Optical in-situ glucose sensor
Steve Taranovich - October 2, 2012
A robust, real time, solid state, in-situ optical sensor has been developed by Stratophase Ltd. that tracks glucose concentration. The sensor has been tested and evaluated successfully for glucose monitoring and control in microbial fermentation providing a critical link in implementing real time feeding control of bioreactors to improve product quality, yield, and facility capacity. Evaluations of the sensor for feedstock control are currently underway in fungal and mammalian cell culture.
The Stratophase sensor utilizes an array of fiber optic Bragg gratings. In a simple fiber Bragg grating, the refractive index of the fiber varies periodically along the length of the fiber Light of a wavelength equivalent to the grating wavelength, Λ, will be reflected toward the source while all other wavelengths will be transmitted.
Schematic diagram of a fiber Bragg grating
When the core of the fiber is exposed to the environment (in the current example application a bioreactor) the source light beam interacts with the molecular layer in contact with the core thereby effectively changing the refractive index response of the grating In short, a Bragg grating can be an effective, miniature, solid state refractive index monitor.
The Stratophase multi element Bragg grating array. Dimensions – 6 x 20 x 1 mm Sensor window 1 x 5 mm
The temperature sensing grating is a critical component of the sensor. It is not in contact with the fluid medium under measurement. It provides real time, sensitive temperature correction of the signal generated by the sensor thus eliminating a major disadvantage of other analytical techniques that rely on refractive index.
The Stratophase probe can be deployed in 3 real time bioprocess control models:
- Golden Batch – Refractive Index monitors chemical composition, making it ideally suited to providing an overview of the chemical changes which occur during a bioprocess. By monitoring a set of fermentations a statistical model is built which represents the Golden Batch., Subsequent processes can be compared in real time. Key points from the process such as maximum rate, point of process completion or significant deviation from the expected trend are provided to the operator. The use of this simple technique during process development, scale-up and production offers increased name plate capacity and greater process flexibility and transferability.
- Controlled Fed Batch - Stratophase’s in-situ probe offers the ability to control the frequency of nutrient addition in real time. Bioreactor control systems are currently capable of in-situ monitoring and closed-loop control of a number of process critical parameters (e.g. pH, DO, etc.). However, feed-rate control to maintain desired nutrient concentration is not automatically controlled by reactor systems. Where nutrient concentration control is required periodic sampling and offline measurements are used, and feed volume is subsequently adjusted. Real time monitoring of the media allows relatively low volume pulses and high frequency feeding events enabling controlled pseudo-continuous feeding. Monitoring metabolic rate enables closed loop control of feeding to maximize nutrient up-take and resultant product evolution whilst ensuring batch-to-batch process quality.
- Predictive Model Control– Multi-variant statistical analysis is a technique for generating process models by taking into account data from multiple sources, both in-situ probes and off line analytical tools. When the model is subsequently applied in real time, it is possible to use the data from the in-situ probes to predict the results which would be expected from the off line analytical tools. Stratophase’s probe data gives a unique overview of all chemical changes which occur within the process enabling more specific and accurate models.
The Stratophase system is fully automated, with simple to use hardware and easy to interpret data regardless of the mode in which it is used. The in-situ probes take readings every 2 seconds which are transmitted back to the management unit for data reduction. This provides real time information output at a rate selectable by the user (typically 30secs), enabling instant decisions to be made by the process control system to optimize the quality and yield of the product.