Chromatography Forum

Hi,

There are two analysts in my work that are disagreeing on how to process HPLC data. I'm not sure with which way to go.

We are analysing an impurity in a solution of various components - acids, methanol etc. Our goal is to determine if the solution contains <1,000ppm of impurity and if so, it's good to be used.

We pipette 0.1ml into a 10ml volumetric flask, record the actual weight of the sample and make up to the mark with ultrapure water and then obtain a ppm result from the HPLC.

Parameters:
Dilution factor: 100
Example sample weight (0.1ml sample): 0.1016g
Example Machine ppm result: 2.41

One of our analysts is saying that all we should do is multiply the PPM result by the 100 dilution factor and report this result as ppm:
- [machine ppm (2.41) * dilution factor (100)] = 241ppm

The other analyst is saying we should account for weight of test sample and apply a correction to bring it per constant weight (0.1g) (?) and report as mg/kg:
- [Correcting to 0.1g (0.1) / sample weight(g) (0.1016) ] = 0.9843
- [0.9843 * machine ppm (2.41)] = 2.3721ppm
- [2.371ppm * dilution factor (100)] = 237 mg/kg

What is the difference between ppm and mg/kg? Can anybody offer any input into clarifying the calculation or how they would go about this?

Thank you.

Since PPM is Parts Per Million you have to know what the one millions parts are. Is it 1 million grams or 1 million milliliters?

Since you are weighing your starting sample then you should report to mg/kg if what you weighed is the product as you will release it. If your product would normally be diluted 1:100 into water then you would report mg/L (ppm) of the final solution.

It seems you are doing the dilution not to produce the final product but to analyze the raw product, so you would want to report it per weight. On the other hand, do you sell the product by weight or by volume? If you sell by volume and you report as ppm then you would want to calculate the impurity as mg/L from the dilution of 0.1ml into 10ml water. This would give the same result as dilution by volume you listed but reported as 241mg/L.

Either way is correct as long as you use the proper units to show if it is per weight or per volume. PPM alone does not give enough information unless you know how it was measured.

In your hplc, are you quantifying using a calibration curve? I assume so. If true, the decision has been taken by whoever made the calibration standards.

The HPLC does not know anything about units. All it's doing is saying that the sample you injected corresponds to 2.41 on the "amount" axis of the calibration curve. It doesn't even know whether you told it amount (mg) or concentration (mg/mL). It certainly doesn't know whether it's ppm by weight or volume. You could set the method up in potatoes per gallon if you want...

You diluted the sample by volume. 1uL of the stuff that came out of the volumetric flask contains 1/100th of what went in. Therefore your first analyst is reporting in the same units as the person who made the calibration curve (they are basically saying: "My method injects 5uL. But That was too much, so I wanted to inject 5/100uL. My autosampler couldn't inject that little, so I diluted the sample so it could inject 5uL but only introduce 5/100uL of the original sample").

Your second analyst is making an additional correction that may or may not be appropriate depending on whether the customer's definition of ppm differs from the definition used by the analyst who prepared the standard for the calibration curve. I personally dislike ppm because it's open to multiple interpretations.

Thank you both for taking the time to reply. It was very helpful.