-
E-mail
inquiry@synvol.com
-
Phone
13851985276
-
Address
No. 2 Jinyuan Road, Jiangning District, Nanjing City
Nanjing Sivo Biotechnology Co., Ltd
Reliability verification of torque and power curves as process endpoint criteria in DIOSNA wet granulation process
Date: 2025-11-21Read: 1
InDIOSNA wet granulationIn the system, real-time monitoring of the stirring blade torque (Torque) and motor power (Power) is the core PAT (Process Analysis Technology) method for determining the endpoint of particle growth. In theory, as the liquid is added, the powder wets, nucleates, and aggregates, the particles gradually grow, the system viscosity increases, and the torque continues to increase; After the particles reach the ideal state, continuing to add liquid will cause excessive wetting, resulting in a decrease or fluctuation in torque. Therefore, the peak torque or plateau period is often used as a criterion for automatically stopping liquid addition.
However, the reliability of this method is affected by multiple factors. Research has found that:
Prescription differences: The torque response of the high cellulose excipient system is slow, and the peak value is not obvious;
Liquid adding rate: rapid spray leads to local over wetting and sudden increase of torque, which does not represent overall uniformity;
Equipment scale: When amplifying from P1 (1L) to P6 (6L), the absolute values of power/torque change, but the normalized trend is consistent.
To verify its reliability, a project conducted multiple batches of tests on the same prescription on the DIOSNA P1-6 equipment, synchronously collecting torque, power, online particle size (FBRM), and final particle size distribution (LD). The results showed that when the torque reached 95% of the maximum value and maintained a stable platform for 30 seconds, the coefficient of variation of particle D50 was less than 5%, which was highly consistent with the trend of FBRM chord length (R ²=0.93). For batches that rely solely on a fixed liquid addition amount, the particle uniformity fluctuates by ± 15%.
The torque and power curves as endpoint criteria have high reliability, but dynamic thresholds need to be set based on prescription characteristics and calibrated with historical batch data. Under the QbD framework, this method can effectively achieve intelligent endpoint control based on processes rather than time, improving batch consistency.