LU506809B1

LU506809B1 - Non-intumescent thermoset polymer compositions for dismantlable composites

Info

Publication number: LU506809B1
Application number: LU506809A
Authority: LU
Inventors: Oussema Kachouri; Abdelghani Laachachi; Samet Ozyigit; Julien Bardon
Original assignee: Luxembourg Inst Science & Tech List
Priority date: 2024-04-05
Filing date: 2024-04-05
Publication date: 2025-10-06
Also published as: WO2025210055A1

Abstract

A heat-debondable thermoset composition comprises a dispersion of non-intumescent phosphonate particles in a thermoset (pre-)polymer resin. The composition has a phosphonate particles content between 5 wt.% and 50 wt.%, preferably between 5 10 wt.% and 45 wt.%, more preferably between 15 wt.% and 40 wt.%. The heat- debondable thermoset composition may be used to enable a method for dismantling a composite material.

Description

DESCRIPTION

NON-INTUMESCENT THERMOSET POLYMER COMPOSITIONS FOR

DISMANTLABLE COMPOSITES

Background of the Invention

[0001] The invention generally relates to thermoset polymer compositions containing a non-intumescent organophosphorus filler, as well as to composites, e.g., carbon- fibre reinforced composites, in which such a composition is used as the binder (matrix or adhesive joint). A further aspect of the invention relates to a method for dismantling composites, in which such a composition is used as the binder (matrix or adhesive joint).

[0002] Efficient disassembly of multi-material products into mono-material constituents is essential for life extension, maintenance, repair, materials recovery, and valorisation in new applications, and represents a key challenge in the life cycle of complex products.

[0003] Composite materials are generally difficult to recycle because the individual components are typically intimately bonded. The most common utilization pathways for composite materials are waste-to-energy (thermal utilization via incineration), chemical separation, and downcycling, e.g., by grinding the composites and incorporating them into products to reduce the amount of virgin materials. None of these approaches is sustainable as currently practiced.

[0004] The paper Bassam J. et al, (‘A Process to Recover Carbon Fibers from

Polymer-Matrix Composites in End-of-Life Vehicles”, 2004, Journal of The Minerals.,

Vol. 56, Issue 8, 43-47) presents a general review of the main processes used to recover fibres from fibre-reinforced polymers. Simple and classic processes are presented therein, such as thermal treatment (heating to degrade the polymer matrix), chemical treatment (solvent, acids, etc.) to degrade the polymer matrix, or thermal shock treatment to break the polymer matrix.

[0005] For decades, the development of composite materials has been driven almost exclusively by performance criteria such as, e.g., high specific stiffness. Only recently, life cycle considerations have become part of the design process of composite-based products, with a gradual increase of recycling efforts, and growing interest for durability analyses. The issues of loop-closing, resource efficiency, waste reduction, and life- extension are only a few facets of the life-cycle engineering concept, developed as an integrated method to design, manufacture, use, and recover materials and products for optimal resources turnover (JP2018199230 (A), KR101763789 (B1),

CN103588989 (A), CN101928406 (A), JP2006218793 (A), JPHO733904 (A),

JP2008081549 (A), US2002101004 (A1), US2019309141 (A1)).

[0006] WO 2021/219736 A1 discloses a composite material comprising at least two substrates onto which a layer of an intumescent composition is set therebetween, the composition comprising a mixture of the following individual constituents: from 5 wt. % to 70 wt.% of at least one carbon source, from 0 wt.% to 50 wt.% of at least one acid source, and from 5 wt.% to 50 wt.% of least one expanding agent, the total being 100 wt.%. The intumescent composition is active at a predetermined temperature in the range from 120°C to 450°C and less than the degradation temperature of the substrates. The intumescent composition is used for disassembling the composite material. The article Kachouri O. et al., “Use of intumescent flame-retardant systems in epoxy adhesives for debonding purpose”, Heliyon 10 (2024) e25240 also investigates these debonding systems. This research has shown that intumescent flame retardants can lower debonding temperatures by weakening joint strength at temperatures significantly below the degradation threshold of unmodified adhesives.

[0007] US 2012/114952 discloses an adhesive composition for detachable adhesive bonds based on an adhesive matrix and an expansion material. The particles of the expansion material are at least partially encapsulated.

[0008] DE 19961940 discloses an adhesive composition, the binder system of which incorporates substances that can be thermally activated and are able to trigger a debonding process. These thermally activatable substances are solid at room temperature and homogenously distributed in the binder matrix in the form of fine particles. When exposed to heat, these substances cause swelling of the binder matrix, undergo a phase transformation, are decomposed and/or produce gases or water vapor.

[0009] US 3615960 relates to an epoxy resin composition that can, after application, be broken up and removed by expanding with the application of heat. The epoxy resin composition comprises a solid, thermally decomposable blowing agent which is decomposable at a temperature higher than the temperature at which the epoxy resin is cured.

[0010] WO 00/40648 describes a heat-debondable adhesive composition. The heat debondable adhesive composition comprises a curable resin and an inorganic heat expandable material.

[0011] WO 2006/042782 relates to a process for recycling electrical or electronic components. The adhesive connection between two elements of a component, formed by an adhesive mass, is broken by expanding particles incorporated into the adhesive mass. The expansion of the particles is initiated by the supply of energy, in particular through heating.

[0012] Intumescent flame-retardant additives generally include “dual molecules” additives such as an acid source (polyphosphate) associated to a foaming agent (ammonium or melamine). The three ingredients, i.e., the acid source, the foaming agent and the carbon source, of an intumescent system have to be in intimate contact to act efficiently. The present invention proposes a solution to the issue of debonding composite materials comprising a thermosetting polymer material as a matrix and/or as adhesive using an additive (preferably a single additive) that results in a more efficient system.

Summary of the Invention

[0013] According to a first aspect of the invention, a heat-debondable thermoset composition is proposed. The heat-debondable thermoset composition may be used to enable a method for dismantling a composite material, which comes under a further aspect of the invention. The heat-debondable thermoset composition comprises a dispersion of non-intumescent phosphonate particles in a thermoset (pre-)polymer resin. The heat-debondable thermoset composition has a phosphonate particles content between 5 wt.% and 50 wt.%, preferably between 10 wt.% and 45 wt.%, more preferably between 15wt% and 40wt%. The heat-debondable thermoset composition may be in the uncured state or in the cured state. The phosphonate is selected such that its activation temperature lies below the decomposition temperature of the polymer resin.

[0014] It may be worthwhile noting that a heat-debondable thermoset composition based on a single molecule additive, i.e., a single phosphonate, is expected to be more efficient than a composition comprising an intumescent system.

[0015] The heat-debondable thermoset composition may have a thermoset polymer resin and, optionally, hardener content from 90 wt.% to 55 wt.%. Preferably, the sum of the phosphonate particles content and the content in thermoset polymer resin and, optionally, hardener, may be 100 wt. %.

[0016] The phosphonate particles may, preferably, have a D95 size in the range from 1 um to 50 um, preferably from 10 um to 50 um, more preferably from 20 um to 40 um, e.g., 30 um.

[0017] According to embodiments, the thermoset polymer resin comprises an epoxy resin. The epoxy resin may, e.g., comprise at least one of a bisphenol-based epoxy resin, a novolak epoxy resin, an aliphatic epoxy resin and a glycidylamine epoxy resin.

[0018] The thermoset composition is preferably halogen-free.

[0019] The non-intumescent phosphonate particles may, e.g., comprise or consist of a biscyclophosphonate, in particular, pentaerythritol spirobis(methylphosphonate). It will be appreciated that, compared to traditional intumescent additives like ammonium polyphosphate (APP) or melamine polyphosphate (MPP), the debonding temperature of a composition comprising pentaerythritol spirobis(methylphosphonate) and epoxy thermoset is much lower (circa 250°C). Furthermore, a composition comprising pentaerythritol spirobis(methylphosphonate) and epoxy thermoset is less sensitive to aggregation and provides a good joint strength even at high loading rates (e.g., 40 wt%), whereas compositions of APP or MPP and epoxy thermoset generally perform less well in terms of joint strength at high loading rates.

[0020] According to a second aspect of the invention, a composite material is proposed. The composite material comprises the heat-debondable thermoset polymer composition as a binder bonding other constituents of the composite material, such as, e.g., fibres or sheets. The binder may take the form of a matrix within which the other constituents of the composite material are embedded. If the composite material comprises a sandwich structure, the binder could also be present in the form of one or more adhesive layers between layers of the other constituents. The other constituents could include carbon fibres, e.g., carbon fibre tow(s) and/or carbon fibre fabric(s).

Alternatively, or additionally, the other constituents of the composite material could include metal filament(s) and/or metal sheet(s), e.g., aluminium foil(s), and/or honeycomb core(s), e.g., aluminium honeycomb or polymer (e.g., polycarbonate, polypropylene, etc.) honeycomb.

[0021] According to an embodiment, the composite material may comprise two or more pre-pregs bonded to one another with the heat-debondable thermoset polymer composition. As used herein, the term “pre-preg” designates a composite-material sheet (e.g., a cloth, a weave), made from "pre-impregnated" fibres, e.g., carbon fibres, and a (partially cured or uncured) polymer matrix.

[0022] In yet a further aspect the invention relates to a method for dismantling a composite material, e.g., a carbon-fibre reinforced plastic or a metal-polymer composite. The composite material to be dismantled comprises a (cured) thermoset polymer material as binder bonding other constituents of the composite material, such as, e.g., fibres or sheets. The thermoset polymer material comprises a dispersion of non-intumescent phosphonate particles in a thermoset polymer, the phosphonate particles representing from 10 wt.% to 45 wt.% of the thermoset polymer material. The method comprises generating or increasing porosity of the thermoset polymer material by heating the composite material to a temperature at which the phosphonate degrades into one or more gases and/or reacts with the thermoset polymer, leading to chemical degradation of at least the phosphonate into one or more gases. When the composite material is heated to the decomposition temperature of the phosphonate, gas-filled pores form, which translates into an expansion of the thermoset polymer material and leads to stress within the thermoset polymer material. As the porosity increases, this may lead to coalescence and/or to the formation of cracks within the thermoset polymer material. The thermoset polymer material thereby loses its integrity, with the consequence that it becomes possible to separate the other constituents of the composite material from the thermoset polymer material. The phosphonate is selected such that its decomposition temperature lies below the decomposition temperature of the polymer resin.

[0023] In the context of the method for dismantling, the thermoset polymer material may be as discussed hereinabove. For instance, the thermoset polymer material may have a content in thermoset polymer resin and, optionally, hardener, from 90 wt.% to wt.%. Preferably, the sum of the phosphonate particles content and the content in thermoset polymer resin and, optionally, hardener, amounts to 100 wt.% of the thermoset polymer material. The phosphonate particles preferably have a D95 size in the range from 10 um to 50 um, more preferably from 20 um to 40 um, e.g., 30 um.

The thermoset polymer material may comprise an epoxy resin, e.g., at least one of a bisphenol-based epoxy resin, a novolak epoxy resin, an aliphatic epoxy resin and a glycidylamine epoxy resin. The phosphonate particles may comprise or consist of, e.g., pentaerythritol spirobis(methylphosphonate).

[0024] According to an embodiment, the composite material may be heated to a temperature between 170°C and 320°C, preferably between 220°C and 320°C to degrade the phosphonate.

[0025] In the present document, the terms “thermoset” and “thermosetting” are used to designate a) a polymer that is obtained by irreversibly crosslinking (hardening, curing) a (typically viscous) prepolymer (resin) or b) a prepolymer (resin) that may be reacted (crosslinked) through catalytic homopolymerization or with a co-reactant (hardener) to form a crosslinked polymer, possibly, but not necessarily, through heating. Unless otherwise dictated by context, both meanings are intended to be encompassed.

[0026] As used herein, the expression “intumescent” designates a substance, composition, or system, that forms an insulating fire-protective carbonaceous layer through carbonization and simultaneous foaming. An intumescent substance, composition or system comprises one or more carbon donors (char-forming compound), one or more inorganic strong acid donors (e.g., polyphosphates) and one or more spumific compounds (expansion agents). Conversely, the expression “non- intumescent phosphonate” designates a phosphonate (organophosphorus) which does not degrade into a carbonaceous form layer because it lacks at least one of the strong inorganic acid donor and the spumific compound.

[0027] When reference is made to particle size, this preferably means the particle diameter measured in accordance with ISO Standard 13320:2020, e.g., using the Mie scattering model. The D95 particle size corresponds to the 95" percentile of the cumulative undersize distribution (weighted by volume). More generally, a “DX” particle size, where X is an integer from 1 to 99 designates the Xt" percentile of the cumulative undersize distribution (weighted by volume).

[0028] In the present document, the verb “to comprise” and the expression “to be comprised of” are used as open transitional phrases meaning “to include” or “to consist at least of’. Unless otherwise implied by context, the use of singular word form is intended to encompass the plural, except when the cardinal number “one” is used: “one” herein means “exactly one”. Ordinal numbers (“first”, “second”, etc.) are used herein to differentiate between different instances of a generic object; no particular order, importance or hierarchy is intended to be implied by the use of these expressions. Furthermore, when plural instances of an object are referred to by ordinal numbers, this does not necessarily mean that no other instances of that object are present (unless this follows clearly from context). When this description refers to “an embodiment”, “one embodiment”, “embodiments”, etc., this means that the features of those embodiments can be used in the combination explicitly presented but also that the features can be combined across embodiments without departing from the invention, unless it follows from context that features cannot be combined.

Brief Description of the Drawings

[0029] By way of example, preferred, non-limiting embodiments of the invention will now be described in detail with reference to the accompanying drawings, in which:

Fig. 1: is a schematic illustration of a method for making a composite according to an embodiment of the invention;

Fig. 2: is a schematic illustration of a method for dismantling a composite according to an embodiment of the invention;

Fig. 3: is an illustration of the layout of the adhesive joints used for conducting pull-off tests with example compositions;

Fig. 4: is a graph showing the joint strengths arrived at with different example compositions (EX1-EX4) in comparison to a comparative example composition (CEX);

Fig. 5: is a graph illustrating the evolution of the joint strengths of the different compositions in the temperature range from 175°C to 400°C, when the temperature was increased by steps of 25°C;

Fig. 6: is a graph illustrating the evolution of the joint strengths of the different compositions after different times of exposure to a fixed temperature of 250°C;

Fig. 7: is a graph of the mass losses of the compositions according to the examples in comparison with PCO 900 powder as a function of temperature;

Fig. 8: is an SEM micrograph of an epoxy composition containing 30 wt.% of PCO 900 filler and a magnified portion thereof,

Fig. 9: is an SEM image of carbon fibre layers of a carbon fibre reinforced epoxy composite according to an example heated to 250°C, after separation;

Detailed Description of Preferred Embodiments

[0030] According to embodiments, the use of a non-intumescent organophosphorus flame retardant, e.g., AFLAMMIT™PCO 900, is proposed for inducing debonding in adhesive joints and polymer matrices. Prior research has shown that intumescent flame retardants like melamine polyphosphate (MPP) and ammonium polyphosphate (APP) can lower debonding temperatures by weakening joint strength at temperatures significantly below the degradation threshold of unmodified adhesives. A non- intumescent organophosphorus flame-retardant influences joint strength through a distinct mechanism.

[0031] The performance of a structural adhesive system modified with

AFLAMMIT™PCO 900 at various concentrations (10, 20, 30, and 40 wt.%) was compared with the unmodified adhesive (ie, the same adhesive without

AFLAMMIT™PCO 900). The mechanical properties and thermal/mechanical degradation of the joints were assessed using pull-off tests. The physicochemical properties of the resin samples were evaluated by thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). Scanning Electron Microscopy (SEM) was used to analyse the dispersion of the flame-retardant agent in the thermoset polymer matrix (an epoxy matrix in the examples). The examples showed that incorporating

AFLAMMITMPCO 900 into thermoset adhesive formulations lowers the degradation temperature and enhances joint strength within the application range. Using a non- intumescent organophosphorus flame-retardant offers a debonding solution at temperatures of 250°C and above. The activation temperature may be in the window from 300°C to 350°C for thermal cycles of 10 min (Fig. 5), and at 250°C for longer thermal cycles (Fig. 6). The activation temperature window may thus be considered to be 250°C-350°C. The technique can contribute to promoting efficient end-of-life recycling and to advancing sustainable manufacturing practices.

[0032] Fig. 1 is an illustration of a method for producing a composite material according to an embodiment of the invention. A thermosetting (pre-)polymer resin 10 is provided. Filler material consisting of non-intumescent phosphonate powder 12 is added to the (pre-)polymer resin. The phosphonate particles preferably, have a D95 size in the range from 10 um to 50 um, more preferably from 20 um to 40 um, e.g., 30 um. The non-intumescent phosphonate particles may, e.g., comprise or consist of pentaerythritol spirobis(methylphosphonate). The thermoset (pre-)polymer resin preferably comprises an epoxy resin, e.g., a bisphenol-based epoxy resin, a novolak epoxy resin, an aliphatic epoxy resin or a glycidylamine epoxy resin.

[0033] The components are mixed until the non-intumescent phosphonate particles are well dispersed in the (pre-)polymer resin. When necessary, hardener may be added to the mixture 14. A sandwich composite 20 is formed by applying the mixture 14 as an interfacial adhesive layer between two sandwiching layers (adherents) 16, 18 and curing the mixture. The sandwiching layers 16, 18 could comprise carbon fibre cloths, metal sheets or plates, etc. Curing may be executed with using any technique compatible with the sandwich composite and its components. When curing involves heating the mixture 14, care should be taken to remain below the temperature at which the filler begins to degrade.

[0034] The mixture 14 forms a heat-debondable thermoset composition having a phosphonate particles content between 10 wt.% and 45 wt.%, more preferably between 15 wt.% and 40 wt.%. The remainder of the mixture 14 preferably consists of the thermoset (pre-)polymer resin and, optionally, the hardener. The thermoset (pre-)polymer resin and, optionally, the hardener thus preferably make up to from 90 wt.% to 55 wt.% of the mixture.

[0035] Fig. 2 illustrates how the sandwich composite of Fig. 1 may be dismantled. The sandwich composite to be dismantled comprises a cured thermoset polymer material as binder bonding the two sandwiching layers 16, 18. Under normal use conditions, the interfacial adhesive layer contributes to the structural integrity of the sandwich composite. To dismantle the sandwich composite 20, it is heated to temperatures at which the phosphonate particles dispersed in the interfacial adhesive layer decompose and release one or more gases 22. The processing temperature is, however, kept below the decomposition temperature of the polymer resin. The degradation of the phosphonate particles leads to the formation of pores 24 within the thermoset polymer material. When the porosity increases, this leads to coalescence and/or to the formation of cracks 26 within the thermoset polymer material. This results in the thermoset polymer material becoming brittle. The constituents of the sandwich composite may thereafter be separated relatively easily, if necessary, by using mechanical action, e.g., crushing, grinding, brushing, blowing, sieving, etc.

Examples 1-4

[0036] Heat-debondable thermoset compositions were prepared from epoxy resin.

Pentaerythritol spirobis(methylphosphonate) particles were used as filler.

[0037] The epoxy resin was the low-viscosity epoxy resin bisphenol-A diglycidyl ether (commonly abbreviated as BADGE or DGEBA), available as D.E.R.332. 4-aminophenyl sulfone (DDS), a high-crosslinking aromatic hardener purchased from

Sigma-Aldrich was used as curing agent for the epoxy resin.

[0038] Pentaerythritol spirobis(methylphosphonate) was purchased as AFLAMMIT™

PCO 900 from THOR GmbH., Germany (hereinafter: “PCO 900”). PCO 900 has a phosphorus content of 24 wt.% and a D95 particle size of 30 um. Before use in the compositions, the PCO 900 filler was dried for 12 hours at 80°C.

[0039] The components of the heat-debondable thermoset compositions were mixed with a Hauschild DAC 400 SpeedMixer™ for 10 minutes at 2350 rpm. The resulting dispersions were uniform and devoid of observable bubbles. The hardener and the

PCO 900 filler were added to the epoxy resin at the same time, and the compositions were mixed at room temperature.

[0040] Example compositions with different PCO 900 loading rates were made (EX1: 10 wt.%, EX2: 20 wt.%, EX3: 30 wt.% and EX 4: 40wt.% of PCO 900 in the composition). A comparative composition (CEX) was made without the addition of

PCO 900.

[0041] The compositions according to examples CEX and EX1-EX4 were subjected to joint strength tests and debonding tests. In order to conduct mechanical testing and debonding experiments, 5 mm-thick aluminium substrates were used. For the joint strength evaluation of modified epoxy joints, 9 cm x 9 cm substrates were selected, on which 9 pull-off joints were arranged as shown in Fig. 3b. For the debonding tests, 4 cm x 4 cm substrates were chosen that supported only one single adhesive joint

(Fig. 3a). All joints consisted of the aluminium substrate on one side, the adhesive, and a 20-mm-diameter aluminium pull-off dolly (compatible with the PosiTest™ A adhesion tester supplied by DeFelsko) on the other side.

[0042] The joints were prepared as follows. The aluminium substrates 28 were cleaned with ethanol. Then, using SiC FEPA 80 sandpaper supplied by Struers United

Kingdom, the substrate surfaces were abraded to a smooth finish by grinding along two mutually orthogonal directions. The substrates were then cleaned again with ethanol to remove abrasion particles. Each adhesive composition was deposited manually between two nylon fishing threads 30 (with 300 um diameter) separated by a distance A =8 mm. The surfaces of the dollies were prepared similarly to the substrate: a first step of mechanical abrasion was applied to both remove any joint residues coming from the former pull-out test and creating a smooth finish for the next test. A second step consisted in cleaning the surface to remove abrasion particles. The adhesive composition was then spread by applying pressure with the dolly so that adhesive spots 32 with 2 cm diameter were formed. The threads 30 ensured that the joint thickness had a reproducible value (300 um), regardless of the pressure applied by the operator and the viscosity of the composition.

[0043] In addition to the joint samples, bulk samples were prepared. To this end, the respective composition was pressed into a steel mould coated with a thin film of release agent. Pressure was applied to remove air bubbles. Slices of the bulk samples were prepared for being used for DSC, TGA, tomography, and SEM.

[0044] Each sample was cured at 150°C for 12 hours and then post-cured at 170°C for 2 hours. The durations and temperatures were chosen so as to ensure full cross- linking without degradation of the PCO 900 filler. The joined dollies were taped to the aluminium substrates to maintain the joint geometry during the curing cycle while the adhesives were not cured.

[0045] The joint strength tests were conducted at room temperature and aimed at gauging the joint strength of the compositions at different PCO 900 filler loading rates.

The joint strength tests were conducted with a pull-off tester (PosiTest™ AT-A). To calculate the standard deviations of the data, each composition was tested six times.

[0046] The results of the joint strength tests are shown in Fig. 4. When PCO 900 was incorporated as a filler into the epoxy resin, a notable increase in adhesive joint strength was observed, compared to the unmodified epoxy (CEX). This enhancement was particularly evident with a 20 wt.% filler concentration, where the joint strength saw an approximately 1.5-fold increase, rising from 7 to 10.7 MPa. It was found that

DEGBA/DDS formulations with a 20 wt.%, 30 wt.%, 40 wt.% of PCO 900 filler consistently outperformed those with a 10 wt.% concentration. It was noted that with a

PCO 900 concentration of 30 wt.% and more, a significant increase in the adhesive’s viscosity was noted. The evaluation of joint strength suggests that the PCO 900 concentration for optimal joint strength (at room temperature) is around 20 wt. %.

[0047] It was found that the addition of PCO 900 filler (made from stiff microparticles), to an epoxy matrix leads to increase of the composite elastic modulus, which may reduce the modulus mismatch at the epoxy-aluminium interface. This could delay interface failure under normal stress conditions and explain the enhanced joint strength. However, when the loading rate is 30 wt.% or more, while this reinforcing mechanism remains valid, the formation of particle agglomerates might lower the joint strength due to increased porosity.

[0048] A first debonding test was used to assess the impact of the PCO 900 filler on the joint strength of samples exposed to high temperatures. During these tests, the joints, initially at room temperature, were exposed to high temperatures for 10 minutes in an oven. These high temperatures were selected to start at 5°C above the curing temperature (i.e. 175°C) and incrementing with steps of 25°C, up to 400°C which corresponds to the decomposition temperature of the reference composition (CEX).

Then the pull-off samples were allowed to cool down to room temperature in lab conditions and the joint strength tests were performed. The debonding temperature of a composition is taken to be the temperature at which the joint stress at failure amounts to 10% of the maximum stress recorded for the composition. Since the measurement was not continuous, the temperature corresponding to the value at 10% of the maximum stress is calculated by linear interpolation. Three pull-off tests were performed with the above-mentioned pull-off tester for each temperature and for each composition. The joint strength was then calculated as the average of the recorded values. Fig. 5 shows the evolution of the joint strength of the different example compositions at the different temperatures.

[0049] In the temperature range from ambient to 25 °C above the curing temperature (175 °C), all modified adhesives and the reference exhibited consistent mechanical properties. As can be seen in Fig. 5, in the range from 175 °C to 225 °C, the joint strength values of the example compositions remained mostly unaffected by the rising temperature. For the higher loaded compositions (Ex2, Ex3, Ex4), the joints strength values tended to slightly decrease with temperature, even though this tendency might not be significant when error bars are considered. This could come from a partial degradation of the PCO 900 filler, but not to an extent sufficient for debonding.

[0050] Excluding CEX, a significant decline in joint strength at 250 °C was observed across all joints. The downward trajectories extended into the temperature range 275- 350°C. The curves indicate that in the temperature range from 250-350°C, addition of

PCO 900 in epoxy undermines joint strength as a function of its loading rate, thereby leading to debonding at high temperatures.

[0051] Table 1 shows the debonding temperatures Ta of the different example compositions.

Table 1

Example | Ta(°C)

[0052] A second debonding test was conducted to ascertain the approximate amount of time needed for the flame retardant to become reactive and initiate the process of debonding. Joints with the different example compositions, initially at room temperature, were exposed to a fixed temperature of 250°C for different times.

Exposure times of 5, 10, 15, 20 and 25 minutes were evaluated. Samples were cooled in laboratory conditions down to room temperature before the pull-out test was performed. Three pull-off tests were performed for each exposure time and for each composition. Fig. 6 shows the evolution of the joint strength of the different example compositions after different times of exposure to a temperature of 250°C. It can be seen that the joint strength of CEX remained essentially constant, whereas the joint strength of the PCO-900-filled compositions decreased with exposure time. Except for

EX1, after 25 minutes, the PCO-900-filled compositions showed no residual joint strength.

[0053] Two primary failure mechanisms were identified through analysis of the examples. Adhesive failure describes a failure which occurs at the interface of the joint with aluminum substrate while a cohesive failure occurs within the adhesive joint.

Mixed failure is defined by a combination of both cohesive and adhesive failure. Prior to reaching the debonding temperature, the observed failure modes were a combination of adhesive failure and cohesive failure. Above the debonding temperatures, the predominant mechanism shifted to interface failure. Adhesives with high PCO 900 loading rates (= 30wt.%) exhibited a propensity to detach from one or both adherends.

[0054] Differential scanning calorimetry (DSC) analyses indicated that introducing

PCO 900 led to a reduction of the glass transition temperature (Tg) of the compositions.

As shown in Table 2, Ty decreases with increasing PCO 900 loading rate. This shows that PCO 900 filler acts as a plasticizer for the epoxy resin.

Table 2

[0055] The influence of the addition of PCO 900 on the degradation temperature of the epoxy resin adhesive system was studied by thermogravimetric analysis (TGA) using a TGA 2 SF apparatus from Mettler Toledo. The testing conditions were inert gas and air under a gas flow of 100 cm? min”! with alumina crucibles (70 ul) containing 15-18 mg of sample. The temperatures ranged from room temperature to 800°C. The runs were carried out in dynamic conditions at a constant heating rate of 10°C min”.

For each composition, the onset temperature corresponding to a 5% mass loss was recorded.

[0056] Fig. 7 shows the TGA graphs of the compositions according to the examples as well as the graph corresponding to the PCO 900 powder. It can be observed that the incorporation of the PCO 900 filler into the epoxy resin shifts the degradation temperature to lower values. The Tonset 5% values were: 292 °C for EX4, 297 °C for EX3 323 °C for EX2, 321 °C for EX1 and 368°C for CEX. The degradation rate of the composition accelerates with increasing PCO 900 filler content.

[0057] Fig. 8 shows an SEM micrograph of a composition according to EX3, including a magnified portion. SEM observations were conducted on bulk composition to investigate the dispersion of the fillers, using a FEI QUANTA FEG 200 environmental scanning electron microscope (ESEM). The SEM micrographs revealed that the PCO 900 filler was homogeneously distributed within the polymer matrix. However, it was also found that the filler tends to agglomerate especially in the compositions with the higher loading rates (EX3 and EX4).

[0058] From the joint strength and debonding tests, one may conclude that the optimal

PCO 900 loading rate for the DGEBA/DDS system falls within the range from 20 to 30 wt.%. Depending on the thermoset system (resin, hardener) and the phosphonate filler, the optimal range could be somewhat different.

[0059] The debonding tests showed that debonding may be achieved at a temperature as low as 250 °C. Under the specific test conditions, exposure times of 15 minutes or more were necessary.

[0060] Thermogravimetric analysis revealed that higher PCO 900 filling rates bring the degradation profile of the composition closer to that of the pure filler. This observation, when coupled with tomographic data showing that micropores appear precisely at the filler locations, shows that PCO 900 decomposes independently from the surrounding polymer matrix. Therefore, the PCO 900 filler does not engage with the resin to form a carbonous layer. Upon activation, the filler begins to degrade, leading to the release of gases such as water vapor, carbon dioxide, and other non- halogenated gases. The transformation of the PCO 900 filler results in the formation of micropores and voids. The pressure generated by the gas release is responsible for the observed volume change, and the pressure exerted on the interfaces of these voids leads to their coalescence, consequently forming cracks and ultimately leading to debonding.

Example 5

[0061] EXS relates to carbon fibre reinforced epoxy composites including 20 wt.% of non-intumescent phosphonate particles in the polymer binder (matrix). Comparative composites (CEX5) have been realized with the same ingredients except for the non- intumescent phosphonate filler.

[0062] Non-intumescent phosphonate filler was PCO 900 as in the previous examples. Epoxy resin was Sika Biresin™ CR131 epoxy resin. Hardener was CH132- 5 amine-based hardener purchased from Haufler Composite Gmbh & Co. Kg.

[0063] Five layers of plain weave carbon fibres (200 g m=, purchased from Haufler

Composite Gmbh & Co. Kg) were cut to a size of 25 cm x 25 cm to produce a reference (comparative) composite plate (CEXS). Epoxy resin and hardener were mixed mechanically in the proper weight ratio (100:28) for 5 min. Subsequently, the homogeneous epoxy mixture was degassed under vacuum for 30 min. The degassed epoxy mixture was applied layer-by layer on the carbon fibres to build a stack of five carbon fibre layers. The resulting plate was consolidated in a vacuum-assisted compression mold at 100°C for 5 hours at a compression force of 25kN.

[0064] A carbon fibre reinforced epoxy composite plate according to the invention (EX5) was made as follows. PCO 900 particles were weighed and dried at 50°C overnight before the addition to the epoxy resin. The dried particles were dispersed in the epoxy resin using probe sonication (Hielscher UP400S) with a duty cycle of 0.5 and 90% power for 40 min. The sonication frequency and full (100%) power were 24 kHz and 400 W, respectively. An ice bath was used to prevent excessive heating during the sonication process. The sonicated epoxy-filler mixture was degassed under vacuum for 1 hour, before hardener was added, and the composition was mixed mechanically for 5 min. The composition was degassed under vacuum for another 30 min. Five layers of plain weave carbon fibres (200 g m2, purchased from Haufler

Composite GmbH & Co. Kg) were cut with to a size of 25 cm x 25 cm. After the degassing step, the epoxy composition was applied layer-by layer on the carbon fibres to build a stack of five carbon fibre layers. The resulting plate was consolidated in a vacuum-assisted compression mold at 100°C for 5 hours at a compression force of 25kN.

[0065] The produced composite plates (EX5 and CEXS5) were cut into test samples of the desired sizes with a waterjet cutting machine (Wazer) : 80 mm x 15 mm for flexural tests or 80 mm x 10 mm for dynamic mechanical analysis .

[0066] Flexural strength and modulus values were analysed to describe the mechanical performance of the composites (EX5 and CEX5). The effect of the PCO 900 particles on the flexural performance of the composites is shown in Table 3.

Table 3

Ex. flexural mod. | flexural flexural | fibre content | normalized (GPa) strength strain in composite | flexural (MPa) (%) structure strength (MPa) (wt. %)

CEX5 |43.83+3.11 618.50+52.32 494.80 32.36+1.20 572.00+25.02 - 61.91 554.35

[0067] The reduction in flexural modulus of EX5 with respect to CEX5 could be referred to increment in the matrix content of the composite and then a reduction in the carbon fibre concentration. Carbon fibres are the main stress carrier in the composite structures and their content directly affects the mechanical performance of the composites. This is why the fibre content in weight percentage (wt%) is also provided in order to compare the composites mechanical properties while taking the change of fibre content into account. Furthermore, for comparison purpose, an indicative value called “normalized flexural strength” is calculated as the ratio of flexural strength over fibre content. When this latter value is considered, the flexural strength shows a slight improvement for EX5 sample compared to CEXS.

[0068] The effect of the PCO 900 particles on Tg (glass transition temperature) and the storage modulus of the composite structure was investigated by DMA. Dynamic mechanical analysis (DMA) tests were performed on the samples with a dimension of 80 mm in length and 10 mm in width using a DMA Q800 (TA Instruments, New Castle,

DE) in dual cantilever mode at a frequency of 1 Hz using a heating rate of 3°C/min in the range from room temperature to 275°C. Table 4 shows how PCO 900 particles influence the thermo-mechanical performance of the composites.

Table 4

Sample Tg (°C) Storage Modulus, E’, at 25°C (MPa)

CEX5 155.1 18971

[0069] The addition of the PCO 900 particles in the composite structure causes Tg to decrease by around 10°C. This reduction can be attributed to a decrement in the cross- linked density of the epoxy matrix after addition of PCO 900. Also, the storage modulus at 25°C decreased due to the increment of matrix content in the composite structure and the possible plasticizing effect of the PCO 900 particles on the cross-linked structure of the polymer.

[0070] The test samples were exposed to heat in an oven at different temperatures between 300°C and 225°C for 25 min to activate the debonding reaction between the carbon fibre layers. The reference (neat epoxy, CEXS samples) composites show evidences of debonding when they were activated at 300°C for 25 min. More precisely, carbon fibres layers could easily (i.e. manually) be peeled off layer by layer when the composite was cooled in laboratory conditions down to room temperature, even though no changes in physical appearance could be observed. It is assumed that this debonding phenomenon is related to the thermal decomposition of neat resin at 300°C.

These composites samples were heated at 275°C, 250°C, and 225°C and no debonding between layers was observed.

[0071] The same procedure was applied also for EX5 samples. The lower debonding temperature was determined to be 250°C, i.e. fibres layers could easily (i.e. manually) be peeled off after heating. No physical changes on samples could be seen after heating at 225°C and 275°C, bubble formation was visually observed after thermal treatments at 275°C and 300°C. Thermally treated EX5 samples preserved their structural consistency: however, they could easily be peeled layer by layer like the

CEX5-samples, but after exposure to lower temperatures : 250° or above for EX5, 300°C for CEXS.

[0072] Fig. 9 shows a SEM image of carbon fibre layers of an EX5 sample heated to 250°C, after separation. It can be seen that micropores were formed in the resin-rich areas and these areas behave as stress intensifier points that trigger the debonding between carbon fibre layers during the thermal treatment. The SEM images were recorded with a FEI QUANTA FEG 200 environmental scanning electron microscope.

[0073] While specific embodiments have been described herein in detail, those skilled in the art will appreciate that various modifications and alternatives to those details could be developed in light of the overall teachings of the disclosure. Accordingly, the particular arrangements disclosed are meant to be illustrative only and not limiting as to the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalents thereof.

Claims

1. A heat-debondable thermoset composition comprising a dispersion of non- intumescent phosphonate particles in a thermoset (pre-)polymer resin, wherein the heat-debondable thermoset composition has a phosphonate particles content between 5 wt.% and 50 wt.%, preferably between 10 wt.% and 45 wt.%.

2. The heat-debondable thermoset composition as claimed in claim 1, having a content in thermoset (pre-)polymer resin and, optionally, hardener, from 90 wt.% to 55 wt.%.

3. The heat-debondable thermoset composition as claimed in claim 2, wherein the sum of the phosphonate particles content and the content in thermoset (pre-)polymer resin and, optionally, hardener, is 100 wt. %.

4. The heat-debondable thermoset composition as claimed in any one of claims 1 to 3, wherein the phosphonate particles have a D95 size in the range from 1 um to 50 um, preferably from 10 um to 50 um, more preferably from 20 um to 40 um,

e.g., 30 um.

5. The heat-debondable thermoset composition as claimed in any one of claims 1 to 4, wherein the thermoset polymer resin comprises an epoxy resin.

6. The heat-debondable thermoset composition as claimed in claim 5, wherein the epoxy resin comprises at least one of a bisphenol-based epoxy resin, a novolak epoxy resin, an aliphatic epoxy resin and a glycidylamine epoxy resin.

7. The heat-debondable thermoset composition as claimed in any one of claims 1 to 6, wherein the thermoset composition is halogen-free.

8. The heat-debondable thermoset composition as claimed in any one of claims 1 to 7, wherein the non-intumescent phosphonate particles comprise or consist of pentaerythritol spirobis(methylphosphonate).

9. A composite material comprising the heat-debondable thermoset (pre-)polymer composition as claimed in any one of claims 1 to 8 as a binder bonding other constituents of the composite material, such as, e.g., fibres or sheets.

10. The composite material as claimed in claim 9, wherein the other constituents include carbon fibres, e.g., carbon fibre tow(s) and/or carbon fibre fabric(s).

11. The composite material as claimed in claim 9 or 10, wherein the other constituents include metal filament(s) and/or metal sheet(s).

12. The composite material as claimed in any one of claims 9 to 11, comprising at least two pre-pregs bonded to one another with the heat-debondable thermoset (pre-)polymer composition.

13. A method for dismantling a composite material, e.g., a carbon-fibre reinforced plastic or a metal-polymer composite, comprising a thermoset polymer material as binder bonding other constituents of the composite material, such as, e.g., fibres or sheets, wherein the thermoset polymer material comprises a dispersion of non-intumescent phosphonate particles in a thermoset polymer, the phosphonate particles representing from 10 wt.% to 45 wt.% of the thermoset polymer material, the method comprising generating or increasing porosity of the thermoset polymer material by heating the composite material to a temperature at which the phosphonate degrades into one or more gases and/or reacts with the thermoset polymer, leading to chemical degradation of the phosphonate into one or more gases.

14. The method as claimed in claim 13, wherein the thermoset polymer material has a content in thermoset polymer resin and, optionally, hardener, from 90 wt.% to 55 wt. %.

15. The method as claimed in claim 14 wherein the sum of the phosphonate particles content and the content in thermoset polymer resin and, optionally, hardener, is 100 wt. %.

16. The method as claimed in any one of claims 13 to 15, wherein the phosphonate particles have a D95 size in the range from 10 um to 50 um, preferably from 20 um to 40 um, e.g., 30 um.

17. The method as claimed in any one of claims 13 to 16, wherein the thermoset polymer resin comprises an epoxy resin, e.g., at least one of a bisphenol-based epoxy resin, a novolak epoxy resin, an aliphatic epoxy resin and a glycidylamine epoxy resin.

18. The method as claimed in any one of claims 13 to 17, wherein the non- intumescent phosphonate particles comprise or consist of pentaerythritol spirobis(methylphosphonate).

19. The method as claimed in any one of claims 13 to 17, wherein the composite material is heated to a temperature between 170°C and 320°C, preferably between 220°C and 320°C, to degrade the phosphonate.