On-line Fermentation Monitoring Research Articles

Domain Adaptation and Federated Learning for Ultrasonic Monitoring of Beer Fermentation

Beer fermentation processes are traditionally monitored through sampling and off-line wort density measurements. In-line and on-line sensors would provide real-time data on the fermentation progress whilst minimising human involvement, enabling identification of lagging fermentations or prediction of ethanol production end points. Ultrasonic sensors have previously been used for in-line and on-line fermentation monitoring and are increasingly being combined with machine learning models to interpret the sensor measurements. However, fermentation processes typically last many days and so impose a significant time investment to collect data from a sufficient number of batches for machine learning model training. This expenditure of effort must be multiplied if different fermentation processes must be monitored, such as varying formulations in craft breweries. In this work, three methodologies are evaluated to use previously collected ultrasonic sensor data from laboratory scale fermentations to improve machine learning model accuracy on an industrial scale fermentation process. These methodologies include training models on both domains simultaneously, training models in a federated learning strategy to preserve data privacy, and fine-tuning the best performing models on the industrial scale data. All methodologies provided increased prediction accuracy compared with training based solely on the industrial fermentation data. The federated learning methodology performed best, achieving higher accuracy for 14 out of 16 machine learning tasks compared with the base case model.

Automated Electrochemical Glucose Biosensor Platform as an Efficient Tool Toward On-Line Fermentation Monitoring: Novel Application Approaches and Insights.

Monitoring and control of fermentation processes remain a crucial challenge for both laboratory and industrial-scale experiments. Reliable identification and quantification of the key process parameters in on-line mode allow operation of the fermentation at optimal reactor efficiency, maximizing productivity while minimizing waste. However, state-of-the-art fermentation on-line monitoring is still limited to a number of standard measurements such as pH, temperature and dissolved oxygen, as well as off-gas analysis as an advanced possibility. Despite the availability of commercial biosensor-based platforms that have been established for continuous monitoring of glucose and various biological variables within healthcare, on-line glucose quantification in fermentation processes has not been implemented yet to a large degree. For the first time, this work presents a complete study of a commercial flow-through-cell with integrated electrochemical glucose biosensors (1st generation) applied in different media, and importantly, at- and on-line during a yeast fed-batch fermentation process. Remarkably, the glucose biosensor–based platform combined with the developed methodology was able to detect glucose concentrations up to 150 mM in the complex fermentation broth, on both cell-free and cell-containing samples, when not compromised by oxygen limitations. This is four to six-fold higher than previously described in the literature presenting the application of biosensors predominately toward cell-free fermentation samples. The automated biosensor platform allowed reliable glucose quantification in a significantly less resource and time (<5 min) consuming manner compared to conventional HPLC analysis with a refractive index (RI) detector performed as reference measurement. Moreover, the presented biosensor platform demonstrated outstanding mechanical stability in direct contact with the fermentation medium and accurate glucose quantification in the presence of various electroactive species. Coupled with the developed methodology it can be readily considered as a simple, robust, accurate and inexpensive tool for real-time glucose monitoring in fermentation processes.

The application of near-infrared spectroscopy in beer fermentation for online monitoring of critical process parameters and their integration into a novel feedforward control strategy

Traditional methods used in the analysis of fermentation media suffer from a number of limitations. The search for more rapid and efficient methods has led to the development and application of near-infrared spectroscopy. Near-infrared spectroscopy has been applied successfully in a variety of industrial processes: agricultural, food, chemical and pharmaceutical, generally in the areas of raw material quality control but also including intermediate and finished product testing. The present research explores its potential for online fermentation monitoring of total cell count (TCC), specific gravity (SG), free amino nitrogen (FAN) and percentage alcohol by volume (% v v−1) in a 300 L pilot-scale validation batch. Models that were generated from three calibration batches for each of these constituents exhibited overall favourable standard error of cross validation (SECV) and fit of predicted vs actual cross validated results (SECV, R2): SG (0.00072, 0.995), ethanol (0.17% v v−1, 0.990), FAN (16.5 mg L−1, 0.886) and TCC (1.24 × 106 cells mL−1, 0.640). The data that was most relevant to cell metabolism was determined to be sugar consumption rate, ethanol production rate, yield of ethanol and fermentation lag time. These ‘critical performance parameters’ were incorporated into a novel feed-forward control strategy where yeast pitching rate was modified based on values of the critical performance parameters from the previous batch. Use of this feed-forward strategy demonstrated how brewers can utilize near-infrared monitoring for quality assurance through early detection of shifts in fermentation performance. Copyright © 2017 The Institute of Brewing & Distilling

Evaluation of an online fermentation monitoring system

The need to introduce promising bioethanol production technologies calls for advanced laboratory techniques to study experiment designs and to obtain their results in a quick and reliable way. Real time monitoring based on general principles of ethanol fermentation, such as effluent CO2 volume, avoids time consuming steps, long lasting analyses and delivers information about the process directly. A device based on the above features and capable for real time monitoring on parallel channels was developed by the authors and is described in this paper. Both for calibration and for fermentation, test runs were carried out on different days and channels. Statistical evaluation was based on the obtained data. According to the t-test (P=0.05) and Grubbs analysis, the calibration method is reliable regardless of the date of calibration. When evaluating the fermentation results by ANCOVA acceptable standard derivations were obtained as impact of channel (58.8 ml), date (82.1 ml) and incorporating all impacts (116....

On Chemical and Ultrasonic Strategies to Improve a Portable FT-IR ATR Process Analyzer for Online Fermentation Monitoring

A setup for online fermentation monitoring was tested with suspensions containing yeast cells. A flow cell was equipped with a horizontal attenuated total reflection (ATR) unit for measurements of mid-infrared spectra. The stopped flow principle was employed to separately assess the liquid and the dispersed cells: dissolved components in the supernatant can be assessed while pumping the fermentation broth through the flow cell. Upon stopping the flow the culture can be monitored as the microorganisms settle onto the horizontal ATR diamond. Due to the surface sensitivity of ATR spectroscopy cleanness of the optical element is of particular importance. For yeast fermentations the formation of biofilms on the ATR surface was identified as limitation in regard to long-term stability. Initial experiments showed that the effective removal of residues was impossible by rinsing with water or a NaHCO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> solution. Therefore, various cleaning agents (2%) have been tested for their ability to clear off the biofilm. The problem of biofilm formation was additionally addressed by the exploitation of forces exerted on suspended particles within an ultrasonic standing wave (USW). The USW ( ~2 MHz) was built up between the ATR element and an ultrasonic transducer facing it. This technique of ultrasonic particle manipulation was applied to actively lift the sedimented material from the ATR after the infrared measurement, thus have the rinsing stream carry it away more effectively. Among the studied reagents surfactants and oxidizing agents showed to be most effective, the use of an USW was found to be feasible to remove a biofilm.

1H NMR Quantitative Assessment of Lactic Acid Produced by Biofermentation of Cane Sugar Juice

ABSTRACT The demand for lactic acid as a raw material for the chemical industry is on the rise. Various lactic acid bacteria (LAB) have been tested for the fermentation of sugar-cane juice and molasses into lactic acid. Industrially, the most commonly used producer strains are from the Lactobacillus genus. We report here on the fermentation of sugar-cane juice and syrup into L-lactic acid using Lactobacillus delbrueckii subsp delbrueckii (NCIM 2365). We present a reliable method to quantitatively assess the amount of L-lactic acid produced during fermentation using 1H NMR spectroscopy. This analytical method offers various advantages over existing ones and can be extended to the on-line monitoring of fermentation of sugars into lactic acid in industry.

Control of alcoholic fermentation in winemaking: Current situation and prospect

Abstract Winemaking fermentations aim to optimise product quality, which is very difficult to quantify. Control over fermentation kinetics generally has direct technological advantages, in terms of tank use optimisation in the winery and control over energy expenses for the regulation of temperature and is generally a prerequisite for controlling the characteristics of the wine. How new technologies can assist the winemakers to elaborate wines with predefined characteristics and improve process performances as well as hygienic properties? On-line fermentation monitoring is one promising way because: (i) it is much more accurate than manual measurements and (ii) it makes possible new control strategies with an optimal control tank by tank. Later, preventive control, based on modelling of the fermentation process, may be possible. New processes, such as cell immobilisation, may also be used as well as methodologies in the domain of data management and knowledge based systems.

Easy Ambient Sonic-Spray Ionization-Membrane Interface Mass Spectrometry for Direct Analysis of Solution Constituents

Using a cellulose dialysis membrane and aqueous solutions of common drugs as a proof-of-principle example, we demonstrate that solid but permeable and flexible membranes can be used as interfaces for the direct analysis of solution constituents via easy ambient sonic-spray ionization mass spectrometry. This new combination of MS techniques, herein termed EASI-MIMS, promotes droplet pick up of the analyte from the external surface of the membrane from where the analyte has selectively permeated for proper mass spectrometry characterization and quantitation. Possible application of EASI-MIMS such as the environmental analyses of effluents, on-line monitoring of fermentation and biotransformations and on-line pharmacokinetic blood analysis are discussed.

An On‐Line Approach To Monitor Ethanol Fermentation Using FTIR Spectroscopy

Fermentation process control is currently limited by its inability to measure parameters such as substrate, product, and biomass concentrations rapidly for consistent on-line feedback. Physical and chemical parameters, such as temperature and pH, currently can be obtained on-line using appropriate sensors. However, to obtain information on the concentration of the substrate, product, and biomass, samples must be taken off-line for measurement. With the use of spectroscopic techniques, real-time monitoring of process constituents such as product and substrate is possible. Spectroscopic techniques are rapid and nondestructive, require minimal or no sample preparation, and can be used to simultaneously assess several constituents in complex matrices. The production of ethanol is the largest fermentation process in terms of production volume and economic value as a result of its prominence in the food, agricultural, and fuel industries. This study attempts to develop an on-line ethanol fermentation monitoring technique using Fourier transform infrared (FTIR) spectroscopy with a flow-through ATR capability. Models developed using multivariate statistics, employed to obtain on-line FTIR measurements, were successfully validated by off-line HPLC analysis and spectrophotometry data. Standard errors of prediction (SEP) values of 0.985 g/L (R2 = 0.996), 1.386 g/L (R2 = 0.998), and 0.546 (R2 = 0.972) were obtained for ethanol, glucose, and OD, respectively. This work demonstrates that FTIR spectroscopy could be used for rapid on-line monitoring of fermentation.

Prediction of organic acids and other quality parameters of wine vinegar by near-infrared spectroscopy. A feasibility study

A method for quality control of wine vinegar has been developed based on near-infrared spectroscopy that can be applied to monitor processes or in quality control in vinegar industries. In vinegar production, stages such as on-line fermentation monitoring, dilution process and quality control of the final product are relevant and necessary. The aim of this study was to achieve and simplify these manufacturing stages through the determination of 14 parameters from a spectral measurement. Total acids, non-volatile and volatile acids, organic acids, l-proline, solids, ash and chloride in vinegar were determined by near-infrared spectroscopy. Different data pre-processing methods such as orthogonal signal correction, multiplicative scatter correction or column centering were used and the final multivariate methods were evaluated and compared. The calibration models were validated and prediction capacity studied in order to examine their practical applicability to the vinegar industry.

On-line fermentation monitoring by mid-infrared spectroscopy.

A new method for on-line monitoring of fermentations using mid-infrared (MIR) spectroscopy has been developed. The method has been used to predict the concentrations of glucose and ethanol during a baker's yeast fermentations. A completely automated flow system was employed as an interface between the bioprocess under study and the Fourier transform infrared (FT-IR) spectrometer, which was equipped with a flow cell housing a diamond attenuated total reflection (ATR) element. By using the automated flow system, experimental problems related to adherence of CO(2) bubbles to the ATR surface, as well as formation of biofilms on the ATR surface, could be efficiently eliminated. Gas bubbles were removed during sampling, and by using rinsing steps any biofilm could be removed from the ATR surface. In this way, constant measuring conditions could be guaranteed throughout prolonged fermentation times (approximately 8 h). As a reference method, high-performance liquid chromatography (HPLC) with refractive index detection was used. The recorded data from different fermentations were modeled by partial least-squares (PLS) regression comparing two different strategies for the calibration. On the one hand, calibration sets were constructed from spectra recorded from either synthetic standards or from samples drawn during fermentation. On the other hand, spectra from fermentation samples and synthetic standards were combined to form a calibration set. Differences in the kinetics of the studied fermentation processes used for calibration and prediction, as well as the precision of the HPLC reference method, were identified as the main chemometric sources of error. The optimal PLS regression method was obtained using the mixed calibration set of samples from fermentations and synthetic standards. The root mean square errors of prediction in this case were 0.267 and 0.336 g/L for glucose and ethanol concentration, respectively.

Monitoring of yeast fermentation by ion mobility spectrometry measurement and data visualisation with Self-Organizing Maps

Ion mobility spectrometry (IMS) measurement combined with unsupervised neurocomputing is considered as a new potential method for on-line monitoring of fermentation and other processes producing volatile compounds that involve micro-organisms. This was demonstrated in a model system in which a strain of brewer’s yeast ( Saccharomyces cerevisiae) was cultivated in a bench-top fermenter. Five phases of yeast growth could be detected from measurements of the exhaust gases from the fermenter, as indicated by the changes in ion mobility spectra analysed by computational methods. The data were first processed using the Self-Organizing Map (SOM) algorithm, the results showing that the phases of fermentation can be detected and identified. The cultivations were also shown by Sammon’s mapping to be comparable to a certain level of accuracy. Contaminated cultivation could be detected by its distinctive ion mobility spectrometry profile.

Dual excitation ratiometric fluorescent pH sensor for noninvasive bioprocess monitoring: development and application.

The development and application of a fluorescent excitation-ratiometric, noninvasive pH sensor for continuous on-line fermentation monitoring is presented. The ratiometric approach is robust and insensitive to factors such as source intensity, photobleaching, or orientation of the patch, and since measurements can be made with external instrumentation and without direct contact with the patch, detection is completely noninvasive. The fluorescent dye 8-hydroxy-1,3,6-pyrene trisulfonic acid was immobilized onto Dowex strongly basic anion-exchange resin, which was subsequently entrapped into a proton-permeable hydrogel layer. The sensor layer was polymerized directly onto a white microfiltration membrane backing that provided an optical barrier to the fluorescence and scatter of the fermentation medium. The ratio of emission intensity at 515 nm excited at 468 nm to that excited at 408 nm correlated well with the pH of clear buffers, over the pH range of 6-9. The sensor responded rapidly (<9 min) and reversibly to changes in the solution pH with high precision. The sterilizable HPTS sensor was used for on-line pH monitoring of an E. coli fermentation. The output from the indwelling sensor patch was always in good agreement with the pH recorded off-line with an ISFET probe, with a maximum discrepancy of 0.05 pH units. The sensor is easily adaptable to closed-loop feedback control systems.

The use of ultrafiltration and column liquid chromatography for on-line fermentation monitoring

The applicability of ultrafiltration and column liquid chromatography for the on-line and automated monitoring of small organic compounds during fermentation processes is reviewed. The influence of the construction of the filtration device, the properties of the membrane, the type of sample and the filtration mode on the sampling efficiency is discussed. In addition, different approaches for determining key components in fermentation samples, including the possibilities to increase the sensitivity and/or selectivity of the analytical procedure, are addressed. To illustrate the usefulness of employing online systems, the monitoring of a number of quite different fermentation processes is shown.

On-line monitoring of mammalian cell and yeast fermentations with a commercial biochemical analyzer

A commercial analyzer was tested for on-line monitoring of fermentations. A new sample block was constructed to effectively degas the fermentation broth. The robust analyzer accurately measured glucose up to 110 mmol/l and lactate up to 21 mmol/l at a frequency of 1 measurement per 2 minutes directly in suspensions of mammalian and yeast cells.

On-line analysis of Alcaligenes eutrophus fermentations

Improvements in the poly(β-hydroxybutyrate) (PHB) productivity of Alcaligenes eutrophus can be achieved by using genetically engineered strains. Knowledge of their metabolic regulation can be obtained from fermentation experiments. Certain events may not be noticed owing to the typically low frequency of sampling. To avoid information gaps, a newly developed fermenter-calorimeter was used for batch experiments allowing comprehensive on-line monitoring of fermentations. PHB-producing strain HI 6 and PHB-lacking mutant PHB−4 were selected, since they are well characterized with respect to the metabolism of PHB. For both strains, the profiles of thermograms, oxygen dissolved in the medium, and carbon dioxide formation were in good agreement with the off-line analysis. The on-line measurements showed characteristic differences between strains PHB−4 and H16. During fed-batch cultivations of PHB-accumulating strain H16, rapid autolysis of the cells was observed.Key words: Alcaligenes eutrophus, PHB, fermentation, calorimetry, on-line analysis, growth efficiency.

A Fiber-Optic Carbon Dioxide Sensor for Fermentation Monitoring

We have developed a fiber-optic chemical sensor for determining dissolved carbon dioxide and assessed its performance for the on-line monitoring of fermentation. The sensor operates on the Severinghaus pCO2 electrode principle; it consists of a pH sensitive dye (hydroxypyrenetrisulfonic acid, HPTS) in an HCO3− buffer solution entrapped in an expanded PTFE support held at the distal end of an optical fiber by a gas permeable membrane. CO2 crossing the membrane produces a pH change in the indicator solution. This change is related to the external CO2 concentration by the Henderson-Hasselbach equation. The sensor has a reversible working dissolved CO2 dynamic range of 0–0.25 atm. The sensor can be auto-claved without affecting its calibration. Results are presented for the on-line determination of CO2 production in beer fermentation.

Development of Enzymatic Autoanalyzers for Monitoring Biotechnological Processes

Original enzymatic methods with amperometnc and potentiometric detection were developed for biotechnological industry. They provide selective determination of various substrate concentrations and can avoid the influence of the buffer capacity of the fermentation broth on the analysis result. Highly stable native enzyme reagents and immobilized enzymes were developed for the methods realization. New enzymatic autoanalyzers based on the methods are proposed for laboratory and on-line monitoring of fermentation processes.The specifications and data on the testing of the autoanalyzers in real fermentation processes are presented.

Fiber-optic Sensor for Continuous Monitoring of Fermentation pH

We have developed a fiber-optic chemical sensor for on-line monitoring of fermentation pH. The sensor is based on a covalently bound fluorescent dye immobilized within a water-permeable polymer layer on an optical fiber. Measurements were performed on a portable fluorimeter and employed a ratiometric approach to account for system instabilities. We show that the use of this fiber-optic sensor provides fast, accurate and reliable measurements during E. coli fermentation in a complex medium.

A flow injection analysis system for fermentation monitoring and control

An automated analysis system for on-line fermentation monitoring is presented. The modular system consists of an in-line sterilizeable crossflow microfilter, a selection valve that allows injection of sample or standards, a degassing unit, a dilution module, and a FIA manifold with a spectrophotometric UV/VIS detector. In the dilution module samples are conditioned and diluted depending upon concentration of analyte and the working range of the analyzer. Methods for the monitoring of glucose, ethanol, ammonia and phosphate are described. Results from the monitoring of glucose and their use in fermentation control are presented. The maximal analysis frequency is 30 samples per hour including the dilution of 1 : 200. Detection limits are 5 mg/L for ethanol and glucose, 1 mg/L for phosphate and 50 mg/L for ammonia.