HI German--

I had problems with my GC column in June (6/22/05, I ended up compacting 
the PLOT column, so the flow was off and peaks were taking forever to 
elute), and got some good tips (Paul and Wolfram, posted below). 

It also sounds like you are having timing issues and perhaps Isodat 
problems--check these first (sorry, I like writing novels).  The latest 
patch, 2.38 (and other ones too?), screws up the on/off, load/inject 
designations for the gasbench switching valve--at least for our set up 
the picture on the computer is opposite of what in going on inside the 
gasbench.  When the screen shows load mode, i.e. it shows the sample 
filling the loop, you can check the "waste" lines coming out of the 
valve itself (remember to have sampling line inserted into a tube ;).  
If He is coming out the waste line by the sample input line, then it's 
NOT filling the loop.  If it's coming out of the waste line next to the 
sample loop (opposite side), then it is filling the loop.

You should be able to fix it by swapping the air lines--just remember 
that's what you did, and to double check things during the next software 
upgrade.  Alternatively, you can modify your programs.  Either way, at 
the end of the run leave the system in the software's version of "load" 
because that's the default starting position for the next line in the 
sequence.

If the load and inject designations turn out to be ok, double check the 
timings for load and inject in the program.  You may need some extra 
time in "load" mode to pressurize the system and push sample through 
loop--not a big deal for 100uL loop, but couldn't get larger loops to 
work well.  Also check the flow through the sampling line (guesstimate 
from bubbling through water out of waste lines from switching valve 
works). 

The nafion tubing can also get ugly and kinked and even clog up a bit.  
They are a pain to get to for cleaning though.

take care

gerry

p.s. plot column notes:

I have noticed that the GSQ PLOT column (J&W, now Agilent)has great stability.
I've been abusing them for years, and have never had one clog up like some
PLOT columns are known to do. Disconnecting a PLOT column without first
bleeding off the head pressure is usually a recipe for disaster, but doesn't
bother the GSQ at all.

Paul Eby
University of Victoria


At the risk of teaching my granny how to suck eggs, you are aware, aren't
you that PoraPLOT columns are exquisitely sensitive to sharp increases in
carrier gas pressure and any other kind of mechanical shock?

When installing a PoraPLOT column, carrier gas pressure should be increased
in 1 psi increments with 2 min intervals, as otherwise one risks the column
bed to collapse / compact (leading to an increase in back pressure and poor
chromatography).

I'm afraid from what you describe it sounds like this happened to your
column.

Allegedly, the new PoraBOND columns do not suffer from this ailment and can
even be used with pressure programmes (manufacturer's claim, not mine).

Chrompack/Varian do as yet not offer PoraBOND equivalents for all the
PoraPLOT columns but as luck would have it, they do a PoraBOND Q.

>From experience, I would stick with Chrompack/Varian when it comes to PLOT
columns.


Cheers,

Wolfram








German Mora wrote:

> Dear all,
>
> We have been having some problems with our GasBench-II system lately.  
> We typically employ our GasBench to determine the isotopic composition 
> of CO2 from pure gas samples and from CO2 equilibrated with water 
> samples. We have noticed that our sample CO2 peaks are progressively 
> arriving later.  Relative to samples ran a month ago, they are 
> arriving 40 seconds later.  Despite their late arrival, the isotopic 
> values were fine. We modified the ISODAT method to make sure that the 
> arrival of the peaks did not interfere with the reference peaks.  The 
> problem is that a couple of weeks ago, there was a gradual change in 
> the intensity of the peaks and their isotopic values.  Under normal 
> circumstances, the sample peaks are equally spaced. During the past 
> two weeks, the second peak in each run is now closer to the first one. 
> The remaining five peaks are however still equally spaced.  Moreover, 
> this second peak is very small (usually half of the intensity of the 
> first peak).  The third peak has a higher peak intensity (and 
> presumably the CO2 abundance) than the second one. After that, the 
> remaining peaks show the usually observed gradual decrease in 
> intensity. Based on CO2 standards of known isotopic composition, the 
> isotopic values for the first three peaks show a large variability 
> (2-3 per mil) relative to its true value. The remaining four peaks 
> show isotopic values that are expected, but the standard deviation is 
> high (~0.2 per mil). It does not appear that there is any leak, we 
> have already conditioned the capillary column (PoraPlot Q, 0.32 mm 
> inside diameter), and checked the gas flow. My impression is that the 
> column is not working properly, which would explain the delay in the 
> peaks.  However, I do not know why the first three peaks are 
> problematic. I was wondering if any of you have encountered this 
> problem before and if the solution is to change the column.  If so, I 
> have read that there are other types of columns out there that are as 
> effective as the PoraPlot, but that are not as sensitive to pressure 
> changes. I was wondering if any of you have any experience with these 
> other type of capillary columns.  Thanks,
>
>  
>
> German
>
>-- 
>Germán Mora
>Assistant Professor
>Depart. of Geol. & Atmosph. Sci.
>Iowa State University
>Ames, IA 50011-3212
>
>Phone: 515-294-7586
>Fax:   515-294-6049
>E-mail: [log in to unmask]
>  
>