In-line control of Granulation at Genzyme

Customer Story published in 2011

LightHouse Probe - Fish eye design

Genzyme reduces probe fouling and improves NIR data using the Lighthouse Probe®

Introduction

In 2011, Genzyme, at the time one of the world’s leading biotechnology companies, was looking for a PAT solution to control the high shear granulation of a development product by measuring a critical product attribute, rather than relying purely on time based processing or impeller current loading.  The company used an optimized Lighthouse Probe® to get representative NIR data of a granulation process. This article explains how the data was then correlated to moisture content, bulk/tapped density and particle size of the final product. 

The reason for choosing a Lighthouse Probe® was to expose as much product to the detector as possible. This was an important consideration for a high shear granulation process where changes in the process can happen quickly. Furthermore the scanning technology used in the spectrometer was FT-NIR, which has a relatively slow scan speed when compared to diode array based spectrometers. 

Also the article describes how probe fouling was overcome, by using the probe in a high shear wet granulation process. The system is capable of being upgraded to a full GMP system for installation in production.

Background

Genzyme was in phase three of the clinical trials of a new drug. The API had a small particle size, and associated poor flow properties. This presented Genzyme with a challenge when forming the dosage for patient administration. To improve the flow characteristics the API was to be formed into granules with other excipient materials by a high shear wet granulation process.

Genzyme chose to use the Lighthouse Probe® from GEA to monitor a high shear granulation process and to measure and quantify critical granule attributes. These measurements enabled the company to monitor the formation of the granules during the process to achieve an optimal granule size distribution.

The experiment used an FT-NIR spectrometer. This technology, whilst having high resolution and low signal-to-noise ratio for spectra, has a relatively slow scanning speed, compared to monochromatic-based spectrometers, such as diode array. Therefore, for the spectrometer to receive as much information as possible on what is occurring during the granulation process, a sample interface with a large window that would not suffer from fouling was required. 

Experimental set-up

The experiment used a Bruker Matrix-F spectrometer, attached to GEA Lighthouse Probe® operating in a PMA-1 10 L Granulator Bowl.  The Lighthouse Probe® was inserted through customized opening in the granulator lid viewing window.

The experiment

The experiment analyzed 20 Batches in a manufacturing process DoE, which included five factors related to the high shear wet granulation step. NIR spectra were collected from the granulation process at a scanning speed of approximately one spectra every five seconds. The last six spectra collected from each granulation step was averaged and correlated against granule attributes, such as water content at the end of granulation, as well as particle size and density of the final blended product. The wet granules were subsequently tray dried, milled and blended with a lubricant.

Principle component analysis was performed on the change in spectra throughout the granulation process to provide an insight into what was occurring during the granulation process.

Results and discussion

Principal component analysis of the spectra highlighted changes in the granulation process and also highlighted the effects of changes in granulation parameters. This can be seen in the Hotelling’s T2 plot (figure 2 - download only), where the first batch in the experiment had a shorter processing time due to less water added and a slower spray rate, relative to the second batch. It is thought these inflexions indicated on the plot could be related to the point where there is sufficient water added to activate the binder in the formulation of the drug product. 

Good correlation was achieved for a dynamic model for correlation of the averaged spectra at the end of the granulation process to:

  • Water by Karl Fischer
  • Particle size D90 and D50 values for final blend  – See figure 3 (download only)
  • Bulk and Tapped Density - See figure 4 (download only)
Conclusions

Genzyme were looking for a PAT solution which would enable control of the high shear granulation step for a development product by measuring of a product attribute, rather than relying purely on time based processing or impeller current loading.   

The Lighthouse Probe® showed that it is well suited to monitoring a high shear wet granulation process when used as the sample interface with a NIR spectrometer. It minimizes the potential for probe fouling based on the 360° window and allows for the maximum absorbance of light from the spectrometer, with which to provide a signal for correlation.

The Lighthouse Probe® produced satisfactory initial models against key granule attributes such as water content, particle size and bulk tapped density. 

This study proved the feasibility of Lighthouse Probe® technology for monitoring and eventual control of a high shear wet granulation process.

Related Articles

If you want to learn more about granulation or the LightHouse Probe®, select one of the links here.

Receive news from GEA

Stay in touch with GEA innovations and stories by signing up for news from GEA.

Need assistance?

We are here to help! With just a few details we will be able to respond to your inquiry.