Saturation may be occuring for several reasons. First, you may have
manually set the integration times to too long. If you are using
auto-integration, either the minimum read time is too long (exceeded the
maximum number of integration replicates allowed), or the sample does not
reach the plasma before the pre-shot takes place. At 14M counts, the
maximum integration time is 65K/14M = 0.0046s. My guess is that you are
set to 0.005s, which is why you are getting intermittant saturations.
The discrepency in the counts per second in the old vs reconstructed
methods could arise from different selected slit widths (resolutions),
plasma temperatures, sample flow rates, and gas flow rates. Before
chasing this part down, make sure you are comparing actual counts, not
counts adjusted against an internal standard. If you can reprocess
spectra before and after into the same dataset, you can easily compare
resolutions (use the scaling features in the display window.)
Best of luck,
Analytical Lab Manager
APL Engineered Materials, Inc.
Urbana, IL 61802
[log in to unmask]
"PLASMACHEM-L: Analytical Chem.(ICP's, DCP's, MIP's)."
<[log in to unmask]> wrote on 01/13/2010 01:02:45 PM:
> Greetings Plasma Chemist,
> I am having an event where my Li wavelength(670.784)is reported
> as saturated in my calibrant. This "error" does not happen everytime
> I calibrate the PE 3000 DV, but it does happen the majority of time,
> about eight out of ten. When Li is successfully calibrated it
> generates approximately 14,000,000 counts. This intensity is higher
> than any other element in the calibration soup. I looked at data
> that was a year old and the counts for LI were 5,000,000 counts. We
> checked the concentration of our Li calibrant on the MS and it was
> within the acceptable range for 1.0 ppm. Our instrument is fifteen
> years old. This started happening after I had to rebuild the method
> after loosing files, so the odds of operator error are high. I have
> looked at the delay times in our documentation and they are the same
> as before the data loss.
> Thankyou in advance for your time and advice.
> Bob Nelles
> Chemist technician
> 1000 NE Circle Blvd.
> Corvallis, Or 97330