January 1, 2018 (Vol. 38, No. 1)

In-Line Dissolved Carbon Dioxide Measurement in Bioproduction

Dissolved carbon dioxide (DCO2) is a critical measurement in bioprocessing. DCO2 concentration plays an important role in many chemical and biological processes. The DCO2 concentration impacts fluxes of key metabolic pathways as well as extracellular and intracellular pH, ultimately influencing growth, productivity, and product quality (Table 1).

There are a number of mechanisms that explain why CO2 is critical in upstream processing.

1) Cell membranes are unable to modulate the in­tracellular CO2 concentration because of the physiochemical properties of CO2. Cells employ many mechanisms to control their intracellular envi­ronment; the cell membrane is an important selective barrier for controlling the intracellular environment. Embedded membrane proteins actively transport large and charged molecules, such as glucose and potas­sium, across the cell membrane. These active trans­ports allow cells to selectively transport molecules in and out of the cell. Although the cell membrane modulates large and charged molecules, it is un­able to modulate small, hydrophilic, uncharged molecules, such as CO2, O2, and N2.

How rapidly CO2, O2, and N2 diffuse across the cell membrane depends on the concentration gradient and on the molecule’s solubility in the boundary layers. The diffusion path into the cell consists of two rate­limiting boundary layers: diffusion from the extracel­lular environment into the lipid bilayer and diffusion from the lipid bilayer into the intracellular environment (Figure 1).

There are many reasons for measuring DCO2 in bioprocessing, and controlling DCO2 during manufacturing is critical for process control and optimization.

High Solubility

The solubility of CO2 in water and oil (similar proper­ties to the lipid bilayer) is high: 20-times that of O2 in water and oil. Since CO2 is highly soluble in the lipid bilayer, this amplifies how rapidly the extracellular CO2 concentration influences the intracellular CO2 concentration; CO2 diffuses rapidly across the cell membrane with little resistance at boundary layers.

2) Dissolved CO2 concentration impacts intracellular and extracellular pH. CO2 reacts with water to form carbonic acid which dis­sociates to produce a hydrogen ion resulting in a pH reduction (see Equation). Intracellularly, this reaction can also be modulated enzymatically. Intracellular and extracellular pH have a significant impact on physical and chemical reactions, impacting metabolism, mol­ecule stability, etc.


The pH-CO2 relationship is very significant in bicar­bonate buffer systems. To control extracellular pH many processes use acids and bases, and to dampen the pH-CO2 relationship many process employ organic buf­fer systems. Even if using an organic buffer system, it is important to limit the use of acids and bases because they can have detrimental effects on culture health either through increased osmolality or toxicity of the chemicals themselves.

Although the extracellular environment can employ organic buffer systems with low pH-CO2 dependency, the intracellular environment can only employ a bi­carbonate buffer system and, therefore, intracellular pH is highly CO2 dependent. Cells expend energy to control intracellular pH either by operating proton pumps or synthesizing enzymes, but despite all the controls a cell has to influence pH, CO2 permeates through the cell wall rapidly and uncontested, influencing the intracellular pH.

This emphasizes the importance of extracellular DCO2 because it has a strong influence on intracellular pH (CO2 acts as a biological Trojan horse). Controlling extracellular dissolved CO2 is one of the few control levers available for influencing intracel­lular pH.

3) Dissolved CO2 is an important metabolite produced and consumed in numerous metabolic reactions. Dissolved CO2 levels can significantly impact overall metabolic reaction fluxes key for growth and productivity. All cells, anaerobic or aerobic, produce and consume DCO2 within many metabolic networks, even though the network as a whole may result in a net gain.

CO2 is a highly active metabolite, involved in more than 200 different meta­bolic reactions: synthesis of amino acids, nucleotides, lipids, fatty acids, cholesterols, etc.

Figure 1. CO2 diffusion across a cell membrane.

Reversible and Diffusion Driven

Many of these reactions, which occur in series, are reversible and diffusion driven. If the CO2 concentration is too high (toxic) or too low (reaction limiting), the CO2 producing or consuming reaction impacts the overall metabolic flux. Metabolism is the engine that drives critical performance and quality attributes that we want to control and understand. Too low can limit synthesis of oxalo­acetate, limiting growth and protein synthesis. Too high can have detrimental effects on proliferations, explained by mitochondrial dysfunction.

The questions arises as to why measure CO2 in-line in the liquid phase? There are three common modes for measuring CO2: off-line liquid sample mea­surement (e.g., BGA), in-line off-gas mea­surement (e.g., mass spec), or in-line liquid mea­surement (e.g.,
lnPro 5000i).

There are a number of disadvantages to using common methods for measuring DCO2 versus the advantages of relying on an in-line technique (Table 2). Measuring in-line pro­vides a continuous measurement for process understanding and control without the very high sample error and error variance from sampling. In-line off-gas analysis can be influenced by changes in the liquid’s solubility (which does change throughout a process due to the many dynamics of a cell culture process) and by changes in the kLa, resulting in a measurement that is not indicative of the dissolved CO2 concentration.

Additionally, the CO2 concentra­tion in the off-gas is low, usually because other gasses such as air and are present in much higher volumetric flowrates, this highly dilutes the CO2 in the off-gas available for measurement.

There are several disadvantages to using conventional methods or measuring dissolved CO2. These methods are typically used in legacy processes that were established prior to reliable in-line technology.

Available Solution

Dissolved CO2 measurement based on the Severinghaus principle is common and previously referenced in the biopharmaceutical industry especially with off-line analyzers such as BGAs. Attempts to apply this principle for inline sensing have previously fallen short. Mettler Toledo’s expertise in glass pH electrode design and construction has eliminated this constraint.

The result is an in-line CO2 sensor for bioprocessing from the bench to manufacturing scale. The design of the InPro 5000i DCO2 sensor is based on the same potentiometric carbon dioxide Severinghaus electrode as used in BGAs, and measures the exact same DCO2 concentration as the microorganisms’ experience. Errors from sampling are eliminated and metabolic changes in DCO2 concentration affecting the biological system dynamics will be reflected in the measurement.

The InPro 5000i (Figure 2) utilizes a bicarbonate buffer system separated from the bioreactor medium by a patented CO2-selective permeable membrane. The DCO2 from the medium diffuses through the membrane until it reaches equilibrium with the buffer solution. A change in the CO2 partial pressure results in a pH change in the electrolyte, which is measured by the inner body pH electrode. The pH value in the buffer system is directly related to the partial pressure of DCO2 present in the media.

Figure 2. Pictured here is an InPro 5000i sensor.

Advanced Sensor Diagnostics

The InPro 5000i features the Intelligent Sensor Management (ISM®) digital technology. ISM offers a number of features that are not possible on analog probes, including a robust digital signal between sensor and transmitter that remains stable over long cable runs and is unaffected by interference from surrounding equipment. Advanced predictive diagnostics calculate when sensor maintenance will need to be performed: preventing a failing sensor from being installed in a bioreactor.

ISM sensors retain their own calibration data, which allows them to be calibrated away from the process in a more convenient location.

Joe Lattari ([email protected]) is U.S. product manager, process analytics, at Mettler Toledo.

Previous articleObesity Leaves Lasting Impression on Blood-Forming Stem Cells
Next articleAdvanced Image of A2aAR Protein Could Lead to Better Drug Design Methods