-
- admin
- Site Admin
-
- Posts: 647
- Joined: Wed Aug 11, 2004 12:19 am
By Sinead Condon on Tuesday, August 3, 2004 - 08:00 am:
I'm working in a pharmaceutical QC lab we are having ongoing problems with loss of % CV between stds.This has happened with different stds on different HPLC systems.
With two weights of std injected the % CV between the 2 stds for the first sample could be fine i.e. 2.0%. This has happened on both isocratic and gradient methods. We are using agilent HP1100 and HP1050 LC systems. We have tried different vials, different caps. We always run a precision at the beginning of the run and this would be fine <2.0%
-------------------------------------------------------------------------------------------------------
By henrik on Tuesday, August 3, 2004 - 11:38 am:
change rotorseal
-------------------------------------------------------------------------------------------------------
By Sinead Condon on Wednesday, August 4, 2004 - 02:34 am:
The rotor seal has been changed and we are still getting this problem.
-------------------------------------------------------------------------------------------------------
By David Blais on Wednesday, August 4, 2004 - 05:04 am:
Sinead,
Since you say you obtain good precision at the beginning of the run, you lead me to believe that you lose precision throughout the analysis...is this correct? Does it appear the overall trend is for the standard areas to increase? Or decrease? Knowing what the trend is will make it easier to help focus the troubleshooting process.
David
-------------------------------------------------------------------------------------------------------
By Sinead Condon on Wednesday, August 4, 2004 - 05:58 am:
David,
The precision is always fine at the beginning of the run (6 injections of std weight 1) then with two samples using an injection sequence as follows:
std weight 1 x 1 injection
sample 1 weight 1 x 1 injection
sample 1 weight 2 x 1 injection
std weight 2 x 1 injection
std weight 1 x 1 injection
sample 2 weight 1 x 1 injection
sample 2 weight 2 x 1 injection
std weight 2 x 1 injection
The % CV between the std weight 1 and 2 would be fine for sample 1. But, the %CV between the std weight 1 and 2 would be failing . It does look like the precision is going throughout the run but we have put on another precision at the end and this would be fine.
We have also reinjected std weight 1 and std weight 2 after this above sequence and the %CV would be fine again.
We are having this problem intermitently over the past year with different samples and different HP1100 and HP1050 systems.
One of the std areas for the second sample would have decreased , would be 80% of the initial area obtained. But , when reinjected would be back up at 100%.
It is like a dud injection. of the std for the second sample.
Sinead
-------------------------------------------------------------------------------------------------------
By David Blais on Wednesday, August 4, 2004 - 11:28 am:
Sinead,
What concerns me most is that you meet precision at the beginning of the run, you lose it during the run, then you regain it again ("We have also reinjected std weight 1 and std weight 2 after this above sequence and the %CV would be fine again."). This leads me to think there is a problem with your autosampler.
If your standard area decreases to 80% and then goes back to 100%, I would say perhaps your HPLC vials are filled too high? If so, the act of piercing the vial and drawing solution can create a tiny vacuum, decreasing the overall amount of solution drawn by the needle. Then, after the volume has dropped the next injection is performed correctly. This is just a guess, however.
In a separate issue, I am a bit concerned that you are running on HP 1100s and are seeing precision around 2%. My experience with these instruments has yielded precision much less than 1%, usually less than 0.5%. Perhaps you simply need complete maintenance on your instrument, especially the autosampler?
These may not be the solution to your problem, but perhaps a place to begin looking. I am curious to see what other, more experience forum users suggest.
David
-------------------------------------------------------------------------------------------------------
By Leadna N Brigola on Thursday, August 5, 2004 - 07:46 pm:
Sir Sinead,
This is just a wild guess: is it possible that there is a carryover from the autosampler or from the sample that is eluted at a much later time? if you try to inject the diluent used in sple preparation after some injections to rinse the needle of the autosampler, will you obtained better results?
Leadna
-------------------------------------------------------------------------------------------------------
By Sinead Condon on Friday, August 6, 2004 - 06:54 am:
David,
We normally do get %CV's of <1.0% for precision on HP1100's but our specification is <=2.0%. The issue with filling the vials too much was considered before and all the analysts know not too fill the vials up. When we get the high %Cv the vial has been checked and it hadn't been filled to high. I appreciate your suggestion it is a good one!!
Sinead
-------------------------------------------------------------------------------------------------------
By s on Friday, August 6, 2004 - 07:00 am:
Leadna,
Normally, the area of the std peak has decreased (to approx. 80% of it's original area) to result in a high CV then on reinjection it would go back up. If there was carryover then the area would be increasing .
Thanks for the suggestion,
Sinead
-------------------------------------------------------------------------------------------------------
By Henrik on Tuesday, August 10, 2004 - 10:25 pm:
Wash your system with 10% HNO3 for half an hour, remember to remove colum an use water before an after HNO3.
Used to work!
-------------------------------------------------------------------------------------------------------
By MikeS on Wednesday, August 11, 2004 - 04:45 am:
If I remember correctly from an LC/GC article, 10% HNO3 can severely damage an HPLC system. The concentration needs to be much higher, at 6N, to avoid damage. Personally, I've never seen a HNO3 wash that accomplished anything positive, but have seen many HPLCs rendered unusable for days while people try to remove it from their system. I'm curious as to why this wash would help the CV issue under discussion
I'm working in a pharmaceutical QC lab we are having ongoing problems with loss of % CV between stds.This has happened with different stds on different HPLC systems.
With two weights of std injected the % CV between the 2 stds for the first sample could be fine i.e. 2.0%. This has happened on both isocratic and gradient methods. We are using agilent HP1100 and HP1050 LC systems. We have tried different vials, different caps. We always run a precision at the beginning of the run and this would be fine <2.0%
-------------------------------------------------------------------------------------------------------
By henrik on Tuesday, August 3, 2004 - 11:38 am:
change rotorseal
-------------------------------------------------------------------------------------------------------
By Sinead Condon on Wednesday, August 4, 2004 - 02:34 am:
The rotor seal has been changed and we are still getting this problem.
-------------------------------------------------------------------------------------------------------
By David Blais on Wednesday, August 4, 2004 - 05:04 am:
Sinead,
Since you say you obtain good precision at the beginning of the run, you lead me to believe that you lose precision throughout the analysis...is this correct? Does it appear the overall trend is for the standard areas to increase? Or decrease? Knowing what the trend is will make it easier to help focus the troubleshooting process.
David
-------------------------------------------------------------------------------------------------------
By Sinead Condon on Wednesday, August 4, 2004 - 05:58 am:
David,
The precision is always fine at the beginning of the run (6 injections of std weight 1) then with two samples using an injection sequence as follows:
std weight 1 x 1 injection
sample 1 weight 1 x 1 injection
sample 1 weight 2 x 1 injection
std weight 2 x 1 injection
std weight 1 x 1 injection
sample 2 weight 1 x 1 injection
sample 2 weight 2 x 1 injection
std weight 2 x 1 injection
The % CV between the std weight 1 and 2 would be fine for sample 1. But, the %CV between the std weight 1 and 2 would be failing . It does look like the precision is going throughout the run but we have put on another precision at the end and this would be fine.
We have also reinjected std weight 1 and std weight 2 after this above sequence and the %CV would be fine again.
We are having this problem intermitently over the past year with different samples and different HP1100 and HP1050 systems.
One of the std areas for the second sample would have decreased , would be 80% of the initial area obtained. But , when reinjected would be back up at 100%.
It is like a dud injection. of the std for the second sample.
Sinead
-------------------------------------------------------------------------------------------------------
By David Blais on Wednesday, August 4, 2004 - 11:28 am:
Sinead,
What concerns me most is that you meet precision at the beginning of the run, you lose it during the run, then you regain it again ("We have also reinjected std weight 1 and std weight 2 after this above sequence and the %CV would be fine again."). This leads me to think there is a problem with your autosampler.
If your standard area decreases to 80% and then goes back to 100%, I would say perhaps your HPLC vials are filled too high? If so, the act of piercing the vial and drawing solution can create a tiny vacuum, decreasing the overall amount of solution drawn by the needle. Then, after the volume has dropped the next injection is performed correctly. This is just a guess, however.
In a separate issue, I am a bit concerned that you are running on HP 1100s and are seeing precision around 2%. My experience with these instruments has yielded precision much less than 1%, usually less than 0.5%. Perhaps you simply need complete maintenance on your instrument, especially the autosampler?
These may not be the solution to your problem, but perhaps a place to begin looking. I am curious to see what other, more experience forum users suggest.
David
-------------------------------------------------------------------------------------------------------
By Leadna N Brigola on Thursday, August 5, 2004 - 07:46 pm:
Sir Sinead,
This is just a wild guess: is it possible that there is a carryover from the autosampler or from the sample that is eluted at a much later time? if you try to inject the diluent used in sple preparation after some injections to rinse the needle of the autosampler, will you obtained better results?
Leadna
-------------------------------------------------------------------------------------------------------
By Sinead Condon on Friday, August 6, 2004 - 06:54 am:
David,
We normally do get %CV's of <1.0% for precision on HP1100's but our specification is <=2.0%. The issue with filling the vials too much was considered before and all the analysts know not too fill the vials up. When we get the high %Cv the vial has been checked and it hadn't been filled to high. I appreciate your suggestion it is a good one!!
Sinead
-------------------------------------------------------------------------------------------------------
By s on Friday, August 6, 2004 - 07:00 am:
Leadna,
Normally, the area of the std peak has decreased (to approx. 80% of it's original area) to result in a high CV then on reinjection it would go back up. If there was carryover then the area would be increasing .
Thanks for the suggestion,
Sinead
-------------------------------------------------------------------------------------------------------
By Henrik on Tuesday, August 10, 2004 - 10:25 pm:
Wash your system with 10% HNO3 for half an hour, remember to remove colum an use water before an after HNO3.
Used to work!
-------------------------------------------------------------------------------------------------------
By MikeS on Wednesday, August 11, 2004 - 04:45 am:
If I remember correctly from an LC/GC article, 10% HNO3 can severely damage an HPLC system. The concentration needs to be much higher, at 6N, to avoid damage. Personally, I've never seen a HNO3 wash that accomplished anything positive, but have seen many HPLCs rendered unusable for days while people try to remove it from their system. I'm curious as to why this wash would help the CV issue under discussion
