More purity testing

[jdlh199]

I'm working on a method for purity testing of Tadalafil. This is a simple RP-HPLC/UV method using an acetonitrile:phosphate isocratic eluant. Calibration of the method is from 20-200 ug/mL. Samples are prepped to 100 ug/mL.

Here's the problem: The client is looking to confirm his product at >98.5% purity. My reference standard has a listed purity of 99.8% on the CofA. However the clients samples are consistently coming out at ~105%.

There is definitely no carryover, the first wash after the high standard gives a back calc of 0.8 ug/mL. Replicate QC samples prepared with the same reference standard as the calibration samples produce perfect data (99-101% recovery). And calibration is perfect with RSQ 0.9999 and no calibrator back calculated conc >5% error.

The clients data has been confirmed through 3 separate analytical runs. Ten duplicate analysese per run gave a result of 105.2% with an %RSD of 1.2%.

To me this simply looks like the clients sample has a higher purity than the reference standard! Do I go back to the supplier and complain and get a new lot, or just give them the data at 105%?

Tadalafil reference standards are not cheap (~$1000 USD/g) and the supplier we are using is the only one we have found that does not have a 1g minimum order, does not custom make it when requested and therefore does not have a 4-6 week delivery estimate!!

Any suggestions?

[juddc]

Might you have a moisture issue with your standard? Try a loss on drying for both your standard and sample if you have sufficient material. Nobody opened the vial straight from the freezer, I hope.

How old is it?

Storage?

How is the manufacturer of your standard determining purity?

In this case, I might pop the $1K for a gm and go get an alternate standard.

I wouldn't report the 105% (officially) without some further explanation.

What's typical USP range for this (if there is one)?

[jdlh199]

We only ordered 250 mg of reference standard, so doing a loss on drying on it will be a little tricky.

However, it's brand new, only a week old and a stock solution (or two or three!) was prepared within the first few days of receiving it! % moisture is listed on the CofA as 0.8%

I don't know how the manufacturer determines purity, I'm waiting to hear back on that one! But when I spoke to them they weren't really too forthcoming!

I would spend an extra $1k on a gram if it wasn't for the 4-6 week delivery time. The client is pretty much waiting for the data. eeek!

I think the USP range is the >98.5% that our client is looking for, but I'm not sure on that!

[juddc]

Re: Purity, I hope they weren't using the old "% of total area" gag...

Can you play with the method (change separation conditions and/or column) & look for other peaks?

Can you get a rough impurity profile? Inject your sample and standard stock solutions and look for a similar pattern of impurities between the two. If you see more junk in your standard stock than in your sample stock, then you might be getting somewhere (assuming your diluents are the same lot, of course).

Are you using a photodiode array? If so, how does peak purity compare?

[Bryan Evans]

Was just going to include 3 possibilities (but looks like 2 were already mentioned):

1. % moisture of reference std
2. Peak purity - evaluate w/ PDA
3. static electricity on balance - how was the humidity in the balance room that day? Was the std weighed with metal spatula or static free spatula?
Was the analyst wearing latex gloves?

[jdlh199]

That's what I think as well. If they just did % of total area, they would never get a >100% number and as long as impurities were minimal would always get good data! Maybe this is why they were pretty apprehensive about revealing too much on the phone!?

The machine it's running on is a tunable but fixed wavelength UV detector. I'm detecting at 280nm for the method. We do have a DAD, but it is taken up on another machine (an LC/MS) which is pretty booked solid for the next few months!

[Dan]

Considerations:

1. Did you correct your standard weights for the purity and 0.8% water from the CofA? (Check to be sure that the labeled purity isn't already corrected for the water content, otherwise you'll be correcting for it twice.)

2. You mentioned stock standard solutions. Are you storing these solutions and are these solutions stable?

3. As someone noted in a recent discussion thread (I can't remember which one; I'll look), is this a matter of "Purity" versus "Assay"? The value of 105% is an Assay value (although it is high). The Purity for a drug substance is often (if not usually) determined by starting with 100% and subtracting out the impurity and solvent results.
i.e. Purity = 100% - total impurities - water content - residual solvents

Where the total impurity result can be obtained by area normalization.

4. Check for those impurity profiles as juddc noted. You can get your total impurities result from those chromatograms.

5. Does the client's sample have a lot of moisture and/or residual solvents?

Regards,
Dan

[JM]

"Purity" of a standard has no use in quantitative analysis. It could be only a chromatographic purity with lots of gunk in it. In addition to what has already been suggested,

I would suggest to run in replicate both standard and sample of same concentration and assuming your "Sample " is 100 % ( you can deduct water + total solvent + residue on ignition from 100 and use that value as your sample assay on as is basis) back calculate the "Assay of your so called pure "standard" using your sample as reference standard.

you got the Assay of your "standard" ( it could be 94-95 % on as is basis). time to talk to your standard supplier.

I suggested this cos you must be having enough quantity of sample to do moisture etc. If your assay method is also selective for impurities you can use chromatographic purity in place of 100 % in calculation.

also see if your standard /samples are salts? if so, free salt concentration difference may be the culprit.

do let us know if this helps!

JM

[DR]

Pretty well boils down to

1) standard is junk (very likely if vendor produces lab scale batches and your material is from larger scale production, new synthesis or purification routes etc.)

2) method used to assay the standard is junk (compared to your method).

as said above - share some of your info w/ the standard vendor and see what they say.
I'm thinking that you have a lot of the material in house... you could just send a sample of it out for further characterization, then subdivide a few grams into little bottles for cold storage and use it as a standard (have it retested periodically). The only way this would not work out well is if what you bought was from the USP (but they're generally more cooperative than what you report).

PS - options 1 and 2 are not mutually exclusive.