Five cases from real life pharma

Case 1: “Insignificant change” in production process had significant effect on particle size

A routine determination of particle sizes of a product that we have analysed for the last 8 years suddenly showed a new distribution – and even though the specifications were still met, it was relevant to investigate this issue further. The customer was very puzzled: Were we absolutely sure that we had not changed something in the experimental procedure? – because they had definitely not! We carefully reviewed the experimental procedure and the customer had a closer look at the processing procedure. We did not find any reasons for the new distribution.  At the end the customer found out that something had actually changed during production: One of the tubes in the production line had been changed – a change that they were absolutely sure would not affect the outcome! Sometimes minimal disturbance of the crystallization conditions might cause large changes in particle sizes – and for the same reason, Regulatory Authorities require verification of the particle size distribution when something in the production line is changed.

Have you experienced such changes? Do you always verify particle size distribution after changes in processing? Do not hesitate to contact us if you have any question or concerns in relation this issue – or if you want to share your knowledge with us. 


Case 2: New peaks on XRD

Recently we ran a routine XRD measurement. The ID test showed that the identity of the crystal form was “as expected” and formally complied to specifications. However, three additional peaks were observed on the diffractogram. Should we care? People working with XRD know that special particle/crystal characteristics can give rise to extra peaks because of preferential orientation. Thus it is tempting to conclude that these extra peaks are  caused by this orientation aspect and do nothing further about the issue. However, a potential risk exists that it is caused by “something else”, for instance, the existence of a new polymorphic form in the sample – or another crystalline impurity. With regard to the regulatory guideline you do not have to examine the origin of these peaks further: The identification test was fine. However, to be sure that you are not about to find a new polymorphic form, that might cause very large problems in later development, the system should be further explored.  In the current case it turned out to be caused by an impurity! It highlighted that XRD should not be used for quantitative analysis unless the method has been validated for this purpose (e.g using an internal standard).

How do you use your XRD results: “Just” for ID or also to determine crystalline purity? Is the method you use satisfactory for your needs? Do not hesitate to contact us to if you want further information.

Håndtegnet graf.

Case 3: Simple explanation to time-consuming experimental problem

A customer had for a year spent time developing a method for determination of particle size of a certain product (of course not full time) – and even though they had obtained a large amount of information about the product, the method kept causing them problems. With an extremely tight timeline we developed and validated a method that fulfilled their needs. During method transfer we were further able to explain the reason for having so much trouble with their method – it was due to some special technical issues related to the instrument that they were not aware of: The instrument needed to “relax” shortly in between measurements (everybody needs a break now and then).

Do you prefer to validate the methods yourself – or do you use an external partner? We would be happy if you would share your thoughts about benefits and drawbacks in relation to your approach. (Share your thoughts by clicking here.)


Case 4: A simple analytical experiment made the need for more equipment redundant

During a meeting with a customer producing API’s we came to talk about storage issues: They had to invest in new drying cabinets in order to have the capacity to dry the material correctly – but they had insufficient laboratory space. After asking a few questions about the process we became aware that the temperature and time for drying of the material was based on “that’s how we usually do”. We decided to perform a single DVS experiment on two of their materials – and realized that the time needed for drying the material was much shorter than expected. It turned out that the knowledge about the reduced time needed for drying solved their capacity problems: They could actually dry much more material than expected in a much shorter time – and the need for extra drying cabinets no longer existed.

How do you determine drying conditions for your products? Could these conditions be optimized? Please contact us if you want further information about examination of water uptake and loss.

INSTRU1 Dessert

Case 5: The water content was actually quite predictable

A customer called us as they had problems interpreting the results from Karl Fischer experiments with a certain product: Large variations were observed between measurements performed on different days, showing water content between 1 and 4%. We suggested a TGA measurement (where the sample is heated under controlled conditions) and a DVS measurement (where the extent of water binding as a function of temperature is examined). These results showed that the product was able to take up around 8% of water at room temperature (=quite hygroscopic), and the uptake was highly dependent on the relative humidity and temperature. Thus, the large variations they observed in their results were caused by differences in the starting conditions of the experiment.

Have you had similar experiences with your products? Please contact us if you need more information about the outcome of these water-uptake and release experiments.