Biodegradable polyester layered silicate nanocomposites based on poly([epsilon]-Caprolactone).

INTRODUCTION

Polymer nanocomposites, especially polymer layered silicate nanocomposites, represent a valuable alternative to conventionally filled polymers (1). Because of the dispersion of nanometer-size silicate sheets, these nanocomposites exhibit markedly improved properties when compared with pure polymers or conventional "microcomposites." Polymer nanocomposites based on layered silicates (i.e., smectite clays such as montmorillonite, MMT) are of current interest because of the fundamental questions they address and potential technological applications. The use of organoclays as reinforcing agents for nanocomposite preparation has been studied for various polymer systems including polypropylene (2), polyamide (3), poly(ethylene terephthalate) (4), polystyrene (5), unsaturated polyesters (6), polyimide (7), polyolefins (8), poly(methyl methacrylate) (9), ethylene-vinyl acetate copolymers (10, 11) and poly(ethylene oxide) (12) among others. This new type of material can be prepared by various techniques including exf oli ation-adsorption, in situ intercalative polymerization, template synthesis and melt-intercalation (11).

Poly([epsilon]-caprolactone)/clay nanocomposites are of particular interest because of the biocompatibilty and biodegradability of the aliphatic polyester matrix and the high property enhancements that could result from the layered silicate dispersion. In this paper, poly([epsilon]-caprolactone) (PCL) layered silicate nanocomposites were prepared by either melt intercalation with molten PCL chains or by in situ ring-opening polymerization of [epsilon]-caprolactone. The clays are usually made hydrophobic by ionic exchange of the sodium interlayer cations with onium cations bearing long alkyl chains. As long as the alkylated surface of the silicate layers is miscible with the polymer, the polymer chains can crawl into the interlayer space and form either an intercalated or an exfoliated nanocomposite. In the intercalated hybrid structure, a single extended polymer chain is intercalated/sandwiched between the silicate sheets, resulting in a well-ordered multilayer of alternating polymer and inorganic sheets. In the exfoliated (or delaminated) hybrid structure, the silicate nanolayers (1 nm thick) are individually dispersed in the polymer matrix. Exfoliation of the silicate layers usually provides nanocomposite materials with improved properties, such as higher Young's modulus and storage modulus, higher thermal stability and flame retardancy, and a more efficient gas barrier (1).

Nanocomposites based on PCL have been prepared by in situ intercalative polymerization of [epsilon]-caprolactone in presence of a protonated [omega]-amino-acid exchanged MMT (13). Another synthetic pathway, recently proposed by Chen et al. (14), produces novel segmented PCL-based polyurethane/clay nanocomposites by step-growth polymerization of diphenylmethane dusocyanate, butanediol and preformed polycaprolactone diol. Messersmith and Giannelis (15) have also reported on the polymerization of [epsilon]-caprolactone directly inside a [Cr.sup.3+]-exchanged fluorohectorite.

In a previous paper (16), we reported preliminary results on the preparation of PCL-based nanocomposites by melt intercalation, a very attractive and environmentally friendly process since no solvent is required. In this article, two methods of PCL nanocomposite formation (i.e., melt intercalation and in situ intercalative polymerization) were examined and compared. Both non-modified clays ([Na.sup.+]-MMT) and silicates modified by either nonfunctional long alkyl chains (MMT-Alk) or chains terminated by carboxylic acid (MMT-COOH) or hydroxyl groups (MMT-[(OH).sub.2]) were studied. Depending on the surface-modification of MMT, nano- or microcomposites were obtained with specific mechanical and thermal properties. Furthermore, it will be shown that the formation of PCL-based nanocomposites depends not only on the ammonium cation and related functionality, but also, for the same cation, on the synthetic route--either melt intercalation or in situ intercalative polymerization.

EXPERIMENTAL

Materials

Commercial poly([epsilon]-caprolactone (CAPA[R] 650) was supplied by Solvay Chemicals sector-SBU caprolactones. The PCL number average molecular weight ([M.sub.n]) was 49 kg/mol with a polydispersity ([M.sub.w]/[M.sub.n]) of 1.4 as determined by size exclusion chromatography (SEC) in THF at 35[degrees]C. [epsilon]-Caprolactone (Fluka was dried over Ca[H.sub.2] and distilled under reduced pressure prior to use. Stannous octoate (Sn[(Oct).sub.2]) and dibutyltin dimethoxide ([Bu.sub.2]Sn[(OMe).sub.2]) were purchased from Aldrich and diluted with dry toluene. Four clays were used in this study. Cloisite[R][Na.sup.+] ([Na.sup.+]-MMT) is a natural purified montmorillonite clay with a cation exchange capacity (CEC) of 92 meq/ 100 g. It is composed of stacks of silicate layers that are 0.96 nm thick and 250 to 500 nm wide with an interlayer distance of 0.26 nm. MMT-Alk (Cloisite[R] 25A), MMT-[(OH).sub.2] (Cloisite[R] 30B) and MMT-COOH (nanofil[R] 820) are montmorillonites, organically modified with dimethyl 2-ethylh exyl (hydrogenated tallow alkyl), with methyl bis(2-hydroxyethyl) (hydrogenated tallow alkyl) ammonium cations and with ammonium bearing 12-aminododecanoic acid respectively. Cloisite[R] [Na.sup.+], Cloisite[R] 25A and Cloisite[R]30B are provided by Southern Clay Products (Texas, USA) and nanofil[R] 820 is provided by SudChemie. The ammonium cations and their content (as determined by TGA and expressed in weight content of organics) in exchanged MMT are listed in Table 1.

Composites Preparation by Melt Intercalation

The PCL-layered silicate composites were prepared by mechanical kneading with an Agila two-roll mill at 130[degrees]C for 10 min. The collected molten materials were compression molded into 3-mm-thick plates by hot-pressing at 10000 for 10 minutes under atmosphere pressure, followed by compression under 150 bars for 10 seconds, and then under 30 bars for 10 seconds, followed by cold pressing at 15[degrees] under 30 bars for 5 minutes.

Composites Preparation by in situ Intercalative Polymerization

Before polymerization, the non-modified [Na.sup.+]-MMT was dried under vacuum at 100[degrees]C for one night and the organo-modified MMT was dried in a ventilated oven at 70[degrees]C for one night. In a polymerization tube, the desired amount of layered silicate was further dried under vacuum at 70[degrees]C for 3 h. A given amount of [epsilon]-caprolactone was then added under nitrogen, and the reaction medium was stirred at room temperature for 1 h. A solution of initiator (Sn[(Oct).sub.2] or [Bu.sub.2]Sn[(OMe).sub.2]) in dry toluene was added to the mixture in order to reach a [[monomer].sub.0]/[Sn] molar ratio of 300. The polymerization was allowed to proceed at room temperature for 24 hours. The inorganic content of each composite was checked by TGA. After polymerization, a reverse ion-exchange reaction was used to isolate the PCL chains from the inorganic fraction of the nanocomposite. A colloidal suspension was obtained by stirring 2 g of the nanocomposite in 30 mL of THF for 2 h at room temperature. Separately, a solution of 1 wt% of LiC1 in THF was prepared. The nanocomposite suspension was added to 50 mL of the LiC1 solution and left to stir at room temperature for 48 h. The resulting solution was centrifuged at 3000 rpm for 30 min. The supernatant was decanted and the solid was washed by dispersing in 30 mL of THF followed by centrifugation. The combined supernatant was concentrated and precipitated from petroleum ether. The white powder was dried in vacuum at 50[degrees]C. The absence of residual PCL in the silicate was checked by FT-IR (absence of carbonyl absorption band at 1727 [cm.sup.-1]).

Characterization

The morphology of the composites was analyzed by X-ray diffracion (XRD) and transmission electron microscopy (TEM). The XRD patterns were collected in a digital form using a Siemens D5000 diffractometer with Cu-[K.sub.[alpha]] radiation ([lambda] = 0.15406 nm) between 1.5 and 30[degrees] by step of 0.04[degrees]. TEM observations were performed with a Philips CM100 apparatus using an acceleration voltage of 100 kV. Ultrathin sections of the composites with a thickness of approximately 80 nm were accordingly cut at -130[degrees]C from the 3-mm-thick plates by using a Reichert-Jung Ultracut 3E, FC4E ultra-cryomicrotome equipped with a diamond knife. Owing to the high electron density difference between the silicate layers and polymer matrix, no sample staining was necessary. Thermogravimetric analysis (TGA) was performed under air flow (74 ml/min) at a heating rate of 20[degrees]C/min by using a Hi-Res TGA 2950 from TA Instruments.

Tensile tests were performed at 20[degrees]C with a Lloyd tensile testing apparatus. Tensile properties were measured at 20[degrees]C with a constant deformation rate of 50 mm/min with a Lloyd LR 10K tensile tester with dumbbell-shaped specimens prepared from compression molded samples according to the 638 type V ASTM norm. Size exclusion chromatography (SEC) measurements were carried out in THF (sample concentration: 1 wt%) at 3500 using a Polymer Laboratory (PL) liquid chromatograph equipped with a PL-DG802 degazer, an isocratic HPLC pump LC1120 (flow rate: 1 mL/min), a Basic-Marathon Autosampler, a PL-RI refractive index detector and three columns: a guard column PLgel 10 [mu]m (50 x 7.5 mm) and two columns PLgel mixed-B 10 [mu]m (300 X 7.5 mm). Molar masses were calculated by reference to a PS standard calibration curve, using the Mark-Houwink-Sakurada relationship [[eta]] = K.[M.sub.[eta].sup.a] ([K.sub.PS] = 1.25 X [10.sup.-4] dl/g, [a.sub.PS] = 0.707, [K.sub.PCL] = 1.09 X [10.sup.-3] dl/g, [a.su b.PCL] = 0.600).

RESULTS AND DISCUSSION

Melt Intercalation

PCL was melt blended in a two-roll mill at 130[degrees]C with a known amount of MMT, either [Na.sup.+]-MMT or MMT previously exchanged by ammonium cations bearing long alkyl chains (MMT-Alk). Ammonium cations containing alkyl chains end-capped by a carboxylic acid (MMT-COOH) or a hydroxyl group (MMT-[(OH).sub.2]) have been also studied. The quantity of filler added is such that the final composition of the composites is 3 wt% of inorganics (layered aluminosilicates). The composites have been analyzed by XRD. Some of the organo-modified MMT (PCLC3: MMT-Alk, PCLC4: MMT-[(OH).sub.2]) show a significant increase in the interlayer distance attesting for the effective polymer intercalation (Table 2). However, the XRD analysis of natural sodium montmorillonite (PCLC 1: [Na.sup.+]-MMT) and montmorillonite organically modified by ammonium cations bearing a carboxylic acid function (PCLC2: MMT-COOH) shows that the interlayer spacing remains unchanged and thus microcomposites are formed rather than nanocomposites (Tabl e 2).

Although PCL nanocomposites cannot be prepared by melt intercalation of PCL within MMT modffied by the protonated 12-dodecanoic acid (PCLC2: MMT-COOH), the in situ intercalative polymerization of [epsilon]-caprolactone In the same organophilic clay was successful (15, 17). It is thus clear that the nature of ammonium cation (and the presence of functional groups on the alkyl chains) used as organic modifier of the silicate layers and the preparation route (in situ intercalative polymerization or melt intercalation) play a key role in the success of the nanocomposite formation. In other words, even if a given organo-modified layered silicate can intercalate a monomer with the formation of an intercalated and/or exfoliated nanocomposite material upon consecutive polymerization, it does not mean that nanocomposites can be generated by melt blending the corresponding polymer and the same organophilic clay. As far as melt intercalation is concerned, the impossibility of forming a PCL nanocomposite with a MMT surfa ce-modified by the protonated 12-dodecanoic acid (PCLC2: MMT-COOH) cannot be explained by the simple presence of polar functionalities attached onto the cations since nanocomposites can be obtained by mixing PCL with hydroxyl-functionalized clays, i.e., PCLC4: MMT-[(OH).sub.2]. It comes out that the type and localization of the functional group on the modifying agent have a tremendous effect on the ability to yield nanocomposites, as will be reported elsewhere (17).

The morphology of the PCL-based composites has been observed by TEM. As expected, a microcomposite morphology is reported for PCLC1: [Na.sup.+]-MMT and PCLC2: MMT-COOH. In both cases, micron-size clay particles are randomly dispersed in the polyester matrix. Consistently with the XRD analysis, PCLC3: MMT-Alk (Fig. 1) or PCLC4: MMT-[(OH).sup.2] shows the typical morphology of a nanocomposite. In addition to small stacks of intercalated montmorillonite, exfoliated silicate sheets are observed, with a preferential orientation perpendicular to the melt compression direction. This type of morphology may be designated as a semi-intercalated/semi-exfoliated structure.

Mechanical Properties

Tensile properties, i.e., Young's modulus (E), stress at break ([[sigma].sub.b]) and elongation at break ([[epsilon].sub.b]), have been measured and are reported in Table 3. These data confirm that PCL is a ductile polymer able to sustain large deformations. Unfortunately, the elastic modulus is rather low, making it useless for any application that requires high rigidity. Thus, the addition of fillers can contribute to improve its stiffness. Even at low filler content. i.e., 3 wt% of inorganics, the elastic modulus is significantly increased from 210 to more than 270 MPa. A modulus as high as 280 MPa is even measured for PCL nanocomposites (PCLC3: MMT-Alk). Interestingly enough, the PCLC3 and PCLC4 samples show a much higher modulus and retain good ductility as confirmed by an elongation at break higher than 500%. Similar observations have been previously reported for nanocomposites based on other thermoplastics, such as polyamides and polyurethanes (1). It is worth noting that an increase of the filling lev el triggers a further enhanced of the material rigidity, with a two-fold increased stiffness with ~10 wt% of MMT-[(OH).sub.2] for instance.

Thermal Properties

The thermal stability of the PCL-based composites has been also examined by TGA with a heating rate of 20[degrees]C/min under air flow. The weight loss due to the formation of volatile degradation products has been monitored as a function of temperature. It has been recently reported that the thermal degradation of PCL fits a two-step mechanism (18). There is first a statistical rupture of the polyester chains by pyrolysis of ester groups with release of C[O.sub.2], [H.sub.2]O and hexenoic acid. In the second step, [epsilon]-caprolactone is formed as result of an unzipping depolymerization process. The PCL filled with MMT-Alk (PCLC3) or MMT-[(OH).sub.2] (PCLC4) shows a substantial improvement of the thermal stability. Indeed, the temperature at which the weight loss is 50 wt%, is shifted towards higher temperature by ~50[degrees]C. In case of the microcomposite (PCL filled with MM-COOH: PCLC2), the gain in stability is less important compared to the parent nanocomposite. Thus, the increase in thermal stabili ty observed for the nanocomposite is to be related to the nanodispersion of the silicate. The silicate layers are thought to oppose an effective barrier to the permeation of oxygen and combustion gas (1). A similar thermal stabilization has been recently reported by some of us for nanocomposites based on ethylene-co-vinyl acetate copolymers (EVA) filled with organo-modified MMT (11). Furthermore, it has been found that PCL nanocomposites exhibit remarkable flame-retardant properties. Although PCL and PCL-based microcomposites (PCLC1-C2) continuously release burning drops able to propagate the fire to surrounding materials when they are exposed to a flame, PCL nanocomposites (PCLC3-C4) show a totally different behavior. Burning drops are no longer formed; rather, an intensive charring of these PCL nanocomposites is observed. The fire-retardant capacity of the PCL nanocomposites is under current investigation by using cone-calorimetry measurements.

In situ Intercalative Polymerization

The PCL nanocomposites were prepared by in situ intercalative polymerization of [epsilon]-caprolactone in the presence of natural [Na.sup.+]-MMT and MMT modified by various alkylammonium cations. The synthetic approach involves dispersion of the silicates in the monomer, followed by polymerization either by thermal activation or by catalytic activation using organometallic compounds.

Nanocomposites were first produced by [epsilon]-caprolactone (CL) polymerization in bulk and by thermal activation at 170[degrees]C during 24 hours. Table 4 presents results obtained for a variety of montmorillonites (3 wt%). Molar masses of poly([epsilon]-caprolactone) matrix were analyzed by SEC after polymer extraction with lithium chloride solution. The polymerization of CL with [Na.sup.+]-MMT gives polymer with a molar mass of 4.8 kg/mol and a narrow distribution (sample PCLC6). As a comparison, in the same conditions but without clay, no polymerization of CL occurs. These results demonstrate the ability of MMT to catalyze and to control CL polymerization at least in terms of molecular weight distribution that remains remarkably narrow. For the mechanism of polymerization, we assume that the monomer is activated through interaction with acidic site on clay surface. In the literature, such an effect was previously reported by Aida et al. (19) for polymerization of lactones in the presence of an aluminosi licate mesoporous zeolite. According to the authors, the polymerization is likely to proceed via the activated monomer mechanism by the cooperative function of Lewis acidic aluminum and Bronsted acidic silanol functionalities on the interior walls. The mesoporous silicate framework is considered to provide an exceptionally wide surface, which is beneficial for the monomer activation and the accessibility of the growing polymer molecules to the monomer activation site.

For the polymerization of CL with the protonated [omega]-amino dodecanoic acid exchanged MMT, the molar mass is 7.8 kg/mol with a high monomer conversion (92%) and again a narrow molecular weight distribution (sample PCLC8).

To determine the degree of nanoparticle dispersion within the polymer matrix, nanocomposites were analyzed by X-ray diffraction. For the natural [Na.sup.+]-MMT and for the protonated [omega]-amino dodecanoic acid exchanged MMT, XRD patterns show the formation of an intercalated structure. It must be noted that such an intercalation has not been previously observed with melt intercalation procedure. Only a microcomposite morphology was reported (16).

PCL nanocomposites were also prepared by catalytic activation of the in situ polymerization with organometallic compounds. By comparison to the thermal activation method, catalyzed ring-opening polymerization allows higher monomer conversion. Several organometallic compounds, such as alkoxides and carboxylates, can be successfully used as catalysts or initiators for ring-opening polymerization of lactones and lactides according to a "coordination-insertion" mechanism (20, 21). The polymerization proceeds through the insertion of the monomer into the "metal-O" bond of the initiator via a selective acyl-oxygen cleavage of the lactone ring. Tin derivatives are the most widely used catalyst in such lactone polymerization.

First, polyester nanocomposites were prepared by polymerization of s-caprolactone in presence of layered silicates (3 wt%) and by polymerization activation with stannous octoate (Sn[(Oct).sub.2]) at 100[degrees]C during 24 hours, results are reported in Table 5. For organomodified clays (samples PCLC12 to PCLC14), nanocomposites are obtained with high conversion (> 90%) and with molar masses between 28 and 60 kg/mol. The polydispersity index is in the range of 1.5-2. The formation of an exfoliated structure for MMT-[(OH).sub.2] is revealed by XRD and TEM (Fig. 2).

Tin catalysts with oxydation degree IV are known to be more efficient than those with oxydation degree II towards [alpha]s-caprolactone ring-opening polymerization. In this purpose, dibutyltin dimethoxide (22), [Bu.sub.2]Sn[(OMe).sub.2] was used for the synthesis of polycaprolactone-based nanocomposites using mild experimental conditions, i.e., at room temperature.

PCL nanocomposites were prepared with various amounts of the modified silicate MMT-[(OH).sub.2] with functional ammonium bearing two hydroxyl groups. The so-obtained nanocomposites exhibited a continuous decrease of molar masses with clay concentration, as indicated in Table 6. This behavior can be explained by the presence of the hydroxyl functions, which can act as co-initiator/chain transfer agent. In the presence of an alcohol, the "coordination-insertion" polymerization with tin alkoxides involves a rapid exchange reaction between tin alkoxides species and ammonium-fixed alcohol molecules, as schematized in Eq 1 (23):

[LANGUAGE NOT REPRODUCIBLE IN ASCII]

In such a system, the propagating species (metal alkoxide) is temporarily converted into a dormant site (alcohol) through the reaction with another alcohol molecule in the reaction medium. This equilibrium between active and dormant species (reversible chain transfer) contributes to the polymerization control. The final degree of polymerization ([DP.sub.n]) is given by: [DP.sub.n] = [[M.sub.o]]/ ([metal [alkoxide].sub.o] + [transfer agent]0). Two types of polymer chains are therefore present: those initiated by the methoxy groups from [Bu.sub.2]Sn[(OMe).sub.2] and those initiated by hydroxyl groups on the clay surface after activation in tin alkoxides through co-initiation and reversible chain transfer reactions. A decrease of polydispersity index together with an increase in clay concentration is also observed for samples prepared with MMT-[(OH).sub.2]. This behavior may be explained by the confining of most of the propagation centers within the clay interlayers that could impose limitations on the mobility of tethered PCL chains, hindering transesterification side reactions.

These results indicate that in the presence of an ammonium bearing initiating groups such as hydroxyl function, the average length of the PCL chains therefore grafted onto the organo-modified layered silicates could be predetermined by taking into account the number of these hydroxyl functions, i.e., the relative content in clay particles.

Concerning the nanocomposite prepared with 3 wt% of the hydroxyl-functionalized clay MMT-[(OH).sub.2], XRD analysis gives evidence for an exfoliated structure, as attested by the absence of a diffraction peak for sample POLO 17 (Fig. 3). The evolution of XRD patterns with the relative clay content is also presented. For high clay concentration (10 wt%, sample PCLC19), a broad peak in the small angle region probably indicates the formation of a partially exfoliated/partially intercalated structure. A decrease in polymer matrix crystallinity with increasing silicate content is also evidenced by the decrease in intensity of PCL-characteristic diffraction peaks at 0.41, 0.37 and 0.33 nm. These observations are in total agreement with data previously published by Giannelis et al. (13).

Thermal Properties

Nanocomposite thermal stability was studied by TGA. Both intercalated and exfoliated structures show higher thermal stability when compared with the unfilled polymer or the corresponding microcomposite. Thermograms of nanocomposites prepared with MMTAIK and pure polymer recovered after clay extraction are presented in Fig 4. The nanocomposite exhibits a 25[degrees]C increase in decomposition temperature at 50% weight loss. The shift of the degradation temperature may be ascribed to a decrease in oxygen and volatile degradation products permeability/diffusivity due to the homogeneous incorporation of clay sheets, to a barrier of these high aspect ratio fillers and char formation. The thermal stability increases with clay concentration. However, above a loading level of 5 wt% (not shown here), the thermal stability is no longer improved. The same effect is visible regardless of the studied clay (24).

At this stage, the tensile properties of the nanocomposites obtained by intercalative in situ polymerization have not been determined. Indeed, the PCL molar mass varies from one experiment to the other and therefore displays a strong effect on the mechanical properties of the resulting compositions. Rather, highly filled (nano) compositions produced by in situ polymerization have been considered as "masterbatch" melt blended with a given commercial PCL. These results, including their mechanical behavior, will be the topic of a forthcoming paper (25).

CONCLUSIONS

Poly([epsilon]-caprolactone) layered nanocomposites were prepared via two synthetic routes: melt-intercalation and in situ intercalative polymerization.

The melt intercalation is a solvent-free approach involving mixing of the layered silicate with molten PCL. Depending on the type of the modifying agent, the stiffness of the PCL nanocomposites can be significantly improved compared to neat PCL. even at a filler content as low as 3 wt% of inorganic layered silicate. Flame retardancy is remarkably improved and related to the deposition of an insulating and incombustible char whenever the PCL nanocomposites are exposed to the flame (16).

By in situ polymerization, nanocomposites are prepared by polymerization of [epsilon]-caprolactone in the presence of layered silicates and by initiation of polymerization with tin derivatives (Sn[(Oct).sub.2], [Bu.sub.2]Sn[(OMe).sub.2]) or by thermal activation. Exfoliated structures have been produced by in situ polymerization in the presence of MMT surface-modified with ammonium cations bearing hydroxyl groups. The lactone polymerization is initiated by all hydroxyl groups available at the clay surface after their activation by the tin (IV) dialkoxide. Size exclusion chromatography measurements of the extracted PCL chains show number average molar masses that can be predetermined and controlled by the amount of dispresed clay. PCL/clay nanocomposites display higher thermal temperature degradation than the corresponding unfilled polymer. Hybrid polyester nanocomposites generated through the covalent grafting of polyester chains onto the filler surface exhibit higher thermal stability than nanocomposites fil led with non-functional clays (24).

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Table 1

Characteristics of Montmorillonites Used.

  Commercial name of
         MMT            MMT code          Cations in MMT

Cloisite[R] [Na.sup.+]  [Na.sup.+]-MMT    [Na.sup.+]
    Nanofil[R] 820      MMT-COOH          HOOC-[C.sub.11][H.sub.22]
                                          [NH.sup.+.sub.3]
   Cloisite[R] 25A      MMT-Alk           [([CH.sub.3]).sub.2]
                                          [N.sup.+]([C.sub.18]
                                          [H.sub.37])([C.sub.8]
                                          [H.sub.17])
   Cloisite[R] 30B      MMT-[(OH).sub.2]  ([CH.sub.3])([C.sub.18]
                                          [H.sub.37])[N.sup.+]
                                          [([CH.sub.2][CH.sub.2]
                                          OH).sub.2]

  Commercial name of    Organic fract-
         MMT            ion (8) (wt%)

Cloisite[R] [Na.sup.+]        --
    Nanofil[R] 820           11.2

   Cloisite[R] 25A           14.1



   Cloisite[R] 30B           20.1




(a) Determined by TGA.
Table 2

Interlayer Spacing Before and After Melt Blending With PCL.

                             Interlayer spacing (a) (nm) in

Sample    Used MMT          MMT               PCL composite (b)

PCLC1     [Na.sup.+]-MMT    1.21                    1.23
PCLC2     MMT-COOH          1.38                    1.37
PCLC3     MMT-Alk           1.86                    2.70
PCLC4     MMT-[(OH).sub.2]  1.84                    3.10

(a) Determined by XRD.

(b) PCL composite filled with 3 wt% of layered silicates (precluding the
organic layer).
Table 3

Tensile Properties of Poly ([epsilon]-caprolactone) and Its Composites.

Sample   Used MMT          Young's modulus (MPa)  [[epsilon].sub.b](%)

PCL (1)  --                   217 [+ or -] 5        745 [+ or -] 45
PCLC1    [Na.sup.+]-MMT       200 [+ or -] 9        715 [+ or -] 45
PCLC2    MMT-COOH             240 [+ or -] 10       710 [+ or -] 40
PCLC3    MMT-Alk              282 [+ or -] 9        530 [+ or -] 60
PCLC4    MMT-[(OH).sub.2]     272 [+ or -] 16       560 [+ or -] 60

Sample   [[sigma].sub.b] (MPa)

PCL (1)      37 [+ or -] 2
PCLC1        35 [+ or -] 3
PCLC2        37 [+ or -] 5
PCLC3        26 [+ or -] 3
PCLC4        25 [+ or -] 4

(1) PCL with a number-average molar mass of 50,000.
Table 4

Thermal Intercalative Polymerization of [epsilon]-caprolactone in
Presence of MMT-[170[degrees]C-24 Hours.

Sample  Used MMT          Filler content (wt%)  Conversion (%)

PCLC5   none                       --                  0
PCLC6   [Na.sup.+]-MMT             3                  52
PCLC7   MMT-Alk                    3                   6
PCLC8   MMT-COOH                   3                  92
PCLC9   MMT-[(OH).sub.2]           3                  30

Sample  [M.sub.n] (1) (kg/mol)  [M.sub.w]/[M.sub.n]

PCLC5             --                    --
PCLC6            4.8                   1.18
PCLC7             --                    --
PCLC8            7.8                   1.15
PCLC9            2.0                   1.12

(1) Number-average molar mass in PCL equivalent.
Table 5

Catalytic Intercalative in situ Polymerization of CL in Presence of
MMT-Sn[(Oct).sub.2]-[[CL].sub.0]/[Sn] = 300-100[degrees] C-24 Hours.

Sample  Used MMT          Filler content (wt%)  Conversion (%)

PCLC10  none                       --                 85
POLC1l  [Na.sup.+]-MMT             3                  72
PCLC12  MMT-Alk                    3                  97
PCLC13  MMT-COOH                   3                  97
PCLC14  MMT-[(OH).sub.2]           3                  97

Sample  [M.sub.n] (1) (kg/mol)  [M.sub.w]/[M.sub.n]

PCLC10           80.1                  2.99
POLC1l           22.9                  1.37
PCLC12           59.9                  2.08
PCLC13           28.0                  1.53
PCLC14           37.2                  1.62

(1) Number-average molar mass in PCL equivalent.
Table 6

Intercalative in situ Polymerization of CL in Presence of
MMT-[(OH).sub.2]-[Bu.sub.2]Sn[(OMe).sub.2]-[[CL].sub.o]/[Sn] = 300-Room
Temperature-24 Hours-Influence of Filler Content.

           Filler                      [M.sub.n,th]  [M.sub.n] (1)
Sample  content (wt%)  Conversion (%)  (2)(kg/mol)     (kg/mol)

PCLC15        0              96            17.1          21.0
PCLC16        1              91            13.7          16.1
PCLC17        3              95            11.2          13.5
PCLC18        5              87             8.8           9.0
PCLC19       10              82             5.6           4.6

        [M.sub.w]/
Sample  [M.sub.n]

PCLC15     2.06
PCLC16     1.95
PCLC17     1.81
PCLC18     1.67
PCLC19     1.57

(1) Number-average molar mass in PCL equivalent

(2) [M.sub.n,th] = [(M).sub.o]/(l)x[M.sub.CL]xconversion considering
that all alkoxides groups are active (coming from
[Bu.sub.2]Sn[(OMe).sub.2] and ammonium cation MMT-[OH.sub.2]).

ACKNOWLEDGMENTS

The SMPC is much indebted to the Region Wallonne and the Fonds Social Europeen for support in the frame of Objectif I-Hainaut: Materia Nova. SMPC and GERM are grateful to the Region Wallanne for support in the frame of the W.D.U. program: TECMAVER, including a grant to N.P and B.L. CERM thanks the "Services Federaux des Affaires Scientifiques, Techniques et Culturelles" for support (PAI 4/11).

REFERENCES

(1.) M. Alexandre and P. Dubois, Mater. Sci. Eng., R28, 1 (2000).

(2.) J. M. Gloaguen and J. M. Lefebvre, Polymer, 42, 5841 (2001).

(3.) L. Liu, Z. Qi, and X Zhu, J. Appl. Polym. Sci., 71, 1133 (1999).

(4.) Y. Ke, C. Long, and Z. Qi, J. Appl. Polym. Sci., 71, 1139 (1999).

(5.) M. W. Weimer, H. Chen, E. P. Giannelis, and D. Y. Sogah, J. Am. Chem. Soc., 121, 1615 (1999).

(6.) D. J. Suh, Y. T. Lim, and 0. 0. Park, Polymer, 41, 8557 (2000).

(7.) A. B. Morgan, J. W. Gilman, and C. L. Jackson, Macromolecules, 34, 2735 (2001).

(8.) N. Hasegawa. H. Okamoto, M. Kawasumi. M. Kato, A. Tsukigase, and A. Usuki, Macromol. Mater. Eng., 280/281, 76 (2000).

(9.) X. Huang and W. J. Brittain, Macromolecules. 34, 3255 (2001).

(10.) M. Zanetti, G. Camino, R. Thornann, and R. Mulhaupt, Polymer. 42, 4501 (2001).

(11.) M. Alexandre, G. Beyer, C. Henrist, R Cloots, A. Rulmont, R. Jerome, and P. Dubois, Macromol. Rapid Comm., 22, 643 (2001).

(12.) R. A. Vaia, S. Vasudevan, W. Kramiec, L. G. Scanlon, and E. P. Giannelis, Adv. Mater., 2, 154 (1995).

(13.) P. B. Messersmith and E. P. Giannelis, J. Polym. Sci., Part A: Polym. Client, 33, 1047 (1995).

(14.) T. K. Chen, Y. I. Tien, and K. H. Wei, J. Polym. Sci., Part A: Polym. Client, 37, 2225 (1999).

(15.) P. B. Messersmith and E. P. Giannelis, Client Mater., 5, 1064 (1993).

(16.) N. Pantoustier, M. Alexandre, P. Degee, C. Calberg, R. Jerume, C. Henrist, R. Cloots, A. Rulmont, and P. Dubois, e-Polymer, 9, 1 (2001).

(17.) N. Pantoustier, M. Alexandre, P. Degee, C. Calberg, R. Jerome, C. Henrist, R. Cloots, and P. Dubois, in preparation.

(18.) O. Persenaire, M. Alexandre, P. Degee, and P. Dubois, Biomacromolecules, 2, 288 (2001).

(19.) a) K. Kageyama, S-I. Ogino. T. Aida, and T. Tatsumi, Macromolecules, 31, 4069 (1998) b) K. Kageyama, T. Tatsumi, and T. Aida, Polym. J., 31, 1005 (1999).

(20.) A. Lofgren. A-C. Albertsson, P. Dubois, and R Jerome, J. Macromol. Sci.-Rev. Macromol. Chem. Phys., C35(3), 379 (1995).

(21.) A. Duda. S. Penczek, A. Kowailski, and J. Libiszowski, Macromol. Symp., 153. 41 (2000).

(22.) H. R Kricheldorf, M. Berl, and N. Scharnagl. Macromolecules, 21, 286 (1988).

(23.) S. Penczek, T. Biela, and M. Duda, Macromol. Rapid. Commun., 21. 941 (2000).

(24.) B. Lepoittevin. M. Devalcknaere, M. Alexandre, N. Pantoustier, C. Calberg, R. Jerome, and P. Dubois, Macromolecules, submitted for publication.

(25.) B. Lepoittevin, N. Pantoustler, M. Devalcknaere, M. Alexandre, C. Calberg, R Jerome, C. Henrist, A. Rulmont, and P. Dubois, Chem. Mater., submitted for publication.

ABBREVIATIONS

a: Mark-Houwink coefficient

[Bu.sub.2]Sn[(OMe).sub.2]: dibutyltin dimethoxide

CEC: cation exchange capacity

CL: [epsilon]-caprolactone

[DP.sub.n]: degree of polymerization

E: Young's modulus (MPa)

[[epsilon].sub.b]: elongation at break (%)

[eta]: intrinsic viscosity (dl/g)

K: Mark-Houwink constant

[Na.sup.+]-MMT: natural montmorillonite with a CEO of 92 meq/100 g

MMT: montrnorillonite

MMT-Alk: montmorillonite organically-modified with dimethyl 2-ethyihexyl (hydrogenated tallow alkyl) ammonium cations

MMT-[(OH).sub.2]: montmorillonite organically-modffied with methyl bis(2-hydroxyethyl) (hydrogenated tallow alkyl) ammonium cations

MMT-COOH: montmorillonite organically-modified with ammonium bearing 12-aminododecanoic acid

[M.sub.[eta]: viscosimetric: average molecular weight

[M.sub.n]: number average molecular weight

[M.sub.w]/[M.sub.n]: polydispersity

PCL: poly([epsilon]-caprolactone)

[[sigma].sub.b]: stress at break (MPa)

SEC: size exclusion chromatography

Sn[(Oct).sub.2]: stannous octoate

TEM: transmission electron microscopy

TGA: thermogravimetric analysis

XRD: X-ray diffraction

Nadege Pantoustier (1,3), Benedicte Lepoittevin (1,3), Michael Alexandre (1,3), Dana Kubies (2,3), Cedric Calberg (2,3), Robert Jerome (2,3), and Philippe Dubois (1,3) *.

(1.) Laboratory of Polymeric and Composite Materials (SMPC) University of Mons-Hainaut Place du Parc 20, 7000 Mans, Belgium

(2.) Center for Education and Research on Macromolecules (CERM) University of Liege Building B6, 4000 Liege, Belgium

(3.) Research Center in Science of Polymeric Materials--CRESMAP, Belgium

* Corresponding author: Prof. Philippe Dubois. E-mail: [e-mail]