Print

Print


Gerry,
OK I will check all ferrules and capillary connections.

Travis,
I think it makes sense... I will check T2.

Thank you so much,

Ayumi


From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf Of Travis Meador
Sent: Monday, July 22, 2019 4:39 PM
To: [log in to unmask]
Subject: Re: [ISOGEOCHEM] how to get rid of m46 peak (1000mV) appearing before CO2, for N2O analysis

Hi Ayumi,

If there is a leak in T2, is it possible that the He Blank injection may purge air that has leaked into the trap space? If so, then N2 injected with air or N2O may not be completely purged from the trap space, thereby giving higher signal with 200 mL injection. Also, the N2 level would presumably decrease with longer transfer time (@ 24mL/min?) as you purge more N2 out of the trap space.

Does the Ar background increase with time if you leave T2 out of the trap?
Injecting Ar or Ar in He (monitoring with Instrument Control) could reveal if the Ar is indeed exiting the vent or accumulating in T2 space.
Have you sniffed for Ar around the T2 trap or considered replacing the capillary in T2?

...thoughts from a Gasbench newbie.

Best of luck,
Travis Meador



From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf Of gerard olack
Sent: Monday, July 22, 2019 4:52 PM
To: [log in to unmask]<mailto:[log in to unmask]>
Subject: Re: [ISOGEOCHEM] how to get rid of m46 peak (1000mV) appearing before CO2, for N2O analysis


HI Ayumi--

You mean I skipped over the sentence where the fully flushed tube does not cause a problem?  Sorry about that ;)

Actually, it does sound like you've isolated the issue--the increase in transfer time to T2.  Why air is in that line, or if there is a little in the He flush line,  I don't know--but you seem to be flushing it out.  Look for He leaks around the ferrules for those connections and for all ferrules in case it's in the previous step.  If the capillary is not fully inserted when the connection is tightened, the ferrule seals on the edge--and I hate when that happens.  The capillary will work it's way out a little, and you'll get a leak.  Snugging it up will help for a little bit, but only for a while.

take care

gerry


On 7/22/2019 9:00 AM, Ayumi Hyodo wrote:
Hi Gerry,

Thank you for your response. Do you think this is a downstream air leak even though running helium in a 100ml container loaded on Precon does not show the m46 peak?

Best,

Ayumi


From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf Of gerard olack
Sent: Friday, July 19, 2019 4:01 PM
To: [log in to unmask]<mailto:[log in to unmask]>
Subject: Re: [ISOGEOCHEM] how to get rid of m46 peak (1000mV) appearing before CO2, for N2O analysis


HI Ayumi Hyodo--

It sounds like a down stream air leak.

I'm presuming if you put the sampling needle into a flushed tube, you see flow out the vent lines on the valco valve--either in load or inject modes.  And I'm presuming you have short pieces of capillaries to serve as vent lines on the valco valve--and it's configured in the standard gasbench manner.

If you just try running a few "samples" without having any sample tube in the gasbench, then there is no sample flow to and thru the valco valve.  The first pulse will be residual air in the sampling loop.  The rest should be small--whatever air can diffuse into the valve thru the capillary lines.  Also, any "air" leaks there have to go thru the GC column, so should take time to show up in the signal.  If there is a pressure effect, i.e. not enough flow thru the open split, then that would show up relatively quickly, say 20 secs or so from valve switching.  You can play with the timings of the valve loading and injecting to make sure what you see is not just a peak from an earlier injection corresponding with the current valve switching.

If you still have big peaks that take time to travel to the open split, then it's probably a problem with the valve seat.  Even with big air leaks, you can use Ar to chase it.  If it's an "immediate" issue when the valve switches, then there could be a problem with flow to the open split.  The nafion tubing just prior to the open split might need to be cleaned, or there could be a partial clog anywhere from valve on down.

If there's not a problem with this test, then the problem is on the sampling side.

take care,

gerry


On 7/18/2019 1:34 PM, Ayumi Hyodo wrote:
Dear Isogeochem,

We had been analyzing O and N isotope ratios of N2O using Precon-GasBench-DeltaV Advantage well, until I left Precon and GasBench off for a half year. When I turned them back on (without changing any plumbing), there is a large (1000-1200mV) m46 peak before the CO2 peak, when 100ml room air is loaded. The N2O and CO2 peak heights are same with what I had been getting before.


*         When blank (helium gas in a 100ml glass container) is loaded, the m46 peak does not appear, indicating the peak is not due to the leakage of the system or the sample container.

*         When 200ml of room air is loaded, the size of m46 is ~2000mV.

*         When I increase transfer time from the sample container to the first LN trap (called T2) from 400 sec to 1200 sec, the m46 peak height is largely reduced to 60mV.

*         Increasing transfer time from T2 to T3 (2nd LN trap) does not change the m46 peak heights.

*         When I increase the helium flow rates from 20 psi to 35 psi (maximum), the size of m46 is not changed so much.

So I thought the transfer time (400 sec) is not long enough anymore for all N2 gas in room air sample to exit through T2, and remaining N2 is causing the m46 peak. Possibly somewhere in Precon (between the sample container to T2 exit) is clogged? However, the flow rates at several fittings of Precon are constant and 23-25ml/min when helium is 20psi. The flow rate at the Valco valve exit is also 24ml/min.

Now I don't know what to do and how to get rid of the m46 peak (except changing the analysis time extremely long). Any comments and suggestions are appreciated.

Additional information:

*         The m46 peak still appears in a blank (helium) analysis after one 100ml air analysis.

*         Baking GC did not help to remove the m46 peak.

*         The m46 peak sizes of 100ml room air, 100ml outside air, and 100ml of 0.3ppm N2O in N2 balance are similar, so it is not due to our lab room air.

*         Replacing chemicals in the chemical trap, and sizes of the chemical trap did not help to solve this issue.

*         If Valco Valve before T2 is not working well, I would think I don't get good N2O and CO2 peaks.

*         (T1 and Oxi combustion furnace are removed from our system.)

Thank you,

Ayumi Hyodo
+++
Research Assistant Professor, Ph.D.
Stable Isotopes for Biosphere Science Laboratory
http://sibs.tamu.edu/
https://twitter.com/sibslab<https://urldefense.proofpoint.com/v2/url?u=https-3A__twitter.com_sibslab&d=DwMD-g&c=u6LDEWzohnDQ01ySGnxMzg&r=l14DqXVYPe0bLgbRzDXNg6ellXdI_3oAURjjGBwv4B0&m=08_znyPFJn7MuTMTlEWSBq4nFJegD3E_-su_Y1dn9aQ&s=TZVz5gs0o0jP2alI7pO_1fauDxkhwH0ZOSB-yVFHHpk&e=>
215A Kleberg Building
2126 TAMU, Texas A&M University
College Station, TX 77843
[Office phone: 979-845-1497]