“I believe that thirty million of these animalcules together would not take up as much room, or be as big, as a coarse grain of sand.”
–Antoine van Leeuwenhoek, 1632 to 1723
The advent of biotechnology marked one of the greatest revolutions in the history of mankind, undeniably the biggest one since the industrial revolution in Europe in the eighteenth century. Reportedly, the Israeli-British scientist Chaim Weizmann designed the first fermentor for producing acetone during the first world war. Various amateur and professional excursions followed this, until the first serious attempt by De Beeze and Liebmann succeeded in the production of yeast in 1944 on a large scale. And how large? It was a 20 litre fermenter!
Since then, the evolution of bioreactors has never seen a slowdown, and it still continues to grow. In this journey, various modifications and improvements accompanied this, such as mechanical agitation, gas sparging, etc. And more importantly, the focus gradually shifted from an empirical to a knowledge-based approach. Understandably, experiments have always remained an integral part of this development; however, a prior knowledge of the equipment behavior is now looked at very seriously, as it saves both time and money, that would have been spent otherwise in trial-and-error.
Technically speaking, the term “bioreactor” is used in the context of mammalian cell cultures, while “fermentors” are used to cultivate bacteria, yeast or algae. Mammalian cells are very sensitive to shear stresses, because they don’t have a cell wall. Practically, however, these terms are used interchangeably. Since the computational techniques and abilities have advanced by the introduction of both, more efficient codes, as well as powerful computers to solve them, stirred tank reactors have always been the centre of attraction, and a natural extension of them are the bioreactors. Typically, they are just like any other stirred tank, and most commonly equipped with more sophisticated controls. However, the design considerations are often significantly different than a normal stirred tank, given the delicate nature of the cells. The challenge here is two-fold:
1. How to accurately predict the operating conditions that should prevail inside a given bioreactor, for a given quality and throughput?
2. How to reliably scale-up/down these bioreactors, with minimum number of experiments?
However, the third, and the most subtle challenge is to take the product to the market as early as possible. With over three decades of constant existence in the field of chemical and biochemical process industries, advanced computational tools, such as computational fluid dynamics (CFD) have already proven its worth, both in process design (R&D) and manufacturing (see adjoining figure for an example of flow pattern visualization using CFD). Yet, it requires a certain level of understanding in order to assimilate the results that the tool gives. It is not surprising that many vendors have intensively started applying these advanced techniques in knowing what goes on inside their reactors. Experiments have their own limitations, especially in case of bioreactors, because a biological process is a potpourri of a number of individual processes such as:
1. Effective transfer of oxygen from air to the liquid
2. Desired transfer of oxygen to the cell mass
3. Uniform mixing of the vessel contents
4. Minimum damage to the cells by agitation
These factors are often counteracting, and hence, it is not possible to ensure maintaining all these factors across all the scales. Process expertise is needed to choose the factor that has the maximum impact on the final product, and use that for the scale-up. A comprehensive and thorough understanding of the hydrodynamics and the process ‘micro environment’ inside these reactors is therefore essential. Traditionally, the easiest approach to scale-up has relied heavily in maintaining the geometric similarities. However, as will be discussed in a series of articles following this one, it is not the smartest way of doing it. It often results in unreasonable values of operating parameters at the production scales, and the last minute adjustments are not possible.
Choosing the right type of impeller is the most critical step in such operations, and unfortunately often an ignored one! These bioreactors are very commonly operated with combinations of such impellers. Hence, it is imperative to understand the function of each of these impellers. E.g. an impeller that is suitable for an efficient gas distribution may not be appropriate for achieving a uniform mixing. Also, a certain type of impeller may be able to achieve the required mass transfer coefficient with minimum damage to the cell mass.
Thus, it is obvious that there is nothing obvious when it comes to designing and scaling up of a biological process. While the process knowledge remains to be critical, selection of a proper reactor configuration has its own important place in this exercise, which cannot be denied. We will discuss this in a greater detail in upcoming articles in this series.
– Tushar Tamhane
