Quality by design (QbD) is an approach to product development where a comprehensive understanding of the effect of various inputs (e.g. process parameters, materials) on the final product (active pharmaceutical ingredient or drug product), leads to a process or product design that ensures reliable manufacture of a high-quality product. Analogous to QbD for product, QbD may also be applied to analytical methods, by defining the required analytical performance and systematically relating how the technique and method parameters relate to analytical performance. Achieving the appropriate level of method understanding typically involves a multifactorial approach. The factors to be studied are identified via a systematic risk analysis, which allows truly impactful parameters to be further investigated in studies applying statistical design of experiments (DOE) or mechanistic models to develop a multidimensional region of experimental space in which the effects of key factors are understood and documented. An operating region can then be defined where there is high confidence that the method will operate reliably. Data generated during use of the method can be monitored over time to ensure continued adequate performance of the analysis. 

Highlights:

• Test parameters of Tanaka and of Snyder showed surprisingly a high correlation.
• USP test is related to Tanaka and Snyder only in terms of hydrophobic characters.
• Some differences in the correlation are shown between aromatic and alkyl phases.
• The hydrophobic-subtraction model is extended to describe calixarene-bonded phases.

This paper presents new reversed phase liquid chro-matographic methods (HPLC-FLD and LC-MS/MS) for the quantification of sulfonamides in spiked and incurred honey samples. The sample preparation was optimized using Oasis HLB (hydrophilic–lipophilic balance) solid-phase extraction (SPE) cartridge. Elution of sulfonamides was carried out under acidic, neutral, and basic conditions using methanol. The sample clean-up was also tested using Strata-XL cartridges. The HPLC-FLD separation was performed using a Varian C18 column and a ternary (methanol-acetonitrile-phosphate buffer, pH 5) mobile phase resulting good selectivity for the determination. The robustness of the ternary gradient method was evaluated by computer simulation with DryLab4. LC-MS/MS separation was carried out on a Kinetex XB core-shell type HPLC column that enabled a low limit of detection (0.01–0.5 µg/kg) and faster separation (6 min). The developed methods were validated in accordance with the European Union Commission Decision 2002/657/EC and were applied successfully for more than four hundred honey samples under a national monitoring program.

A new type of method development that uses modeling by DryLab4 to find the "best" separation for high performance liquid chromatography (HPLC) was investigated. Principles of Quality by Design (QbD) were followed when planning the investigation. A simple Design of Experiments (DoE) with only three measured factors — gradient time, pH, and temperature — was used with different columns. The basic experiments were saved in an electronic file with the peak tracking results. After calculating a Design Space, the best prediction was compared with a confirmation run. The process delivered precise results and the method was able be transferred to a routine quality control (QC) laboratory.

Highlights:

• The influence of changing %B (or tG), °C, and solvent type on selectivity is reexamined.
• These changes can be as effective as changing pH or column type when adjusting selectivity.
• These changes can mimic changes in pH, reducing or eliminating the need to optimize pH.
• Such continuous changes can be more effective than a step-wise change in column type

An older method for amlodipine was reworked with the goal to reduce the analysis time of 60min below 6min. To select the best column for short and robust analysis, 9 different UHPLC column chemistries were investigated using 3-dimensional resolution spaces based on 12 experiments using DryLab4 modelling software. The main variables used were gradient time (tG), temperature (T) and the pH of eluent A. The best critical resolution was calculated and located in a 3-dimensional space in an automated fashion and the corresponding best experiments were carried out. The work (9×12=108 runs) for DryLab4 modeling was finished with an UHPLC instrument in less than 24h. The comparison between predictions and real experiments showed an excellent correlation with differences typically less than 0.04min (<3s) in average, although the set points were located at quite different conditions on gradient times, pH's and temperatures for the individual columns. With the support of DryLab4 all columns could perform the required baseline separation at their individual best working points with satisfactory results.