quantification of residual solvents in botanical extracts

[mallory]

I'm trying to teach myself how to measure residual solvents in botanical extracts, but am unsure about a few things when it comes to the calculations and methods. Any input and corrections would be greatly appreciated.

I'm using a shimadzu GC-14
restek rt Q Plot 30 m column.
detector 220
inj 200
initial temp 45
initial time 2
prog rate 10
final temp 160
final time 6.5
total time 20


I have two different type of standards that I purchased. The first is a gas cylinder that contains 1000 ppm each of butane, ethane, hexane, methane,pentane, & propane.

The second standard contains 100 ppm of acetone, acetonitrile, etoh, 2 propanol, meoh, propane, butane, isobutane, hexane, and heptane in DMSO.

The samples that I will be testing will be herbal extracts. I will weigh out ~50mg into a 40ml voa, heat the voa in a heating block, and using a gas tight syringe I will pull up 500ul of headspace and manually inject into the gc.

For my curve if I use the gas cylinder standard I have been with drawing 500ul directly into a syringe and injecting that directly into the gc.

Using peak simple I set the area of each peak to equal 1000ppm

To calculate the result of a sample that has a butane peak I am doing the following calculation





Does this seem correct?


My other question is if I use the standard that is dissolved in dmso instead of the gas cylinder, would I inject it in the liquid state, or would I treat it like a sample and put it in a headspace voa, heat it, and then inject it. If I were to put it into a voa, how much would I put into the voa? I believe that I would want the injection amount to stay the same as with a sample, but im not sure what the ppm of standard would be once placed in the voa.

Am I making any sense at all?

[rb6banjo]

This is tricky business. You're mixing your units and your sampling modes. When you calibrate with a gas standard, analysis of the gas above a condensed-phase sample in the vial will only tell you what the concentration of your analytes are in the gas above that condensed-phase sample. Not how much is in the sample. If you want to know how much is in the sample, you need to add the analytes to the sample, allow them to partition out of that sample at your sample-incubation temperature, then analyze the headspace above the sample.

To use a gas standard in this mode to determine the analytes in the sample, you need to make sure that you get exhaustive extraction of the analytes from the sample. You might take a look at literature that describes a technique called "multiple-headspace extraction". That can help you get an estimate of how much of your analytes are in the sample.

Are your extracts that you're putting in your vials liquids? You can perhaps use your liquid standard and the "method of standard addition" to determine your response factors. If they're homogeneous liquids, you will likely have good success. If not, it might be a bit more problematic (solids in the sample can act as sorbents and you may not get good recoveries of your added analytes).

You just need to be very careful how you interpret the data obtained by merely analyzing the gas above a condensed phase sample when you're calibrating with a gas-phase standard.

Additionally, it is very important to understand that gas-phase "ppm" is not the same as condensed phase "ppm". Gas phase ppm comes from the ideal gas law. You can think of it as µL analyte/L of ideal gas or µmole analyte/mole ideal gas. The ratios are the same because of the gas law. Personally, I like µmole/mole because I can use that to convert from that to a mass. Condensed phase "ppm" is relating a mass of analyte to a mass of condensed phase (µg/g). Many of us use µg/mL as ppm but it's strictly true when 1 mL of the solvent = 1 g of the solvent (e.g., water).

[mallory]

The extracts are a liquid and when incubated essentially go completely into the gas phase. The solution is etoh based, and Im wanting to verify that it does not contain any of the solvents that are in my liquid standard.

Would standard addition of the smso standard be the best way to test the unknown sample, or could I put the liquid sample into a voa and incubate it and inject the headspace in the same manner as the unknown sample. If so would I use the same calculation as the one above to determine the concentration of each solvent in the standard based on the weight of standard added to the voa?

[rb6banjo]

Dissolve some of you DMSO standard in clean ethanol and run it through your proposed sampling procedure. If the amount you recover compares favorably to the amount in the gas standard, then you know that you are completely vaporizing the sample in your preparation.

If that checks out, and if you treat your samples the same way and compare that to your gas-phase standard, you should be able to relate your gas-phase ppm (µmole/mole) to a condensed-phase ppm (µg/g) if you know the mass of sample to start and the molecular weight of the analyte in question.

Take butane for an example. Say you measure 450 ppm butane in the gas phase and you started with 62 mg of sample (0.5 mL injection at atmospheric pressure and 20 °C lab temperature, sample is in a 20 mL headspace vial).

PV/RT = (1 atm)(0.02 L)/(0.0821)/(273.15 + 20) = 8.3 x 10^-4 mole

450 µmole butane/mole gas x 8.3 x 10^-4 mole gas x 58 µg Butane/µmole x 1/0.062 g = 349 µg/g

The key will be to demonstrate that indeed you are vaporizing all of the sample in the headspace vial.