Process and quality control device for halogenated vinyl polymer latex Field of the Invention The present invention relates to a process for monitoring the quality of latexes of halogenated vinyl polymers. It relates more particularly to a process for controlling the quality of latexes of said polymers during the radical polymerization in aqueous emulsion of halogenated vinyl monomers. It also refers to a device for the control of the quality of aqueous emulsions. BACKGROUND OF THE INVENTION Radical polymerization in aqueous emulsion typically employing, in addition to an aqueous polymerization medium, emulsifying agents and radical initiators, constitutes a polymerization technique particularly suited to the manufacture of aqueous emulsions of halogenated vinyl polymers, generally referred to as latexes. , such as for example poly (vinyl chloride) latex or vinylidene chloride copolymers. The latexes thus produced finally contain elementary polymer particles having very low average diameters, which may be from about 10 to about 4000 nm (nanometers). For certain applications, such as for example in the field of paints or coatings in general, these latexes find direct application without prior separation of the polymers from their polymerization medium. In other applications, such as for example PVC-based plastisols, the polymers are isolated from their polymerization medium and dried before use. In all cases, the quality of the products formed depends, mainly, on the average diameter of the polymeric elementary particles of the latex and on the distribution of the diameters of the particles. This is why it is considered appropriate to have an efficient and rapid control procedure for the quality of the latexes during the aqueous emulsion polymerization in order to be able to ensure a very high reproducibility of the quality of the latexes and polymers produced. The control of the development of a polymerization, by measures that reflect the properties of the polymer in formation, makes it possible in fact to regulate the polymerization conditions so that polymers having predetermined properties with better reproducibility are produced. It has already been proposed to control the properties that characterize the molecular structure of the polymers, such as the viscosity index or the density of the polymers, by measuring the absorption spectrum in the near infrared region (hereinafter "NIR"). ) of polymer samples taken during the polymerization and previously separated from their polymerization medium and, in particular, from olefin polymer powders (patent application EP-A-0328826). The object process allows a regulation of the polymerization acting on the conditions of the polymerization when a predetermined difference appears between the properties measured in the isolated polymer and those to be achieved. However, the process according to EP-A-0328826 requires the regular taking and analysis of polymer samples previously separated from their polymerization medium and therefore not adapted to an "in situ" control during the polymerization. Furthermore, the measurement (in relation to a contrast) of the spectral transflectance in the near infrared region, ranging from 900 to 1200 nm, of the light transmitted through latex of styrene and methyl acrylate copolymers has been described, with 30% by weight of dry materials, containing polymer particles whose average diameter is raised to approximately 60 to 80 nm (Paul D. Gossen et al., Applied Spectroscopy, Vol. 47, No. 11, 1993, pp. 1852-1870 ). These measurements have allowed establishing certain correlations between the transferability of specific wavelengths and certain latex properties of copolymers of styrene and methyl acrylate, such as the average diameter of the particles and the solids content of the latexes. However, this "off-site" contrast technique on the light transmitted through the latex is discouraged for latexes containing polymer particles whose average diameters rise to approximately 150 to 200 nm (and above). Therefore it is not applicable to the control of industrial latex whose concentration can exceed 50% at the end of the polymerization, nor to the determination of average diameters and more generally to the distribution of polymer particle diameters in aqueous emulsions consisting of one or several families of particles whose final average diameter is generally greater than about 200 nm. SUMMARY OF THE INVENTION The present invention relates to a process for the quality control of latexes of halogenated vinyl polymers that do not have any of the aforementioned drawbacks. For this purpose, the invention relates to a process for quality control of latexes of halogenated vinyl polymers during the radical polymerization in aqueous emulsion of halogenated vinyl monomers, characterized in that a luminous radiation is induced in the aqueous emulsion, is captured and the diffuse light reflected by the polymer particles of the latex in the spectral band between 1100 and 2500 nm is transmitted to a near-infrared spectrophotometer. The reflected diffuse light of certain specific wavelengths located in this spectral band and correlated is measured with specific properties of the latex with the help of correlation equations stored in a memory, the measured values are entered in said memory and the specific controlled property is calculated. The wavelengths chosen to effect the correlation between the reflected diffuse light (hereinafter briefly referred to as "reflectance") and a specific property of a polymer latex will generally be a function of the polymerization conditions, the nature of the latex and the halogenated vinyl polymer. Therefore, they must be previously determined experimentally. Advantageously, this contrast is made by statistical analysis of a large number of latex samples. The establishment of the correlation equations is advantageously carried out by statistical regression. The correlation equations can then be entered into the memory of a computer, coupled to the spectrophotometer, or in the spectrophotometer itself, and the properties directly calculated.
As non-limiting examples of latex grades which can be controlled during the polymerization according to the process of the invention, the average particle diameter, the total content of dry matter and consequently the conversion rate can be mentioned
(or index of progress of the polymerization reaction), the distribution of the diameter of the particles (by families) and therefore the detection of the different families of particles or even the total surface of the particles. As examples of wavelengths correlated by statistical regression with, for example, specific properties of a vinyl chloride homopolymer latex, there can be mentioned: for the total content of dry materials: 1602, 1610, 1692, 1714, 1886, 2394 nm - for the average diameter of the particles: 1332, 1382, 1390, 1816, 2102, 2180 nm - for the family of particles with a diameter of 500 to 900 nm: 1394, 1402, 1410, 2094, 2488 nm, and - for the total surface area of particles: 1190, 1198, 1206, 1282, 1404, 2230 nm The measurement of the reflected diffuse light (reflectance) of specific wavelengths is generally carried out at least a certain number of times during the polymerization; for example at least at the beginning of the polymerization in order to follow the initiation and at the end of the polymerization in order to verify the conversion rate before degassing the reactor. According to a particular embodiment of the method of the invention, the quality control of the halogenated vinyl polymer latex is carried out in an aqueous emulsion sample taken during the polymerization. This sampling can be carried out, for example, in a by-pass located in the polymerization reactor. According to another particular embodiment which is given preference, the control method according to the invention is carried out directly in the aqueous emulsion in the polymerization reactor. In this preferred embodiment, no latex sampling is necessary and the response to any deviation of a property from a preset value is even faster. In addition, this embodiment has the advantage of allowing continuous monitoring of the quality of the latexes during the polymerization. By halogenated vinyl monomer, it is intended to designate for the purposes of the present invention any radical polymerizable monomer possessing a terminal olefinic unsaturation and substituted with at least one halogen atom. Preferably, these monomers are selected from the substituted derivatives of ethylene and propylene and comprise no more than two or three carbon atoms, respectively. Preferably, the halogen atoms are chosen between chlorine and fluorine atoms. As non-limiting examples of such monomers, mention may be made of vinyl chloride, vinyl bromide, vinyl fluoride, vinylidene chloride, vinylidene fluoride, chlorotrifluoroethylene, tetrafluoroethylene and hexafluoropropylene. The invention is particularly adapted to the control of the radical polymerization in aqueous emulsion of vinyl chloride, vinylidene chloride and vinylidene fluoride. By polymerization, it is intended to designate for the purposes of the present invention both homopolymerization of halogenated vinyl monomers and their copolymerization with each other or with other monomers copolymerizable therewith. As examples of the latter, mention may be made of vinyl esters, such as vinyl acetate, acrylic esters, such as methyl acrylate and methyl methacrylate, unsaturated nitriles, such as acrylonitrile and methacrylonitrile, allyl esters, such as allyl acetate, derivatives styrenics and alpha-olefins, such as ethylene and propylene. However, the invention is preferably applied to latex control of polymers containing at least 50 mole% and preferably at least 80 mole% units derived from halogenated vinyl monomers. It is particularly applicable to the quality control of vinyl chloride polymers latex, more especially vinyl chloride homopolymers, vinylidene chloride copolymers and polymers, homo- and co-polymers, of vinylidene fluoride. As non-limiting examples of vinylidene chloride copolymers, mention may be made of copolymers containing about 70 to 96% by weight of vinylidene chloride, the remainder consisting essentially of alkyl acrylate, such as methyl acrylate, or even sodium chloride. vinyl. As non-limiting examples of vinylidene fluoride copolymers, mention may be made of copolymers with chlorotrifluoroethylene or with hexafluoropropylene containing at least 75% by weight of vinylidene fluoride. By radical polymerization in aqueous emulsion, it is intended to designate for the purposes of the present invention any radical polymerization process that is carried out in an aqueous medium in the presence of emulsifying agents and radical initiators. This definition specifically includes the polymerization in aqueous emulsion called "classical" in which water-soluble radical initiators are used, as well as the polymerization in microsuspension called even in homogenized aqueous dispersion, in which oil-soluble initiators are used and an emulsion of droplets is made of monomers thanks to a powerful mechanical agitation and the presence of emulsifying agents. The nature of emulsifying agents and initiators, hydro- and oleo-soluble, is not critical. The control process according to the invention is therefore applied to any process of radical polymerization in aqueous emulsion of halogenated vinyl monomers in the presence of initiators and emulsifying agents customary in this type of polymerization. It is particularly suited to the control of "classical" aqueous emulsion polymerization employing emulsifying agents and water-soluble initiators. The method according to the invention has the advantage of allowing the monitoring of the quality of the latexes, such as, for example, monitoring the conversion rate and / or the average size of the polymer particles, during the entire course of the polymerization, say both for the low conversion rates, (at the beginning of the polymerization) when the latex is poor in solids and contains only particles of very small dimension, as well as for high conversion rates, (at the end of the polymerization) when latex is rich in solids (several tens percent) and contains mainly large diameter particles. The process according to the invention also has the appreciable advantage of allowing the direct measurement during the polymerization of certain properties of the latex without prior separation of the polymer from its polymerization medium (water in this case). In fact, it allows a very fast response in case of deviation of a property of the latex in relation to a pre-established value, that is to say a rapid intervention in case of necessity of the parameters of the polymerization that has an influence on the property "deviated" and consequently a regulation of the polymerization able to guarantee a very high reproducibility of the quality of the produced latexes. As non-limiting examples of the parameters well known to those skilled in the art having influence on the properties of the polymer latexes during the polymerization, the polymerization temperature, the stirring speed of the polymerization medium, the program of injection of the reactants, such as the emulsifier and the initiator, or even the introduction during the polymerization of one or more polymer latexes.
The invention also relates to a device for controlling the quality of latexes of halogenated vinyl polymers in a reactor during polymerization. The device according to the invention consists essentially of a measuring probe arranged in the reactor, an optical link between said probe and a near-infrared spectrophotometer located outside the reactor, said spectrophotometer being provided with means for emitting luminous radiation and for capturing the diffuse light reflected in the spectral band between 1100 and 2500 nm and a computer that have in memory the equations that establish a correlation between the value of diffuse light reflected from specific wavelengths and the specific properties of latex. Measuring probes of this type, called "NIR probes" can be found in commerce. The optical link between the measurement probe and the near-infrared spectrophotometer is advantageously made by means of optical fibers. It is recommended to arrange the measuring probe at the bottom of the reactor and, in addition, arrange it above the agitator, checking that it is placed in such a way that it is covered at the beginning of the polymerization (ie when the reactor contains the complete polymerization charge). Under these conditions, the control measures carried out are representative of the quality of the latex. It is intended that the probe disposed in the reactor be able to withstand the high pressures developed during the polymerization and the depressions (voids) applied prior to the polymerization in order to remove the oxygen from the reactor. It is also intended that the probe be perfectly sealed in order to exclude any leakage of gaseous monomer. For this purpose, it is recommended to use a probe provided with two watertight barriers: a first watertight barrier on the side of the reactor, for example a sapphire crystal, and a second watertight barrier on the outside, for example an O-ring type union, and keep the interior of the probe under a fixed pressure of a neutral gas. The control of this pressure will detect a failure of the barrier on the side of the reactor due to an increase in pressure during polymerization (or a drop in pressure during the vacuum phase) and a failure of the barrier on the outside by a drop in pressure. In order to provide maximum security and to be able to detect the simultaneous eventual defect of the two watertight barriers, it is also possible to provide a monomer detector, for example vinyl chloride, at the place where the probe enters the polymerization reactor. In the hypothesis of a simultaneous defect of the two watertight barriers, even a supplementary double protection can be envisaged, that is, the closing of a rotary valve that divides the optical fiber attached to the measuring probe and ensures the sealing of the seam in the enclosure, as well as the closure (after having separated the sectioned optical fiber) of a second rotary valve installed in series with the first in order to provide an additional security of total tightness of the seam of the enclosure. The device according to the invention can be adapted to the quality control of the latex of halogenated vinyl polymers in aqueous emulsion in any polymerization reactor usually used for the aqueous emulsion polymerization of said monomers. These reactors are generally provided with means for the introduction of the reactants and for the discharge of the polymer latex, a stirrer and a double shell for the evacuation of the polymerization heat. Detailed Description of the Invention The following examples are intended to illustrate the method and device according to the invention.
They refer to the homopolymerization in aqueous emulsion of vinyl chloride to produce a latex of poly (vinyl chloride) controlling directly in the reactor the evolution of the content total dry matter of the poly (vinyl chloride) latex and the average diameter of the polymer particles during the development of the polymerization. The measurements of the reflectance in the spectral band between 1100 and 2500 nm of wavelengths specifically correlated with these two properties of a vinyl chloride homopolymer latex (given above) are carried out every two minutes from the initial filling to the end of the polymerization reaction and degassing of the residual vinyl chloride. The examples are carried out in a laboratory reactor of 6, 6 liters, provided with a double envelope through which circulates a heat carrier fluid and a conventional stirrer of stainless steel blades. The reactor is also provided with a measuring probe in the near infrared region (NIR) fixed hermetically through the double envelope with its measuring window arranged flush with the wall of the reactor. This probe is located at the bottom of the reactor below the initial filling level. The correlation equations and the calculator are incorporated into a computer attached to the spectrophotometer. EXAMPLE 1 This example relates to the control of the total content of dry matter and the average diameter of the particles during the polymerization in aqueous non-micellar emulsion of vinyl chloride to produce a monomodal latex, constituted by a single family of particles, whose average diameter end rises to approximately 300 nm. By non-micellar aqueous emulsion polymerization, it is intended to designate the polymerization in the presence of an amount of emulsifier whose concentration does not exceed at any time the critical micelle concentration of the emulsifier. 2675 g of demineralized water and 0.0075 g of copper sulfate, 5 aq. Are successively introduced into the reactor at room temperature. (ie 7.5 cm3 of a solution of 1 g / 1). The reactor is closed and the agitator is started at 225 revolutions / minute. The temperature of the reactor content is brought to 30 ° C. Once stabilized at this value, the vacuum is made twice
(at 180 irati of absolute mercury) and, between the two operations, the reactor is swept with technical nitrogen at the pressure of 600 mm of absolute mercury. Then 2180 g of vinyl chloride are introduced and the temperature of the reactor content is progressively brought to 51 ° C. At the time when 51 ° C is reached, 1.30 g of ammonium persulfate (ie 26 cm 3 of a 50 g / 1 solution) is introduced. Six minutes later, 2.16 g of ammonia (i.e. 16 cm3 of a solution of 135 g / 1) are introduced. This moment is considered as the time zero of the polymerization (to). Between + 0 h 30 min and at + 4 h 15 min, 21.8 g of myristic acid (in the form of ammonium salt) is introduced progressively (ie 242 cm3 of a solution of 90 g / kg of myristic acid) . After a pressure drop of 2 bar, the temperature is brought to 80 ° C. The stirring speed is reduced to 50 revolutions / minute and the residual vinyl chloride is removed by degassing and separation of boiling volatiles. As a control of the values obtained from the reflectance measurements, a sample of polyvinyl chloride latex is taken from 0 to 30 min. And at regular intervals of approximately 30 minutes. The samples taken are released from the residual vinyl chloride by degassing and separation of boiling volatiles. In these samples, the content of dry matter (by densitometry) and the average diameter of the particles is determined by photosedimentometry. In figure 1, appended, the total dry matter content of the controlled latex is expressed in ordinates, expressed in%. In Figure 2, attached, the mean diameter of the particles expressed in micrometers (μm) is carried in ordinates. In figures 1 and 2, the time is taken in abscissa, expressed in minutes, counted from to. In these figures, the empty squares correspond to the measurements made every two minutes according to the method of the invention and the circles filled to the measurements made every 30 minutes in samples taken during the polymerization. Example 2 This example relates to the control of the total content of dry matter and the average particle diameter during the polymerization in aqueous, seeded, non-micellar emulsion of vinyl chloride to produce a monomodal latex, consisting of a single family of particles, whose final average diameter is raised to approximately 600 nm. In the reactor, 2520 g of water, 0.0070 g of copper sulfate, 5 aq. (ie 7 g of a solution of 1 g / 1) and 295 g of polyvinyl chloride in the form of latex produced in example 1. The reactor is closed and the stirrer is started. Thereafter, the procedure of Example 1 is carried out in its entirety, the amount of ingredients (expressed in g) being the following: vinyl chloride: 1975.00 ammonium persulfate: 0.70 ammonia: 2.00 myristic acid (in salt form): 11.80 As in Example 1, the reflectance value at the aforementioned wavelengths is measured directly in the aqueous emulsion by means of the probe every two minutes and is also taken regularly every 30 minutes. minutes latex samples that after degassing are analyzed as a control. The results of the measurements are shown in figures 3 and 4, appended, in which the abscissas also represent time, expressed in minutes, counted from to. In figure 3, the total dry matter content of the controlled latex is shown in ordinates, expressed in% and in figure 4 the average diameter of the particles expressed in μm. The comparison of the results represented in figures 1 to 4 demonstrates the reliability and reproducibility of the latex quality control method according to the invention.