Monitoring the wetting phase of fluidized bed granulation process using multi-way methods: The separation of successful from unsuccessful batches

Matero, Sanni; Poutiainen, Sami; Leskinen, Jani; Рейникайнен, Сату-Пиа; Ketolainen, Jarkko; Järvinen, Kristiina; Poso, Antti

doi:10.1016/j.chemolab.2009.01.003

Cited by 17 publications

(17 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In summary, by utilizing the idea of AE spectra to be a function of present granule size distribution together with multi-way data analysis that enables the handling of the batch-to-batch variation [27], we were able to predict the granule size distribution of the end product by examining the AE spectra acquired from the nucleation phase. In the further examination, the nucleation phase ends around 450 s for each batch.…”

Section: Discussionmentioning

confidence: 99%

“…The earlier study [27] exposed the applicability of the multi-way methods in modeling batch data with varying outcome. Moreover, they fit nicely with the 3-way data acquired with AE as revealed by the present results.…”

Section: Discussionmentioning

confidence: 99%

“…The 25% quartile of the cumulative distribution of granule sizes as a descending order of size portion ( Figure 4A) was chosen since it provides information about the granule size distribution. Since a badly performed batch has a wider size distribution with a high amount of coarse particles, as well as fine ones [27], the descriptive quartile should separate these batches from the well-performed counterpart. As an example of comparative analysis between different quartiles, the 65% quartile is illustrated in Figure 4B.…”

Section: The Y Matrixmentioning

confidence: 99%

“…This segmentation was necessary since AE spectra are otherwise very noisy and as shown in the earlier study the segmentation does not remove any critical information of the process [16]. The data set to be analyzed consisted of 26 granulated batches of which 10 were deliberately over-wetted which also contributed to an undesired granule size distribution [27]. The batches were analyzed using Tucker3 and N-way partial least squares (N-PLS) methods.…”

Section: The X Matrixmentioning

confidence: 99%

See 3 more Smart Citations

Estimation of granule size distribution for batch fluidized bed granulation process using acoustic emission and N‐way PLS

Matero

Poutiainen

Leskinen

et al. 2010

Journal of Chemometrics

Self Cite

View full text Add to dashboard Cite

Fluidized bed wet granulation of a pharmaceutical mixture is an unpredictable and complex process. Batch-to-batch variation, i.e. output material inconsistency, is still an unmanaged problem in fluidized bed granulation. The variation in outcome with different size distributions and yield arises from process conditions, which until now have proved difficult to control. The good quality batch is usually one with a high yield of the desired narrow granule size distribution. Thus, for successful granulation the size distribution needs to be controlled throughout the process in order to prevent the heterogeneity of different batch runs and inconsistent batch products. This paper introduces a method for determining end-product granule size distribution in the early phase of granulation, based on the observation that the quality of the early phase nucleation strongly affects the quality of the end product. In the quantitative model, the information gained from the process stream acquired with an acoustic emission (AE) transducer was combined and modeled using an N-way PLS method. Reference sieve analysis for the granule size distribution was performed off-line and was used as the response variable. AE monitoring is a non-invasive technique for evaluating granulation process performance by detecting the sound of particle impacts. AE caused by the particle-chamber wall interaction is influenced by the size of the granules; thus AE contains information about the granule size distribution. This study shows for the first time that the nucleation phase of granulation can be detected using AE techniques, and thus the possibility of predicting end product granule size distribution by means of AE measurement during the nucleation phase.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: The Y Matrixmentioning

confidence: 99%

Section: The X Matrixmentioning

confidence: 99%

See 2 more Smart Citations

Estimation of granule size distribution for batch fluidized bed granulation process using acoustic emission and N‐way PLS

Matero

Poutiainen

Leskinen

et al. 2010

Journal of Chemometrics

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, and since freeze-drying is a batch wise process, also the batch-to-batch variation has to be addressed. The differentiation between good and bad batches in the early process phase is a major concern in the pharmaceutical industry since batch-to-batch variability can be unpredictable [13]. The unpredictability of batch variation can lead to quality problems in the final product (e.g.…”

Section: Introductionmentioning

confidence: 99%

A batch modelling approach to monitor a freeze-drying process using in-line Raman spectroscopy

Sarraguça

Beer

Vervaet

et al. 2010

Talanta

View full text Add to dashboard Cite

34Freeze-drying or lyophilisation is a batch wise industrial process used to remove water 35 from solutions, hence stabilizing the solutes for distribution and storage. The objective 36 of the present work was to outline a batch modelling approach to monitor a freeze-37 drying process in-line and in real-time using Raman spectroscopy. A 5% (w/v) D-38 mannitol solution was freeze-dried in this study as model. The monitoring of a freeze-39 drying process using Raman spectroscopy allows following the product behaviour and 40 some process evolution aspects by detecting the changes of the solutes and solvent 41 occurring during the process. Herewith, real-time solid-state characterization of the final 42 product is also possible. 43The timely spectroscopic measurements allowed the differentiation between batches 44 operated in normal process conditions and batches having deviations from the normal 45 trajectory. Two strategies were employed to develop batch models: Partial least squares 46 (PLS) using the unfolded data and parallel factor analysis (PARAFAC). It was shown 47 that both strategies were able to developed batch models using in-line Raman 48 spectroscopy, allowing to monitor the evolution in real-time of new batches. However, 49 the computational effort required to develop the PLS model and to evaluate new batches 50 using this model is significant lower compared to the PARAFAC model. Moreover, 51 PLS scores in the time mode can be computed for new batches, while using PARAFAC 52 only the batch mode scores can be determined for new batches. 53 54

show abstract

Core consistency diagnostic in PARAFAC2

Kamstrup-Nielsen

Johnsen

Bro

2013

Journal of Chemometrics

View full text Add to dashboard Cite

PARAFAC2 is applied in multiple research areas, for example, where data containing shifts are analysed, but it is a challenge to determine the appropriate number of components in the model. In this paper, it is hypothesized that the core consistency diagnostic, which is currently applied in, for example, PARAFAC1 can be used to determine model complexity in PARAFAC2. Theoretically, a PARAFAC1 model is fitted ‘inside’ the PARAFAC2 algorithm, and it should therefore be possible to apply the core consistency diagnostic from PARAFAC1 in PARAFAC2. To support this hypothesis, three different datasets, as well as simulated datasets, have been evaluated by means of PARAFAC2, and the core consistencies have been investigated. There is a general trend that if the core consistency is low, the model is overfitted as in PARAFAC1. Also, core consistency captures the true variation in the data, whereas small peaks are easily overlooked by visual inspection of noisy models. However, for determining the number of components in a PARAFAC2 model, we suggest usage of the core consistency in combination with other model parameters such as residuals, loadings, and split‐half analysis. Copyright © 2013 John Wiley & Sons, Ltd.

show abstract

Monitoring the wetting phase of fluidized bed granulation process using multi-way methods: The separation of successful from unsuccessful batches

Cited by 17 publications

References 31 publications

Estimation of granule size distribution for batch fluidized bed granulation process using acoustic emission and N‐way PLS

Estimation of granule size distribution for batch fluidized bed granulation process using acoustic emission and N‐way PLS

A batch modelling approach to monitor a freeze-drying process using in-line Raman spectroscopy

Core consistency diagnostic in PARAFAC2

Contact Info

Product

Resources

About