Method for measuring the conversion and continuous control of a polymerisation reaction and related sensor system implementing said method"

Abstract

Method for measuring the conversion and continuous control of a polymerisation reaction, comprising the following steps: a) carrying out a polymerisation reaction in a batch reactor (3) by reacting at least one analyte monomer in the presence of an initiator, in a solvent, at a temperature greater than or equal to the activation temperature of the initiator, in a controlled atmosphere; b) sending the gaseous mixture, coming from step a), in a liquid-gas equilibrium, to an inert chamber (7), said inert chamber (7) comprising at least one semiconductor metal oxide (MOS)-based gas sensor, each of said at least one sensor being able to progressively change, in the presence of at least one gaseous analyte monomer, at least one physical parameter, with respect to a reference condition, and of outputting a respective electrical response signal, proportional to the extent of change in said parameter; c) sending said electrical response signals, coming from step b), to electronic means (8) that are able to acquire and process said response signals in real time, producing analytical data as output, depending on the extent to which the physical parameter of each sensor changes. A further object of the present invention is a sensor system (1), configured to implement the method according to the present invention.

Description

[0001] TITLE: “ Method for measuring the conversion and continuous control of a polymerisation reaction and related sensor system implementing said method"

[0002] DESCRIPTION

[0003] FIELD OF THE INVENTION

[0004] The present invention lies in the field of methods for measuring and controlling the course of chemical reactions and their implementation in sensor devices.

[0005] STATE OF THE ART

[0006] An electronic nose is an instrument that exploits a set (array) of partially specific gas sensors and a suitable olfactory fingerprint recognition system (pattern recognition), capable of recognising both simple and complex odours, generally by exploiting a machine learning method.

[0007] One of the main directions in the development of electronic noses is the attempt to form multiple gas detection devices that are sufficiently sensitive and focus on the use of appropriate software / artificial intelligence to improve their qualitative and / or quantitative ability to detect complex odours.

[0008] Sensitive materials suitable for such purposes include metal oxide semiconductors (MOS), among others. MOS sensors generally consist of three layers: a substrate of ceramic material; a heating wire; a semiconductor film based on metal oxides (e.g. Sn, Zn, Co, ...).

[0009] MOS sensors are generally sensitive to many gaseous species, including hydrocarbon species (usually saturated hydrocarbons), alcohols, aldehydes and ketones.

[0010] The operating principle of MOSs is based on conductivity changes induced by surface reactions, due to adsorption of the analyte gas. In particular, the reaction mechanism underlying the operation of MOSs is based on an exchange of oxygen between the volatile molecules and the metal film, which causes a change in the recorded resistance that correlates with the adsorbed amount of the analyte compounds. WO2021147274 Al describes a method for online monitoring of the fermentation process and malodorous pollution using a system employing a gas-sensitive electronic nose and gas chromatography. This system thus provides a method for detecting multiple state parameters in fermentation processes and malodorous pollution by means of an electronic nose device combining gas sensitivity and gas chromatography, for solving online long term detection problems. However, this system is rather complex, since it involves coupling with a chromatographic column, to conduct the necessary separation of the gas mixture leaving the fermenter including the gaseous analytes, with an electronic nose. Consequently, this system provides for the analysis of the analytes by the sensor array to be conducted on the individual gaseous compounds, provided they have been separated beforehand by chromatography. Since this system is not easy to use, it requires the intervention of specialised personnel for commissioning.

[0011] CN106769978 B describes an automatic fermentation apparatus and an online process control detection device, in particular an online fermenter tail gas analyser. CN111208241 B describes a method for predicting the quality of frying oil, based on the use of an electronic nose combined with an artificial neural network.

[0012] SUMMARY OF THE INVENTION

[0013] In this context, the Applicant developed a method for conversion measurement and continuous control of a polymerisation reaction, comprising the following steps: a) carrying out a polymerisation reaction in a batch reactor 3 by reacting at least one analyte monomer in the presence of an initiator, in a solvent, at a temperature greater than or equal to the activation temperature of the initiator, in a controlled atmosphere; b) sending the gaseous mixture, coming from step a), in a liquid-gas equilibrium, to an inert chamber 7 comprising at least one semiconductor metal oxide (MOS)-based gas sensor, each of said at least one sensor being able to progressively change, in the presence of at least one gaseous analyte monomer, at least one physical parameter, with respect to a reference condition, and of outputting a respective electrical response signal, proportional to the extent of change in said parameter; c) sending said electrical response signals, coming from step b), to electronic means 8 that are able to acquire said response signals in real time, producing as output analytical data, depending on the extent to which the physical parameter of each sensor changes, and processing said output data by means of artificial intelligence.

[0014] A further object of the present invention is a sensor system 1, configured to implement the method according to the present invention, said sensor system comprising:

[0015] A. a polymerisation section 2 comprising a batch reactor 3 wherein a polymerisation reaction is carried out in an aqueous solution, preferably in an aqueous emulsion, said batch reactor 3 being provided with an inlet 4, into which an inert gas flow enters, and an outlet 5, out of which an inert gas flow comprising at least one nonreacted analyte monomer in the gas phase, in a liquid-gas equilibrium, exits;

[0016] B. a control section 6, arranged downstream of the polymerisation section 2, comprising: i) an inert chamber 7, suitable for receiving as input the inert gas flow exiting the reactor 3, said inert chamber 7 comprising at least one semiconductor metal oxide (MOS)-based gas sensor, each of said at least one gas sensor being able to progressively change, in the presence of the monomer analyte, at least one physical parameter, with respect to a reference condition, and to output a respective electrical response signal proportional to the extent of change in said parameter; ii) electronic processing means 8, configured to receive said response signals as input, producing analytical data of the response signals as output, depending on the extent of change in the physical parameter of each sensor, and process said analytical data by means of artificial intelligence.

[0017] Advantages of the invention

[0018] The method in accordance with the present invention enables the indirect measurement and continuous control of polymerisation reactions by analysing the volatile component emitted during the process, associated with the higher volatility of the monomers used, with respect to the corresponding polymers produced. In particular, advantageously, the method according to the present invention allows the indirect measurement of conversion by characterising the gas phase in equilibrium with the condensed phase by means of a specific gas detection sensor system combined with signal processing software.

[0019] Still advantageously, unlike direct methods, which involve chemical analysis on the analyte component and which consequently require a certain specificity and the use of related apparatus, as well as specialised personnel, the method according to the present invention is simpler and more cost-effective.

[0020] Again advantageously, the method in accordance with the present invention allows real-time control of polymerisation reactions, since it exploits real-time processing of signals related to changes in the properties of the sensors used, modified during interaction with the analytes.

[0021] Advantageously, the method according to the present invention is non-invasive, since it provides that, during the course of the polymerisation reaction, the analyte gas is preferably withdrawn at a flow rate comprised between 0.05 and 0.20 L / min, preferably 0.10 L / min, i.e. in such an amount that the reaction conditions are not significantly altered.

[0022] Finally, the invention allows the indirect measurement of conversion by operating on the gas phase, which is always optimally mixed and, given its low mass content, is also typically low in fouling the sampling ducts.

[0023] DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 : diagram of a preferred embodiment of the sensor system according to the present invention. Note that the polymerisation section and the control section, indicated by numerical references 3 and 6 respectively, are shown with a dashed outline.

[0025] Figures 2a, 2b and 2c: graph showing time-dependent resistance signal variation curves obtained by the method according to the present invention for polymerisation reactions of ethyl acrylate (EA) [Figure 2a], butyl acrylate (BA) [Figure 2b], and a 1 : 1 mixture of EA and BA [Figure 2c], respectively. Note that the curves shown as “sO-s5” in Figures 2a-2c refer respectively to each individual MOS sensor of the sensor device array employed according to the present invention. In particular, the curves sO and si are identified by black and white circular markers respectively, the curves s2 and s3 are identified by black and white square markers respectively, the curves s4 and s5 are identified by black and white triangular markers, respectively.

[0026] Figure 3: score plot (regression variance: 93.61%) of the principal component analysis (PCA), obtained by the method in accordance with the present invention, for various tests of conducting polymerisation reactions of mixtures of EA and BA in a 1 : 1 weight ratio.

[0027] Figure 4: comparison graph of two reagent percentage conversion curves as a function of acquisition time, for polymerisation reactions of EA and BA mixtures in a 1 : 1 weight ratio. This graph compares a curve (shown as a continuous line) of the aforementioned conversion, obtained by the method in accordance with the present invention, and a percentage conversion curve (or gravimetric conversion), obtained by gravimetric analysis on liquid samples taken from the reactor, during the aforementioned polymerisation reaction.

[0028] DETAILED DESCRIPTION OF THE INVENTION

[0029] Definitions

[0030] For the purposes of the present invention, the definition “comprising” does not exclude the possibility of further components, in addition to those expressly listed after that definition; conversely, the definition “consisting of’ excludes the possibility of further components, in addition to those expressly listed after that definition.

[0031] A “monomer analyte” is defined as a monomer or, in a broader sense, a substance derived from its degradation or reaction with other substances present, capable of activating a measurable response in MOS sensors.

[0032] A “monomer-additive” is defined as a monomer capable of polymerising with the analyte monomer according to reproducible stoichiometric ratios. A “polymerisation reaction” is defined as a polymerisation that can take place in bulk, solution, emulsion or dispersion.

[0033] A “controlled atmosphere” is defined as a gas phase whose mass oxygen content is less than 200 ppm, preferably 100 ppm, more preferably 10 ppm.

[0034] Method for continuous measurement and control of a polymerisation reaction

[0035] As described above, the object of the present invention is a method for the measurement and continuous control of a polymerisation reaction comprising the steps described below. a) Carrying out a polymerisation reaction in a batch reactor

[0036] In step a) of the process according to the present invention, a polymerisation reaction is carried out in a batch reactor 3, as shown in Figure 1, by reacting at least one analyte monomer in the presence of an initiator, in a solvent, at a temperature greater than or equal to the activation temperature of the initiator, in a controlled atmosphere.

[0037] Preferably, in step a) an inert gas flow is continuously fed into the batch reactor 3, at a flow rate comprised between 0.05 L / min, more preferably comprised between 0.05 and 0.20 L / min, more preferably 0.10 L / min.

[0038] Preferably, the inert gas has an average residence time (or dwell time) in the reactor comprised between 0.2 and 50 minutes, preferably comprised between 1 and 11 minutes.

[0039] It should be noted that average residence time values within the above-mentioned limits are an indication that step a) is carried out operating with certain inert gas flow values in relation to the free volume of the reactor.

[0040] By operating within these average residence time values, it is ensured that the polymerisation reaction takes place under liquid-gas equilibrium conditions of the analyte monomer, the monomer-additive and the solvent, while excluding the polymers formed during the reaction, which have zero volatility. Still preferably, the inert gas flow is fed in step a) using common flushing techniques known to a person skilled in the art, and measured by means of a flowmeter placed downstream of the reactor 3.

[0041] Preferably, the atmosphere consists of nitrogen (N2), argon (Ar), helium (He), or mixtures thereof with oxygen (O2).

[0042] Preferably, the polymerisation reaction in step a) is carried out in emulsion, preferably in aqueous emulsion, or in solution, preferably in aqueous solution.

[0043] In accordance with a preferred embodiment, the polymerisation reaction is conducted in aqueous emulsion and step a) comprises the following sub-steps: al) preparation of the aqueous emulsion comprising adding the at least one analyte monomer, the monomer-additive and a surfactant in an aqueous solvent; a2) adding a polymerisation initiator to the aqueous emulsion.

[0044] Preferably, the at least one analyte monomer is selected from ethyl-acrylate (EA), butyl-acrylate (BA), or mixtures thereof, preferably a mixture of EA and BA in a 1 : 1 weight ratio.

[0045] Preferably, the monomer-additive is methacrylic acid (MAA) or acrylic acid.

[0046] In accordance with a preferred embodiment, the at least one monomer-analyte and the monomer-additive are in a weight ratio comprised between 18.5: 1 and 19.5: 1, respectively, preferably 19: 1.

[0047] Preferably, the solvent is an aqueous solvent, preferably water, or an organic solvent, preferably selected from toluene, ethyl acetate, methylene chloride, cyclohexane, n- hexane, decene, isodecane, propylene carbonate, and mixtures thereof.

[0048] Still preferably, the surfactant is sodium lauryl sulphate (SDS) or sodium lauryl ether sulphate (SLES).

[0049] Preferably, the analyte monomer and surfactant are added in a weight ratio comprised between 22: 1 and 26: 1 respectively, preferably 24: 1.

[0050] According to a preferred embodiment, the polymerisation initiator is a cationic azoinitiator, preferably selected from commercially available ones such as, V-50, VA- 086, VA-044, V-501, VA-057, preferably it is V-50 more preferably in diluted form in water in weight ratios comprised between 1:50 and 1 : 150, preferably 1 : 100. Further initiators suitable for the purposes of the present invention are, for example, persulphates, such as ammonium persulphate, or metabisulphites, such as sodium metabisulphite.

[0051] Preferably, the analyte monomer and polymerisation initiator are added in a weight ratio comprised between 180: 1 and 200: 1, preferably 190: 1.

[0052] It should be noted that, preferably, the polymerisation initiator is added in step a) in the form of a solution, preferably an aqueous solution, 4-6%, preferably 4.8% by weight.

[0053] Preferably, step a) is carried out at a temperature comprised between 35°C and 45°C, preferably 40°C.

[0054] Note that, preferably, in step a) the temperature is regulated and kept under control by means of a thermostatic bath 10. b) Sending a gaseous mixture to an inert chamber

[0055] In a step b) of the method according to the present invention, the gaseous mixture, coming from step a), in liquid-gas equilibrium, is sent to an inert chamber 7 comprising at least one, preferably 6, semiconductor metal oxide (MOS)-based gas sensor(s), each of said at least one sensor being able to progressively change, in the presence of at least one gaseous analyte monomer, at least one physical parameter, with respect to a reference condition, and of outputting a respective electrical response signal, proportional to the extent of change in said parameter.

[0056] Preferably, in step b) the at least one unreacted analyte monomer in the gas phase is condensed in a condenser 9, arranged downstream of the reactor 3 and upstream of the inert chamber 7, and said condensed analyte monomer is recycled in the reactor 3.

[0057] It should be noted that, advantageously, the condenser 9 allows the condensation of most of the analyte monomer exiting the reactor 3 throughout the polymerisation process, thus guaranteeing its recycling and preventing the alteration of reaction conditions, such as the liquid-gas equilibrium established by the analyte monomer itself. c) Sending electrical response signals to electronic means

[0058] In a step c) of the method according to the present invention, electrical response signals from step b) are sent to electronic means 8 that are able to acquire said response signals in real time, producing analytical data as output, depending on the extent to which the physical parameter of each sensor changes, and processing said output data by means of artificial intelligence.

[0059] Note that, preferably, the electrical response signals from step b) are signals associated with the change in electrical resistance of the sensors, and the analytical data output from the electronic means 8 are linearised according to the following relationship:

[0060] Where Rsis the point value of linearised resistance, Rsis the point value of measured resistance and Rrif is the resistance value recorded following the addition of the analyte monomer into the reactor, Rmaxis the maximum measured resistance value.

[0061] Sensor system

[0062] A further object of the present invention is a sensor system 1, configured to implement the method according to the present invention, comprising:

[0063] A. a polymerisation section 2 comprising a batch reactor 3 wherein a polymerisation reaction is carried out in an aqueous solution, preferably in an aqueous emulsion, the batch reactor 3 being provided with an inlet 4, into which an inert gas flow enters, and an outlet 5, out of which an inert gas flow comprising at least one non-reacted analyte monomer in the gas phase, in a liquid-gas equilibrium, exits;

[0064] B. a control section 6, arranged downstream of the polymerisation section 2, comprising: i) an inert chamber 7, suitable for receiving as input the inert gas flow exiting the reactor 3, and comprising at least one semiconductor metal oxide (MOS)-based gas sensor, each of said at least one gas sensor being able to progressively change, in the presence of the monomer analyte, at least one physical parameter, with respect to a reference condition, and to output a respective electrical response signal proportional to the extent of change in said parameter; ii) electronic processing means 8, configured to receive said response signals as input, producing analytical data of the response signals as output, depending on the extent of change in the physical parameter of each sensor, and process said analytical data by means of artificial intelligence.

[0065] Preferably, the MOS sensors are sensors based on tin oxide (SnO), zinc oxide (ZnO), cobalt oxides, or combinations thereof, preferably they are sensors based on SnO.

[0066] It should be noted that electronic processing means 8 are preferably combined with artificial intelligence techniques known to a person skilled in the art, capable of correlating the modification of sensor parameters with the course of the desired polymer formation reaction.

[0067] By way of example, commercial MOS sensors suitable for the purposes of the present invention are selected from: Figaro TGS2600, Figaro TGS2610, and / or Figaro TGS 2620.

[0068] Preferably, the MOS sensors are installed on the upper wall 11 of the inert chamber 7, and are preferably arranged in such a way that each of them is tangentially lapped by the flow of inert gas exiting the reactor 3.

[0069] Still preferably, the inert chamber 7 comprises or consists of an inert polymer, preferably polytetrafluoroethylene (PTFE).

[0070] Preferably, the polymerisation section 2 comprises a condenser 9, arranged downstream of the reactor 3 and upstream of the inert chamber 7, configured to receive the inert gas flow exiting the reactor 3 and condensing the solvent contained in said inert gas stream.

[0071] Still preferably, the inert chamber 7 also comprises a first and second sensor, configured for temperature and humidity readings, respectively. EXAMPLES

[0072] By way of non-limiting example, the following are examples of the method and related sensor system according to the present invention.

[0073] EXAMPLE 1

[0074] The structure of a sensor system in accordance with the present invention is described below. This system comprises:

[0075] • a polymerisation section, comprising: a batch reactor with a capacity of 1 L, equipped with a mechanical paddle stirrer; a thermostatic bath to regulate the reactor temperature; a flow meter to regulate the flow of nitrogen into the reactor at a rate of 0.1 L / min; a continuous condenser to recover the reagent stripped from the inert gas during operation. Note that water at a temperature of 25°C is used in the condenser as a cooling fluid;

[0076] • a control section, comprising: an inert chamber with 6 metal oxide semiconductor (MOS) sensors, one temperature and one humidity sensor, operating at a temperature of 80°C controlled by a hot-wire heating system; integrated electronic means for signal acquisition and pre-processing; computer for processing and displaying both acquired and processed signals.

[0077] This sensor system was initially trained for various emulsion polymerisation reactions of acrylates. Below, Table 1 shows the quantities used for the 3 different polymerisation reactions considered: polymerisation of ethyl acrylate (EA); of butyl acrylate (BA); of a mixture of EA and BA.

[0078] Table 1: Polymerisation reactions considered

[0079] The tests for each of the above polymerisation reactions, summarised in Table 1, were carried out according to the following experimental protocol: i) maintaining the system in an inert gas atmosphere (nitrogen at a flow rate of 0.40 L / min) for all subsequent steps, filling the batch reactor (volume: 1 L) with water and heating the system to 40°C, using a thermostatic bath; ii) sampling the inert gas exiting the reactor for 15 min to establish the baseline of the sensor signals; iii) adding the analyte monomer, additive monomer and surfactant under anchor stirring at a speed of about 100 rpm; iv) sampling the inert gas comprising the analyte monomer in liquid-gas equilibrium at the reactor outlet for 5 minutes to assess the sensitivity of the sensors with respect to the analyte monomer; v) adding the polymerisation initiator to the reaction mixture; vi) sampling the inert gas comprising the analyte monomer in liquid-gas equilibrium at the reactor outlet for 420 minutes to monitor the course of the polymerisation reaction.

[0080] In order to train the control system, the signals recorded during the tests (such as signals associated with changes in electrical resistance) were combined with the results of the gravimetric characterisation of liquid samples, taken at regular time intervals of 30 min from the reaction flask. Conversion prediction by analysing signals associated with changes in electrical resistance

[0081] Figures 2a-2c illustrate the trends of the signals recorded by the sensor system, associated with the variation of the electrical resistance of the 6 MOS sensors in the array, for tests 1, 2 and 3 respectively, shown in Table 1. The three figures show a very similar trend, characterised by an initial stationary phase followed by a progressive increase in electrical resistance, associated with a reduction in the monomer content in the gas phase as a result of its conversion to a less volatile polymer.

[0082] However, this comparison also shows subtle differences in the duration of the stationary phase between tests 1, 2 and 3, associated with the different sensitivity of the sensors to the different monomer-analytes. The different polymerisation kinetics of the different reagent mixtures tested is, however, manifested by the sensor system signals, through the subtle differences in the slope of the curves.

[0083] In order to implement models for the prediction of reaction conversion, the resistance signals recorded during the polymerisation process (Rs) were normalised according to the following equation:

[0084] Where $ is the point value of linearised resistance, Rsis the point value of resistance and Rrif is the resistance value recorded following the addition of the analyte monomer into the reactor.

[0085] Signal processing

[0086] To confirm the sensor system’s ability to monitor the progress of the monomer-to- polymer conversion process in real time, signal processing was carried out. Data processing makes use of artificial intelligence techniques, specifically those pertaining to multivariate analysis. This method consists of several algorithms, the most commonly used of which are principal component analysis (PCA) and Support Vector Regression (SVR). Test 3, relating to the polymerisation of the EA + BA mixture, is summarised in the PCA score plot shown in Figure 3. This score plot consists of a two-dimensional visualisation of the dataset in 6 dimensions (one dimension for each sensor) relative to the polymerisation of the EA + BA mixture, and is obtained by projection of the raw signals of the sensor array into the reference system, defined by the principal components PCI and PC2, determined by application of Principal Component Analysis (PCA), which express approximately 94% of the variance of the dataset.

[0087] The points in the score plot are identified on the basis of the reaction conversion determined by gravimetric analysis of the liquid samples taken during the experiments, i.e. 20% conversion (identified with white rhomboid markers), 40% conversion (black triangular markers), 60% conversion (white square markers), 80% conversion (black circular markers).

[0088] Analysis of the PCA score plot in Figure 3 shows that there is a correlation between the distribution of points along the PCI axis and the reaction conversion: points with 80% conversion cluster to the right in the graph; conversely, points with 20% and 40% conversion cluster in the central area of the graph.

[0089] In order to provide an estimate of the conversion, the linearised signals were used to implement a regression model. As an example, the SVR algorithm was used to build the model. Note, however, that the correlation between the variables can be obtained with any of the known regression techniques commonly used in artificial intelligence.

[0090] In the case of the different polymerisation reaction scenarios considered, i.e. EA, BA, and mixture of EA and BA, the implemented models proved to be effective tools for the real-time prediction of the conversion estimate, with Root Mean Square Error (RMSEP) and Mean Absolute Error (MAE) ranging between ±0.5% and ±3%.

[0091] Estimated v.v. gravimetric conversion

[0092] To evaluate the effectiveness of the implemented model, new control tests were carried out on the polymerisation reactions considered, for which the control system provided a prediction of the reaction conversion.

[0093] In relation to these new tests, the comparison curves are shown in Figure 4, summarising the results of the conversion predicted by the sensor system of the invention (“estimated conversion” curve) with that obtained by gravimetric analysis (“gravimetric conversion” curve), in relation to Test 3, i.e. the polymerisation of the mixture of EA and BA.

[0094] The results obtained prove the potential of the invention as a tool for online control of the polymerisation process: the conversion estimates obtained from the analysis of the volatile component are comparable to the data obtained by gravimetric characterisation, as can be seen from the substantial superposition of the two relative graphs shown in Figure 4.

Claims

CLAIMS1. Method for the conversion measurement and continuous control of a polymerisation reaction, comprising the following steps: a) carrying out a polymerisation reaction in a batch reactor (3) by reacting at least one analyte monomer in the presence of an initiator, in a solvent, at a temperature greater than or equal to the activation temperature of the initiator, in a controlled atmosphere; b) sending the gaseous mixture, coming from step a), in a liquid-gas equilibrium, to an inert chamber (7) comprising at least one semiconductor metal oxide (MOS)- based gas sensor, each of said at least one sensor being able to progressively change, in the presence of at least one gaseous analyte monomer, at least one physical parameter, with respect to a reference condition, and of outputting a respective electrical response signal, proportional to the extent of change in said parameter; c) sending said electrical response signals, coming from step b), to electronic means (8) that are able to acquire said response signals in real time, producing as output analytical data, depending on the extent to which the physical parameter of each sensor changes, and processing said output data by means of artificial intelligence.

2. Method according to claim 1, wherein the polymerisation reaction in step a) is carried out in the presence of a monomer-additive.

3. Method according to claim 1 or 2, wherein step a) is carried out in an aqueous emulsion at a temperature of between 35 and 45°C, preferably of 40°C.

4. Method according to any one of claims 1 to 3, wherein step a) comprises the following sub-steps: al) preparation of the aqueous emulsion comprising adding the at least one analyte monomer, the monomer-additive and a surfactant in an aqueous solvent; a2) adding a polymerisation initiator to said aqueous emulsion.

5. Method according to any one of claims 2 to 4, wherein the at least one analyte monomer is selected from ethyl acrylate (EA), butyl acrylate (BA), or mixturesthereof, preferably a mixture of EA and BA in a 1 : 1 weight ratio, and the monomeradditive is methacrylic acid (MAA) or acrylic acid.

6. Method according to any one of claims 2 to 5, wherein the at least one analyte monomer and the monomer-additive are in a weight ratio of between 18.5: 1 and 19.5: 1 respectively, preferably equal to 19: 1.

7. Method according to any one of claims 1 to 6, wherein in step b) the solvent in the gas phase is condensed in a condenser (9), arranged downstream of the reactor (3) and upstream of the inert chamber (7), and the condensed solvent is recycled in step a) back into the reactor (3).

8. Sensor system (1), configured to implement the method in accordance with any one of claims 1 to 7, comprising:A. a polymerisation section (2) comprising a batch reactor (3) wherein a polymerisation reaction is carried out in an aqueous solution, preferably in an aqueous emulsion, said batch reactor (3) being provided with an inlet (4), into which an inert gas flow enters, and an outlet (5), out of which an inert gas flow comprising at least one non-reacted analyte monomer in the gas phase, in a liquid-gas equilibrium, exits;B. a control section (6), arranged downstream of the polymerisation section (2), comprising: i) an inert chamber (7), suitable for receiving as input the inert gas flow exiting the reactor (3), said inert chamber (7) comprising at least one semiconductor metal oxide (MOS)-based gas sensor, each of said at least one gas sensor being able to progressively change, in the presence of the monomer analyte, at least one physical parameter, with respect to a reference condition, and to output a respective electrical response signal proportional to the extent of change in said parameter; ii) electronic processing means (8), configured to receive said response signals as input, producing analytical data of the response signals as output, depending on the extent of change in the physical parameter of each sensor, and process said analytical data by means of artificial intelligence.

9. Sensor system (1) according to claim 8, wherein the polymerisation section (2) comprises a condenser (9), arranged downstream of the reactor (3) and upstream of the inert chamber (7), configured to receive the inert gas flow exiting the reactor (3) as input and to condense the gas phase solvent contained in said inert gas flow.

10. Sensor system (1) according to claim 8 or 9, wherein the inert chamber (7) further comprises a first and a second sensor, respectively configured for reading temperature and humidity.

Images