CA2199616A1

CA2199616A1 - Layered silicate-epoxy nanocomposites

Info

Publication number: CA2199616A1
Application number: CA002199616A
Authority: CA
Inventors: Phillip B. Messersmith; Emmanuel P. Giannelis
Original assignee: Individual
Current assignee: Cornell Research Foundation Inc
Priority date: 1994-09-12
Filing date: 1995-08-29
Publication date: 1996-03-21
Also published as: NO971109L; EP0782593A1; MX9701845A; EP0782593A4; US5554670A; NO971109D0; AU3419195A; JPH10505871A; BR9508914A; KR970706328A; WO1996008526A1

Abstract

An epoxy-silicate nanocomposite is prepared by dispersing an organically modified smectite-type clay in an epoxy resin together with diglycidyl ether of bisphenol-A (DGEBA), and curing in the presence of either nadic methyl anhydride (NMA), and/or benzyldimethyl amine (BDMA), and/or boron trifluoride monoethylamine (BTFA) at 100-200 ~C. Molecular dispersion of the layered silicate within the crosslinked epoxy matrix is obtained, with smectite layer spacings of 100.ANG. or more and good wetting of the silicate surface by the epoxy matrix. The curing reaction involves the functional groups of the alkylammonium ions located in the galleries of the organically modified clay, which participate in the crosslinking reaction and result in direct attachment of the polymer network to the molecularly dispersed silicate layers. The nanocomposite exhibits a broadened Tg at slightly higher temperature than the unmodified epoxy. The dynamic storage modulus of the nanocomposite was considerably higher in the glassy region and very much higher in the rubbery plateau region when compared to such modulus in the unmodified epoxy.

Description

2 ~ 9 9 6 1 6 PCT/US95/10956 LAYERED SILICATE-EPO~Y NANOCOl\IPOSITES

Field of ~nV~ntion 5 This invention relates generally to mineral-polymer coln~osil~ materials, and more specifically relates to an epoxy-~m~ctite nanocomposite and method of pre~alillgsame.

~ ~ l~round of ~nv~ntiQn Particulate minerals such as kaolins, talcs, calcium carbonate, calcium sulfate and various micas have long been utilized as inert extenders or fillers in polymers or similar matricies. Aside from providing economic advantages in ~xlP~ g the more costly polymeric material, such fillers serve in many ;I~ res to illl~rove the 15 prop~llies of the res~llt~nt plastics with respect to such ~al~lR~ as th~rm~l ~Xp~n~iQn coefficient, ~lirr..~ss and creep resi~ r-e.

It is also well known in the prior art to render fillers of the folegoing type of ascd compatability with the polymer matrix to illlpiOVe the interfacial adhesion20 of the mineral to the matrix. Thus, for example, in Pa~los. U.S. Patent No.

3,227,675, kaolin clays are described, the surfaces of which are modified with organofunctional silanes. The kaolin clays so modified are used as fillers for natural and synthetic rubbers and the like. Additional lerelellces of this type include ~nnir~lli, U.S. Patents Nos. 3,290,165 and 3,567,680. Similarly, in U.S.25 Patent No. 4,789,403, a method is disclosed for producing a layered lattice silicate which is surface modified with an organic material. The layered lattice silicate is contacted with an organic monolllel, comonolllel~, or a pre-polymer, and surfacepolyllRl~lion or reaction in situ is effected in the yrestllce of a gaseous hydrogen atmosph~re. Among the organic monomers ~at can be used in the process are 30 various precursors of nylon.

More lecellLly, processes have been disclosed which are said to be useful in producing composite m~tPri~l~ composed of a polymer and a ~ c~ile-type clay mineral, with the mineral being co~ Le~ to the polymer through ionic bonding.

W096/08526 . 2 1 9 9 6 1 6 PCT/US9S/1095~

For example, in K~wasllmi et al., U.s. Patent No. 4,810,734 a process is disclosed wherein a sm~ctite-type clay mineral is contacted with a swelling agent in the presence of a dispersion m~ lm thereby forming a complex. The complex co~ the dispersion ",~-li"." is mixed with a monomer, and the monomer is 5 then polyllleliLed. The patent states that the swelling agent acts to expand the interlayer ~i~t~n~e of the clay mineral, lll~r~y pe. ",i~ g the clay mineral to take monomers into the interlayer space. The swelling agent is a compound having a onium ion and a functional ion capable of lea.;lillg and bonding with a polymer compound. Among the polymers utilizable are polyamide resins, vinyl polymers, 10 thermosetting resins, polyester resins, polyamide resins and the like. Related disclosures are found in U.S. Patents Nos. 4,739,007 and 4,889,885.

The swelling agents used in the K~".~.."~i et al. and related patents cited above, technic.~lly qualify as organoclays. In the present invention as well, org~ni- ~lly 15 modified ~mPctit~-type clays, h~leil arlel referred to as "olg~nophilic" or "organoclays", are used as the mineral col~ollellL of the colllposil~. In general, organoclays leplesent the reaction product of a ~mPctitÇ-type clay with a higheralkyl cont~ining ammonium colll;)uulld (often a qll~t~. "c, y), and have long been known for use in gelling of organic liquids such as lubricating oils, linseed oil, 20 toluene and the like and for use as rheological additives in a variety of organic based liquid ~y~ s and solvents. The general ploce-lules and chemic~l reactions pursuant to which these organoclays are plepalcd are well known. Thus under al)proplia~e conditions the organic compound which contains a cation will react by ion exchange with clays which contain a negative layer lattice and exchangeable 25 cations to form the ol~,.ncl~y products. If the organic cation contains at least one alkyl group coll~ ing at least ten carbon atoms then the reslllt~nt organoclays will have the ~lopelly of swelling in certain organic liquids. Among the prior art patents which discuss at length aspects of the plt~alalion and properties of organoclays are U.S. Patent Nos. 2,531,427, 2,966,506, 3,974,125, 3,537,994, and30 4,081,496.

As utilized in the present s~ecirlcalion, the term "~m~ctitç"or "~ c~;le-type clays"
refers to the general class of clay minerals with ~xl.~n~ crystal lattices, with the W O 96/08526 2 1 9 9 6 1 6 PCTrUS95/10956 exception of v~....i~ .~lite. This inrl ldes the dioctAhf rlral ~ f ~;1;l~ s which consist of montrnorillonite, beidellite, and nollllo~ e, and to the trioctAhP~lral ~...f~Cl;lPs, which inrll~des sapolliLe, hect~ e, and sauconite. Also enro~ a~sed are sllleclile-clays plepalcd synthf-tir-Ally, e.g. by hydlull. - ...Al processes as r1i~rlose-l in U.S. Patents 5 Nos. 3,252,757; 3,586,468; 3,666,407; 3,671,190; 3,844,978; 3,844,979;
3,852,405; and 3,855,147.

The phase dispersions exhibited by the composite materials thus far discussed are relatively coarse, and differ m~teri~lly in this respect from nanocomposites. The latter 10 are a relatively new class of m~tPri~l~ which exhibit ultrafine phase ~limf n~jons, typically in the range 1-lOOnm. Experimental work on these m~tçri~l~ has generally shown that virtually all types and classes of nanocomposites lead to new and improved propcllies when CUlllpal~d to their micro- and macrocomposite Coulllcl~
While the number of nanocomposites based on smectite-type clays and linear therrnoplastics is growing, little work has been devoted to cros~linkf cl polymeric systems such as epoxies. Recent reports of particulate-based epoxy composites suggest that the dimensional stability, conductivity, mech~nical, thermal and other 20 properties may be modified due to the incol~uldLion of filler particles within the epoxy matrix. For the most part, however, the improvements in plopcllies observed with these conventionally plcpaled composites are modest when cûnlpared (on an equal volume basis of particulate filler) to those that have been established for various polymer-ceramic nanocomposites.
Previous work by the present inventors on poly(imide), and poly(~-caprolactone) have demo..~l . al~d the feasibility of di~l,elsing molecular silicate layers within a lllaclulllolecular matrix, which results in ~ignifi~nt i,ll~rovclllents in physical plo~cllies with only modest particulate contents (< 10% by volume).
Wang and Pinnavaia have lecelllly reported del~Tnin~tion of an org~ni~lly modified r in an epoxy resin by heating an onium ion exchallged form of montmûrillonite with epoxy resin to lelllpel~lules of 200-300~C. Chf .~ of WO 96/08526 2 1 9 9 6 1 6 PCT/US95/1095~

Materi~l~, vol. 6, pages 468474 (April, 1994). X-ray and electron lmCl~OSCO~y studies of the composite suggested de!~min~tion of the silicate layers, ~lth. ugh phase segregation of the polyether-coated ~m~ctit~ from the epoxy matrix was observed. Fu~ lore, the product of the high le~ c,d~ule curing reaction is an 5 intractable powder rather than a continuous solid epoxy matrix.

In accordance with the fofego,l~g, it may be regarded as an object of the present invention to provide a ~ clil~-epoxy nanocomposite which can be mixed, applied in various forms (e.g. as adhesive films, coatings, or c~ctings), and cured by 10 coflvc;l,lional means;

A further object of the invention is is to synth~si7e a polymer-ceramic nanocomposite in which ~.-.P~ -type org~nt cl~ys individual layers with a thi~nPss of lOA and a high aspect ratio (100-1000) are dispersed within a cros~lin~o(l epoxy l 5 matrix.

A yet fur~er object of the invention, is to provide a process for the plcpalalion of a e-epoxy nanocomposite which fulfills the above re4uile,l,~"~, and is processed using co"~ lional epoxy curing agents at ~ cl~.lul~,s signifir~ntly 20 lower than those previollsly ~ltili7~

A still further object of the invention, is to provide a process for pftp~illg alile-epoxy composite, in which the resnlting composite e~ molecular dispersion of the silicate layers in the epoxy matrix, good optical clarity, and25 si~nifi~ntly improved dynamic "~rhAnil~l propc,lies cOl~al~1 to the unmodified epoxy.

Sl.. ~.~ of the Inventio~

30 Now in acco,dance with the present invention, a method for prep~ g an epoxy-~m~cthe-type clay ~,ocolll~osite is provided, accof~ulg to which there is di~e,~ed in an epoxy resin a dry ~...~clile-type-clay which has been m-~ifi~d to an or~~nncl~y by ion e~rc~nge with an alkyl anllllolliulll salt, together with diglycidyl ether of bi~henol A (DGEBA). The positive ion of the salt is of the general form+NH3R" +NH2R2R3, +NHR4R5R~, or +NR7R8RgRlo, whtlcin R, through Rlo are organic radicals; and wllclc~ R~, at least one of R2 and R3, at least one of R4, R5, and Rs~ and at least one of R" R8, R9 and Rlo, contain a functional group capable 5 of reacting and bonding with the epoxy upon crosslinkin~ of same, such as hydroxy or epoxy, or carboxylic. Preferably an ~mmonillm salt is used which has at leastone aLkyl ammonium chain having a tellllinal hydroxyl group. A particularly plercllcd ammonium salt culll~ises a bis(2-hydroxyethyl)methyl tallow allyl ammonium salt. The ll~i~Llule is cured in the prcsellce of a curing agent which either 10 cross-links the DGEBA in the ~lcsellce of the organoclay, reacts dh~ectly with the organoclay, or catalyzes the crosslinkin~ reaction btlween the organoclay and DGEBA. This enables di~ei~ion of the organoclay in the dry state, and enables curing of the nanocomposite to occur at much lower t~ eraluies than in the priorart. In addition, formation of c1l~rniral bonds b~lvve,_~ the crosslink~d ll~lwol~ and 15 the silicate l~y~licles results in direct Att~ m~nt of the epoxy matrix to the silicate layers, thereby m~xi~ g a&esion belweel1 the two phases. Curing is typically carried out at telll~tl~lul~s in the range of 100 to 200~C. The ~ clile most preferable for use in the invention is montmorillonite, the structure of which consi~L~ of layers made up of one oct~h~dral al~min~ sheet sandwiched between two 20 tectrahedral silica sheets. The curing agent may be sele-cte~l from one or more melllbel~ of the group co~ g of nadic methyl anhydride (NMA), benzyl~ r~ ylamine (BDMA), and boron trifluoride monoethylamine (BTFA).

Brief Descliy~ion of Drawi~s In the drawings appended hereto:

FIGURE 1 depicts the XRD diffraction paLLerns of a dry organoclay pûwder and the uncured org~nt c!~y/DGEBA llli~lule.
FIGURE 2 depicts XRD p~ll. "c of org~nncl~y/DGEBA ll~i~lure (4% MTS
by volume) heated in situ to various l~ eralules. The spectra are displaced vertically for clarity, with scan l~lllyClalul~,S (in ~C) from bottom to top as follows:

wo s6/oss26 2 1 9 9 6 1 6 PCT/USg5/l0956 20; 50; 70; 90; 100; 110; 120; 130; 140; 150. The dashed lines in~ir~e the location of the silicate (001) and (002) reflections at 20~C.

FIGURE 3 is an XRD pattern of fully cured og~nocl~y/DGEBA/MDA
5 composite cont~ining 2% OMTS by volume. The silicate (001) reflection corresponds to a layer spacing of 36~.

FIGURE 4 depicts XRD p~ s of organoclay/DGEBA/BDMA Il~ixLule (4% organoclay by volume) heated in situ to various telll~elalules. The spectra are 10 displaced vertically for clarity, with scan l~ ldlules (in C~) from bottom to top as follows: 20; 40; 50; 60; 70; 80; 90; 100; 110; 130; 150. The dashed lines inrlir~tto the location of the silicate (001) and (002) reflections at 20~C.

FIGURES 5 and 6 respecLivt;ly depict XRD p~ s of fully cured 15 organoclay/DGEBA/BDMA and org~nocl~y /DGEBA/NMA nanocomposites cont~inin~ A: 0.4% B:1.2% C: 2% D: 4% org~nocl~y by volume. Spectra are displaced vertically for clarity.

FIGURE 7 shows TEM micrographs of thin sections of fully cured 20 oganoclay/DGEBA/NMA nanocomposite cont~ining 4% MTS by volume Dispersed silicate layers are viêwed edge-on and are clearly visible as dark lines of thirl~nP~ ap~o~ill~lely loA, with 80-120A of epoxy matrix sep~alillg neighboringsilicate layers. Scale bars =a) 100nm and b) 10nm.

FIGURE 8 depicts FT-IR spectra of A)DGEBA/BDMA and B)organoclay/DGEBA/BDMA (4 vol% MTS) resin llli~lUleS taken at various élll~eralures during heating in situ at 0.5~C/min.

FIGURE 9 depicts DSC curing scans of organoclay/DGEBA/BDMA (dashed line) and DGEBA/BDMA (solid line).

FIGURE 10 shows DSC curing scans of organoclay/DGEBA/NMA (dashed line) and DGEBA/NMA (solid line); and W0 96/08526 2 1 9 9 6 1 6 PcrluS95/10956 FIGURE 11 depicts lc~ ,cld~ule depçn-lenr-e of E' (open symbols) and tan 8 (shaded symbols) for fully cured DGEBA/BDMA (~,-) and organoclay/DGEBA/BDMA (o,-) cont~ining 4% MTS by volume.

S De~ lion of Fr.f~ Embodiments The ~yll~esis ~locedul~ used for nanocomposite plcl)~dlion involves di~ sion of the organoclay in a suitable mol~lller, followed by polymerization. Under properconditions ~ min~tion of the organoclay into individual silicate layers occurs, 10 which llltim~tely become dispersed within the macromolecular matrix. In a typical procedure mixing of the organoclay and DGEBA is carried out at lelll~)cldlufes in the range of 20 to 150~C, followed by sonication, addition of curing agent, and curing of the l~C~WOl~ at a prescribed set of telllyel~lulcs. Initial mixing of the organoclay and DGEBA is more preferably pelrolllled at about 90~C to ensure low 15 resin visco~iLy. Following addition of small amounts of the clay ( 0.1 to 10% by weight), the resin viscosity is only slightly increased. However, samples sonicated briefly (1-2 ~ es) e~elie~c a signifir~nt increase in resin viscosity at relatively low shear rates while turning from opaque to semi-Llansp~elll during sonication.Organoclay loadings above about 10% (w/w) begin to result in strong gel formation 20 during sonication, even after rchf~ to telllpeldLul~,s at or above 100~C. Theobserved increase in resin viscosity following sonication may be due to the dis~elsion of high aspect ratio (100-1000) silicate layers within the epoxy resin and is due to formation of a so-called "house of cards" ~llu~;lule, in which edge-to-edge and edge-to-face interactions belween dispersed layers form percolation structures.
25 Similar rheological cha~ges have been observed when org~nncl~ys are dispersed in various organic media and attributed to the formation of the "house-of cards"
structure.

The invention is further illustrated by the following Example, which is to be 30 considered as illustrative and not delimiting of the invention otherwised disclosed:

W096/08526 2 1 9 ~ 6 1 6 PCT/US95/10956 Example Synthesis of Nanocomposite Sam~les 5 The organoclay used in this Example was p,~p~,d by Southern Clay Products, Inc. of Gonzales, Texas by an ion-ex~h~nge reaction from Na-montmorillonite and bis(2-hydlo~y~Lllyl) methyl tallow-alkyl allUllOlliUlll chloride (Ethoquad T/12, Akzo ChPmir~l~) as shown in equation 1, 10 Na+-.Sm~ctite + (HOCH2CH2)2R'R"NfCl- _ (HOCH2CH2)2R'R"N+-SmPctite +
NaCl (1) where R' is pre~lo.~ ..lly an octadecyl chain with smaller amounts of lower homologues (approximate colll~osilion: Cl8 70%, C,6 25% and Cl4 4% and Cl2 15 1%) and R" is a methyl group. The dry organoclay powder was added with stirring to diglycidyl ether of bisphenol A (DGEBA, Dow ChPmir~l's DER 332, epoxide equivalent weight = 178) and cured by addition of either nadic methyl anhydride (NMA, Aldrich), boron trifluoride monoethylamine (BTFA, Aldrich), benzyl-limPthylamine (BDMA, Aldrich), or methylene ~ nilinP (MDA, 20 Aldrich). The amount of curing agent used for each formulation was as follows:
DGEBA/NMA: 87.5 parts NMA per hundred resin (phr), with or without 1.5 phr BDMA. DGEBA/BDMA: 1.5-10 phr BDMA. DGEBA/BTFA: 3 phr BTFA. DGEBA/MDA: 27 phr MDA. Org~nocl~y/DGEBA ll~Lul~s were held at 90~C with stirring for one hour, then sonicated for 1-2 ll~illules while hot 25 using a Fisher Model 300 Sonic Di~mPmhrator (Fisher Scientific, Itasca, IL).
Following sonication samples were cooled, curing agent was added with thorough rnixing, and then loaded into disposable ~ylinges. Samples were centrifuged in the syringes for 30 seconds at 3000rpm to remove bubbles, and then di~ellsed into rectangular teflon molds with ~ llp~ ons 20mm by lOmm 30 by 1.5mm thick, or casted as free-st~n-1in~ films with thil~nPs~es of 0.1-0.3mm.
All samples were cured at 100~C for 4 hours, 150~C for 16 hours, and 200~C
for 12 hours (in vacuo).

219~616 Ch~ tion of Nano~ sj~

X-ray diffraction (XRD) eA~Clllll~ i were pclro..lled dhe~;lly on the nanocomposite samples using a Scintag Pad X diffra~;lu--RIer with Cu ( 5 =1.54A) or Cr ( =2.29A) irradiation. In-situ, hot-stage XRD t~A~,lilllCnl:i were contlllct~pd using a special thPrm~ ,P~I which allowed samples to be heated to a llUll~el of dirrelenl Itlll~alules without removing the sample from the diffractometer. Samples were ramped at 10~C/min between the set lel~alul~es, and sc~nnPd after a 10 minute isoll.c... al equilibration. The 10 exothPrmir epoxy curing reaction was followed by dirre.~nlial sc~n~-in~
calolilllclly (DSC) using a du Pont 9900 thPrm~l analyzer. Spectra were obtained under flowing nitrogen at a sc~nning rate of 10~C/min. In-situ infrared curing studies were pc.rol...ed on a Mattson Galaxy 2020 Series FT-IR
using a pro~ l~l----able variable le...~eralu.e heating cell (Model HT-32, Spectra-15 Tech, Inc.). Spectra were collected at a resolution of 4 cm-'. Co~ osile l~icrosLIucture was imaged using Ll~ ion electron microscopy (TEM) on carbon coated 100nm thick sections of the con~po~ile using a JEOL 1200EX
sion electron microscope at an accelclali..g voltage of 120kV. Dynamic mPch~nir~l analyses (DMA) of the cured composite films were pelÇol.lRd on a 20 Rheovibron DDV-II-C viscoe~ o..~ t~_l (Toyo Baldwin Co., Japan) opelaling at a driving frequency of 110 Hz and a ~...pclalu.e sc~n~-ing rate of 1~C/min.

Delamination of Org~noclays 25 XRD analysis was used to follow the progress of organoclay dispersion during mixing with DGEBA and subse~lucnt curing reactions. Figure 1 shows the XRD
patterns of the dry organoclay and the uncured organoclay/DGEBA llli~lUlC. I'he top scan was obtained at room t~llllp~ ldlule following heating of the organoclay/DGEBA mixture at 90 ~C for one hour. The XRD pattern of the 30 organoclay powder shows a primary silicate (001) reflection at 2 =4.8~, with a low intensity shoulder at roughly 2 =5.8~. The main silicate reflection in organoclay cc,--e~ollds to a layer d-spacing of 17~ which l~ S~ an increase of approximately 7~ from the van der Waals gap of Na-montmorillonite.

WO 96108526 2 1 9 9 6 1 6 PCT/US95/1095~

Following mixing of the organoclay and DGEBA at room te~ cldlwc, an additional reflection centered at 2 =2.5~ emerges which col.c~onds to intercalated organoclay/DGEBA. As is known, organoclays can readily intercalate various small organic molecules from either the vapor or liquid phase.
5 The second peak at 2 =5~ collc~onds to the coexistence of unilll~rcalated (d~","-17A) and intercalated (d,0"-17.5~) organoclay. The persi~t~nce of some elcalated organoclay at room temperature can also be seen by the small ~clllllant shoulder at 2 =5.8~. In contrast, mixing of DGEBA and the organoclay at 90~C results in only DGEBA intercalated organoclay (d,~"-35A) with no 10 residual organoclay peaks observed, as shown in the top trace of Figure 1. The reflections observed at 2 =2.5~,4.9~, and 7.6~ co.lc~l,ond to the (001), (002), and (003) reflections of the DGEBA intercalated phase, respectively. Further evidence for the presence of only intercalated organoclay/DGEBA comes from the disal~pe~dllce of the organoclay shoulder at 2 =5.8 ~, which is no longer masked15 by any of the silicate (001) reflections.

The XRD results discussed relate to resin samples cooled to room te~llp~al after mixing at 90~C and, therefore, do not necessarily replcsellt the structures present at the mixing and curing telllpcldLw~,s. Dynamic high telll~CldLlllC, in-situ 20 XRD c~l cl--llents were used to ~letennine the exact structure of the resin llllXLWCS
at elevated temperatures. Samples were pl~aled by mixing organoclays and DGEBA in a vial at 90~C, and cooling to room ~lllpcldlulc before transferring todiffractometer chamber. Shown in Figure 2 are a series of XRD scans of the organoclay/DGEBA mixture previously heated to 90~C taken at various intervals 25 bdweell room ten~eldLulc and 150~C. The low tempcld~ , scans exhibit three orders of reflections indicating the e~i~t~nre of DGEBA intercalated organoclay with d~ =36A~ With increasing tclll~cldLLue a gradual increase in d~ from 36A toappro~illlately 38A was observed, although the constant hllensily of the peaks suggests that little or no del~min~tion occurs at or below 150~C. With the 30 observation that intercalation but not del~min~tion of the organoclay occurs in the ~ presence of DGEBA, the i.. ve--lo.~ sought to identify potential epoxy curing agents which would produce both ~lel~min~tion of the organoclay and cro~linking of the epoxy resin. It was found that the choice of curing agent was critical in w096/08526 2 1 996 1 6 PCT/US95/10956 det~rmining d~l~min~tion and optical clarity.

Selection of Curin~ A~ent 5 A survey of common epoxy curing agents revealed that many curing agents studied resulted in little or no increase in layer separation, reslllting in composites with silicate d-spacings of 30-40A or less. An example of this behavior is shownin Figure 3 for methylene tli~niline (MDA) cured organoclay/DGEBA composite.
This composite was prepared by adding MDA to the organoclay/DGEBA mixture, 10 which resulted in immediate clouding ofthe resin. Int~le;,lingly, all bifunctional primary and secondary amine curing agents used were found to have this effect and resulted in opaque composites, in conlld~l to the transparent composites following del~min~tion of organoclay. One explanation for this behavior might bethe bridging of the silicate layers by the bifunctional amine molecules, which 15 prevents further expansion of the layers. Another possibility is that the N-Hgroups in the plilll~ y and secondary amines are sufficiently polar to cause reaggregation of dispersed silicate layers. Others have observed similar degellation (deexfoliation) of organoclays dispersed in organic solvents upon the addition of polar additives.
Pursuant to the present invention, curing agents have now been found (NMA, BDMA, BTFA, and combinations thereof), which result in organoclay del~min~tion during heating of the reaction llli~Lule. Shown in Figure 4 are in-situ X~D scans of the organoclay/DGEBA/BDMA lllixlu~e illu~lldlillg the 25 del~min~tion of the organoclay on heating from room l~nl~eldlu e to 150~C. Asbefore, the sample was prepared by mixing the organoclay and DGEBA in a vial at 90~C, cooling to room temperature, and mixing in BDMA immediately before tr~n~ferring to the diffractometer chamber. The mixing of BDMA into the organoclay/DGEBA resin at room telllp~ldlule resulted in an intercalated system 30 with d"""-39A (slightly ~p~n~l~d from the d~ -36A observed with organclay/DGEBA). Furthermore, in contrast to what is observed in the absence of a curing agent, heating of the organoclay/DGEBA/BDMA llli~ resulted in substantial ~ m~tion of the peak at 2 =2.3 ~. This peak almost disappeared by wo 9~ 2 1 9 9 6 1 6 Pcr/uS95/logs~

150~C (top of Figure 4), with only a trace le.,.~ g at 2 =3 ~. The virtual disappearance of the organoclay (001) reflections clearly indicates del~min~tion of the organoclay has taken place.
.

5 XRD analysis of completely cured nanocomposite samples also lacked silicate (001) reflections as shown in Figures 5 and 6 for organoclay/DGEBA/BDMA and organoclay/DGEBA/NMA, re~e~ ely The absence of silicate (001) reflections in the cured nanocomposites shows that the del~nnin~tion and dispersion of the silicate layers within the epoxy matrix is retained after complete curing of the10 epoxy. The exfoliation of the silicate was further confirmed using TEM. The micrographs of the BDMA-cured composite are shown in Figure 7. These micrographs show quite clearly the existçn~e of well-dispersed individual silicate layers (dark lines in Figure 7) of thickness lnm embedded in the epoxy matrix.
Some areas of the epoxy matrix appear to contain oriented collections of 5-10 15 parallel silicate layers. These domains of parallel layers are presumably re~ ls of organoclay tactoids, but with substantial expansion of the gallery beyond that collG*,("lding to an intercalated silicate phase (see for example Figs. 1 and 3).
Close çX~ tiQn ofthese ~lom~in~ reveals consistent layer spacings of approximately 1 OOA or more, with the intervening galleries between layers filled 20 with crosslinked epoxy matrix . It is particularly interesting to note that the samples are mostly homogeneous with no phase separation between the silicate layers and the epoxy matrix. In fact, c~ tion of the micrographs shows excellent al"~osilion between the clay layers and the polymeric matrix.

25 Curin~ Reactions In contrast to the work of Wang and Pinnavaia, where no curing agent was added, a curing agent is used in the present invention that either crosslinks DGEBA in the presence of the organoclay, reacts directly with the organoclay, or catalyzes the 30 crosslinking reaction between organoclay and DGEBA. The benefits of this approach are first, curing of the nanocomposite occurs at much lower telllpclalulcs than reported previously, and second, formation of chemical bonds between the crosslinked network and the silicate nanoparticles results in direct ~tt~rhnl~nt of the epoxy matrix to the silicate layers, thereby m~imi7ing interfacial adhesion between the two phases.

One plcr~ d curing agent is BDMA, which can catalyze the homopolymçri7~tion 5 of DGEBA, but is also capable of catalyzing the reaction bclweell hydroxyl groups of the organoclay alkylammonium ions and the oxirane rings of DGEBA.
Curing conditions of the composite resin may have an effect on the reaction meçh~ni~m For example, increasing the telllpcl~lwe ofthe organoclay/DGEBA/BDMA and DGEBA/BDMA mixtures from 20~C to 250~C
10 at slow rates ( 0.5 ~C/min) resulted in little difference in curing behavior between the composite and unmodified epoxy as shown by col"pil, ;"g the coll~ollding infrared spectra (Figure 8). Both series of spectra show a gradual disappearanceof the epoxy band at 918 cm-' at telllpelalulcs between 80~C and 150~C. The extent of DGEBA reaction as given by the intensity of the epoxy peak is roughly 15 equivalent for both compositions (organoclay/DGEBA/BDMA and DGEBA/BDMA).

At higher heating rates, however, a dirr~l~llce in curing behavior is seen. Figure 9 shows DSC scans of the organoclay/DGEBA/BDMA and DGEBA/BDMA curing 20 reactions at a sc~nning rate of 10~C/min, showing a strong exotherm associated with curing between 100 and 150~C for organoclay/DGEBA/BDMA. That the DSC scan of the DGEBA/BDMA llli~lule shows a considerably smaller exotherm over the sarr~e t~lllp~ e range, suggests that the organoclay plays a catalytic role in the base-catalyzed homopolymerization of DGEBA, or that the reaction 25 proceeds by an altogether dirr,v~ .ll meçh~ni.cm in the presence of the organoclay.
One possibility as shown in eq. 2 involves the base-catalyzed oxirane ring-opening reaction between hydroxyl groups of the organoclay and DGEBA
reslllting in formation of I, an organoclay-glycidyl ether of bisphenol A oligomer.

2 1 9 9 6 1 6 PCTrUS95/10956 MTS'R3N~'--oH ''' ~'~'~~ ~1~
ORGANOC:I,AY DGEBA

MTS-R3N~, OH
(MTS=smectite-type clay) (2) I can subsequently react with free DGEBA via similar base-catalyzed oxirane ringopening to build up the crosslinked epoxy network. It is hllhesLi~lg to note that the tel,ll,cldlule at which curing occurs (approx. 100~C ~ shown by the exothermin Figure 9) co.~ onds to the same temperature that del~min~tion of the 10 organoclay occurred (see Figure 4). The telllpc~aL~lre coincidence of curing and lçl~min~tion makes intuitive sense, since del~min~tion e~oses the hydroxyl groups of the alkyl ~mmonillm chains in the interlayer to DGEBA and BDMA.

The participation of the hydroxylated org~n~ çl~y alkyl~mmo~ium ion in the 15 curing reaction is more clearly ill~l~ted with the org~nocl~y/DGEBA/NMA
system. Interestingly, full curing of the DGEBA/NMA lni~ e did not occur in the ~bsenre of organoclay, regardless of heating rate. Shown in Figure 10 are DSC scans ofthe or~nocl~y/DGEBA/NMA curing reaction. During dynamic curing of this formulation two distinct exotherms are observed; a weak one at 20 1 80~C followed by a strong exotherm at 247~C. Although the complete sequenceof reactions has not yet been de~ ;..e(l a possible sequence might first involvethe reaction of organoclay hydroxyl groups with NMA to form the monoester, II, as shown in eq. 3.

M'r~R3N'~ ~~C'~'C~~ ' MTS-R3N' OJ~R'J~OH

ORGANOCI.AY NMA Il (3) Nascent carboxylic groups of II can subsequently react with the epoxide resulting in formation of the diester, III, according to eq. 4.

J~ 1 ~

Il DGEBA
O O
3 ~ O~lRJ~O~o{~--OH
II~

(4) 15 Further reaction of III with DGEBA results in epoxy network formation. This reaction sequence results in c~mic~l bonding b~,Lv~ the org~nocl~y and the epoxy network. It is clear from the data shown in Figure 10 that in the absence of the organoclay, the DGEBA/NMA formulation does not result in curing under the conditions used in this e"~e ;~ nt This provides further evidence that the 20 organic component of the organoclay participates in the curing reaction.

wo96/08526 2 1 9 9 6 1 6 Pcr/uss5/los5~

Mechanical Properties of the Nanocomposite The effect of molecular dispersion of the silicate layers on the viscoelastic properties of the crosslinked polymeric matrix was probed using DMA. This 5 experiment involves applying an oscillatory strain to a sample while moniL~ g the resultant stress, which consists of both in-phase and out-of-phase components.
These stresses can then be used to calculate the in-phase (E' ) and out-of-phase(~n ) components of the modulus. The ratio E"IE' = tan~ is a measure of the ratio of energy lost to energy stored per cycle of deformation, and typically goes 10 through a maximum at the glass transition (T~) of the polymer. At T, there is a substantial drop in E, with a peak in tan~ indicating viscous damping due to segment~l motion in the polymer. For crosslinked polymers, both En and T, generally increase with crosslink density.

15 Shown in Figure 11 are the telllyGl~l~G dependencies of the tensile storage modulus, E', and tano of the organoclay/DGEBA/BDMA composite co..~ .g 4% silicate by volume, and the DGEBA/BDMA epoxy without any silicate. The shift and brofldening of the tan o peak to higher temperatures indicates an increase in nanocomposite T,and broadening of the glass transition. The shift in T, 20 as measured by the tan o peak m~xi",ll... is on the order of only a few degrees (4~C for the sample shown in Figure 11) and cannot account for the significant increase in plateau modulus. Furthermore, since the extent of curing is comparable in both samples (as measured by DSC), the increase cannot be attributed to variations in curing. BroA~lenin~ and increase of T, have been 25 observed in other organic-inorganic nanocomposites and are generally attributed to restricted segmentAl motions near the organic-inorganic interface. Chemical bonding at the interface of the silicate and epoxy matrix could lead to hinderedrelaxational mobility in the polymer segment~ near the interfAce, which leads tobro~clçning and increase of T,.
Below T" both samples exhibit high storage modulus, with a slight decrease in E ' with increasing tellly~ ~e. Notably, E ' in the glassy region below T, is approximately 58% higher in the nanocomposite colllpal~ed to the pure epoxy (2.44 lO~oco~ edto 1.55 lO~dyne/cm~at40~C). Evenmorestrikingisthelarge increase in E ' at the rubbery plateau of the nanocomposite as shown in Figure 11.
The nanocomposite exhibits a plateau modulus approximately 4.5 times higher than the unmodified epoxy (5.0 10 compared to 1.1 10~ dyne/cm1 at 150~C).
5 These changes are considerable, particularly in view of the fact that the silicate content is only 4% by volume. In this context, it is h~ lg to compare these results with reports of viscoelastic properties of conventionally pl~ared epoxy composites co"~ g micron or larger size filler particles. Typically, the conventional filled epoxies do not exhibit ~ub~ lial changes in E' at the filler10 volume contents (<10%) used in this study.

Theoretical ~.es~ions have been derived by Halpin and Tsai (Halpin, J.C.;
Kardos, J.L. Polym. Eng. Sci. 1976, 16, 344) to calculate elastic modulus of a composite con~i~ting of lmi~xi~lly oriented particles of filler suspended in a 15 continuous matrix. For composites with platelike particles, these equations predict a strong dependence of composite elastic modulus on filler aspect ratio.Solving the simultaneous Halpin-Tsai equations with the experimental dynamic storage modulus data in the glassy and the rubbery region yielded an ap~
aspect ratio of 43. It is clear from the TEM micrographs shown in Figure 7 that 20 some relatively unmodified epoxy matrix exists between the domains of 5-10 del~min~ted silicate layers. As a result, the effective aspect ratio of the silicate-rich domains could be much lower than the 100-1000 predicted for fully ~lel~min~ted and dispersed silicate layers.

25 While the present invention has been particularly set forth in terms of specific embotlim~nt~ thereof, it will be understood in view of the present disclosure, that numerous variations upon the invention are now enable to those skilled in the art, which variations yet reside within the scope of the instant te~chings. Accordingly, the invention is to be broadly construed, and limited only by the scope and the 30 spirit of the claims now appended hereto.

Claims

WE CLAIM

1. A method for preparing an epoxy-smectite-type clay nanocomposite, comprising dispersing in an epoxy resin a dry smectite-type clay which has been modified to an organoclay by ion exchange with an alkyl ammonium salt, together with diglycidyl ether of bisphenol A (DGEBA); the positive ion of the salt being of the form +NH3R1, +NH2R2R3, +NHR4R5R6, or +NR7R8R9R10, wherein R1 through R10 are organic radicals; and wherein R1, at least one of R2 and R3, at least one of R4, R5, and R6, and at least one of R7, R8, R9 and R10, contain a functional group capable of reacting and bonding with the epoxy upon crosslinking of same; and curing in the presence of a curing agent which either cross-links the DGEBA in the presence of said organoclay, reacts directly with the organoclay, or catalyzes the crosslinking reaction between theorganoclay and DGEBA.

2. A method in accordance with claim 1, wherein said functional goup is selected from one or more members of the group consisting of hydroxyl, epoxy,or carboxylic groups.

3. A method in accordance with claim 1, wherein said ammonium salt has at least one alkyl ammonium chain having a terminal hydroxyl group.

4. A method in accordance with claim 1, wherein said ammonium salt comprises a bis (2-hydroxythyl) methyl tallow alkyl ammonium salt.

5. A method in accordance with claim 1, wherein said curing agent is selected from one or members of the group consisting of nadic methyl anhydride (NMA), benzyldimethylamine (BDMA), and boron trifluoride monoethylamine (BTFA).

6. A method in accordance with claim 1, wherein said smectite-type clay comprises a montmorillonite.

7. A method in accordance with claim 1, wherein said curing agent is capable of catalyzing the homopolymerization of DGEBA and catalyzing the reaction between said organoclay terminal hydroxyl group and the oxirane rings of said DGEBA.

8. A method in accordance with claim 7, wherein said curing agent comprises BDMA.

9. A method in accordance with claim 5, wherein said quaternary ammonium salt comprises a bis (2-hydroxyethyl) methyl tallow alkyl ammonium salt.

10. A method in accordance with claim 5, in which said curing is conducted at temperatures in the range of 100 to 200°C.

11. A method in accordance with claim 5, wherein dispersing of the dry smectite in the epoxy resin is carried out by mixing the smectite and DGEBA
at temperatures in the range of 20° to 150°C, following by sonification.

12. A polymer-mineral nanocomposite comprising an organically modified smectite-type clay which is molecularly dispersed within a crosslinked and continuous solid epoxy matrix.

13. A nanocomposite in accordance with claim 12, in which the phase dimensions are in the range of 1-100nm; and wherein the smectite nanoparticles are chemically bonded to the crosslinked network.

14. A nanocomposite in accordance with claim 12, in which the smectite has layer spacing of at least 100.ANG..

15. A nanocomposite in accordance with claim 12, wherein the smectite-type clay has been modified by ion exchange with an alkyl ammonium salt having at least one alkyl ammonium chain having a terminal hydroxyl group.

16. A nanocomposite in accordance with claim 15, wherein said quaternary ammonium salt comprises a bis(2-hydroxyethyl) methyl tallow alkyl ammonium salt.