Bioequivalence and Bioavailability Forum

Hi Ionsource!

❝ I want to know about the causes of inconsistent recovery specially in liquid extraction […]

Without knowing details of the method we have no clue. Sometimes the most weird things happen.
I give you an example (my personal translation):^*

Practical example 14: Special case phenol in LLE
We had to analyze samples of several BE studies of silibinin within a certain time interval. The first study was a resounding success. In revalidation for the next study there were always outliers toward lower results in calibration and QC samples. Once a method works, one relieved can pass it to employees and finally want to enjoy the fruits of – in this case – two years of development. And now those results! Employees were closely monitored then in their operation and pH values of buffers critically examined (essential in this method) – and no discrepancies were found. Then you analyze a batch with all your own personal experience and find outliers as well. What’s to blame? To minimize suspense: A small detail had been changed by an employee without being noticed as important. For practical reasons we removed the stoppers from centrifuge vials before freezing out samples during extraction. In contrast to the first study, where all results had been fine, the employee did not use a small beaker (150 mL) with acetone / dry ice (space for about six samples and reaching up to half the height of centrifuge tubes), but a larger beaker (250 mL) – which could hold twice as many samples. The beaker was higher and had about 80% of the height of the centrifuge tubes. So where was the problem? While the samples cooled down, some of the dry ice evaporates from the acetone / dry ice mixture as CO₂ gas, which is relatively dense and slowly trickles down over the beaker rim. With an occasional little air movement CO₂ gas in the 250 mL beaker could flow into the centrifuge tube. CO₂ was decanted along with the solvent and thereby changing the pH of the buffer (pH 10.5) slightly downwards. Since the pKa of phenols is about 9.5-9.8, one immediately sees that the back-extraction performed at pH values of 10 or just below was clearly impaired and gave worse recoveries.

Shit happens. Sometimes it is very difficult to find out why. My friend Hermann Mascher has close to 40 years of experience in bioanalytics. I can image his feelings.

---

• H Mascher
  Klinische Analytik mit HPLC. Ein Ratgeber für die Praxis
  Wiley-VCH, Weinheim, pp 111–2 (2010)
  I can recommend the new English edition:
  HPLC Methods for Clinical Pharmaceutical Analysis
  Wiley-VCH, Weinheim (2012)
  ISBN: 978-3-527-33129-1