WO2014087374A1 - Procédé et sonde de surveillance de la fermentation alcoolique avec spectroscopie uv-vis-swnir - Google Patents

Procédé et sonde de surveillance de la fermentation alcoolique avec spectroscopie uv-vis-swnir Download PDF

Info

Publication number
WO2014087374A1
WO2014087374A1 PCT/IB2013/060677 IB2013060677W WO2014087374A1 WO 2014087374 A1 WO2014087374 A1 WO 2014087374A1 IB 2013060677 W IB2013060677 W IB 2013060677W WO 2014087374 A1 WO2014087374 A1 WO 2014087374A1
Authority
WO
WIPO (PCT)
Prior art keywords
fermentation
vis
swnir
spectra
probe
Prior art date
Application number
PCT/IB2013/060677
Other languages
English (en)
Portuguese (pt)
Inventor
Rui Miguel DA COSTA MARTINS
António Augusto MARTINS DE OLIVEIRA SOARES VICENTE
Luís Fernando DE SOUSA FERREIRA DA SILVA
Eurico Augusto RODRIGUES DE SEABRA
José António COUTO TEIXEIRA
Original Assignee
Unicer - Bebidas, S.A
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Unicer - Bebidas, S.A filed Critical Unicer - Bebidas, S.A
Publication of WO2014087374A1 publication Critical patent/WO2014087374A1/fr

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12G—WINE; PREPARATION THEREOF; ALCOHOLIC BEVERAGES; PREPARATION OF ALCOHOLIC BEVERAGES NOT PROVIDED FOR IN SUBCLASSES C12C OR C12H
    • C12G1/00—Preparation of wine or sparkling wine
    • C12G1/02—Preparation of must from grapes; Must treatment and fermentation
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12C—BEER; PREPARATION OF BEER BY FERMENTATION; PREPARATION OF MALT FOR MAKING BEER; PREPARATION OF HOPS FOR MAKING BEER
    • C12C11/00—Fermentation processes for beer
    • C12C11/003—Fermentation of beerwort
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12G—WINE; PREPARATION THEREOF; ALCOHOLIC BEVERAGES; PREPARATION OF ALCOHOLIC BEVERAGES NOT PROVIDED FOR IN SUBCLASSES C12C OR C12H
    • C12G3/00—Preparation of other alcoholic beverages
    • C12G3/02—Preparation of other alcoholic beverages by fermentation
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12M—APPARATUS FOR ENZYMOLOGY OR MICROBIOLOGY; APPARATUS FOR CULTURING MICROORGANISMS FOR PRODUCING BIOMASS, FOR GROWING CELLS OR FOR OBTAINING FERMENTATION OR METABOLIC PRODUCTS, i.e. BIOREACTORS OR FERMENTERS
    • C12M41/00—Means for regulation, monitoring, measurement or control, e.g. flow regulation
    • C12M41/48—Automatic or computerized control
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/85—Investigating moving fluids or granular solids
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/02—Food
    • G01N33/14—Beverages
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
    • G01N21/314—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry with comparison of measurements at specific and non-specific wavelengths
    • G01N2021/3155—Measuring in two spectral ranges, e.g. UV and visible
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2201/00—Features of devices classified in G01N21/00
    • G01N2201/12—Circuits of general importance; Signal processing
    • G01N2201/129—Using chemometrical methods

Definitions

  • This invention relates to the monitoring and diagnosis of yeast metabolism allowing the quality management of beverages and fermented products produced based on the spectral information collected by the spectroscopy probe and analysis system.
  • This invention describes a system and method for monitoring beer fermentation based on UV-VIS-SWNIR (short-wave near-infrared / visible / near-infrared) spectroscopy, and how to obtain and process information on the physicochemical characteristics of musts and beer from this data.
  • This method is characterized in that in a preferred embodiment it comprises the following steps:
  • UV-VIS-SWNIR short-wave near-infrared / visible / near-infrared
  • Yet another preferred embodiment of the present invention has the feature that, based on spectral variance, further comprises one or more of the following steps: - non-directed chemical detection of yeast metabolic abnormalities by spectral information of the fermenting must;
  • UV-VIS-SWNIR ultraviolet-VIS-SWNIR (ultraviolet / visible / near near-infrared near-wavelength) spectrophotometer, in particular using wavelengths between 200 and 1200 nm;
  • a preferred embodiment of the present invention is further characterized by: light source, in particular halogen source;
  • Fiber optic probe for diffusive reflectance measurement adapted to brewer's wort measurements, namely by temperature measurement, cleaning system included, standardized connections for the factory environment.
  • control set-points are predefined and proportional to the processing, with automatic monitoring of temperature, optical density, pH and pressure as online monitoring parameters. These allow for some remotely processed automatic diagnostics on important factors affecting beer quality, such as the aroma, taste, color and foam consistency.
  • a bubble column is a very simple reactor as it has no moving parts and consists essentially of a reaction vessel. In the fermentation process, this dispenser is used in the initial aeration of the must. In the case of beer production, fermentation produces large CO2, generating an essentially vertical gas flow which in turn causes flow patterns to occur in the liquid. Relevant are both the mechanisms of liquid occlusion in the vortices caused by rising gas bubbles and the circulatory flows generated under most operating conditions. These mechanisms affect the flow patterns of the liquid within the bioreactor by influencing the type and degree of mixing that can be obtained.
  • Spectroscopy is a technique used in analytical chemistry for the identification and quantification of chemical compounds based on the principle that each type of molecule is associated with a specific frequency response (spectrum) when illuminated with a controlled light source.
  • UV-VIS-SWNIR short-wave near-infrared / visible / near-infrared
  • spectroscopy records the sample spectrum typically in the 200 to 1200 nm wavelength range. This range corresponds to the electronic transitions between energy levels of molecular orbitals and the transitions associated with molecular vibration.
  • Many organic molecules, which make up yeast metabolism, have groups of chromatophores (color) and fluorescence that enhance the application of this technology for identification and quantification of these compounds.
  • molecules indicative of fermentation status such as fructose, glucose, ethyl alcohol or acetaldehyde, have functional groups that can be recorded by UV-VIS-SWNIR spectroscopy.
  • Wavelength ranges between 700 nm and 1200 nm provide information on the overtones (harmonics) associated with the molecular vibration frequencies of metabolites such as water, proteins, lipids or carbohydrates.
  • UV-VIS-SWNIR spectroscopy Recent technological developments in UV-VIS-SWNIR spectroscopy have led to the emergence of miniaturized devices with high portability and high spectral resolution. These new devices have enabled the development of new measurement techniques that can be applied directly under the "in vivo" sample.
  • the joint development of new mathematical techniques for frequency signal analysis allows the use of spectroscopy for simultaneous monitoring of a high range of metabolites from a single spectrum.
  • the existing mathematical methods for spectra processing are based on the factorization of the spectra in a new base system (usually a subspace) to obtain a more interpretable representation and extract the parts of the spectrum with most relevant systemic information.
  • the methods used are grouped into two classes: a) latent variable methods for systematizing the information collected by the spectra. This class consists of mono-block decomposition techniques such as principal component analysis (PCA) or non-negative matrix factorization (NMF). b) latent variable methods for multivariate regression. This class includes multi-block decomposition techniques such as partial least squares regression (PLS), truncated total least squares, principal component regression (PCR) or supported vector machines ').
  • PCA principal component analysis
  • NMF non-negative matrix factorization
  • Spectroscopy equipment shall have the following characteristics: (a) permit the acquisition and recording in digital format of the frequency response (spectrum) of the sample reflectance, in particular in the 200 to 1200 nm wavelength range, using a adequate and calibrated light; (b) a device to warn the operator that he already has the necessary data at a given sampling point; and (c) have a communication interface enabling the equipment to be ordered to acquire a spectrum at a given time and to store this information in a file, eg. text in digital format, in particular with the following information: (i) the parameters of the relevant apparatus for defining the conditions for spectrum acquisition; and ii) a table with the numerical value of the wavelength and the value of the intensity of radiation detected at that wavelength.
  • mathematical models for the control charts should be provided.
  • the models should have the characteristics described in the previous section and should be implemented to provide in particular five operative functionalities: a) obtaining the latent variables from a vector containing a spectrum; b) obtaining an estimate for the spectrum given a vector containing the latent variables; c) obtaining the variance-covariance matrix of the latent variables; d) obtaining the percentile corresponding to a given significance level and referring to the statistical variable defined by the reconstruction squared error (sum of squares of the differences between the spectrum and the estimated spectrum from the latent spectrum variables); and, e) obtaining the percentile corresponding to a given level of significance and referring to the Hotelling T square statistic for the selected latent variables.
  • a mathematical model is preferably provided which reflects the corresponding control chart.
  • the system uses a mathematical model obtained based on principal component analysis with eg. two, latent variables of greater variance and applied to the average-centered spectra.
  • Each calibration mathematical model should have the characteristics described in the previous section and be implemented to provide five operative functionalities: a) obtaining the estimate for the metabolite concentration given a vector with a spectrum; b) obtaining the upper and lower limit for the confidence interval given a vector with a spectrum and a given confidence level; c) obtaining latent variables given a vector with a spectrum; d) obtaining an estimate for the spectrum given a vector containing the latent variables; and e) obtaining the percentile corresponding to a given level of significance and referring to the statistical variable defined by the reconstruction squared error (sum of squares of the differences between the spectrum and the estimated spectrum from the latent spectrum variables).
  • This invention provides a method for monitoring and supporting beer fermentation management based on UV-VIS-SWNIR spectroscopy.
  • the method is structured in three parts: i) data collection and validation; ii) mathematical treatment of the collected data; iii) presentation of information for monitoring and support to production management;
  • the conditions and parameters necessary for the operation of the spectroscopy data collection equipment should be optimized prior to the data collection procedure and kept unchanged throughout the procedure. These conditions and parameters should be optimized for the different types of musts, keeping the signal within the optical receiver's linear signal pickup range, ie above the lower sensitivity limit and below the sensor's maximum saturation limit, thus obtaining different spectra for different operating conditions.
  • the number of spectra that need to be performed is set automatically by following the procedure: i) collecting a number, eg. 20 spectra; ii) in the obtained spectra, correct the signal strength and variance in the sampling procedure; (iii) perform a principal component analysis with all corrected and averaged spectra; iv) to verify the number of samples within the 95th percentile of the Hotelling T square and reconstruction square error statistics, for the principal component model with the, eg. two, latent variables of greater variance obtained in point iii); (v) if the number of spectra in the preceding point is no higher than a given number of spectrum threshold, eg. 17, then collect a further number of spectra, e.g. 10, spectra and return to point ii). If the total number of spectra collected exceeds a certain number of spectra, eg. 100, the spectrum acquisition equipment may be flagged for verification as there may be a malfunction.
  • This section describes the procedures involving the mathematical treatment of the collected spectra, addressing: i) the preprocessing of the spectra (see eg PT 104566, block p.3); (ii) the calculation of multivariate control parameters based on the model (s) provided (see eg PT 104566, block p.4); and (iii) the calculation of estimates for metabolite concentration, validity and prediction error, based on the calibration model (s) provided (see eg EN 104566, block p.5).
  • Preprocessing of the collected spectra see i) the preprocessing of the spectra (see eg PT 104566, block p.3); (ii) the calculation of multivariate control parameters based on the model (s) provided (see eg PT 104566, block p.4); and (iii) the calculation of estimates for metabolite concentration, validity and prediction error, based on the calibration model (s) provided (see eg EN 1045
  • each model of control chart will have all the necessary parameters to allow the visualization of the respective chart and the identification of sampling points with abnormal characteristics.
  • This section describes all procedures that involve interaction with the system operator or manager, addressing: i) anomaly detection (block o .2); (ii) presentation of information concerning multivariate control charts (block o.1); and (iii) presentation of information on calibration models (block o.3).
  • Anomaly detection depends on the results obtained in the mathematical treatment modules (block p.3, p.4, p.5).
  • This module implements various tests for detecting anomalous behavior as a way of: i) validating the entire data set collected; (ii) detect sampling times that have characteristics different from the nominal characteristics described by the control chart models; (iii) recommend recalibration of a given model of control chart; iv) detect which sampling times have different characteristics from the nominal characteristics described by the calibration models; (v) detect which sampling times a given calibration model should not be applied to; and vi) recommend recalibration of a given calibration model.
  • Detection of data pre-treatment anomalies This detection procedure cannot identify the nature of the cause (s) for the abnormal characteristics observed. These may be due to causes such as sudden change in the operation of the acquisition equipment (eg breakage of the optical fiber of the equipment or occlusion of the spectroscopy sensor) or sudden changes in the fermentation process. Therefore all sampling points detected as anomalous should be investigated by the system operator / manager to identify the cause (s).
  • the procedure of this operation is described in PT 104566, which is incorporated herein by reference, especially with regard to detecting anomalies resulting from data pretreatment.
  • control card models There are four tests for detecting anomalies based on control card models: (i) applicability test of control card models; (ii) testing the need for recalibration of control chart templates to be applied to each of the control charts provided to the system and whenever a new set of sampling points is processed; iii) test for determination of anomalous sampling points; and iv) testing for the determination of sampling points that are outside the control zone defined in the control chart template.
  • Tests involving sampling points are not unique and a given sampling point can be detected as anomalously independently by multiple tests.
  • PT 104566 The procedure for testing the applicability of a control card template is described in PT 104566, which is incorporated herein by reference.
  • the test procedure for detecting anomalous sampling points for a control chart model is also described in PT 104566, which is incorporated herein by reference.
  • Sampling points outside the control zone should be further analyzed to identify the cause (s) of the observed characteristics. It will be the responsibility of the system operator / manager to identify and further define the corrective actions to be taken in the areas defined by the sampling points outside the control zone.
  • test for anomaly detection based on calibration models There are four tests for anomaly detection based on calibration models: (i) applicability test of calibration models; (ii) testing the need for recalibration of calibration models to be applied to each of the calibrations provided to the system and whenever a new set of sampling points is processed; iii) test for validation of the predictions of a calibration model; and iv) testing for prediction reliability at a given sampling point.
  • Each test should be applied independently to each of the calibration templates provided to the system. Tests involving sampling points are not unique and a given sampling point can be detected as anomalously independently by multiple tests.
  • the reliability assessment of the calibration model estimate shall be defined by the model operator / manager based on these three criteria provided by the system and at each sampling point.
  • the temporal synchronization is performed according to the following procedure:
  • the R (i) series represents a reference fermentation process from the beginning - R (l) to the end - R (i m ax) f
  • P (j) represents the fermentation process to be controlled from the beginning to the moment current process.
  • V Use the regression model to predict the fermentative path of P (k + n) using as input the model R (k + n), where n is the number of points to be predicted;
  • the number of threads is set to minimize synchronization errors but avoid overfitting.
  • THE Said optimization can be achieved by any suitable method, such as local minimization methods, such as simulated annealing or one of several genetic algorithms.
  • Said synchronization may also be accomplished by the latent variables representative of the fermentation space, or by the fermentation control parameters and / or metabolite concentration.
  • the reference fermentation can be chosen manually or statistically as representative of fermentation histories. Another possibility is to make an average, median or other estimate of the central value of a set of reference fermentation histories. Yet another possibility is to synchronize individually with each of a set of reference fermentation histories and postprocess the results obtained to obtain more likely values and confidence intervals.
  • the following preferred embodiments are based on research carried out by the inventors on analyzes of alcoholic brewing fermentations.
  • the microorganism used during the fermentation process was Sacharomyces pastorianus under the following conditions: (i) A so-called universal probe was developed, comprising a fiber optic sensor for diffusive reflectance measurement, a cleaning system and a sensor for temperature. This probe makes it possible to tie and secure all sensor systems required for the monitoring and determination of fermentation status, as well as evaluating process deviations and beer production defects.
  • This universal measuring probe allows the collection of data necessary for, for example, to forecast fermentation times and to check the bioreactor wash cycles.
  • One of the main hygiene problems in the brewing industry is the formation of biofilms, causing production interruption, increased amount and frequency of cleaning agents, increased monitoring of process conditions or decreased beer quality.
  • the mooring system shown and described later in this document was designed, which is generally constructed of stainless steel, e.g. AISI-304, with seals, e.g. in EPDM and shutter e.g. in PTFE. a) Dimensions
  • This universal probe with a shape eg. cylindrical, eg. the clamping dimensions of 205 mm in length and 35 mm in diameter without the cleaning system connection racords eg. performed with CO2.
  • b) Preferred characteristics of materials used All components of the probe are eg. made of AISI 304 stainless steel, except for the following components: 6 (a conical plug) eg. PTFE specially designed for this application; 8 (two standard shutter guiding seals and CO2 seal, eg in EPDM), 9 (two standard seals for C02 pipe connection fittings, eg in EPDM), 11 (body seal mooring system) and 12 (optical probe hole seal), these last two seals in e.g. EPDM designed specifically and custom-made for this application.
  • 6 a conical plug
  • PTFE specially designed for this application
  • 8 two standard shutter guiding seals and CO2 seal, eg in EPDM
  • 9 two standard seals for C02 pipe connection fittings, eg in EPDM
  • the universal probe is comprised in a preferred embodiment of 12 types of components:
  • Figure 4 shows a drawing of the constructed system showing all 12 types of assembled components.
  • Fermentations according to the brewing process (a) Preparation of the must:
  • the fermentation temperature was 12 ° C
  • the initial concentration of yeast cells to begin fermentation was 16x106 cells / mL
  • the must type was always the same throughout all pilot scale fermentations. Fermentations began with the addition of the inoculum to the malt must previously placed inside the bioreactor and cooled to 12 ° C.
  • the inoculum was composed of Saccharomyces pastorianus (carsbergensis) (brewer's yeast) with a concentration of 12 to 16x10 6 cells / ml must.
  • the inoculum was prepared as follows: The microorganisms were incubated in must (previously aerated) under anaerobic conditions on a rotary shaker (120 rpm) and at a temperature of 27 ° C for 72 h. After this time, the yeast was collected from the wort by centrifugation (5000 rpm for 10 minutes) in a Sigma 4K15 centrifuge (Sigma, 2008). It was then diluted in NaCl solution (0.9% v / v) to a ratio of 4 ml / g yeast to give a paste.
  • the fermenter used was a cylindrical type, with a maximum volume of 1.9 L and a useful volume of 1.4 L.
  • a cooling coil was placed. involving the cylindrical part of the bioreactor, allowing to reach a complete fermentation in 5 days. For each sample taken during fermentation the following parameters were determined: i) number of yeast cells, ii) dry extract; iii) total sugar (percentage by weight), and iv) the pH.
  • the following at-line sampling plan was developed for the quantification of the following parameters and metabolites: cell number / Biomass (cfu / ml); brix grade; pH; alcoholic content (% v / v); dry extract (g / l); Er (%); And the (%); RDF (%); ADF (%), n-propanol (mg / l), isobutanol (mg / l); total amyl alcohols (mg / l); ethyl acetate (mg / l); amyl acetate (mg / l); acetaldehyde (mg / l); dimethyl sulfate (mg / l); diacetyl (mg / l).
  • the spectroscopy equipment used is a dispersive spectroscopy equipment, together with the universal probe, shown in Figures 2 and 3.
  • the operating parameters of the equipment were optimized to obtain the spectrum with a maximum signal-to-noise ratio. background noise for variations in reactor type, initial must and development of the fermentation process concerned.
  • a preferred embodiment comprises monitoring the metabolic profile - central metabolism. Data collected during the 16 fermentations were processed according to PT 104566, which is hereby incorporated by reference, and calibration models were developed for monitoring central yeast metabolism as well as for brewing quality parameters:
  • Table 2 shows that for both the UV-VIS and Vis-SWNIR spectral ranges the pilot scale calibrations are valid for the industrial scale since the mean correlation coefficients between the predicted data for the Industrial-scale validation lot is 0.95. This type of analysis demonstrates the applicability of industrially accurate monitoring of the above parameters and can be used in real time estimation to monitor, diagnose and control the fermentation process.
  • a preferred embodiment comprises monitoring the metabolic profile - Aromatic profile generated by alcohols and esters.
  • Data collected during the 16 fermentations were processed in accordance with PT 104566, which is hereby incorporated by reference, and calibration models were developed for monitoring central yeast metabolism as well as for aromatic quality parameters:
  • n-propanol (mg / l), iso-butanol (mg / l); total amyl alcohols (mg / l); ethyl acetate (mg / l); amyl acetate (mg / l) and acetaldehyde (mg / l).
  • the patent system can be used as a complementary method to the on-line laboratory method as well as to reduce the number of laboratory samples required. during the industrial process.
  • Table 4 demonstrates that for both UV-VIS and Vis-SWNIR spectral ranges the pilot scale calibrations are valid for the industrial scale since the average correlation coefficients between the predicted data for the industrial scale validation lot is of 0. 95 This type of analysis demonstrates the applicability of monitoring of parameters listed above with industrial precision, which may be their real-time estimation used for the monitoring, diagnosis and control of the fermentation process.
  • FIG. 7 shows the metabolic profiles of one of the fermentations recorded at-line and online prediction using the patent system. It is possible to observe that for all parameters the measured and predicted values are statistically similar (p ⁇ 0.001), allowing online and real-time measurement of central yeast metabolism: n-propanol, iso-butanol; total amyl alcohols; ethyl acetate; amyl acetate and acetaldehyde.
  • a preferred embodiment comprises monitoring using the fermentative spectral profile.
  • the electromagnetic spectrum recorded by the patent system is a spectral signature of the chemical and physical properties of must throughout the fermentation process.
  • the spectral surface shown in Figure 6 (c) shows the typical spectral pattern of an industrial scale fermentation (fermentations 12 and 13), which have a relatively similar profile to each other, given the strict procedural control applied on this scale. It is also possible to verify that for different fermentation phases different spectral profiles are observed, which correspond to different physicochemical properties of the must.
  • the band combination presented describes: i) the latency phase (period of acclimation to the fermentative medium before the beginning of cell growth) and initial sedimentation of particles; ii) exponential growth (primordial sugar depletion, pH lowering and ethanol formation); iii) phase of total consumption of the sugar source in the must (change in metabolism, flocculation, start of acetaldehyde consumption and formation of alcohols and esters due to catabolism); and iv) stabilization with the anabolic consumption of amino acids and metabolites resulting from cell death in older yeast populations.
  • Metabolic changes are recorded in the relationships between spectral bands along the fermentation spectral surface. Spectral variations chemically reflect the atomic and molecular orbitals of the compounds in solution, and their interference patterns are too complex to be visually interpreted, and latent variable-based mathematical models are required to interpret the observed patterns (see EN 104566).
  • Figure 8 shows the unsynchronized intensity value at 700 nm, demonstrating the kinetics of the data collected in: i) pilot scale bioreactor (fermentations 8 to 11); and ii) industrial scale bioreactor (fermentations 12 and 13). It is possible to observe (central metabolism - glycolysis and TCA) that pilot scale fermentations occur at higher rates, resulting in a shorter duration of each batch. For example, while a pilot scale fermentation takes an average of 200 hours, on an industrial scale we observe an average duration of 400 hours.
  • the graph also shows that all fermentations have common (standard) information that is proportional to the metabolites produced and consumed in the central metabolism, presenting kinetic particularities due to the different yeast cell states in the different operating conditions of each one. one the fermentations tested. We can then consider that the fermentations performed are not exactly the same, either in terms of dynamics or metabolic variability. This is one of the main difficulties of the fermentation process and the industry whose processing depends on alcoholic fermentation: how to diagnose, control and direct a fermentation process to a desired metabolic target at the end of fermentation.
  • the type profile of a fermentation represents the most characteristic spectral vector of a matrix in the latent variable space as presented in section 2.3.4.
  • Figure 8 (a) shows synchronized fermentations 8 to 12. Observing the figure shows that all fermentations follow the common pattern of metabolism. It is possible to synchronize the common dynamic behavior between different fermentations. As expected, it is possible to distinguish two distinct fermentative patterns between the pilot scale and the industrial scale.
  • the metrics of different fermentation zones can also be compared. For example, taking into consideration the carbon source consumption phase (up to 100 hours and 200 hours, pilot scale and industrial scale, respectively), this presents an average spectral correlation of 0.90 (p ⁇ 0.05), which allows us to conclude that the pilot scale process is statistically similar to the industrial scale process in terms of spectral signature up to this point of fermentation. The proceedings differ from this point.
  • a preferred embodiment comprises the multivariate fermentation process control charts based on the spectral profile.
  • Figure 8 a) demonstrates the synchronized processes 8 to 12, using moments (1) to (4) for compare the state of the fermentation process in relation to the standard production process. In this case study, it is considered that fermentation 12 follows the statistical specifications of the fermentation process pre-established in the production process of alcoholic fermentation.
  • the synchronization process between fermentations utilizes the statistical similarities of the spectral information pattern between the two processes. It can be seen from Figures 8 (b) to (e) that each moment has a high level of spectral similarity, that is, there is no a priori for any of the fermentations, a prominent spectral pattern.
  • Figure 8 (f) shows that at the time of analysis (1) it can be seen that fermentations 9 to 11 are within the T-Hotelling confidence limits 0.90 and 0.95, with fermentations 8 and 13 showing procedural deviations, respectively. It is possible to diagnose the cause of these deviations using the original graph, where fermentation 8 presents very high values of signal intensity, ie delay in the beginning of fermentation in relation to the standard process, due to the prolongation of the latency phase. Fermentation 13 has the opposite effect: a very early exponential phase precocious. In terms of fermentative control, the operator should raise the operating temperature in fermentation 8 and lower it in fermentation 10, taking into account new spectral values.
  • Fermentation 13 the dynamic one that required the greatest corrections, is also the one with the most pronounced deviations at the end of the fermentation. However, if no control were applied, this fermentation would be much further away and could be considered defective. The same kind of consideration applies to fermentation 8.
  • a preferred embodiment comprises multivariate fermentation process control charts based on the predicted composition.
  • the foregoing embodiment portrays the use of the spectral standard for fermentative diagnosis and applies corrective measures. However, it is not possible to know directly the metabolic cause of procedural deviation; For this, it is necessary to use the spectral processing and calibration models presented in the metabolic profile monitoring (aromatic profile generated by alcohols and esters) and the monitoring using the fermentative spectral profile, respectively.
  • Figure 9 presents the compositional predictions for moments (1) to (4) for the following parameters: pH; alcohol (% v / v); Er (%); And the (%); color; RDF (%); n-propanol (mg / l); iso-butanol (mg / l); amyl alcohol (mg / l); ethyl acetate (mg / l); acetaldehyde (mg / l); and diacetyl (mg / l).
  • moments (2) and (4) were considered as an exemplary embodiment, with the respective multivariate control charts and contribution to the deviations of the analyzed samples shown in Figure 10.
  • Figure 10 (a) demonstrates the control chart for momentum (2). It can be observed that at this time fermentations 11 and 13 are outside the established chemical parameters. In the case of fermentation 11 the failure can be diagnosed as attributed to the following variables: i) pH (4,265); ii) n-propanol (5.82 mg / l); iii) amyl alcohols (38.47 mg / l); ethyl acetate (5.24 mg / l). In the case of fermentation 13 the variables responsible for fermentation failure are ethyl acetate (38.0 mg / l) and amyl alcohols (28.0 mg / l) (see Figures 10 (b) and (c), respectively).
  • Fermentation 11 may be said to have below normal pH and n-propanol values and high amyl alcohol values; Fermentation 13 already has very high values of amyl alcohols and ethyl acetate, as it is at the moment (2) "advanced" in relation to the other fermentations. In these cases it is recommended to take samples for laboratory characterization in order to deepen the biochemical causes of these deviations.
  • Figure 10 (d) shows the control chart for moment (4) near the end of the fermentation process. It can be observed that at this time fermentations 9 and 11 show slight deviations from the expected composition. In the case of fermentation 9, this is due to the following variables: RDF (30.61%) and amyl alcohols (81.27 mg / l); In the case of fermentation 11, The main responsible for the deviation is the amyl alcohol content (119.5 mg / l).
  • a preferred embodiment comprises predictive diagnosis.
  • One of the biggest difficulties in the study of alcoholic fermentations is having the ability to predict the fermentation path, given the different operating conditions between processes.
  • the must and inoculum preparation procedures as well as the operating conditions being very manageable, which gives some dynamic stability in yeast behavior and therefore reproducibility of the yeast. processes (within certain limits).
  • the method presented in section 2.3.4 exemplified in the flow chart of Figure 11, was used.
  • the present system receives a time series of spectra. This series is used to be synchronized against the standard process time series of spectra in order to know in which fermentative region the current process is. Once synchronized, prediction can be made by the end of the regression fermentation against the standard process as explained in section 2.3.4. procedures described in the monitoring achievements using the fermentative spectral profile and multivariate control charts (based on the spectral profile or predicted composition) with the predicted data.
  • Figure 12 presents the prediction of the fermentation path based on the described method, using the spectral series based on the first 20, 40 and 60 sample points, respectively, for fermentation 13.
  • the results of the algorithm point to an average correlation between estimated data and data obtained from 0.97, thus being statistically similar (p ⁇ 0.05).
  • the results demonstrate (notably in Figs 12a-12i) that, since the fermentation process is within the expected fermentation pattern, it is possible to obtain a statistically robust prediction of the fermentative path, making it possible to perform monitoring achievements using the profile fermentative spectral and multivariate control charts (based on spectral profile or predicted composition) with predicted data so that predictive diagnosis can be made at the following points:
  • transmittance or reflectance spectra are equivalent, provided proper care is taken in adapting the transmitter and signal sensor, which varies depending on the signal quality to be captured and the parameters, including opacity, of the wort.
  • Figure 4 Universal probe assembly drawing (containing cleaning system and placement locations for optical sensor and temperature sensor).
  • Figure 5 Normalized chemical data obtained during two fermentations ( ⁇ Biomass x 2x108 cells / mL must; A Extract x 0.1382 g / mL must, pH; and total sugar, mass percentage in must x 13.5). Black lines represent fermentation 1; and gray lines represent fermentation 2.
  • Figure 6. Example of profile synchronization of the spectral profile of alcoholic fermentations: (a) unsynchronized fermentations at 700 nm; (b) synchronized fermentations at 700 nm; (c) spectral profile of unsynchronized fermentations 12 and 13; (d) Spectral profile of synchronized fermentations 12 and 13.
  • Figure 8 Spectral comparison between synchronized fermentation points (for fermentations 8 to 12): (a) synchronized spectra and comparison times (1) to (4) at different fermentation stages; spectra collected at moments - (b) (moment 1), (c) (moment 2), (d) (moment 3), (e) (moment 4); and their multivariate control charts - (f) (moment 1), (g) (moment 2), (h) (moment 3), (i) (moment 4).
  • Figure 9 Compositional prediction from spectral models for synchronized moments (1) to (4) in fermentations 8, 9 and 11 - 13: (a) pH; (b) alcohol (% v / v); (c) Er (%); (d) Ea (%); (e) color; (f) RDF (%); (g) n-propanol (mg / l); (h) iso-butanol (mg / l); (i) amyl alcohol (mg / l); (j) ethyl acetate (mg / l); (k) acetaldehyde (mg / l); and (1) diacetyl (mg / l).
  • Figure 10 Compositional prediction from spectral models for synchronized moments (1) to (4) in fermentations 8, 9 and 11 - 13: (a) pH; (b) alcohol (% v / v); (c) Er (%); (d) Ea (%); (e) color; (f)
  • Multivariate control charts and contributions of sample deviation from standard fermentation process (a) MCC at time (2); (b) deviations of variables in fermentation 11; (c) deviations of variables in fermentation 13; (d) CCM at time (4); (e) deviations of variables in fermentation 9; (f) deviations of variables in fermentation 11.
  • a preferred embodiment comprises an alcoholic fermentation monitoring method characterized by the following steps: a) defining the sampling frequency for predicting composition at key points of the fermentation process;
  • a preferred embodiment comprises an alcoholic fermentation monitoring method characterized by the following steps: a) validating spectrum collection, repeating measurement if necessary;
  • a preferred embodiment comprises a universal alcoholic fermentation monitoring method and probe which further comprises the following steps: a) adapting the universal probe to the universal sockets used in the biotechnology industry
  • optical sensor adaptable to the detection range of the ultra violet to infrared ⁇ ;
  • a preferred embodiment comprises a universal fermentation monitoring method and probe which further permits: a) keeping the optical sensor surface clean of any CO2 biofilm, particles or bubbles that affect the spectral properties of the measurement while maintaining the values predicted by the system in optimal operating condition;
  • the washing system starts in pre-programmed cycles optimized for the different production processes, or at the operator's request after analyzing the spectral data. c) ensure the quality, robustness and reproducibility of spectral data collected at different times and fermentation processes.
  • a preferred embodiment comprises an alcoholic fermentation monitoring method characterized by the following steps: a) verifying the minimum of samples within the confidence interval given by Hotelling's quadratic T and within the validity range of the main component model of the control chart;
  • a preferred embodiment comprises an alcoholic fermentation monitoring method characterized by the following steps: a) analyzing the projections of latent variables through partial least squares regression models;
  • a preferred embodiment comprises an alcoholic fermentation monitoring method characterized by quantification of the yeast central metabolism in real time and non-invasively and non-destructively in terms of biomass, brix-grade, pH, alcohol content, dry extract, Er, Ea, RDF, ADF and color.
  • a preferred embodiment comprises an alcoholic fermentation monitoring method characterized by the quantification of the secondary metabolism of the yeast of alcohols and esters in real time and non-invasively and non-destructively in terms of n-propanol, iso-butanol, amyl alcohols, ethyl acetate, amyl acetate, acetaldehyde and diacetyl.
  • a preferred embodiment comprises alcoholic fermentation monitoring method characterized by the following steps: (a) detect yeast metabolic abnormalities through the spectral information of the must and its fermentation;
  • a preferred embodiment comprises an alcoholic fermentation monitoring method which also allows for the comparison of the fermentative performance of: a) different batch processes;
  • a preferred embodiment comprises an alcoholic fermentation monitoring method which also allows for the comparison of predictive fermentation performance of: a) different batch processes;
  • a preferred embodiment comprises a method for monitoring alcoholic fermentation also for predictively diagnosing fermentations in terms of: a) detecting yeast metabolic abnormalities through the spectral information of the must and its fermentation; b) record the fermentative history in an "untargeted" mode allowing to simulate on-line fermentation; c) diagnose the fermentation process by its various spectral weights;
  • f) define the optimum times for changing operating conditions, depending on the data collected and process characteristics; (g) predict the time of the batch fermentation process in order to efficiently manage the fermentation plant, production and washing cycles; (h) predict in advance whether a batch process will irreparably deviate from the standard process in order to abort its production, and diagnose the causes of catastrophic control letter failures as well as further detailed laboratory analysis;

Landscapes

  • Chemical & Material Sciences (AREA)
  • Organic Chemistry (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Wood Science & Technology (AREA)
  • Zoology (AREA)
  • General Engineering & Computer Science (AREA)
  • Genetics & Genomics (AREA)
  • Analytical Chemistry (AREA)
  • Physics & Mathematics (AREA)
  • Food Science & Technology (AREA)
  • Pathology (AREA)
  • Immunology (AREA)
  • General Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Computer Hardware Design (AREA)
  • Sustainable Development (AREA)
  • Microbiology (AREA)
  • Biotechnology (AREA)
  • Biomedical Technology (AREA)
  • Medicinal Chemistry (AREA)
  • Apparatus Associated With Microorganisms And Enzymes (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

Le présent brevet concerne la gestion de la qualité du processus de fermentation alcoolique, par exemple de la bière, par utilisation de technologies à haut débit de spectroscopie UV-VIS-SWNIR (longueurs d'onde comprises entre 200 et 1200 nm), le processus étant surveillé au moyen des données en ligne recueillies. Cette invention décrit un procédé et une sonde universel pour la surveillance de la qualité de la bière pendant la fermentation, faisant intervenir la spectroscopie UV-VIS-SWNIR (ultraviolet/visible/proche infrarouge à onde courte). Un système de traitement de données est conçu pour calculer et exécuter un modèle prévisionnel multivariable, à partir des spectres capturés, de paramètres de régulation de fermentation et/ou de concentration de métabolites.
PCT/IB2013/060677 2012-12-05 2013-12-05 Procédé et sonde de surveillance de la fermentation alcoolique avec spectroscopie uv-vis-swnir WO2014087374A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
PT10668212 2012-12-05
PT106682 2012-12-05

Publications (1)

Publication Number Publication Date
WO2014087374A1 true WO2014087374A1 (fr) 2014-06-12

Family

ID=49958511

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB2013/060677 WO2014087374A1 (fr) 2012-12-05 2013-12-05 Procédé et sonde de surveillance de la fermentation alcoolique avec spectroscopie uv-vis-swnir

Country Status (1)

Country Link
WO (1) WO2014087374A1 (fr)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104198431A (zh) * 2014-09-01 2014-12-10 河南科技学院 鸟苷发酵过程红外指纹图谱的构建方法
CN104297191A (zh) * 2014-09-01 2015-01-21 河南科技学院 利用鸟苷发酵液紫外指纹图谱检测鸟苷产量模型的构建方法
CN106872362A (zh) * 2017-01-18 2017-06-20 浙江大学 用于可见近红外光谱检测的led光源装置及其应用
US10570357B2 (en) 2015-06-17 2020-02-25 University Of Northern Colorado In-line detection of chemical compounds in beer
WO2020047474A1 (fr) * 2018-08-30 2020-03-05 Brewjacket, Inc. Dispositif automatisé de brassage de bière
US20230166766A1 (en) * 2021-12-01 2023-06-01 International Business Machines Corporation Hybrid challenger model through peer-peer reinforcement for autonomous vehicles

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6873409B1 (en) * 1998-11-17 2005-03-29 Kaiser Optical Systems Optical probe with sampling window cleaning configuration
PT104566A (pt) * 2009-05-12 2010-11-12 Univ Do Minho Método e dispositivo de monitorização da produção de uva com espectroscopia uv-vis-swnir

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6873409B1 (en) * 1998-11-17 2005-03-29 Kaiser Optical Systems Optical probe with sampling window cleaning configuration
PT104566A (pt) * 2009-05-12 2010-11-12 Univ Do Minho Método e dispositivo de monitorização da produção de uva com espectroscopia uv-vis-swnir

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
JOÃO S SILVA ET AL: "CHARACTERIZATION OF WINE FERMENTATIONS USING FIBER OPTIC LWUV-VIS-SWNIR SPECTROSCOPY", 24 October 2011 (2011-10-24), XP055109254, Retrieved from the Internet <URL:http://www3.dsi.uminho.pt/seeum2011/cd/Resumos/A604.pdf> [retrieved on 20140321] *
JUAN FERNÁNDEZ-NOVALES ET AL: "FEASIBILITY OF USING A MINIATURE FIBER OPTIC UV-VIS-NIR SPECTROMETER TO ASSESS TOTAL POLYPHENOL INDEX, COLOR INTENSITY AND VOLUMIC MASS IN RED WINE FERMENTATIONS", JOURNAL OF FOOD PROCESS ENGINEERING, vol. 34, no. 4, 2 August 2011 (2011-08-02), pages 1028 - 1045, XP055109250, ISSN: 0145-8876, DOI: 10.1111/j.1745-4530.2009.00530.x *
MCLEOD G ET AL: "A comparison of variate pre-selection methods for use in partial least squares regression: A case study on NIR spectroscopy applied to monitoring beer fermentation", JOURNAL OF FOOD ENGINEERING, BARKING, ESSEX, GB, vol. 90, no. 2, 1 January 2009 (2009-01-01), pages 300 - 307, XP024097768, ISSN: 0260-8774, [retrieved on 20080709], DOI: 10.1016/J.JFOODENG.2008.06.037 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104198431A (zh) * 2014-09-01 2014-12-10 河南科技学院 鸟苷发酵过程红外指纹图谱的构建方法
CN104297191A (zh) * 2014-09-01 2015-01-21 河南科技学院 利用鸟苷发酵液紫外指纹图谱检测鸟苷产量模型的构建方法
US10570357B2 (en) 2015-06-17 2020-02-25 University Of Northern Colorado In-line detection of chemical compounds in beer
CN106872362A (zh) * 2017-01-18 2017-06-20 浙江大学 用于可见近红外光谱检测的led光源装置及其应用
CN106872362B (zh) * 2017-01-18 2023-12-12 浙江大学 用于可见近红外光谱检测的led光源装置及其应用
WO2020047474A1 (fr) * 2018-08-30 2020-03-05 Brewjacket, Inc. Dispositif automatisé de brassage de bière
US20210324310A1 (en) * 2018-08-30 2021-10-21 Brewjacket, Inc. Automated beer brewing device
US20230166766A1 (en) * 2021-12-01 2023-06-01 International Business Machines Corporation Hybrid challenger model through peer-peer reinforcement for autonomous vehicles

Similar Documents

Publication Publication Date Title
WO2014087374A1 (fr) Procédé et sonde de surveillance de la fermentation alcoolique avec spectroscopie uv-vis-swnir
CN109668858A (zh) 基于近红外光谱检测发酵过程生物量和组分浓度的方法
Mercier et al. Multivariate data analysis as a PAT tool for early bioprocess development data
JP5520519B2 (ja) 自動分析装置及び分析方法
CN101308086B (zh) 基于近红外光谱技术的水果内部品质在线检测装置
CN100561194C (zh) 一种利用aotf近红外光谱仪检测中药中微生物的方法
Tosi et al. Assessment of in‐line near‐infrared spectroscopy for continuous monitoring of fermentation processes
Kourti Process analytical technology beyond real-time analyzers: the role of multivariate analysis
US9360421B2 (en) Use of nuclear magnetic resonance and near infrared to analyze biological samples
US20090104594A1 (en) Bioreactor Process Control System and Method
WO2012008324A1 (fr) Dispositif d&#39;analyse automatique, procédé d&#39;analyse et dispositif de traitement d&#39;information
CN102876816A (zh) 基于多传感器信息融合的发酵过程状态监测与控制方法
US11561182B2 (en) Method for detecting quality of cell culture fluid based on Raman spectral measurement
CN109540836A (zh) 基于bp人工神经网络的近红外光谱糖度检测方法及系统
Alves-Rausch et al. Real time in-line monitoring of large scale Bacillus fermentations with near-infrared spectroscopy
CN103454335A (zh) 一种多通道集成嗅觉模拟仪器和生物发酵过程在线分析方法
Ferreira et al. Monitoring complex media fermentations with near‐infrared spectroscopy: Comparison of different variable selection methods
Cozzolino State-of-the-art advantages and drawbacks on the application of vibrational spectroscopy to monitor alcoholic fermentation (beer and wine)
Graf et al. A novel approach for non-invasive continuous in-line control of perfusion cell cultivations by Raman spectroscopy
CN115985404A (zh) 监测和自动化控制生物反应器的方法和装置
CN110084420B (zh) 浓香型白酒发酵中黄水的总糖、总酸和酒精度的检测方法
CN116925908A (zh) 一种细胞培养过程在线智能监测系统和方法
CN116952896A (zh) 一种啤酒厂糖化麦汁中总氮的近红外快速检测方法
CN108796005A (zh) 一种实时监测谷氨酸棒杆菌发酵过程的方法
Bonk et al. In-situ microscopy and 2D fluorescence spectroscopy as online methods for monitoring CHO cells during cultivation

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13821523

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13821523

Country of ref document: EP

Kind code of ref document: A1