Chromatography Forum

we just installed a new capillary column with integral guard column in our 5890

we prepared some new standards and injected them, we see no retention whatsoever, everything goes through the column with the heptane peak, then baselines

We have narrowed it down t one of two potential issues:
1. There is a small possibility we put the column in backwards.
2. Our 1/4" FID ferrule is cracked, this connection is a bit loose. I tightened it with the FID hot, and it is not spinning, but i dont know if that means it is sealing or not.

We have new ferrules on the way.

Would the lack of retention be symptomatic of either of these conditions, or should we look elsewhere.

What would be the difference in results of injecting with no guard column, if it was backwards?

How sensitive is the detector end to leaks? would this cause us not to see peaks?

We are using manual injection, BTW.

also, we switched from a JW 0.32mm fused silica column to a Restek 0.53mm metal column. Our inlet pressure before was in the 7-9psi range, now we are in the 1-2psi range. I know the larger diameter column will flow more at a given pressure, but should the difference be that dramatic?

Having the column in backwards will make no difference to the retention.

I presume that you are seeing a solvent (heptane) peak, and then nothing else. While this suggests that you do have gas flow through the column you need to verify tis by removing the column form the detector and dipping the end in some clean volatile liquid - you want to see a stream of bubbles.

Depending on temperature programmes and gas flow rates there is a vague possibility that you have too much retention, not too little, and that the peaks are not eluting within the run time. Try increasing the oven temperature.

Inject some methane (lighter gas will do) - what is the retention time ?

A leaking detector base might reduce sensitivity enough for you to see the solvent but not the analytes.

Peter

JF,

What are you shooting and how much? Since you went to a 0.53 metal, I am guessing something that has to be very hot to get off the column, like biodiesel?

Also, what is the flow rate through the column? Did this application work with the 0.32?

Best regards.

"A leaking detector base might reduce sensitivity enough for you to see the solvent but not the analytes. " - Peter

I'm guessing that this is your situation. I remember about 25 years ago a PhD in a different department was trying to get their GC working, and couldn't get any peaks. My boss asked me to help him out, and I found he had a leak at the column inlet (packed column then) the size of Baltimore, so nothing was going through the column. With capillary columns, if you have inlet pressure, you get flow, but likely most is getting out the leak and all you see is the solvent.

As to your question what a corresponding inlet pressure should be with your new column to match the linear velocity of your old column: I'd download Agilent's flow calculator http://www.chem.agilent.com/cag/servsup ... s/GCFC.htm
You may need to register with Agilent online (they have a fair amount of useful help) and that program works with Windows 2000 and XP as well, we use it here.

i work with jf,

previously, we had injected a couple of calibration standards and one actual sample and saw no retention. we thought it was the flow rate of the carrier gas being too high, adjusted it, then saw no peaks at all on the next run, not even a heptane solvent peak.

we took the unit apart and found a chunk of septa in the liner. jl ran some blanks and found a broader heptane peak which made sense for the corrected flow rate, but with much more tailing. there was also a ghost peak, which suggests retention was established, but the baseline also rose significantly and did not go down.

we do have a ferrule to the detector that is cracked. we ordered some new ones, but it does not actually connect the column but the adapter to the 1/4" nut (we have an agilent 5890). all the other connections are sound.

the other problem is if we did install it backwards, what are the consequences of that? i can see a problem if we used it long term without the guard or with the guard on the detector end!

How have you set up the flow for the column. The 1 to 2 psi is low and could be giving you problems. I suggest that you measure the flow at the exit of the fid, with hydrogen and makeup gas flowing. If the fid jet is partially clogged, you could have a back pressure across the jet when you have all gases flowing through it that is equal to your inlet pressure. This will virtually give you no flow in the column.

Another thing to look at is the zero level for the signal. If your old column was bleeding a lot you may have a high zero set up. If the new column bleed signal is well below the old value, your peaks must first come higher than the set zero value. If the concentrations are low, they could be below the zero value.

Gasman

Gasman

I doubt it. We use a bubble flow meter and we are getting really high flow rates above 4 psi. My boss and I are both engineers he did a quick pressure-flow calculation and we found that 2 psi is appropriate for 3 mL/min thorugh a 14 m long w/ a 2 m guard .53 um diameter column, where as before we had 8-10 psi with a .32 m column. When we hooked up the bubble flow meter, it gave us 3 mL/min as the method says.

We're definately getting flow through the column b/c, a) we see flow through the bubble flow meter at the detector and b) when I detached the detector end we submerged it in some methanol and saw bubbles.

As for resetting the zero point we have already adjusted that as well, but there was no need. In the beginning, during conditioning and during some initial blank runs the baseline was at the currently set zero point anyway.

Thanks

3 ml/min is too low a volume flow rate for the diameter of column, and with the detector back pressure issues raised by gasman, and a good chance that the pressure regulator is not working very well at such low pressure, the flow through the column needs to be set by injecting lighter gas and adjusting the inlet pressure to give you a mean linear flow rate through the column of about 50 cm/s for hydrogen carrier gas, 35 cm/s if you are using helium and 15 cm/s for nitrogen. In other words measure the retention time of the lighter gas and divide the column length by the time to get the linear gas velocity.

Peter

the astm method specifies 3 ml/min and it doesn't specify the column length/diameter. why is 3 ml/min too low a flow rate for that diameter?

the peaks i have been getting (including the heptane peak) have been very sharp and narrow, which i attributed to the flowrate being too high. when i lowered the pressure even more, the heptane peak achieved some of the broadness that it had before.

why would you inject lighter gas? you mean like methane or propane? if it's a gas, how is it going to retain onto anything? it seems like it would simply flow through the column, as the carrier gas would. so...why not simply measure the carrier gas flow rate directly and do a pressure drop calculation using the length/diameter of the column, then determine if it fits that linear velocity that way? if retention time is necessary factor, why not just insert a blank solvent and determine it's retention time along with another blank solvent and determine it's retention time and average them? Then you can determine the mean linear gas flow rate for the flow rates for the substances you are determining?

i suppose it is a shorter column with a larger diameter than we were using before, but the method, although it is poorly written, says use 3 ml/min as the carrier gas flowrate and doesn't specify the column length/diameter as a corresponding value.

it makes sense that at that pressure the EPC is not able to control the pressure adequately at that low of a pressure, i've seen that i've had to make a lot of adjustments and it does not adjust for approximate values when it is that finite.

it also makes sense that the mean linear velocity of the gas through the column is incorrect, and that the method was simply poorly written and assumed all columns were the same length/diameter and so did not specify that. even the standard deviation calculation for the calibration has a typo in it that renders it incorrect, so i believe it is possible that they simply did not take carrier gas flow rate as it relates to column diameter, pressure drop, and column length into account.

if the flow rate through the column is too high, what typically happens? do the peaks get narrower as i predicted, or is it something else? can the detector still pick up the peaks at the appropriate concentrations when the flow rate is too high?

-thanks all for your input. sorry for my doubts, i tend to question everything. i am a beginner with gc. thanks

For any given carrier gas there is a flow rate that gives the best separations in the minimum time (exactly what that flow is is somewhat controversial, run a search on the forum for some different views). Taking as a given that optimum separations are obtained when hydrogen flows along the column at 50 cm/s, helium at 35 cm/s and nitrogen at 15 cm/s the corresponding volume flow rates are 6.6 ml/min, 4.6 ml/min and 2 ml/min. BTW what carrier gas are you using ???? Unless you are using nitrogen carrier gas the 3 ml min is too slow.

If you do the calculations you will find that a volume flow of 3 ml min gives very high linear flow rates in narrower columns, so a method that specifies a volume flow rate without the corresponding column dimensions has something missing.

The simplest way to unequivocally determine the linear velocity of the carrier gas with the column installed and the instrument running is to inject a plug of an unretained substance and measure how long it takes to get to the detector. Unless you have a very thick stationary phase in the column, low MW compounds like methane are not retained by the stationary phase so they move at the same velocity of the carrier gas. The benefit of measuring linear flow like this over measuring volume flow and then doing the calculations is that you cannot measure volume flow with the detector gasses running - and with the very small pressure drops that you get with short megabore columns the back pressure at the detector can make a significant difference to the flow through the column. To borrow a phrase from elsewhere on this forum; don't calculate, measure !

At higher flow rates peaks typically get narrower (which all else being equal is a good thing) but there are all sorts of other things besides carrier gas flow rate that can affect peak width, and so it is not a good indicator that flow rate is correct or incorrect.


Peter

For any given carrier gas there is a flow rate that gives the best separations in the minimum time (exactly what that flow is is somewhat controversial, run a search on the forum for some different views). Taking as a given that optimum separations are obtained when hydrogen flows along the column at 50 cm/s, helium at 35 cm/s and nitrogen at 15 cm/s the corresponding volume flow rates are 6.6 ml/min, 4.6 ml/min and 2 ml/min. BTW what carrier gas are you using ???? Unless you are using nitrogen carrier gas the 3 ml min is too slow.

If you do the calculations you will find that a volume flow of 3 ml min gives very high linear flow rates in narrower columns, so a method that specifies a volume flow rate without the corresponding column dimensions has something missing.

The simplest way to unequivocally determine the linear velocity of the carrier gas with the column installed and the instrument running is to inject a plug of an unretained substance and measure how long it takes to get to the detector. Unless you have a very thick stationary phase in the column, low MW compounds like methane are not retained by the stationary phase so they move at the same velocity of the carrier gas. The benefit of measuring linear flow like this over measuring volume flow and then doing the calculations is that you cannot measure volume flow with the detector gasses running - and with the very small pressure drops that you get with short megabore columns the back pressure at the detector can make a significant difference to the flow through the column. To borrow a phrase from elsewhere on this forum; don't calculate, measure !

At higher flow rates peaks typically get narrower (which all else being equal is a good thing) but there are all sorts of other things besides carrier gas flow rate that can affect peak width, and so it is not a good indicator that flow rate is correct or incorrect.


Peter

The ASTM method D6584 does to a point define the column length and diameter. The method quotes " Columns, either 10m or 15m in length, with a 0.32mm internal diameter, amd a 0.1um film thickness have been found satisfactory". However, the next sentence states "Any column with better or equivalent chromatographic efficiency and selectivity can be used". The 3ml/min are for the 0.32mm column, and as Peter points out, you would need to use a different flow to get the 'equivalent chromatographic efficiency" with the 0.53m column.

If you check most of the commecial offerings for a biodiesel analyzer for the ASTM D6584 method, they are using the 0.32mm column. The Achilles heel of the analyzer is the connector between the 0.53mm guard column and the 0.32mm analytical column, and probably for this reason Restek are offering the one piece 0.53mm column with built in guard column.

Gasman

gasman, you are correct on the method. thanks, it makes more sense now! helium is our carrier gas btw.

i can see the benefit of doing this, although it seems somewhat redundant to just calculating it. it is another piece of equipment we have to purchase for the sake of determining linear flow as opposed to volumetric flow, which it seems you can determine empirically no matter what the pressure drops are...

the pressure drop is still going to effect the flow of an unretained gas to the detector, so i don't see how it makes much difference how it is measured volumetrically or otherwise if it is still going to flow at the same rate whether it is measured on a volumetric basis or a retention time to the detector basis. one does not seem to be more precise than the other.

in other words, is this necessarily a more precise way of determining linear flow rate?

Welcome to the world of "interesting" ASTM, USP, and other "official" test procedures....

The only equipment that you need to determine the linear flow rate by injecting lighter gas is a microsyringe (which you surely have already since you are in a GC lab) and a cigarette lighter. Suck some of the gas into the syringe, inject, a peak will appear on the chromatogram, the software will tell you the retention time. To measure volume flow you need a flow meter - there might be one built into the GC EFC, but the problem that you describe could be due to the EFC malfunctioning so you need an independent measurement.

I cannot think of any way of accurately volumetrically measuring a volume flow of 3 ml/min through a column that is connected to a detector with the detector gasses flowing to match the back pressure that you get when running an analysis. This is the advantage of measuring linear flow rate in terms of the retention time of an unretained compound - you can do it with the column installed to the detector and under exactly the same conditions as you use for the actual analysis.

Peter