Back to the Top
Hello,
Did anyone observed peak at the Rt of analyte in the matrix blank or in the
reagent blank, when used gradient program and disappeared when isocratic elution
mode was used.
If yes, what could be the reasons and possibilities?
In both the cases baseline are negligible.
Regards,
Sudipto Basu
Back to the Top
This is most likely due to the mobile phase contamination. The analyte gets
stuck at the front end of the column under initial gradient conditions, then
elutes when you run the gradient. Very easy to check - just connect the pump
directly to the column (bypassing the injector), and run the gradient cycle. If
you get a peak - your mobile phase is contaminated. Note that even the smallest
contamination will give a peak, as the volume you are pumping through the column
is huge compared to the normal injection volume.
Hope this helps.
Andrew Volosov
Shire HGT
Back to the Top
Sudipto,
This sounds like what is called "carry-over", and results from a tiny amount of
the target analyte being retained on an instrument component and then injected
into the blank, (or other run).
This can be difficult to track down because not absolutely all of the instrument
component surfaces are washed during the usual cycles, especially for gradient
modes.
The usual suspects are the "needle", (which isn't an actual needle on a modern
LC), and the gaskets/seals of the injector. I mention these pieces of hardware
because many vendors offer these components made of various materials in an
attempt to avoid this issue.
Of course, various wash cycles can be tried, but may use additional run time.
Regards,
Frank
Back to the Top
The following message was posted to: PharmPK
You did not mention detection mode or whether the retention time and/or peak
changed.
Run a 'dummy' run, i.e. the same gradient without actually injecting a sample;
this will allow you to narrow down the location of the 'contamination'.
You may also want to change your guard-column and/or pre-column filters if
applicable.
HTH
Frederik Pruijn
Back to the Top
The following message was posted to: PharmPK
As mentioned by others, what is the detection mode? How is the blank
gradient looks, when you bypass injector? Generally, end of the
gradients will have higher absorption (UV) or higher counts (MRM). If
your compound eleutes on this gradient spike, you will deceptively think
there is a blank carryover. Since it is disappearing in isocratic mode,
I suspect it is a gradient spike and not carry over.
Trouble shoot in following sequence:
Clean complete system, prepare fresh mobile phases with newly opened
bottles- check response in your original condition
If problem still persists, change gradient condition, solvent quality,
brand etc
If problem still persisits, is gradient a must? IF not, settle with
isocratic
Regards,
Vinayak Nadiger
Vinayak Nadiger
Senior Scientist
Forma Therapeutics(Singapore)
11,Biopolis Way ,Helios # 08-05
Singapore 1386607
Back to the Top
Thanks Andrew and all. However I'm still wondering about mobile phase
contamination as in both the cases baseline is very negligible (< 20 cps), which
is not possible in case of contaminated mobile phase (LC-MS/MS).
Vinayak, Thanks for your advise to switch to a isocratic mode, but I want to
know the route cause of this problem.
Regards,
Sudipto Basu
Back to the Top
The following message was posted to: PharmPK
Dear Sudipto,
There are some reasons like carryover (it may be from sample injection needle or
from column), matrix component, water quality which is used in mobile phase and
needle wash.
At initial gradient there is higher proportion of aqueous which should not favor
in proper ionization and that's why we can observe low baseline, but when the
proportion of gradient increases it favors ionization and we got a peak (mostly
in fast gradient)and sometimes its RT matches with analyte RT. If it is not
carry over then the peak could separate from analyte peak or disappear by using
slow gradient (50-70 org. to 90 org.) or isocratic method instead of fast
gradient (10 org. to 90 org.) with same run time, we can also play with slope
and steps of gradient.
Best Regards,
Dipak Modi
Back to the Top
A simple way to test whether the interference/carryover is from the injector or
column is to run the gradient several times from a single injection. If you do
not "see" the peak it is probably from the injector. If you do see the peak, it
is from accumulation on the column or contamination in the mobile phase.
Back to the Top
Hi Sudipto,
It is in fact very possible that the baseline will be negligible. Consider how
much you are pumping through the column compared to normal injection volume.
Let's say the injection volume is 10 microliters, and the flow rate is 0.3
ml/min - under these conditions, you will have 150 injection volumes pumped
through the column in just five minutes. So if your contamination is at the LOD
level, you will see a peak 150 times higher than the LOD signal. One telling
sign of contaminated mobile phase - the peaks get larger the longer you wait
between injections.
There is no magic - the peak is coming from somewhere. Just try to simplify the
system as much as possible, then add componets one by one. Start with connecting
the pump directly to the column; if this gives you a peak when you run the
gradient cycle, there is no doubt your mobile phase is contaminated. If not, add
the injector and inject from an empty vial - this will tell you if simply
turning the injector generates a peak. Then add the injection of clean mobile
phase. Always do multiple injections, to make sure things are reproducible.
Somewhere along the way you will figure it out.
Andrew
Back to the Top
Hi Sudipta,
I agree with Dipak. In my experience, the interference has got a lot to do with
the gradient programme.. more importantly when the gradient switches onto a
greater proportion of organic mode, we see a hump or a peak which will be
consistent in all the blank injections (which is when most of the analytes elute
also). The best way to come out of it is to slow down the gradient/ check out
the slope options.
Hope it helps
Regards,
AB Rao
|
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Gradient versus isocratic peak" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)