WO2010112602A1

WO2010112602A1 - Branched poly (hydroxy acid) and production process thereof

Info

Publication number: WO2010112602A1
Application number: PCT/EP2010/054459
Authority: WO
Inventors: Henri Wautier
Original assignee: Solvay Sa
Priority date: 2009-04-02
Filing date: 2010-04-02
Publication date: 2010-10-07
Also published as: US8987343B2; US20120027973A1; US10035876B2; CN104725623A; CN102414244B; JP5631970B2; FR2944021B1; FR2944021A1; EP2414428A1; JP2012522859A; CN104725623B; US20150183926A1; CA2757239A1; CA2757239C; CN102414244A

Abstract

Process for manufacturing a polymer by polycondensation of a hydroxy acid, said polymer comprising at least 80% by weight of units that correspond to the hydroxy acid, according to which at least one polyfunctional reactant capable of giving rise to the formation of a three-dimensional polymer network is mixed with the hydroxy acid, and according to which the mixture is subjected to temperature and pressure conditions and for a duration which are all suitable for giving rise to the formation of the network. Poly(hydroxy acid) (PHA) obtainable by such a process.

Description

BRANCHED POLY(HYDROXY ACID) AND PRODUCTION PROCESS THEREOF

The present application claims the benefit of the French application no. 0952113 filed on April 2, 2009, herein incorporated by reference.

The present invention relates to a branched poly (hydroxy acid) (PHA), to a process for the manufacture thereof, and to a film or hollow body comprising at least one layer based on such a PHA.

Hydroxy acids may be polycondensed in order to form polymers and some of them (glycolic acid (GA), lactic acid (LA), etc.) have been the subject of a resurgence of interest in recent years due to their bio-sourced nature.

The conventional synthesis of polyglycolide (PGA) adopts the same philosophy as the synthesis of polylactide. Firstly, a polycondensation of glycolic acid is carried out in order to obtain a low molecular weight oligomer. Then, at high temperature and low pressure, this oligomer is depolymerized with a view to distilling mainly the cyclic diester, glycolide. A relatively large number of purification steps follow in order to obtain an ultrapure (> 99.90%) glycolide, which will then be subjected to a ring-opening polymerization (in accordance with a procedure similar to the polymerization of the caprolactone monomer, for example). The polymer obtained is linear and has a high molecular weight. Its cost price is high due to the cost of the depolymerization reaction and of the purification of the glycolide. In fact, the first step of this process does generally not make it possible to obtain high enough molecular weights (this is why they were described as oligomers above), hence the need to resort to the subsequent steps. Takahashi et al., (Polymer, 41 (2000), 8725-8728: Melt/solid polycondensation of glycolic acid to obtain high molecular weight poly(glycolic acid)) describes a process that makes it possible to obtain higher molecular weight PGA by melt polycondensation in the presence of certain catalysts such as Zn acetate (hydrated or non-hydrated, the first being preferred). Although the molecular weights that are reported therein are relatively high, the Applicant, having reproduced the examples that were described therein and measured the melt viscosity (according to a procedure which will be described below), found very low values (of the order of 1 Pa. s) and observed an almost negligible melt strength, probably linked to the fact that the molecules obtained are linear. However, melt strength is essential for processing operations of the type: blow moulding of film, extrusion of a sheet or manufacture of a multilayer bottle by extrusion-blow moulding. Since PGA has barrier properties that are comparable to those of EVOH or PVDC, its use in a PET/PGA/PET bottle structure would be very advantageous for packaging carbonated beverages that are oxygen-sensitive such as beer for example.

It is known from the literature to add polyfunctional agents to the polycondensation medium of certain polyesters in order to modify the structure thereof. Thus, Kim et al. (Journal of Applied Polymer Science, Vol. 80, 1388- 1394 (2001)) describe the use of trimethylolpropane (TMP) in the polycondensation medium of succinic acid and 1,4-butanediol in order to form a branched PBS (poly(butylene succinate)). The Applicant has attempted to apply this teaching to hydroxy acids such as GA and LA to produce PGA and PLA, without succeeding in obtaining polymers that have a sufficient melt strength for the aforementioned types of processing operations.

The objective of the present invention is to solve this problem and to provide (a process for manufacturing) polycondensed hydroxy acid polymers having an improved melt strength allowing in particular a processing operation as described above. For this purpose, the present invention relates to a process for manufacturing a polymer by polycondensation of a hydroxy acid, said polymer comprising at least 80% by weight of units that correspond to the hydroxy acid, according to which at least one polyfunctional reactant capable of giving rise to the formation of a three-dimensional polymer network is mixed with the hydroxy acid, and according to which the mixture is subjected to temperature and pressure conditions and for a duration which are all suitable for giving rise to the formation of the network.

In this way, the oligomers obtained during the polycondensation are combined together to form said network, which has the effect of increasing the molecular weight and of providing a novel melt viscosity behaviour compared to the known conventionally linear polyglycolide (PGA from PURAC).

The present invention may be applied to all hydroxy acids capable of polycondensing, i.e. of forming a macromolecule by condensation (chain addition of monomers with removal of water). It gives good results, in particular with glycolic acid (GA) and lactic acid (LA). In general, hydroxy acids that have a primary alcohol are preferred as they are more reactive. Glycolic acid (GA) is very particularly preferred. In one variant of the present process, the hydroxy acid is bio-sourced, that is to say derived from a natural and renewable raw material, as opposed to a fossil raw material. The use of bio-sourced hydroxy acids allows the synthesis of "green" polymers, that is to say polymers synthesized from renewable raw material. The process according to the present invention is suitable for the manufacture of polymers containing at least 80% by weight of units corresponding to the initial hydroxy acid, preferably at least 90% by weight, in particular at least 95% by weight.

According to the invention, polyfunctional molecules (multifunctional organic compounds) are added to the hydroxy acid before its condensation, these molecules being capable of leading to the formation of a three-dimensional network generally based on oligomers of the acid joined to one another by nodes constituted by the reaction sites. The term "polyfunctional" is understood to mean that these molecules contain at least 2 reactive groups, preferably at least 3, or even 4.

In order to be able to produce a three-dimensional network, the molecules added generally contain functional groups capable of reacting with the acid functional group of the hydroxy acid, functional groups capable of reacting with the alcohol functional group of the hydroxy acid and functional groups capable of making the oligomers being formed react with one another. It should be noted that a given functional group may be involved in 2 or even 3 of these types of reactions, and that, in addition, these functional groups may be present on one and the same reactant (variant 1) or be present on different reactants that are used as a mixture (variant X). Examples of reactants that give good results within the context of the 1^st variant are multifunctional silanes that have, besides the silane functional group, either at least two other groups capable of reacting with the acid functional group and with the hydroxide functional group of the hydroxy acid, or at least one group capable of reacting with the acid or the hydroxide while generating a new active functional group capable of reacting with a new acid functional group (for example: an epoxide group). The advantage of silanes is that they can react with the water of condensation to create a network of -Si-O- linkages and thus participate in the branching of the polymer. Silanes of this type which give good results are epoxysilanes and in particular, glycidyloxypropyltrimethoxysilane. Other reactants that give good results within the context of the 1^st variant are polyepoxides such as bisphenol A diglycidyl ether (BADGE) or epoxidized oils such as epoxidized soybean oil or epoxidized linseed oil. - A -

Examples of reactants that give good results within the context of the 2^nd variant (active functional groups present on at least 2 molecules of different types) are the mixtures of at least one polyol with at least one polyacid, of which at least one of the polyol or of the polyacid is trifunctional, preferably tetrafunctional. Advantageously, the polyol and the polyacid are each at least trifunctionalized, for example tetrafunctionalized. Alcohols that give good results are ethylene glycol, trimethylolpropane and pentaerythritol; examples of acids that give good results are malic acid, succinic acid, tricarballylic acid and benzenetetracarboxylic or butanetetracarboxylic acid. A mixture that gives good results, in particular with GA and LA, is a stoechiometric mixture of pentaerythritol (= tetrol) and butanetetracarboxylic acid. Another mixture that gives good results, in particular with GA and LA, is a stoechiometric mixture of trimethylolpropane and tricarballylic acid.

The amount of reactant(s) is generally adapted to their "effectiveness" in terms of functionalities, a too low crosslinking leading to polymers that are too fluid and a too high crosslinking leading to polymers that are no longer able to be melt processed. In practice, good results have been obtained with GA and LA by using a few mol% of active functional groups per mol of GA or LA. In general, the amount of reactant(s) is of about 1 to 100 mmol of active functional groups per mol of GA or LA, most often of 2 to 50 mmol, in particular of 3 to 35 mmol, for example of about 3 to 12 mmol. Depending on the reactant(s) selected, the amount of reactant(s) may be of about 0.2 to 9% by weight of GA or LA, in particular of 0.5 to 6%, for example, depending on the nature of the reactant(s), of about 0.7% or else of 2.5%. The polyfunctional reactant(s) may be added to the hydroxy acid in any known manner. Since hydroxy acids are generally sold as an aqueous solution, it is generally advantageous to dissolve the reactants in this solution, optionally by means of heating.

Preferably, the polycondensation reaction is carried out at least partly at a temperature that is high enough so that the reaction takes place in a reasonable time, but that is not too high, in order to avoid degradation (and the associated coloration problems). The duration of the polycondensation reaction is not critical and may be of about 2 to 24 h, most often of about 5 to 12 h, depending on the temperature. In practice, good results have been obtained with glycolic acid and lactic acid at a temperature between 160 and 24O⁰C. In particular, for LA, the temperature is advantageously from 170 to 23O⁰C, preferably from 180 to 21O⁰C. For GA, the temperature is advantageously above 200⁰C and below 24O⁰C, in particular from 205 to 23O⁰C, for example in the vicinity of 215 or 22O⁰C. Such temperatures are within their melting/crystallization range so that, during the reaction, the crystallization of the polymer obtained is often witnessed. If the temperature during the polycondensation reaction is too low, the polymer obtained will have a melt viscosity and melt strength that are too low relative to what is desired for a processing operation of the type: blow moulding of film, extrusion of a sheet or manufacture of a multilayer bottle by extrusion-blow moulding.

In a first variant of the present process, once the polycondensation temperature is reached, it is kept as constant as possible throughout the whole of the polycondensation step, which corresponds to a single temperature plateau. According to the present invention, the expression "temperature plateau" means that the temperature is kept substantially constant for at least 5 minutes.

In a second variant, the temperature profile during the polycondensation step may be such that it includes more than one temperature plateau. Preferably, the various temperature plateaus are between 160 and 24O⁰C. In particular, for LA, the various temperature plateaus are advantageously within the range of 170 to 23O⁰C, preferably from 180 to 21O⁰C. For GA, the various temperature plateaus are advantageously above 200⁰C and below 24O⁰C, in particular within the range of 205 to 23O⁰C. In this second variant, the temperature difference between the various plateaus may vary from 5 to 3O⁰C, in particular may be of about 10 to 2O⁰C.

In another variant of the present process, that can possibly be combined with the two preceding variants, the polycondensation step may be followed by or end in a plateau at a lower temperature, in particular at a temperature of 10 to 7O⁰C below the temperature of the highest temperature plateau reached during the polycondensation step, for example at a temperature of aboutl50 to 19O⁰C, preferably of 160 to 18O⁰C. In this third variant, the lowest temperature plateau is generally maintained for 1 to 24 h. Very particularly preferably, the polycondensation is initiated at a lower temperature (below the melting/crystallization temperature of the polymer), in order to favour the formation of oligomers and the removal of the water of reaction, and to prevent the distillation of the monomer when the temperature is increased. One way of proceeding consists in leaving the reaction mixture for a few hours, for example between 5 and 100 h, in particular between 12 and 80 h, in an oven at a temperature of 100 to 12O⁰C for GA, and of 50 to 8O⁰C for LA, before driving the temperature to a value above the melting temperature of the polymer that is desired to be obtained. The oven is usually a ventilated oven.

Hence, according to one preferred variant of the invention, the polymer is a semicrystalline polymer and the polycondensation is initiated over several hours at a temperature below the melting/crystallization temperature of the polymer (phase A of the polycondensation), and is then continued at a temperature close to the melting/crystallization temperature of the polymer until crystallization of said polymer (phase B of the polycondensation).

With semicrystalline polymers such as PGA (or PLA obtained from a single isomer of LA), it has been observed that these crystallize very rapidly, often even before the polycondensation reaction has finished. In this case, the Applicant has observed that it was preferable to carry out what is referred to as a solid post-condensation (SPC) step. This may take place by leaving the solid to rest, typically under vacuum, for several hours or even several days, at a temperature above its glass transition temperature but below its melting/crystallization temperature. Typically, such a SPC step may be carried out at a temperature of 140 to 200⁰C, in particular of 150 to 19O⁰C, for example at around 170-180⁰C and at a pressure below 10 mbar. Depending on the nature of the hydroxy acid, on the nature of the reactant(s), on their proportions and on the duration, on the temperature and on the pressure during the polycondensation step, the duration of the SPC step may be a few hours to 1 week, in particular from 6 to 150 h, for example of about 10 to 60 h. It should be noted that a too high temperature during the SPC step may also result in a coloration due to the thermal degradation of the polymer. A long duration does not, on the other hand, have a negative influence on the polymer obtained.

Preferably, phase B of the polycondensation and the SPC step take place under vacuum in order to evaporate the water of reaction and prevent the latter from hydrolyzing the polymer chains being formed. Very particularly preferably, phase B is initiated at atmospheric pressure and the vacuum is applied gradually until a pressure of the order of a few mbar, in particular less than 10 mbar, for example from 2 to 8 mbar, is achieved. The SPC step is typically carried out at a pressure of about 0.01 to 10 mbar, in particular of 0.05 to 5 mbar, for example of about 0.1 mbar.

Alternatively to a heat treatment under vacuum or SPC step, the solid obtained may be subjected to a post-condensation in a microwave device, preferably under conditions such that they cause the melting of the solid. In practice, durations of aboutlO to 30 min at a power of 350 W gave good results. In the process of the present invention, a milling step is advantageously carried out between the polycondensation step and the SPC step. Such a milling step may be carried out by any means known to a person skilled in the art, for example by milling in a rotary mill such as the Pulverisette® from FRITSCH. In one variant, a granulation step may be carried out at the end of the melt phase polycondensation in order to carry out the SPC step on granules. This granulation may especially be carried out at the outlet of the reactor on rods cooled in an air stream then introduced into a granulator. Such a granulation or milling is advantageous since it increases the surface area of the solid resulting from the polycondensation step, which allows an easier evaporation of the residual water present in the medium. Furthermore, the milled or granulated product is easier to handle.

In the process of the present invention, a polycondensation catalyst may optionally be added to the reaction medium. Such a catalyst is usually added in an amount of about 0.01 to 2%, in particular of about 0.1 to 1%. Such polycondensation catalysts are well known to a person skilled in the art and may be selected, for example, from tin (II) chloride, stannous octoate, zinc acetate, zinc lactate and methane sulphonic acid, methanesulphonic acid being preferred. In the process of the present invention, an antioxidant may optionally be added to the reaction medium. Preferably, such an antioxidant is added between the hydroxy acid polycondensation step and the SPC step. Such an antioxidant is typically added in an amount of about 0.01 to 1% by weight of the medium, in particular of about 0.1 to 0.5%. Such antioxidants are well known to a person skilled in the art and may be selected, for example, from hindered phenols and hindered phosphites. Bis(2,4-di-£-butylphenyl)pentaerythritol diphosphite sold under the name ULTRANOX 626® by CHEMTURA and bis(2,6-di-f-butyl-4- methylphenyl)pentaerythritol diphosphite sold under the name ADK STAB PEP 36® by ADEKA PALMAROLE are particularly preferred.

The present invention also relates to a poly(hydroxy acid) (PHA) that can be obtained by the process described above. For this purpose, it relates to a PHA comprising poly(hydroxy acid) chains bridged into a three-dimensional network by polyfunctional molecules. The polymer of the invention is typically constituted of at least 80% by weight of units corresponding to the hydroxy acid, in particular of units corresponding to glycolic acid or lactic acid. These units corresponding to the hydroxy acid are generally present in the form of linear oligomer chains, said oligomer chains being joined to one another in a three- dimensional network by one or more multifunctional organic compounds. The polymer of the invention preferably comprises at least 90% by weight of units corresponding to the hydroxy acid, in particular at least 95% by weight. The polymer of the invention is preferably constituted of at least 95% of poly(hydroxy acid) (preferably glycolic or lactic acid) in the form of linear oligomer chains, and of one or more multifunctional organic compounds that join these linear chains in a three-dimensional network. The polymer of the invention is preferably a polyglycolide or a polylactide containing at least 80% by weight of units corresponding to glycolic acid or lactic acid respectively, preferably at least 90%, in particular at least 95%. The structure of the PHA that can be obtained by the process described above will depend on the nature of the multifunctional organic compounds used. Thus, without wishing to be tied to any one theory, the PHAs obtained according to the 1^st variant of the invention in the presence of polyepoxides such as bisphenol A diglycidyl ether (BADGE) or epoxidized oils will probably exhibit an open branched structure in which two linear oligomer chains made from polycondensed hydroxy acid are attached to one original epoxide functional group. The PHAs obtained according to the 1^st variant of the invention, in the presence of epoxysilanes will probably exhibit a star-type structure. Specifically, the silane functional groups may react with the water of condensation to create a network of -Si-O- linkages. The linear oligomer chains made from polycondensed hydroxy acid would therefore be arranged radially (or in a "star") around nodules constituted of a network of -Si-O- linkages. Finally, the PHAs obtained according to the 2^nd variant of the invention in the presence of polyols and of polyacids, at least one of which being trifunctional, will probably exhibit a structure that comprises loops, each loop comprising at least two linear oligomer chains made from polycondensed hydroxy acid, one molecule of polyol and one molecule of polyacid, these polyfunctional molecules making the link between oligomer chains within the loop and corresponding to branching points to other chains and/or loops. Generally, these PHAs are polymers which are difficult or even impossible to put into solution, even in extremely active solvents such as HFIP (hexafluoroisopropanol) .

These PHAs are generally characterized by a viscoelastic behaviour in the melt state that is non-Newtonian, i.e. that at a low velocity gradient (for example: at a frequency between 1 and 100 rad/s during ARES measurements), their viscosity is not constant but decreases when the frequency increases. Typically, at 24O⁰C, this viscosity is of the order of tens or even hundreds of Pa. s at 100 rad/s and of the order of tens (or even hundreds) of thousands of Pa. s at 1 rad/s. Usually, it is thus possible to obtain, at 24O⁰C, a viscosity of about 50 to 500 Pa. s at 100 rad/s and of about 1000 to 60 000 Pa. s at 1 rad/s.

The present invention finally relates to a film or hollow body comprising at least one layer of a hydroxy acid polymer as described above, preferably in combination with at least one layer of polyester, for example of polyethylene terephthalate (PET), or of polyethylene naphthalate (PEN). For example, the hydroxy acid polymer as described above may be used for the manufacture of a multilayer container comprising at least one layer of said hydroxy acid polymer, in particular of PGA or of PLA and at least one layer of a thermoplastic polyester, in particular of PET or of PEN.

The present invention is illustrated non-limitingly by the following examples. In these examples, the dynamic viscosity measurements were carried out using an ARES rheometer, under the conditions (of temperature and velocity gradient) indicated. The graphs that result therefrom illustrate the change in the viscosity (Eta, in Pa.s) as a function of the frequency (F in rad/s). Example 1 : comparative example outside of the invention

17.2 g of pure (99%) solid glycolic acid were introduced into an 80 ml BUCHI round-bottomed flask were and said flask was placed in a ventilated oven at HO⁰C overnight, i.e. for around 15-17 h.

The next day, the mixture was subjected to the following temperature and pressure profile: 180°C (l h) at Patm 200°C (l h) at 200 mbar 200⁰C (30 min) at 80 mbar 200⁰C (1 h) at 60 mbar 200⁰C (2 h) at 3 mbar

The product was crystallized in the BUCHI round-bottomed flask and was then subjected to a post-condensation step overnight at 19O⁰C and under 3 mbar. Example 2: comparative example outside of the invention

The purpose of this example is to reproduce the teaching of Takahashi explained above.

17.1 g of pure (99%) solid glycolic acid and 0.1 g of Zn acetate dihydrate were introduced into an 80 ml BUCHI round-bottomed flask; the whole assembly was placed in a ventilated oven in order to melt the catalyst (while being mixed) and left at HO⁰C overnight. The next day, the mixture, which was almost completely crystallized, was subjected to the following temperature and pressure profile: 19O⁰C (1 h) at Patm (melting) 19O⁰C (1 h) at 200 mbar 19O⁰C (4 h) at 40 mbar (start of crystallization) 19O⁰C (1 h) at 3 mbar

24O⁰C Patm for 20 min for complete melting 19O⁰C (22 h) at 3 mbar (rapid crystallization).

An ARES measurement gave a dynamic viscosity of 0.6 Pa. s at 100 rad/s and of 3.2 Pa.s at 1 rad/s, which indicates a very fluid polymer having insufficient melt strength. Example 3: test according to the invention, with addition of epoxysilane to GA

17.77 g of pure (99%) solid glycolic acid were introduced into an 80 ml BUCHI round-bottomed flask, then 7.7 g of water were added and the acid was left to dissolve in the water by heating at 50-80⁰C in order to favour the dissolution.

Next, 0.46 g of glycidoxypropyltrimethoxysilane was added (i.e. around 2.6% by weight of GA or 8.3 mmol of epoxide functional groups and 25 mmol of alcoholate functional groups (that can be hydrolyzed to alcohol functional groups) per mol of GA). Everything was dissolved and the solution obtained was perfectly clear.

The mixture was placed in a ventilated oven at HO⁰C overnight. The next day, the mixture was subjected to the following temperature and pressure profile: 205⁰C (I h) at Patm

205⁰C (1 h) at 200 mbar 205⁰C (30 min) at 80 mbar 205⁰C (1 h) at 60 mbar 205⁰C (2 h) at 3 mbar The product was crystallized in the BUCHI round-bottomed flask. Next, it was subjected to a post-condensation step overnight at 205⁰C and under 3 mbar.

Dynamic viscosity measurements were carried out on the product from Example 3 (E3), and also on that of counterexample 1 (El) and of counterexample 2 (E2), on a commercial grade of PGA (polyglycolide PURASORB PG20 from PURAC) and on a standard grade of PET (at 24O⁰C for all the products, except at 27O⁰C for PET), the result of which is illustrated in the appended Figure 1. It can be seen therein that only the product according to Example 3 exhibits non-Newtonian behaviour.

Example 4: test according to the invention, with addition of tetrol and of tetracarboxylic acid to GA 17 g of pure (99%) solid glycolic acid were introduced into an 80 ml

BUCHI round-bottomed flask, then 7.5 g of water were added and the acid was left to dissolve in the water by heating at 50-80⁰C.

The following were then added sequentially: 0.087 g of pentaerythritol (i.e. 0.5% by weight of GA or 11.5 mmol of OH functional groups per mol of GA) and 0.15 g of butanetetracarboxylic acid (i.e. 0.9% by weight of GA or 11.5 mmol of COOH functional groups per mol of GA).

Everything was dissolved and the solution obtained was perfectly clear.

The mixture was placed in a ventilated oven at HO⁰C overnight.

The next day, the mixture was subjected to the following temperature and pressure profile:

205°C (l h) at Patm 205°C (l h) at 200 mbar 205⁰C (30 min) at 80 mbar 205⁰C (1 h) at 60 mbar 205⁰C (2 h) at 3 mbar

The product was crystallized in the BUCHI round-bottomed flask. Next, it was subjected to a post-condensation step overnight at 205⁰C and under 3 mbar. Example 5: test according to the invention, with addition of BADGE to GA

17.1 g of pure (99%) solid glycolic acid were introduced into an 80 ml BUCHI round-bottomed flask, then 0.9 g of BADGE was added (i.e. 5.2% by weight of GA or 23.5 mmol of epoxide functional groups per mol of GA).

The whole assembly was placed in an oven in order to melt and mix, then the mixture was left in a ventilated oven at HO⁰C overnight.

215°C (l h) at Patm 215⁰C (I h) at 200 mbar 215⁰C (30 min) at 80 mbar 215⁰C (1 h) at 60 mbar 215⁰C (2 h) at 3 mbar

The product was crystallized in the BUCHI round-bottomed flask. Next, it was subjected to a post-condensation step overnight at 18O⁰C and under 3 mbar. Example 6: test according to the invention, with addition of tetrol and of tetracarboxylic acid to LA

20 g of lactic acid in aqueous solution at a concentration of 80% (16 g of lactic acid per 4 g of water) were introduced into an 80 ml BUCHI round- bottomed flask, then the following were added sequentially: 0.136 g of pentaerythritol (i.e. 0.85% by weight of LA or 22.5 mmol of OH functional groups per mol of LA) and 0.232 g of butanetetracarboxylic acid (i.e. 1.45% by weight of LA or 22.5 mmol of COOH functional groups per mol of LA).

Everything was solubilized and the solution obtained was placed in an unventilated oven at 5O⁰C for one night (loss of 1.4g), then one night at 75⁰C (loss of 2.3 g) then one weekend at 7O⁰C (loss of 4.2Ig).

The mixture thus treated was subjected to the following temperature and pressure profile: 205⁰C (3 h) at Patm 205⁰C (1 h) at 800 mbar 205⁰C (1 h) at 600 mbar 205⁰C (1 h) at 200 mbar 205⁰C (30 min) at 80 mbar 205⁰C (30 min) at 60 mbar 205⁰C (30 min) at 20 mbar 205⁰C (1 h) at 8 mbar

The product was crystallized in the BUCHI round-bottomed flask. Next, it was subjected to a post-condensation step overnight at 16O⁰C and under 8 mbar.

Dynamic viscosity measurements were carried out on the products from Examples 4 to 6 (curves E4, E5 and E6), and also on a standard grade of PET and on the counterexample 1 (El) (at 24O⁰C for all the products, except at 27O⁰C for PET), the result of which is illustrated in the appended Figure 2.

It can be seen therein that only the products according to Examples 4 to 6 exhibit non-Newtonian behaviour. Example 7: test according to the invention, with addition of triol and triacid to GA

765 g of pure (99%) solid glycolic acid were introduced into a 1.5 1 reactor, to which were added 2.31 g of trimethylolpropane (i.e. 0.3% by weight of GA or 5.1 mmol of OH functional groups per mol of GA) and 3.03 g of tricarballylic acid (i.e. 0.4% by weight of GA or 5.1 mmol of COOH functional groups per mol of GA). Next, 270 g of water were added. The mixture was heated at 50- 8O⁰C, with stirring, until a perfectly clear solution was obtained. Then 2.31g of methanesulphonic acid (i.e. 0.3%) were added as a polycondensation catalyst. The medium was stirred in order to homogenize it, then the mixture was heated at 110-115⁰C in the reactor, with stirring and a stream of nitrogen, for 64 h. Next, the mixture was subjected to the following temperature and pressure profile:

215°C (l h) at Patm

215°C (l h) at 200 mbar

215⁰C (30 min) at 80 mbar 215⁰C (1 h) at 60 mbar

215⁰C (1 h) at 30 mbar

215⁰C (10 min) at 8 mbar

215⁰C (2 h) at 2.5 mbar

215⁰C (1 night) at 2.5 mbar The reactor was then cooled to ambient temperature. The product, crystallized in the reactor, was milled in the form of grains having a diameter of around 0.5mm in a Pulverisette® mill from FRITSCH.

The powder obtained was reintroduced into the reactor and the product was subjected to a solid post-condensation step at 17O⁰C under 0.1 mbar for 138 h. Dynamic viscosity measurements were carried out (at 245⁰C) on the product before the solid post-condensation step (E7/1) and after 68 h (E7/2) and 138 h

(E7/3) respectively of solid post-condensation. The result of these measurements is illustrated in the appended Figure 3.

Example 8: test according to the invention, with addition of triol and triacid to GA

20 g of pure (99%) solid glycolic acid were introduced into a 80 ml

BUCHI round-bottomed flask, to which were added 0.06 g of trimethylolpropane

(i.e. 0.3% by weight of GA or 5.1 mmol of OH functional groups per mol of GA) and 0.08 g of tricarballylic acid (i.e. 0.4% by weight of GA or 5.1 mmol of COOH functional groups per mol of GA). Next, 7.07 g of water were added. The mixture was heated at 50-80⁰C, with stirring, until a perfectly clear solution was obtained.

Then 0.06g of methanesulphonic acid (i.e. 0.3%) were added as a polycondensation catalyst. The mixture was stirred in order to homogenize it. The mixture was then placed in a ventilated oven at 11O⁰C overnight.

The next day, the mixture was subjected to the following temperature and pressure profile: 215°C (l h) at Patm

215°C (l h) at 200 mbar

215⁰C (30 min) at 80 mbar

215⁰C (1 h) at 60 mbar 215⁰C (1 h) at 30 mbar

215⁰C (10 min) at 8 mbar

23O⁰C (2 h) at 3 mbar

17O⁰C (1 h 30 min) at 3mbar

The product was crystallized in the BUCHI round-bottomed flask. The product was then left overnight at 22⁰C at 2.5 mbar.

Next, the product was milled, then reintroduced into the BUCHI round-bottomed flask in order to be subjected to a solid post-condensation step at 17O⁰C under

0.1 mbar for 72 h.

Dynamic viscosity measurements were carried out on the product from Example 8 (curve E8), and also on a commercial grade of PGA (polyglycolide

PURASORB PG20 from PURAC) (at 245⁰C for the product E8 and at 24O⁰C for the PGA from PURAC), the result of which is illustrated in the appended

Figure 4.

It can be seen therein that only the product according to Example 8 exhibits non-Newtonian behaviour.

Example 9: test according to the invention, with addition of triol and triacid to

GA

This test was carried out under the same conditions as Example 8 except for the amounts of triol and of triacid added, which were halved: 0.03 g of trimethylolpropane (i.e. 0.15% by weight of GA or 2.6 mmol of OH functional groups per mol of GA) and 0.04 g of tricarballylic acid (i.e. 0.2% by weight of

GA or 2.6 mmol of COOH functional groups per mol of GA).

After 2 h at 23O⁰C and 3 mbar, the product was left overnight at 18O⁰C and

3 mbar (and not 1 h 30 min at 17O⁰C and 3 mbar, then 1 night at 22⁰C and 2.5 mbar as in Example 8).

The product was crystallized in the BUCHI round-bottomed flask. The product was milled, then reintroduced into the BUCHI round-bottomed flask in order to be subjected to a solid post-condensation step at 18O⁰C under 0.1 mbar for 53 h. Dynamic viscosity measurements were carried out on the product from

Example 9 (curve E9) at 245⁰C, the result of which is illustrated in the appended

Figure 4. It can be seen therein that the product according to Example 9 exhibits non-Newtonian behaviour.

Example 10: test according to the invention, with addition of epoxysilane to GA 255 g of pure (99%) solid glycolic acid were introduced into a 1.5 1 reactor were. Next, 90 g of water was added and the acid was left to dissolve in the water by heating at 50-80⁰C, with stirring, in order to favour the dissolution. The solution obtained was perfectly clear.

Next, 3.83 g of glycidoxypropyltrimethoxysilane were added (i.e. 1.5% by weight of GA or 4.8 mmol of epoxide functional groups and 14.5 mmol of alcoholate functional groups per 1 mol of GA), then 0.77 g of methanesulphonic acid (i.e. 0.3%) was added as a polycondensation catalyst, and the medium was stirred in order to homogenize it.

The mixture was heated at 110-115⁰C in the reactor, with stirring and a stream of nitrogen, for 65 h. Next, the mixture was subjected to the following temperature and pressure profile:

215°C (l h) at Patm

215°C (l h) at 200 mbar

215⁰C (30 min) at 80 mbar 215⁰C (1 h) at 60 mbar

215⁰C (1 h) at 30 mbar

215⁰C (10 min) at 8 mbar

215⁰C (2 h) at 2.5 mbar

17O⁰C (1 night) at 2.5 mbar The next day, the reactor was cooled to ambient temperature. The product, crystallized in the reactor, was milled in the form of grains having a diameter of around 0.5mm in a Pulverisette® mill from FRITSCH.

The powder obtained was reintroduced into the reactor and the product was subjected to a solid post-condensation step at 17O⁰C under 0.1 mbar for 66 h. Dynamic viscosity measurements were carried out (at 24O⁰C) on the product before (ElO/1) and after (E10/2) the solid post-condensation step (66 h).

The result of these measurements is illustrated in the appended Figure 5.

Example 11 : tests according to the invention, with addition of triol and triacid to a GA/LA (95/5) mixture 19.09 g of pure (99%) solid glycolic acid and 1.25 g of lactic acid as an

80% solution in water, which corresponds to a 95/5 GA/LA mixture, were introduced into an 80 ml BUCHI round-bottomed flask. The following were added: 0.06 g of trimethylolpropane (i.e. 0.3% by weight of the GA/LA mixture or 5.1 mmol of OH functional groups per mol of the GA/LA mixture) and 0.08 g of tricarballylic acid (i.e. 0.4% by weight of the GA/LA mixture or 5.1 mmol of

COOH functional groups per mol of the GA/LA mixture), then 7 g of water. The mixture was heated at 50-80⁰C, with stirring, until a perfectly clear solution was obtained.

Then 0.06g of methanesulphonic acid (i.e. 0.3%) were added as a polycondensation catalyst. The mixture was stirred in order to homogenize it. The mixture was placed in a ventilated oven at HO⁰C for around 40 h. Next, the mixture was subjected to the following temperature and pressure profile:

210°C (l h) at Patm

210°C (l h) at 200 mbar

21O⁰C (30 min) at 80 mbar 21O⁰C (1 h) at 60 mbar

21O⁰C (1 h) at 30 mbar

21O⁰C (10 min) at 8 mbar

21O⁰C (2 h) at 5 mbar

17O⁰C (1 h 30 min) at 5 mbar 16O⁰C (1 night) at 3 mbar

The product was crystallized in the BUCHI round-bottomed flask.

The next day, the reactor was cooled to ambient temperature. The product was milled, then reintroduced into the BUCHI round-bottomed flask in order to be subjected to a solid post-condensation step at 16O⁰C under 0.1 mbar for 114 h. Dynamic viscosity measurements were carried out (at 22O⁰C) on the product after 44 h (El 1/1) and 114 h (El 1/2) respectively of the solid post- condensation step. The result of these measurements is illustrated in the appended Figure 6.

Example 12: tensile strength Prior to the manufacture of the test specimens, the extruder (DSM) was preheated to 245⁰C, the (dumbbell type) tensile test specimen mould was heated to 8O⁰C, and the injection ram was heated to 22O⁰C.

0.15 % of ULTRANOX 626® stabilizer from CHEMTURA was added to

20 g of polyglycolide prepared according to Example 7 (after 138 h of solid post- condensation step), previously dried under vacuum at 8O⁰C for 2 h. The mixture was kneaded at a rotation speed of 20 rpm for around 4 minutes in the extruder.

The mixture was then injection moulded in order to form the tensile test specimens. The injection moulding pressure was 5 bar and the hold pressure was 7 bar for 10 seconds. The test specimens were dried for 2 h under vacuum at 8O⁰C and stored away from moisture. The test specimens had a width of the order of 5 mm and a thickness of the order of 2 mm.

This procedure was repeated with a commercial grade of linear PGA (polyglycolide PURASORB PG20 from PURAC).

The tensile strength of these test specimens was measured using a ZWICK universal testing machine (model Z020) equipped with a Multisens extensometer. The gap between the jaws of the machine was 58 mm. The gap between the arms of the Multisens extensometer (displacement sensor making it possible to measure the strain during the test) was 25 mm. The measurements were carried out at a temperature of 22⁰C. The results are given in the following table.

The term "Mod " denotes the segment modulus between 0.05 and 0.25% strain, that is to say the slope of the curve between 0.05 and 0.25% strain. The higher this value, the more rigid the material.

The term "σRupt" denotes the stress (force/cross-sectional area of the test specimen) just before the test specimen breaks. The higher this value, the more resistant the material is to rupture.

The term "εRupt (extenso)" denotes the strain in % measured by the gap between the arms of the Multisens (displacement of the Multisens (mm)/25mmxl00) when the test specimen breaks. The higher this value, the more deformable the material before rupture.

Should the disclosure of any of the patents and patent applications that are incorporated herein by reference conflict with the present disclosure to the extent that it might render a term unclear, the present disclosure shall take precedence.

Claims

C L A I M S

1. - Process for manufacturing a polymer by polycondensation of a hydroxy acid, said polymer comprising at least 80% by weight of units that correspond to the hydroxy acid, according to which at least one polyfunctional reactant capable of giving rise to the formation of a three-dimensional polymer network is mixed with the hydroxy acid, and according to which the mixture is subjected to temperature and pressure conditions and for a duration which are all suitable for giving rise to the formation of the network.

2. - Process according to Claim 1, wherein the reactant is chosen from epoxysilanes and polyepoxides.

3. - Process according to Claim 2, wherein the reactant is chosen from glycidyloxypropyltrimethoxysilane and bisphenol A diglycidyl ether (BADGE).

4. - Process according to Claim 1, wherein the reactant is a mixture of at least one polyol with at least one polyacid, of which at least one of the polyol and of the polyacid is trifunctional.

5. - Process according to Claim 4, wherein the reactant is a stoechiometric mixture of pentaerythritol and butanetetracarboxylic acid.

6. - Process according to Claim 4, wherein the reactant is a stoechiometric mixture of trimethylolpropane and tricarballylic acid.

7. - Process according to any one of Claims 1 to 6, wherein the polymer is a semicrystalline polymer, and wherein the polycondensation is initiated during several hours at a temperature below the melting/crystallization temperature of the polymer (phase A of the polycondensation), and is then continued at a temperature close to the melting/crystallization temperature of the polymer until crystallization of said polymer in order to obtain a solid (phase B of the polycondensation) .

8. - Process according to Claim 7, wherein phase A takes place at atmospheric pressure and phase B is initiated at atmospheric pressure then continues under a progressive vacuum until a pressure of less than 10 mbar, in particular of 2 to 8 mbar is attained.

9. - Process according to Claim 8, wherein the solid is kept under vacuum and is left to rest for several hours at a temperature above its glass transition temperature but below its melting/crystallization temperature (solid post- condensation (SPC) step).

10. - Process according to Claim 7 or 8, wherein the solid obtained is subjected to a post-condensation in a microwave device for a duration and under conditions such that they cause the melting thereof.

11. - Poly(hydroxy acid) (PHA) obtainable by a process according to any one of Claims 1 to 10 and comprising poly (hydroxy acid) chains bridged into a three-dimensional network by polyfunctional molecules.

12. - PHA according to Claim 11, constituted of at least 80% by weight of units that correspond to glycolic or lactic acid in the form of linear oligomer chains, said oligomer chains being joined to one another in a three-dimensional network by one or more multifunctional organic compounds.

13. - PHA according to Claim 11 or 12, characterized in that it is insoluble in hexafluoroisopropanol (HFIP).

14. - PHA according to any one of Claims 11 to 13, characterized by a non- Newtonian viscoelastic behaviour in the melt state, preferably having a melt viscosity at 24O⁰C of the order of tens or even hundreds of Pa.s at 100 rad/s and of the order of tens (or even hundreds) of thousands of Pa.s at 1 rad/s.

15. - Film or hollow body comprising at least one layer of a PHA according to any one of Claims 11 to 14, preferably in combination with at least one layer of polyester, in particular polyethylene terephthalate (PET).