Decrease in peak response

[Krishnamurthy]

We are working with a molecule which is a diethylamine adduct. As soon as the molecule comes in contact with the water it degrades i.e. cleavage of adduct will takes place. So the only option to analyze this molecule is by non aqueous method.

The compound is basic in nature and its molecular weight is about 500. The aim of the method is estimate the compound as such and to separate the impurity (which forms due to the cleavage of adduct). We have tried with different column chemistries like C8, Cyano, Diol (different make) and silica and found that Diol column (YMC /Betasil) is more suitable than other columns. The mobile phase used is Acetonitrile with 0.01% ammonium hydroxide and Isopropyl alcohol with 0.01% ammonium hydroxide in the ratio of 85:15. The detection wavelength is 254nm. The RT of analyte is about 6 min and impurity is at about 4 min. sample prepared in pure Acetonitrile.
We are facing the following problems with this method.
1. Area of the analyte peak is decreasing continuously. Some times it is working fine but many times % RSD is failing for standard injections.
2. Compound is stable in Acetonitrile, so there is no chance of the degradation. The compound is highly soluble in Acetonitrile so there are no chances for the precipitation to occur.
3. Some times the response is nil.
4. We have saturated the column by injecting the concentrated solution of the analyte by repeatedly injecting about 5-7 injections. Even then we are facing the same problem. Why the response is decreasing.5. If we inject the same sample next day, the response will come but it goes on decreasing up on injection to injection.

Please suggest some way to resolve this.

[zokitano]

We are working with a molecule which is a diethylamine adduct. As soon as the molecule comes in contact with the water it degrades i.e. cleavage of adduct will takes place. So the only option to analyze this molecule is by non aqueous method.

The compound is basic in nature and its molecular weight is about 500. The aim of the method is estimate the compound as such and to separate the impurity (which forms due to the cleavage of adduct). We have tried with different column chemistries like C8, Cyano, Diol (different make) and silica and found that Diol column (YMC /Betasil) is more suitable than other columns. The mobile phase used is Acetonitrile with 0.01% ammonium hydroxide and Isopropyl alcohol with 0.01% ammonium hydroxide in the ratio of 85:15. The detection wavelength is 254nm. The RT of analyte is about 6 min and impurity is at about 4 min. sample prepared in pure Acetonitrile.
We are facing the following problems with this method.
1. Area of the analyte peak is decreasing continuously. Some times it is working fine but many times % RSD is failing for standard injections.
2. Compound is stable in Acetonitrile, so there is no chance of the degradation. The compound is highly soluble in Acetonitrile so there are no chances for the precipitation to occur.
3. Some times the response is nil.
4. We have saturated the column by injecting the concentrated solution of the analyte by repeatedly injecting about 5-7 injections. Even then we are facing the same problem. Why the response is decreasing.5. If we inject the same sample next day, the response will come but it goes on decreasing up on injection to injection.

Please suggest some way to resolve this.

As you said about the analyte being degraded when it comes in contact with aquous medium, the observed decreasing of the analyte peak during analyses might be a process driven by the on-column or (HPLC)in-tube decomposition of your analyte when it comes in contact with the mobile phase. Eventaully if the any compound in the mobile phase or active surface speeds up the reaction of degradation, you'll probably see continuous decreasing of your analyte peak during subsequent injections.

Just a thought

Regards

[mohan_2008]

Please check if your molecule degrades by water contact. There could be many degradation sources - the temperature & light being more common.

However, if the molecule is sensitive to water contact - this is a big, big problem. In other words, even if you use NPC, there is ambient moisture that can get into the column (no matter how careful your are!) and can start accumulating on the column over a period of time.

And your molecule being water sensitive can start degrading right away as soon as it hits the column.

Please check immediately after each injection and monitor relative peak areas of the degradant & main peak. Is the degradant (assuming as the only degradant) area increasing proportionally (or making any sense!) with the decreasing/disappearing main peak. Maybe, this can give us an idea if the molecule is disintegrating while traversing down the column.

But, again please check what is the real source of degradation.

Also, method-wise:

1. Pretty aggressive combination of mobile phase. Amm.hydroxide+ IPA can start eating up the column. Acetonitrile is pretty localizing for an NPC method.

2. It could be: since you are using a highly localizing mixture of IPA + ACN, your compound may not be able to equilibrate with your column effectively. The IPA and/or ACN can compete with your molecule in sticking to the stationary phase - and may result in drifting RTs' & Peak areas.

3. And additionally, high pH mode will give you extensive silanol interactions.

4. Check if your compound is soluble in hexane. If so, use hexane as the diluent (gives the molecule some immunity from ambient moisture). Use a hexane-IPA mixture (85:15) with 0.1% TFA. I hope, this low pH combination should not affect the resolution. If so, use a mid pH range with an optimum buffer.

Try this experiment and see if this helps.

[unmgvar]

When you say that you re-inject, is it from the same vial or do you fill a new vial with the same sample preparation?

[Krishnamurthy]

When you say that you re-inject, is it from the same vial or do you fill a new vial with the same sample preparation?

It is from same vial

[unmgvar]

Krishnamurthy

when you reinjet the next day, the area comes back to the initials values and then start to decrease again? or is it lower then the initial area in the first injection always?

what is your injection volume?
do you use a non slitted septa?

[Krishnamurthy]

Yes, Area comes back to the initial values and then starts decresing.
Injection volume is 20 microlitres. We use non slitted septa

[Peter Apps]

When you inject the next day and the area comes back to its initial value, is this from the same vial that you used the day before, or a fresh vial ?

Peter

[zlb215]

"Area comes back to the initial values and then starts decresing", I'd like to give my suppose: your hplc system may be unstable, and the mostly reason, I think is that the energy of D2 lamp in the detection is insufficient, so the response decrease as operating time goes by.You can check this by injecting another sample.

[unmgvar]

i would look a little at the samples behavior like Mohan said and see if you get a compound decrease in peak area with an increase of impurity or both of them decrease.

i believe that the second one is the case, therefore your problem is maybe the autosampler settings for the injection.

change them so that the injection process is slower.
make the syringe speed slower and if you can let the needle stay inside the vial longer.
how to actually do it depends on the autosampler and software type.

[Krishnamurthy]

"Area comes back to the initial values and then starts decresing", I'd like to give my suppose: your hplc system may be unstable, and the mostly reason, I think is that the energy of D2 lamp in the detection is insufficient, so the response decrease as operating time goes by.You can check this by injecting another sample.

All my HPLC systems are brand new. They are 3 months old.

Lamp energy is OK

we have injected another sample also. But the same problem.

We can rule out the system related problems.

[Krishnamurthy]

i would look a little at the samples behavior like Mohan said and see if you get a compound decrease in peak area with an increase of impurity or both of them decrease.

i believe that the second one is the case, therefore your problem is maybe the autosampler settings for the injection.

change them so that the injection process is slower.
make the syringe speed slower and if you can let the needle stay inside the vial longer.
how to actually do it depends on the autosampler and software type.

All my systems are brand new. I am not getting this problem due to the autosampler settings.

We have not observed this problem with other molecules. (I might be working with atleast 20 molecules).

I think it is something to do with the molecule nature

[HW Mueller]

Krishnamurthy, you apparently did not answer the only relevant question, namely that of Peter Apps. Without knowing that answer one can only speculate wildly.

[unmgvar]

Carefull reading will show that he already did answer that question to my request right before that.
he injects the same solution from the same vial.

as for the instrument being new is irrelevant since the problem might be a physical issue.
are all your molecules being disolved in 100% ACN?
which instrument do you use.

it is definetly not an issue of the sample degrading in the vial, but it could be an issue of less sample going inside the instrument, maybe sampler settings.

[Peter Apps]

HI unmgvar

I'm not sure that my question had been answered, that's why I asked it. We know that injections on the same day are from the same vial, I want to be sure that the same applies to injections on different days. If injections on different days are from the same vial I would want to know what happens to the HPLC and to the sample vial overnight in terms of conditioning the column, storage temperature etc.

I cannot agree that the age of the instrument is irrelavent, since a new instrument will not have an old, weak lamp for instance.

Problems with autosamplers typically manifest themselves as eratic variability, the problem here is a progressive decline in peak area over multiple injections from the same vial. I cannot think of a way that slowing down the autosampler could solve such a problem.

Krishnamurthy is almost certainly right that the problem is due to the moisture sensitivity of his analyte - now how do we solve the problem ?

Peter