Biotech processes hinge on delivering sufficient oxygen to growing cells in a bioreactor. However, oxygen transfer can be surprisingly complicated, especially when transitioning from small-scale lab conditions to large industrial fermenters.

Recognizing this challenge, two mass transfer specialists from the Delft University of Technology—Lars Puiman and Rik Volger—recently discussed how understanding kLa (the volumetric mass transfer coefficient) can help scientists optimize cell growth and product formation. 

Drawing on their experience in computational fluid dynamics (CFD) modeling, large-scale reactor design, and fundamental chemistry, they highlighted critical factors that affect oxygen transfer, ways to troubleshoot at the lab scale, and strategies for scaling up to commercial volumes.

The Importance of Oxygen in Bioprocessing

All aerobic organisms need oxygen to drive respiration and maintain proper metabolic function. In an industrial bioprocess, cells rely on oxygen to convert substrates into products such as antibodies, enzymes, or other valuable compounds. 

Bioreactors try to satisfy this need by sparging air or pure oxygen into the culture. Yet, oxygen transfer from bubbles into the liquid phase (and subsequently to cells) remains one of the most common bottlenecks in commercial-scale fermentation. 

According to Puiman and Volger, scientists frequently face unanticipated oxygen gradients, foam formation, and local dissolved oxygen (DO) depletion when they scale from small shake flasks or lab reactors to multi-thousand-liter systems.

Defining kLa

kLa is a key parameter that describes a bioreactor's ability to transfer oxygen into the liquid phase. 

It consists of two primary components:

• kL: A mass transfer coefficient that quantifies the "speed limit" of gas diffusion at the bubble–liquid interface
• a: The interfacial area available for transfer, often visualized as the total bubble surface area

Multiplying kL by a yields kLa, which can be conceptualized like a highway's capacity: kL represents how fast cars (oxygen molecules) travel, and a represents how many lanes are available. A higher kLa means oxygen can diffuse more effectively into the liquid phase, boosting cell growth and productivity.

Why kLa Matters

• Oxygen-Limited Growth: If the culture is starved for oxygen, cells may shift to unwanted metabolic pathways, reducing product yields or causing undesired by-products.
• Process Predictability: Understanding kLa helps engineers design more consistent processes. Without reliable oxygen transfer, results can vary widely among batches.
• Scale-Up Strategy: The lab scale is often well-mixed and easy to control, masking the complexities that arise at larger volumes. Tracking kLa across scales avoids costly trial-and-error at the pilot or commercial scale.

Fundamental Challenges in Mass Transfer

Puiman and Volger emphasize that oxygen solubility in water is inherently low, particularly under standard operating conditions. This limitation drives engineers to optimize mixing, bubble formation, pressure, and other parameters that accelerate oxygen dissolution.

Major Influences on kLa

They outline multiple factors that can either boost or hinder oxygen transfer:

1. Agitation (Stirring Speed): Increasing the stirring rate generally forms smaller bubbles and raises the interfacial area. However, this also raises power consumption, which can become prohibitively expensive at large volumes.
2. Gas Flow Rate: Sparging additional air (or pure oxygen) provides more oxygen. Yet, introducing more gas may lead to the coalescence of bubbles, changes in foam levels, or mechanical stresses on the cells.
3. Pressure and Temperature: Higher operating pressures increase gas solubility, helping oxygen diffuse. However, they also add complexity to vessel design and gas supply systems. Temperature can affect oxygen solubility and cell metabolism, so the net effect on kLa may vary.
4. Broth Composition: Nutrients, salts, cell debris, foam-control agents, and final products affect bubble behavior. For instance, ethanol can reduce bubble coalescence, thus forming smaller bubbles and enhancing mass transfer. By contrast, many antifoam agents lead to larger bubbles and reduced oxygen absorption.
5. Sparger Design: Various sparger types—coarse bubble spargers or microbubble diffusers—can produce drastically different bubble sizes. Microbubble spargers are highly effective in increasing kLa but require more energy to push the gas through fine openings.

There is a point with the spargers that make smaller bubbles is that in general those need higher air pressures to get the air into the broth. So while you will increase your mass transfer at the same time you do need a larger compressor to get the air in. So with that you probably increase the cost of your run.

Consequences of Altering kLa

Manipulating one parameter frequently imposes trade-offs. For example, accelerating agitation can elevate shear forces that may damage sensitive mammalian cells. 

Adding antifoam might mitigate an overflowing fermenter but can impair oxygen transfer by enlarging bubbles and reducing the interfacial area. A well-designed process balances these competing demands to maintain high productivity with acceptable energy and resource costs.

Lab-Scale vs. Industrial-Scale Realities

While measuring and controlling Dissolved Oxygen (DO) in a 1- to 5-liter bench reactor is relatively straightforward, scaled-up systems can exceed thousands—or even tens of thousands—of liters. In these giant vessels, the fluid dynamics become more complex, and local variations in mixing or aeration often lead to heterogeneous environments.

Differences in Larger Volumes

1. Energy Costs: Running high RPMs in a tiny lab reactor is trivial. However, replicating similar power-per-volume intensities in a 20,000-liter fermenter may be prohibitively expensive or technically infeasible.
2. Sensor Limitations: A single DO probe in a large bioreactor provides only localized data. Cells near the probe might see ample oxygen, but cells elsewhere in the vessel could experience oxygen scarcity or completely different conditions.
3. Local Concentration Gradients: CFD simulations often reveal stark gradients in dissolved oxygen, substrate concentrations, and pH. Many large-scale reactors do not perfectly homogenize the culture, forcing cells to move through microenvironments of distinct oxygen or nutrient availability.

Scale-Up Hazards

• Inconsistent Cell Performance: Cells grown in microenvironments with poor oxygen transfer might partially shift to anaerobic metabolisms, leading to side products or inconsistent yields.
• Unpredictable Productivity: Observing robust cell growth in the lab does not guarantee similar growth in a large tank if oxygen demand surpasses the available mass transfer.
• Delayed or Failed Projects: If developers do not account for these challenges early on, they may need repeated scale-up runs, driving up costs and prolonging development timelines.

Strategies for Managing Mass Transfer Across Scales

Puiman and Volger encourage a more integrated approach, combining rational design, modern computational tools, and careful bench-scale experimentation.

Using CFD and Downscaling

One of the most important techniques to forecast large-scale behavior is computational fluid dynamics (CFD). 

By modeling a reactor's geometry, inlet gas flow, and fluid properties, engineers can predict local DO levels, bubble sizes, and mixing times:

• CFD Insights: This tool identifies regions in the tank that might be oxygen-starved and calculates how often cells experience low-oxygen microenvironments.
• Downscaled Experimentation: CFD results can inspire advanced "downscaled" lab tests. For instance, operators might program the stirrer in a small reactor to vary over time, artificially simulating the dynamic oxygen changes observed in a large bioreactor.

Multi-Sensor Approaches

In many industrial processes, relying on a single DO sensor is insufficient. Cells at different heights or agitation zones might see radically different oxygen conditions. 

Strategies to combat this include:

• Additional DO Probes: Multiple probes will be placed across the vessel to track local fluctuations.
• Off-Gas Analysis: Monitoring the oxygen content in the exhaust gas to estimate overall oxygen uptake rates (OUR) and confirm whether oxygen supply meets demand.
• Periodic Recalibration: Since DO probes can drift over long runs, calibrating sensors regularly during a continuous or extended batch campaign ensures accurate readings.

Designing Robust Smaller-Scale Systems

While microtiter plates, single-use mini-bioreactors (e.g., Ambr 15), and microfluidics can accelerate screening, often fail to replicate large-scale hydrodynamics. In response, some labs develop specialized downscale reactors that deliberately introduce oxygen gradients or intermittent aeration to mimic real-world challenges:

• Variable Stir Speed: Program small reactors to alternate high and low agitation, approximating how cells might encounter inconsistent mixing in large tanks.
• Microfluidics for Fluctuations: In advanced scenarios, microfluidic channels can introduce cells to rapidly changing oxygen levels. This technique reveals how cells adapt—or fail to adapt—when they traverse macro-scale gradients.

Measuring kLa in Practice

Beyond CFD-based predictions, direct measurement of kLa remains a best practice for verifying oxygen transfer performance. In the lab, a common technique is the static gassing-out method, where the vessel is flushed with nitrogen to remove oxygen before reintroducing air and logging how quickly DO returns to saturation. 

This method works well for water-based systems without biomass. At larger scales or with cells present, dynamic gassing-out can factor in oxygen consumption by the culture, making it more accurate for real-time conditions.

Practical Pitfalls and Considerations

Although kLa is critical, measuring or controlling it can be fraught with errors. Puiman and Volger highlight typical stumbling blocks that hinder successful scale-up:

• Probe Drift and Calibration: DO sensors may drift over time, especially in lengthy or harsh fermentation processes. Rigorous calibration is essential for confidence in results.
• Fluid Complexity: Foaming, antifoams, metabolic by-products, and high cell densities dramatically change bubble behavior. Basing scale-up designs solely on water-based kLa data ignores these complexities.
• Energy-Intensive Fixes: Adding pure oxygen or drastically upping agitation can salvage mass transfer problems but at a high energy or cost burden. Overuse of such measures reduces economic viability.
• Inadequate Documentation: For instance, if operators do not accurately track the amounts of antifoam or acid/base added during a run, they may be unable to replicate results or interpret anomalies in oxygen transfer.

Future Directions in Mass Transfer Optimization

The conversation around advanced technologies—AI, machine learning, or next-generation sensors—naturally arises. Puiman and Volger see promise in these tools but also caution that fundamental engineering principles often deliver the most significant gains:

• Expanded CFD Models: With more computational power, future CFD models may incorporate detailed kinetics, real-time environment changes, and morphological cell data, lending even better predictions.
• AI-Driven Process Control: Machine learning systems could integrate sensor data, feed rates, and sparger adjustments to automatically optimize oxygen supply.
• Enhanced Sensor Technologies: Novel miniaturized sensors placed at multiple points in large reactors could offer a richer picture of local DO and bubble distribution, supporting more nuanced control schemes.

However, the experts emphasize that no advanced technology replaces a fundamental understanding of how stirring speed, bubble formation, and solutes govern oxygen transfer. Turning up the RPM or adding more air may look good in a small system but cause massive scale-up headaches unless carefully balanced with cost and potential cell damage.

To have a good idea of what's happening inside of your vessel and to have a good predictive power, it's also important to know what you've added. So proper quantification of the amounts added of these is essential and I think still often lacking.

Re-Focusing on "Bubbles and Broth"

Referring to a recent review article titled "Bubbles and Broth," Puiman and Volger reiterate that mass transfer can shift dramatically once a real fermentation medium is in play. 

They advocate for:

• Realistic Broth Testing: Avoid generalizing results from water or simplistic media. Complex media can create smaller or bigger bubbles, alter bubble coalescence, and make kLa measurements inaccurate if not accounted for.
• Process-Specific Focus: Each new fermentation target—a monoclonal antibody or a novel synthetic biology product—could generate unique by-products and cell densities that transform oxygen-transfer characteristics.

Key Takeaways for Biotech Scientists

Biotech scientists can adopt several immediate practices to reduce scale-up surprises tied to oxygen transfer:

• Measure and Monitor:
  • Treat kLa as a dynamic parameter, recognizing that it shifts during the run as cells grow and media conditions change.
  • Use multiple DO sensors or off-gas analysis for a more holistic view of larger vessels.
• Design Downscale Studies Thoughtfully:
  • Simulate large-scale conditions by introducing variable mixing or microfluidic-based fluctuation patterns.
  • Use complex media or at least account for by-products and foaming agents.
• Use CFD Where Possible:
  • Invest time in computational modeling to identify potential problem areas.
  • Rely on empirical correlations or partial modeling only when computational resources are limited, considering the reduced predictive power.
• Document Process Additives and Conditions:
  • Track how much antifoam, acid/base, or other additives go into the broth.
  • A thorough process record ensures repeatability and helps interpret anomalies in oxygen supply or cell behavior.
• Balance Technology with Basics:
  • Novel sensor systems, advanced AI algorithms, and microfluidic platforms are valuable but cannot replace a fundamental understanding of kLa.
  • Evaluate your cost, complexity, and development timelines before fully committing to high-end solutions.

Final Remarks

Oxygen transfer shapes much of an aerobic bioprocess's success or failure. Despite the availability of refined lab reactors, single-use vessels, and high-throughput screening systems, bridging the gap to large industrial fermenters still poses formidable technical challenges. 

Mass transfer experts like Lars Puiman and Rik Volger stress that the first step is a solid grasp of core engineering principles: bubble size, stirring speed, broth composition, and vessel design converge to form a unique kLa profile. From there, biotech scientists can deploy computational fluid dynamics, multi-sensor approaches, or well-planned downscaling experiments to proactively address oxygen-limitation pitfalls.

Looking to the future, advanced computational models and data-driven automation promise to refine mass transfer control further. Yet, as Puiman and Volger underscore, "back to the basics" remains the best starting point—measuring real kLa under realistic conditions, tracking fluctuations in industrial processes, and never assuming that a single DO probe or water-based test can represent a true large-scale fermentation. 

By systematically tailoring oxygen transfer strategies, the biotech community can more efficiently bring life-saving therapies and innovative products to market while minimizing the expensive, time-consuming missteps that have long plagued commercial scale-ups.

About 

Lars Puiman is a mass transfer and fermentation scale-up expert, postdoctoral researcher at the University of Queensland, and former PhD Student at the Delft University of Technology.

Rik Volger is a computational biotechnologist, mass transfer expert, and PhD student at the Delft University of Technology.

David Brühlmann is a strategic advisor who helps C-level biotech leaders reduce development and manufacturing costs to make life-saving therapies accessible to more patients worldwide.

He is also a biotech technology innovation coach, technology transfer leader, and host of the Smart Biotech Scientist podcast—the go-to podcast for biotech scientists who want to master biopharma CMC development and biomanufacturing.  

Hear It From The Horse’s Mouth

Want to listen to the full interview? Go to Smart Biotech Scientist Podcast. 

Want to hear more? Do visit the podcast page and check out other episodes. 
Do you wish to simplify your biologics drug development project? Contact Us