[HPLC] drug in oil

[milkvetch1234]

I'm new to pharmaceutical analysis and I'm finding it challenging to analyze drugs in oil. I have a steroid hormone in oil, and I modified a literature method using 45% ACN: 55% H20 (the analyte is eluted in this ratio) followed by a gradient run to 95% ACN: 5% H2O to remove oil in the column (after the analyte is eluted).

I dilute the formulation sample with ACN 20x (conc. 0.005, 5ppm) before HPLC analysis (Reverse-phase C18, PDA ) in order to extract the API. The problem is that this detection method has a relatively low response factor, as well as low signal-to-noise ratios. I used methanol for diluent and mobile phase (e.g. 80% MeOH: 20% H2O), but the noise and signal levels did not improve. No problem on testing the standard solution (100ppm API in ACN) with relatively low response factor.

How do I analyze this steroid hormone in oil? Is there a way to optimize the HPLC method (e.g. column, mobile phase? What is the common HPLC detection method for analyzing drugs in oil? Is UV-Vis Spectrophotometer a better option?

Many Thanks!

[vmu]

I'm new to pharmaceutical analysis

I guess you are new to HPLC.

I modified a literature method

What was the original method? What is the reason to modify it? What did you modify?

I dilute the formulation sample with ACN 20x (conc. 0.005, 5ppm) before HPLC analysis (Reverse-phase C18, PDA ) in order to extract the API.

If you dilute the sample, you do not "extract" anything. You just... dilute. Do you do one dilution (from 100 ppm in the drug to 5 ppm in the final solution)? What is "conc. 0.005"? Why do you dilute 20-fold? What about 10- or 5-fold dilution?
If you use a one-point calibration, it is necessary to use the standard solution with the concentration close to the expected concentration of the API in the sample solution. Otherwise, you must prove the linearity in the range from 5 ppm (in the sample) to 100 ppm (in the standard).

The problem is that this detection method has a relatively low response factor, as well as low signal-to-noise ratios.

The latter quantity (signal-to-noise ratio) includes the former one (response).
To increase the S/N you have to increase the S and/or to decrease the N.
To increase the signal:
- increase the injected amount (use higher injection volume and/or less diluted sample solution);
- use the appropriate wavelength for your analyte (What is the WL? What is the analyte? How does its spectrum look like?);
- consider changing other detection parameters (e.g., the data sampling frequency if it's too low);
- consider changing chromatography (the narrower the peak (in volume units), the larger its height).
To reduce the noise:
- use the appropriate wavelength (not too low);
- use the appropriate data sampling frequency (not too high);
- use a VWD instead of a PDA detector.

Is there a way to optimize the HPLC method (e.g. column, mobile phase?

Anything that leads to a narrower and higher peak increases the signal. The available options for the mobile phase depend on the wavelength.

What is the common HPLC detection method for analyzing drugs in oil?

There is no common detection method. The method is specific to the task. ELSD, CAD, and RID are the typical alternatives to the UV detection.

Is UV-Vis Spectrophotometer a better option?

A PDA detector is a UV-Vis Spectrophotometer.

[milkvetch1234]

Thank you for your reply! To make it clearer, let me explain

I'm new to pharmaceutical analysis

I guess you are new to HPLC.

I modified a literature method

What was the original method? What is the reason to modify it? What did you modify?

The original method uses an isocratic method, 45% ACN: 55% H2O, but it's for urine samples, not oil samples. Later, we found out that the HPLC system was overpressured because the column was clogged, so we added a gradient step to remove residual oil from the column after the analyte was eluted.[/color]

I dilute the formulation sample with ACN 20x (conc. 0.005, 5ppm) before HPLC analysis (Reverse-phase C18, PDA ) in order to extract the API.

If you dilute the sample, you do not "extract" anything. You just... dilute. Do you do one dilution (from 100 ppm in the drug to 5 ppm in the final solution)? What is "conc. 0.005"? Why do you dilute 20-fold? What about 10- or 5-fold dilution?

We prepared 3 samples for HPLC.
1) 100ppm standard solution, dissolved in ACN. - noise was relatively low
2) 5ppm standard solution, dissolved in ACN. - noise is relatively low, but the signal-to-noise is not very distinct
3) Pretreated, 5ppm of analyte solution, (original sample is 0.1mg/mL API, dissolved in oil, oil is a must vehicle for animal test dosing) - noise is relatively high, some interference peaks were quite next to the target peak.


Pretreatment of sample 3) : Of course we performed dilution for samples 3), as we can't inject original "oil" samples in HPLC. HPLC system will be overpressure. Spin down the "oil sample" and kept the supernatant (oil at the bottom of centrifuge tube) and followed by 0.45um filtration .

Analyte in oil has a poorer baseline than analyte in ACN. Therefore, we guess the "noise" was due to anything was "extracted" from the oil
so 20x dilution was performed to minimize possible interference in oil extracted by ACN, if I use methanol instead...No improvement in baseline

The problem is that this detection method has a relatively low response factor, as well as low signal-to-noise ratios.

The latter quantity (signal-to-noise ratio) includes the former one (response).
To increase the S/N you have to increase the S and/or to decrease the N.
To increase the signal:
- increase the injected amount (use higher injection volume and/or less diluted sample solution);
- use appropriate wavelength for your analyte (What is the WL? What is the analyte? How does its spectrum look like?);
245nm. Noise peaks found at 0 min to 3.8min. Retention emT of analyte is round 4.5min


- consider changing other detection parameters (e.g., the data sampling frequency if it's too low);
- consider changing chromatography (the narrower the peak (in volume units), the larger its height).
To reduce the noise:
- use appropriate wavelength (not too low);
- use appropriate data sampling frequency (not too high).

Is there a way to optimize the HPLC method (e.g. column, mobile phase?

Anything that leads to a narrower and higher peak increases the signal. The available options for the mobile phase depend on the wavelength.

What is the common HPLC detection method for analyzing drugs in oil?

There is no "common" detection method. The method is specific to the task. ELSD, CAD, and RID are the typical alternatives to the UV detection.

Is UV-Vis Spectrophotometer a better option?

PDA is a UV-Vis Spectrophotometer.

I mean if I pick UV-Vis Spectrophotometer. No dilution is required for my analyte sample in oil, but the baseline caused by oil may still the issue.
What we have in our lab is HPLC-PDA detector, HPLC-FLD and UV-Vis spectrophotometer.

[vmu]

I mean if I pick UV-Vis Spectrophotometer. No dilution is required for my analyte sample in oil, but the baseline caused by oil may still the issue.

The main issue with the spectrophotometry (without HPLC) is that the spectrophotometry is much less selective than HPLC-UV. All the substances absorbing at the used wavelength contribute to the signal. The same solvent for the sample and for the standard may be required as well.