bad recovery!

[Nour]

I have the follwing situation:

1- Fentanyl citrate injection solution (containing NaCl and pH 4.7)
2- Standard: fentanyl citrate prepared in water
Result of the Assay by HPLC: 108%

when standard is prepared in NaCl solution at pH 4.7 instead of water the result of the assay by HPLC is 100%

The same phenomenon was observed for a perfusion solution of Fluconazole:

1- Fluconazole perfusion solution diluted in mobile phase 1:1
2- standard prepared in mobile phase
Pesult of the Assay by HPLC: 106%

a lab formulation was prepared and we obtained the same result 106%
means that if the lab formulation is used as standard the assy will be 100%

shall we prepare the standard by adding NaCl at the same concentration of the sample? or make lab formulation to control the unknown sample?

we changed the column
we changed the whole method (using another mobile phase etc..)
the result are the same

we get the methods from some scientific papers and the authors attest that the methods are validated...

i didn't give the detailed chromatographic conditions used because in both cases we tried different conditions.

has anyone an idea why this occurs? or has someone experienced such phenomenon?

thank you for your help!

[tom jupille]

"Matrix effects" are not at all uncommon and can result in subtle differences in peak shape that can throw off your integration parameters (thus the anomalous recovery) A good rule is that calibrators should be prepared to be as similar to samples as possible.

[Nour]

something else i observed while validating the method for impurities. I injected the sample as is (solution containing fentanyl citrate, NaCl pH 4.7) and the 1% dilution of the sample (1ml was diluted to 100 ml using water)
Result: the peak area of the 1% solution was 20% less than expected i obtained the expected area only when the dilution was made with a NaCl solution at pH 4.7

[Nour]

oops thank you, Tom for the fast reply and the helpful suggestions.

[Uwe Neue]

Is it possible that your analyte sticks to the wall of the vial, and different conditions prevent this from happening?

[Stryder08]

Fentanyl likes to stick to plastic, very much so.

[mohan_2008]

Tom Jupille has made the point.

I had a similar compound behave exactly in the same way.

1.The matrix effect (in this case the high ionic strength, NaCl solution) will give a better Peak shape and hence, better recovery than the pure water.

2. In my case, it even remarkably increased the solubility of the solute in the NaCl solution than in Water (I had to sonicate in water).

What I did is, make all my standards & sample solutions in the buffer at pH 4.0 rather than in pure water.

Result, no problems, considering the length of the study.

[Bruce Hamilton]

As noted above, use what works, as is also best practice. . I'm a little surprised that both compounds would give similar issues.

However, if you're seeking another explanation, you might like to consider the polymorphic ability of both compounds, as well as where your detector is reading on the peak.

Please keep having fun,

Bruce Hamilton

[Nour]

Indeed, I'd agree with Tom and mohan.
in the second case of fluconazole, the dilutions were made in mobile phase (buffer and ACN) the only difference between sample and standard is the NaCl (the sample is madeup by pipetting 25 ml of the perfusion solution and diluted to 50 ml using the mobile phase) I am thinking to dissolve the standard in 25 ml of NaCl solution and complete to 50ml using mobile phase this will definetly work. I just wonder why in the methods I have read there are no indications to that and no one had outlined the effect of NaCl ions.
In the case of fentanyl the API and one of its impurity behaved the same; i had to makeup my solutions (samples and standards) in the same manner.

[Nour]

thank you bruce
I am surprised too... that's why I wanted to share my observations with chromatographers in here. the wavelength was 206nm for fentanyl and 260nm for fluconazole.