Help choosing an internalstandard

[Banzo]

Hello,

I'm in need of a bit of help choosing an internal standard to use with a GC-MS.
I'm developing a method to analyse a lot of different drugs. It is impossible to get separate internal standards for each drug because of the high costs this will bring.
It's also impossible to get one internal standard which behaves almost the same as all the drug I want to analyse.
So it should just be an internal standard with a high purity and it should not be used in the drugs (as an cutting agent or drug itself).
It does not have to be an deuterated internal standard because it will not react the same as all the drug anyway.
Also it should not react with bstfa:tcms but should be soluble in it.
Can you make me a suggestion(s) on which internal standard to use? Preferably not an to expensive one.

Best regards,
Banzo

[Peter Apps]

With these criteria the only analytical variation that the internal standard will correct for is injection volume, which is very likely to be a minor contributor to the uncertainty in your results. You can use anything stable that elutes more or less in the middle of the chromatogram and does not overlap with any other peaks. Check out chromatograms on manufacturers' websites to get an idea of what has the right retention time.

Peter

[Banzo]

What would you suggest to correct for more analytically variation?
I would like to get as little variation as possible.
The only thing I'll be doing as preparation is dissolving a weighed amount of sample in a liquid and adding an IS before measuring with the GC-MS.
A few substances need to be derivatized. Which is done by adding BSTFA:TCMS, heating and measuring with the GC-MS.
Also can you show me where you can find these retention times because I can't find them.

Banzo

[Peter Apps]

What would you suggest to correct for more analytically variation?
I would like to get as little variation as possible.
The only thing I'll be doing as preparation is dissolving a weighed amount of sample in a liquid and adding an IS before measuring with the GC-MS.
A few substances need to be derivatized. Which is done by adding BSTFA:TCMS, heating and measuring with the GC-MS.
Also can you show me where you can find these retention times because I can't find them.

Banzo

For an IS to compensate for more variation in the sample preparation it has to match the properties of the analyte - which you have already excluded.

With a process as simple as your I would bother with an internal standard.

To find retention times go to a column manufacturers website, choose the type of column that you will be using and look up the applications that have been run on that column. This will bring up examples of chromatograms. NB that the conditions probably will not match yours so you will have to make some adjustments.

Drugs are a pretty well researched area, are you sure that you need to develop a new method rather than using an existing one ?

Peter

[Banzo]

I thought an Internal standard should always be used to have a constant factor to which you can quantify the other compounds.

I know this area is well researched therefore I will be building a method on the basis of an existing method.
I need a screening method to identify several drugs and then I'll use the corresponding method to quantify them (if it can't be done in the screening method).

The method I'm using is from an article. Are there other ways to obtain methods rather then manually putting them in from articles?

I´m new so I´m not yet completely familiar with GC-MS methods etc. But I know how to set the machine and maintain it.

Banzo

[Don_Hilton]

You can include an internal standard - or as long as your final dilution of the sample is to a contant volume, you can assume the injection volume to be that set for the autosampler and you depend on the accuracy of weighing, diluting, and transferring for your accuracy of result.

For an internal standard. It looks like you have a collection of methods for various cmopunds, should you find them in your screening technique. Any internal standards in the methods? You can even make a mixture that contains multiple internal standards and apply the appropriate internal standard to each analyte.

Do you have a specific list of compounds for which you will be screening? If so, and if you can share it, perhaps some of us can offer some specific reccomendations.

[Banzo]

The final volume is a constant volume (1ml or 10ml) it will either be done by pipetting 1ml of liquid to the solids or by dissolving the solids and then adjusting the volume to 10ml.
I think the second option is probably the best option since the volume will be the same each time. But the accuracy of the pipet is more accurate then the glassware used to adjust it to 10ml. Which one would you choose?

In the method I use for the screening method I use a method wherein a lot of different drugs and related substances are analyzed (200+). There is an internal standard used in the screening method but since it is only 1 standard it can't correct for all the substances. Also the standard used is pretty expensive and since we would be using it a lot it would cost us a fair amount of money per analysis. So I thought about using another internal substance which is less expensive.

The current list is:
Cocaine,
Methamphetamine,
Amphetamine,
MDMA
and LSD

But in the future we would like to expand the list when needed with more exotic substances such as Jwh's (synthetic cannabinoids), 2C's (phenetylamines) etc.
Do you have some recommendations?

[Don_Hilton]

My preference for putting samples into solution is dilute to the mark on a volumetric flask. With class A glassware, you get pretty accurate. Dissolve then dilute. The volume of solution is sometimes a bit surprising.

And I assume that this final volume is with the MSTFA added?

Starting the thinking process - you have some amines, phenols, and other large groups of compounds. My first thoughts are some kind of large phenolic compound (yes it has an active hydrogen) with a fluorine on it somewhere that would not be related to one of the target compounds. An amine marked in a similar manner, and so on. And put in an ether similar to the large phenolic compound. Picking several classes, you have an internal standard for each. And the phenolic compound becomes a reference point to check derivitization of your standards (and by inference your analytes). From a brief look at your list, you should be pretty well able to cover it with a few compounds. But, as I write, I realize a weakness that needs to be considered.

Keep in mind that once you pick out a compound in the unknown sample using this screen, you have to develop a calibration for that compound - and you need to demonstrate that you can quantify it in the matrix that you have without losses due to "stuff" in the mix. This may be best performed as a standard addition experiment - and the internal standard is easily selected as you start spiking with a standard mix, making the standard additions. Also, you will need to run the samples and something related to the calibration at the same time - because instrument sensitivities change over time as a function of the age of inlet liners, age of electron multipliers, source fouling, etc. And with derivitization, what you do with a sample one day may give one response, but doing what should be the same thing the next may give another. (Reactions may not run to completion when expected.)

[Banzo]

For accuracy I'll use a 10ml volumetric flask because the error of the 1 and 10ml flasks is the same (+-0,025ml).

For LSD it gets a bit tricky because a sample of LSD doesn't contain a lot of LSD.
I'm not sure how I would do this. We have BSTFA:TCMS for the derivatisation and I've tested the derivatisation and it works.
But I did the reaction in a 250ul vail insert with 100ul of BSTFA:TCMS (without IS) and I only got a small peak but was not to hard to find though.
But if I would put it in a 10ml flask it would be diluted by a factor 100 and the peak would be very hard to find.
What do you suggest regarding this?

I forgot to mention Heroin as a substance we want to analyse.
Can you give me an example of a substance you might think to be suitable as IS?

I'm aware of the proccess needed to add another compound to the list but there isn't any other option I guess.
I need to make a calibration line and need to confirm I can quantify the substance in the matrix.

I know instrument sensitivity changes over time but how can I compensate for this? I assume the area of the IS will change in the same manner as the analyte due to this?
The only thing is that the LOD (and LOQ) will get higher.

Regarding the derivatisation you can track the progress of derivatisation with an IS right? Or will one compound be derivatizated faster than the other? How is it that these derivatizations will not always work as you would expect?

[Don_Hilton]

I'll start at the bottom. Most compounds derivatize fairly quickly. But some derivatize more quickly than others. And derivatized compounds can rearrange over time - sugars, with their many hydroxyl groups and steric issues show this. What you have listed so far should not be a problem. But if the drug is cut with something, will the sample change over time? If so is that controlled for in the first derivitization you do with a sample?

Sensitivity changes over time: depending on the history of the liner in the GC and activity on the head of the GC column, the instrument may discriminate against various classes of compounds. This can change as column activity changes or as an inlet liner becomes dirty. With this the sensitivity of the instrument for one compound relative to another can change. With pure compounds, this is less likely to be a problem. But depending on what is used to cut drugs, you have the potential for the presence of compounds that would foul the GC inlet. Depending on what is going on with the instrument the LOD may go lower or higher. Often LODs become worse over time as the GC inlet becomes dirty. But, there are situations where a compound will show poorly for the first few injections and the show much better on the instrument as active sites become filled.

The problem you are trying to solve is not an uncommon one - and it is one faced by many law enforcement laboratories around the world. And, as I consider what I would do with the sample that I imagine that you would have, it crosses my mind that I would want to talk with someone who has experience doing forensic analysis - particulalry with GC methods and the types of compounds you want to look at. And, where you need to go with the accuracy of your answer depends on what you are going to do with it. If you need to know what a patient consumed to begin treatment, that is one measurement. If you need to be able to show that a suspect had at least a certain mass of controlled substances – with a number that will stand up to a defense attorney, that may be another. (And both do have to be good numbers.) The best source for internal standards would probably be out of the forensic literature, but if a I get a few minutes while I am work today, I’ll pick up a catalog and see if I spot some good ideas.

[Banzo]

Do you know how I can come in contact with a forensic analyst with experiance with GC (MS)?
Perhaps you know someone in this area?

In order to pevent a lot of fouling of the GC inlet I think it would be wise to filtrate the sample before adding it to an autosampler vail and injecting it. Since pills almost always have some insoluble matter in them like fillers etc. This way I can prevent most of the fouling (due to insoluble matter) I think. Regular liner changes will also be done but I can't compensate for all of the factors with an IS which is not 100% identical as the analyte right? But it would take a lot of time and money to make everything 100%. Since I should do an analysis figure out what the substance is select the according method and run it again with the corresponding IS.
Besides the analysis we do don't have to stand strong in court nor does it have lives depending on it. We just need to know the composition (quantitative and qualitative) of pills and powders etc. preferably as accurate as possible without spending too much time and money.
Good to know i'm not the only one

[Don_Hilton]

For contacting someone with forensic background - try your local chemical society meeting and ask around. I don't know where you are in the US, but in much of the country there are active American Chemical Society meetings - which are a good place to make contacts. Some areas also have chromatography discussion groups or other interst groups which draw chemists. And the trick ot making and using such contacts is to find out who knows someone who might know who to talk to. (Note that you are looking for a reference to someone two places beyond the person you are talking to. People may not know the person who you want to talk to, but often, when pressed, can think of someone who might give you good guidance.)

In thinking about your analysis, choice of solvennt becomes important - as you may have a compound that is "happier" sticking to the insoluable filler than going into solution. Would it make sense to mix the unknown with a minimum quantity of solvent (like DCM), inject a microlter - and if there is nothing, remove the solvent and go to MTBE and so on until chromatographic peaks show up. Then with an identified peak take a second portion of the sample weigh it, extract with an appropriate solvent and quantify? I don't know that I would filter. If anything, cetrifuge and take the sample to inject from the supernatant. I can envision some compounds partitioning into carbohydrate materials (like corn starch) until the solvent becomes farily polar. Or is there a single solvent that you already know works well for this kind of applicatoin.

[Banzo]

Most substances I've tested to well in Methanol since it's an polair and organic solvent. But there are some cases where a substance will not dissolve like verry apolair (or big) substances. Then I'll change the solvent to tolueen (by evaporation). And if that doesn't get me any peaks I'll use BSTFA to derivatisize the substances and I should get a peak with at least one of the solvents. Then I'll prep again using the correct solvent and quanitfy. This isn't a real problem yet because the substances I want to quantify at fist all dissolve readily in methanol except LSD which is quite easy to distinguish because most of the times it comes as blotters.

I'm not in the US I'm from the Netherlands. I'll be reaching out to my former school because I think hey have good contact or might help me further themselves.