Cholecalciferol Method Development Issues

[jmmunson]

I am currently working on developing a method for Cholecalciferol in the Vehicle of Lactose NF Anydrous. The cholecalciferol is mixed with maltodextran prior to being compounded with the placebo. The problems that have arisen are as follows.

1. I have not be able to find a diluent that will achieve 100% recovery of the material other than 100% MeOH, which does not match up with the mobile phase (65% ACN, 35% H2O * currently) it was (90% ACN/10% H2O).

2. The lower percentage of organic used the lower the response of the API as well as a much broader peak is achieved.

3. I have tried both C18 and CN columns with different dimensions to ensure that resolution can occur between any degradant peaks. On either column I still run into the first 2 issues. I achieved better resolution with higher water content but I sacrifice peak shape. I have also tried gradients that have been found to not work as well.

4. A quick test was ran to see if the API was dissolved and diluted in 100% MeOH first, then diluted 1:1 with water (Final Diluent - 50% MeOH/50% H2O) would give better peak characteristics. I have ran this test multiple times and have continuously ran into the issue of the peak shape deteriorating if lower organic was used. This time when the organic % was lowered the peak shape looked great for both diluents, but now I am seeing a strange shift/jump or drop in the baseline after 20 minutes of run time. This strange shift only occured with the samples that were prepared in 100% MeOH but was not consistant, while the 1 sample with 50% MeOH/50% H2O as its final diluent gave a flat baseline.

I am at the point to where I do not know what to try next, Any thoughts or ideas that may help?

[JMB]

The sample prepn & HPLC method given here,

2.2 Sample Preparation

Tablets were stored at 60 C/ amb. RH/ open dish for 7 month. Ten tablets were combined and extracted by sonication with 100 mL of water-methanol (5:95, v/v) for
0.5 h, followed by stirring for 3 h. A portion of the solution was centrifuged at 14,000 rpm for 5 min prior to analysis by LC/UV and LC/MS.

2.3 LC/UV Analysis

Aliquots (100 L) of the solution (sample tray at 5 C) were injected onto an HP 1100 liquid chromatograph. Data were collected by a VG Multichrom system. Gradient chromatography was performed on a Platinum EPS C18 column (150 x 4.6 mm; 3 m particles) held at 17 C, with UV detection at 265 nm. The organic component of the mobile phase was acetonitrile, while the aqueous component was 0.1% phosphoric acid. Gradient program: %-organic (time, min.), 4 (0), 4 (7), 80 (20), 85 (48), 95 (50), 95 (60), 100 (61), 100 (68), 4 (75).

gives the chromatogram not shown below; I cannot paste it in here !!!

However,

major excipient RT ~21 min
pre-vit. D3 ~38 min
vitamin D3 ~42 min

Baseline is flat and level after 10 min; 0.1 % H3PO4 in aqueous was found to help peak shape.

Let us know if this works for your formulation.

[JMB]

Here is what I could not paste before [done by my son]




The small peaks between RT 55-65 min are esters of vit. D3 and pre-vit. D3

[jmmunson]

I greatly appreciate the method that you advised me to use, however it did not work my application. The only variation from the method that you recommended was the column. I am unable to get my hands on a Platinum EDS, so I used a Luna C18 with the exact same dimensions. With that said, the retention time was ~68 min. The main degradants I have had the trouble with are the 2 small peaks that are on either side of the API in the chromatogram that you uploaded. If you have any other suggestions I am definitely open to them.

Thank you again.

[HPLCaddict]

Luna C18 and Platinum EPS are tremendously different columns, although formally both are "C18". Luna is a densely covered, hydrophobic column, while Platinum EPS ("Extended Polar Selectivity") has a deliberately low coverage. So it's not really a surprise that these two column yield different chromatograms.

[carls]

The problem is not clear.

In isocratic elution peak width increases and peak height decreases with increasing retention time. The peak area, however, is unaffected by changes in retention time. Is this what you see?

In gradient elution peak width and peak height remain relatively constant throughout the run for peaks that are reasonbly retained.

[JMB]

As HPLCaddict said, the Luna and EPS column packings are totally different.

Although I did not perform the HPLC method development, I was responsible for the LC/MS work on the project and did sit in at the group meetings. MANY, many HPLC columns and chromatographic conditions were evaluated to obtain the given LC/UV profile.

We were particularly interested in the late-eluting ester peak doublets, so you can modify the gradient if they are not present in your samples.

However, the devil truly is in the details (column temp.; aq. 0.1% phosphoric acid etc.....), and the column itself was the major contributor to the overall HPLC performance.

This work was done about 10-12 years ago, so there may now be newer C18 columns of the EPS type that will give equivalent or better performance.

Please let us know the outcome.

[jmmunson]

I apologize for such a late response. I ended up going with the original method I had found in an article online. The column I ended up using was a Waters Spherisorb ODS2. I was able to get 100% recovery using 100% Methanol.

Thank your for your suggestions and help.