Benzene/Toluene in water: what internal standard?

[RobinDelBosco]

Good morning all

I have to measure the concentration of benzene, or of toluene, in water with a GC, FID detectors. I put benzene in water (not necessary distilled water; drinking water is a problem?) with an adsorbent and after hours I've to test the concentration of benzene. Common practice to test BTEX is use solvent and internal standard for calibration. Can I do the calibration with only benzene (or toluene) and the internal standard and when I'll have to test the water concentration, take a sample made of 50% water (+benzene) and 50% internal standard?
If someone is able to clarify me the use of the solvent or give me advices related to the procedure..thank you!
My column is capillary, non polar, suitable for aromatics and pesticides. Phase: bonded; poly(5%diphenil/95%dimethylsiloxane).

[rb6banjo]

What is your technique for capturing the the headspace of your sample? Knowing this may help us to give you better advice on how to attack your calibration problem.

Lots of times, the internal standard is only added to improve the precision of your measurement. The theory is that if there's variability in the sampling by your instrument, then on a relative basis it will be the same for the internal standard as it is for the analyte. Another is that sometimes, you can get by with a 1-point calibration for a given sample - AFTER you've demonstrated linearity of your detector response with concentration for both your analytes and your internal standard. Once you know the response factor of your analyte relative to the internal standard, if you merely add the IS to the sample at a known concentration in the linear range and analyze it, you can use the data to calculate the concentration of your unknown analyte.

If you have a very reliable analysis, you don't necessarily NEED an internal standard. I have had pretty good luck merely picking something that is well behaved from a sampling and chromatographic standpoint which does not elute with anything else in my sample. In general, my samples are very complex so sometimes it's tough to find one.

[RobinDelBosco]

Thanks,
I try to describe better. In the lab we've already methanol and toluene, so I'm thinking to use toluene as pollutant for the adsorption test and methanol for the calibration. I also have dichloromethane but I probably is too volatile to be suitable. So I want conduct the calibration using only with 1,5,10,100,200 ppm of toluene in methanol, with methanol as internal standard.
I don't want volatilisation, but I want to conduct the tests at room temperature. The adsorption test will be in vials closed (by PTFE/silicone septa?), continuously stirred and I want to full the vials with water and toluene, with no headspace. Immediately prior to the GC injection, six times in 24 hours, I want remove the cap and take out half microliter of water from the sample with a syringe, I'll full the other 50% of the syringe with the substance that I've used for the calibration (methanol) and I'll inject all in the GC.
Finally, I'd like to understand if this is or not a good procedure or if improvement are needed. And also if methanol is ok for my column.
All advices are welcome!
Thank you

[rb6banjo]

So you are trying to evaluate a sorbent for BTEX (toluene in particular). Is that correct? Is your sorbent suspended in the water or does it sit like a lump in the bottom of the vial?

Mixing your internal standard in the syringe is probably not a good idea. I fear that your precision will not be very good. Methanol is likely fine for your column but the water may lead you to some problems in your analysis (improper refocussing of the toluene at the head of the column).

The best way to do this would be to put your sorbent in the vial along with your toluene-containing sample. Close it and stir. Then, at your prescribed time intervals, collect a sample of the headspace above the water. Perhaps you could use multiple vials of the same mixture and then analyze them using static-headspace sampling? In my lab, I would use solid-phase microextraction to sample the headspace. SPME doesn't take much out of the sample so you could likely analyze the same vial multiple times. Headspace analysis gets rid of your water-in-the-chromatogram problem.

[gdmfsob]

The adsorption test will be in vials closed (by PTFE/silicone septa?), continuously stirred and I want to full the vials with water and toluene, with no headspace

It sounds like he is using liquid injection?

So I want conduct the calibration using only with 1,5,10,100,200 ppm of toluene in methanol, with methanol as internal standard.

Maybe he is saying he wants to use the solvent peak as the I.S.?

mmediately prior to the GC injection, six times in 24 hours, I want remove the cap and take out half microliter of water from the sample with a syringe, I'll full the other 50% of the syringe with the substance that I've used for the calibration (methanol) and I'll inject all in the GC.

This makes it sound like you want to dilute your sample inside the syringe you are going to inject with.

My advice is dilute your water sample (containing benzene) and your solvent (Methanol) with your internal standard..... try Heptane? and mix, than take your syringe and pull off your microliter and inject.

Hope I was helpful!

[dblux_]

Good morning all

I have to measure the concentration of benzene, or of toluene, in water with a GC, FID detectors. I put benzene in water (not necessary distilled water; drinking water is a problem?) with an adsorbent and after hours I've to test the concentration of benzene. Common practice to test BTEX is use solvent and internal standard for calibration. Can I do the calibration with only benzene (or toluene) and the internal standard and when I'll have to test the water concentration, take a sample made of 50% water (+benzene) and 50% internal standard?
If someone is able to clarify me the use of the solvent or give me advices related to the procedure..thank you!
My column is capillary, non polar, suitable for aromatics and pesticides. Phase: bonded; poly(5%diphenil/95%dimethylsiloxane).

Seems that your post is suitable to be moved to Student Project subforum.
If you want to determine benzene or toluene in water with GC/FID you need to use headspace, purge and trap or at least liquid-liquid extraction.
Internal standard as a rule should be similar in nature to your analyte, so methanol is not a good choice.

[RobinDelBosco]

rb6banjo, gdmfsob and dblux
thanks. I'm spending always hours evaluating the literature before give you questions, ask is always the final step but the only way that remains to improve my knowledge, so thank your advices.
I'd like discuss the solution evaluating your answers together.
rb6banjo and gdmfsob: yes, you understood perfectly.
dblux, liquid liquid extraction seems the simpler method. But previously, I need to undestand:

1 Why not methanol as solvent: I've already it in the lab, if it works will be great. But if not, what solvents and internal standards you recommend? In the literature I found hexane and dichloromethane. I've already dichloromethane in lab, but a document says that is not a good choice because too volatile. Your advice?
2 I've to analyze the adsorption of benzene in water and of toluene in water (different samples), so could be a good idea use benzene as int stand for toluene adsorption test and toluene as internal standard for benzene adsorption test? I found studies in which is used as int st for benzene toluene or dichlorobenzene. What others int st could be suitable?
3 the method suggested bj gdmfsob is the simplest, because it requires calibration with int stand and solvent, but avoids liquid liquid extraction. Why don't use it?

[dblux_]

...
1 Why not methanol as solvent: I've already it in the lab, if it works will be great. But if not, what solvents and internal standards you recommend? In the literature I found hexane and dichloromethane. I've already dichloromethane in lab, but a document says that is not a good choice because too volatile. Your advice?
2 I've to analyze the adsorption of benzene in water and of toluene in water (different samples), so could be a good idea use benzene as int stand for toluene adsorption test and toluene as internal standard for benzene adsorption test? I found studies in which is used as int st for benzene toluene or dichlorobenzene. What others int st could be suitable?
3 the method suggested bj gdmfsob is the simplest, because it requires calibration with int stand and solvent, but avoids liquid liquid extraction. Why don't use it?

If liquid-liquid extraction has to be used then solvent shouldn't dissolve in water. Commonly used: n-pentane or hexane. You won't probably get properly shaped methanol peak on your DB-5 column so it's not a good choice for internal standard. Fluorobenzene is a good choice.
But you may use methanol as a solvent for benzene and toluene to prepare standard solutions in water.

[rb6banjo]

What is the starting concentration of toluene in your solution? This will give you an indicator as to whether "dilute-and-shoot" will work for you. If your starting concentration is too small, you could dilute yourself right out of being able to detect the toluene at all. Also, I've never had much success reliably diluting aqueous solutions of insoluble things like toluene. The precision is usually terrible. A solvent-solvent extract of the water with methylene chloride (or heptane) would perhaps be better. That also depends on your starting concentration of toluene in the water.

[RobinDelBosco]

Dear all,
thanks. I'd like to be useful for you one day (I'm an environmental engineer without experience, but..who knows)
Just to be sure, my standard solutions = solutions that I use only for calibration, because in the adsorption test I use only water and benzene or toluene in form of mixture-emulsion. I try to describe the procedure, hoping that you can correct errors and give me advice.
I can prepare the solutions for the calibration for toluene with solvent, toluene and benzene as int standard and with solvent, benzene and toluene as internal standard for benzene.
Without water, correct? And I can try to use these internal standards (benzene for toluene and toluene for benzene), or I've to use fluorobenzene or dichlorobenzene?
About the solvent in the calibration solutions, it will be methanol if I don't do liquid liquid extraction and I follow the procedure initially suggested by gdmfmsob; it will be the same solvent (heptane or hexane) that I'll use in liquid liquid extraction if I'll do extraction. So, we've to evaluate if liquid liquid extraction is required or not. Your advice?
The target is testing different concentration in water samples from 400 or 200 to 1 ppm.

[rb6banjo]

From an analytical standpoint, that is a lot of toluene. I don't think you'll have trouble extracting it from the water with some type of solvent and still seeing it in your analysis. I do think you will have a very difficult time diluting the water/toluene mixture in a manner that is reproducible. Those concentrations are very large for a typical headspace analysis. Extraction followed by direct injection of the extract is probably your best approach.

[RobinDelBosco]

Thanks.
extraction is the best approach, but the main doubt remains: can I avoid it? Extract means stop stir the sample and do the extraction 6 times in 24 hours, with volatilisation during the procedure and other issues. Why not this faster procedure, (gdmfsob suggestion) with less volatilisation:
prepare a solution of solvent and internal standard and solutions of solvent and benzene, in different concentrations. Do the calibration. Start the adsorption test and every two hours extract with a sirynge 0.3 microliter from the sample of water and benzene, fill the others 0.7 microliter of the syringe with the solution of internal standard and solvent and do the injection (I'm quite sure that that quantity of water is not a problem for the column).

[rb6banjo]

You can certainly try it but I fear that your precision will not be good. Why can you take a sample out of the stirring solution (with a glass pasteur pipette) then transfer it to a small vial for extraction? methylene chloride will sink to the bottom, you can reach down to the bottom with your syringe and pull your samlpe. Dissolve your internal standard in the methylene chloride. Just choose one that doesn't transfer to the water to any great extent.

[Peter Apps]

Thanks.
extraction is the best approach, but the main doubt remains: can I avoid it? Extract means stop stir the sample and do the extraction 6 times in 24 hours, with volatilisation during the procedure and other issues. Why not this faster procedure, (gdmfsob suggestion) with less volatilisation:
prepare a solution of solvent and internal standard and solutions of solvent and benzene, in different concentrations. Do the calibration. Start the adsorption test and every two hours extract with a sirynge 0.3 microliter from the sample of water and benzene, fill the others 0.7 microliter of the syringe with the solution of internal standard and solvent and do the injection (I'm quite sure that that quantity of water is not a problem for the column).

The repeatability with which you can fill 0.3 ul and then 0.7 ul into the syringe will be the limiting factor on the repeatability and accuracy of your analysis. I doubt that you will get it any better than an rsd of 30%. You do not have to take my word for it. You can test the technique by spiking some water with benzene or toluene (with none of the adsorbent you plan to use) in a septum-topped vial with 2 - 5 ml of headspace, than take 0.3 ul from the liquid phase followed by 0.7 ul of the standard (if your sample is benzene use toluene standard, and if you sample is toluene use benzene standard). Do six replicates, calculate the sample to standard peak area ratio for each, and then calculate the rsd of the ratios over the six replicates.

Peter

[RobinDelBosco]

Thanks for all the useful advices.
Finally I'll edsorption experiment with benzene. I'll use:
hexane as solvent and ethylbenzene or toluene as internal standard
or
dichloromethane as solvent and dichlorobenzene as int stand.
Extraction: I'll centrifuge water and benzene+solvent at 600 rpm for 15 minutes, I'll put the vial at 4 degrees and when the vial reach the 4 degrees I take out 0.5 microliters of solvent+benzene with a syringe, I fill the other 0.5 microliters of the syringe with solvent + internal standard and I inject in the GC.This procedure is to reduce volatilisation and to know the partition of benzene between water and solvent.
Is it a good way? How know the partition?
The GC column is capillary, non polar, suitable for aromatics and pesticides. Phase: bonded; poly(5%diphenil/95%dimethylsiloxane). I'd like to know if the substances mentioned above are dangerous for the column and how set this program in the best way:

Hydrogen 30 ml/min (fixed)
Air 300 ml/min (fixed)
Carrier gas: He 1 mL/min
Detector temperature: 180°C or 250?
Injector temperature: 180°C 250 or 250?
Oven temperature program: I know three program used with other column to determinate BTX, but I need to identify only benzene. What is the best program?
35°C for 5min , to 70 °C at 5 C/min, to 180°C at 15C/min and hold 10 minutes at 180°C
50°C holding time 1 min, to 70°C at 5/min, to 200 at 20 °C/min and holding 200°C for 20 min
40°C for 5 min, to 48°C at 2 C/min and to 120 at 15 °C/min (no holding time at 120°C)
Injection: splitless.
Thanks!

Andrea