtmPress/(AtmPress + pg)).Fout

where AtmPress = 1 atm = 14.7 psi.


Your Example:

Column and setpoints:
Column 30m-0.25mm
Oven temperature 180C
Gas helium
Outlet flow, Fout = 1 mL/min
Inlet temperature: Tin = 175C

Calculated parameters of helium flow:
Inlet pressure: pg = 15.1 psig
Average velocity: 38.1 cm/s
Inlet flow, Fin = ((273+175)/(273+0)) ((14.7/(14.7+15.1)) 1 mL/min = 0.809 mL/min

Comment:
It appears from the calculation of Inlet flow in the example (and from several other test of FlowCalc that I made) that the reference temperature, RefTemp, for the measurement of Outlet flow in the FlowCalc is 0C. On the other hand, as far as I know, in Agilent 6890 GC, in Agilent 6850 GC, and in the instruments from other vendors, the column flow rate is assumed to be measured at 25C. Keep that possible discrepancy in mind.

[JI2002]

lmb, thanks for the clarification. I did the calculation based on 25C.

[Ron]

I wouldn't worry too much about the theoretical optimal flow rate. I had a 9 component sample I was running on a 15 m x 0.25 mm id column, and close to the optimal theoretical flow rate the run took about 20 minutes using the suggested temperature program. Modifying the temperature program and increaing the flow to 10 mL/min allowed analysis in less than 2 minutes, with a minimum resolution between adjacent peaks of at least 4. If all your peaks are well resolved you can increase the flow significantly to speed up the analysis. Theory is fine, throughput is usually better.

[lmb]

Dear Ron,

If the throughput is important for you, the following theoretically developed and practically verified actions let you further reduce the analysis time without loss in resolution that you get from 15 m x 0.25 mm column at 10 mL/min (helium?):

1. Cut the column to 4.5 m (make three columns from one 15 m column)
2. Use recommended flow rate (2.0 mL/min for helium, 2.5 mL/min for hydrogen)
3. Speed-up the temperature program in proportion with the reduction in hold-up time (1.423 times for helium). (Try GC Method Translation freeware from Agilent, http://www.chem.agilent.com/cag/main.html, instead of actually doing these calculations)

I guarantee that, if you do that, the analysis time will be 1.41 min instead of 2 min – exactly as theory predicts.

Comments:

1. There is no need to choose between the throughput and the theory (as there is no need to choose between going and knowing where you go). I am not aware of any recommendation for the ultimate performance of a chromatographic system in any respect (resolution, throughput, detection limit, etc.) that did not come from theory. Thus, the use of the super-optimal flow (like 10 mL/min rather than 2 mL/min for helium) is known as one of many ways to trade excessive resolution for speed. However, the theory shows (and experiments confirm) that, in many cases (yours is one of them), cutting the column gives better results.

2. It follows from the theory that increasing the flow rate from optimal (2.0 mL/min?) to 10 mL/min should have reduced the analysis time from 20 min to only 6.9 min (by the factor of 2.914). The fact that you were actually able to achieve the 10-fold reduction in the analysis time suggest that, either in the original 20 min analysis or in the 2 min analysis or in both, the flow rate or the heating rate or both were not optimal. If that was the case then further reduction in the analysis time (beyond 1.41 min) might be possible. I can not be more specific without knowing more about the actual method parameters.

lmb

[Ron]

Unfortunately the 4.5 m column would require a negative head pressure for a flow of 2 mL/min into my mass spectrometer, so I need at least 10 m of column to control flow properly. In most cases the method does not need to be totally optimized, and I am not saying that theory has no plece in figuring out what to do. I have seen too many cases where people get so caught up in trying to develop the perfect method that they are still trying to optimize the method two weeks after the project could have been completed.