Druckfreigabe/approval for printing

Without corrections/ ` ohne Korrekturen j After corrections/ 277 nach Ausfçhrung ` 1 der Korrekturen 2 3 Date/Datum: ...... 4 Signature/Zeichen: ...... 5 6 9 7 Biodegradable Polymer-based Nanocomposites: 8 9 Nanostructure Control and Nanocomposite Foaming 10 with the Aim of Producing Nano-cellular Plastics 11 Masami Okamoto 12 13 14 9.1 15 Introduction 16 17 The increased importance of renewable resources for raw materials and recyclability 18 or biodegradability of the material at the end of its useful life is demanding 19 a shift from petroleum-based synthetics to agro-based materials in industrial 20 applications. Increased social awareness of environmental problems posed by 21 the non-degradable, non-recyclable content of their products is forcing manufac- 22 turers to enhance the biodegradable content, which in turn favors a switch to 23 biomaterials [1]. 24 Hence, with this background, there is an urgent need to develop renewable source- 25 based environmentally benign polymeric materials (biopolymers [2]). Such materials 26 would not involve the use of toxic or noxious components in their manufacture, and 27 could allow composting to naturally occurring degradation products. 28 In today’s commercial environment, biopolymers have proven to be relatively 29 expensive and available only in small quantities. This has led to limitations on their 30 application to date. However, there are signs that this is changing, with increasing 31 environmental awareness and more stringent legislation regarding recyclability and 32 restrictions on waste disposal. Cargill Dow has a polylactide (PLA) in production 33 (with the Trade Name Natureworks). Metabolix has been working on polyhydroxy- 34 alkanoate (PHA) (Trade Name Biopol) [2]. 35 Thus, the increasing application of the various intrinsic properties of bio- 36 polymers, coupled with the knowledge of how such properties can be improved 37 to achieve compatibility with thermoplastics processing, manufacturing, and en- 38 d-use requirements, has fueled technological and commercial interest in 39 biopolymers. 40 Of particular interest are the recently developed nanocomposites consisting of a 41 polymer and layered silicate because they often exhibit remarkably improved 42 mechanical and other properties [3] when compared with pure polymer or conven- 43 tional composites (both micro- and macro-composites). A primary progress in 44 45 Bio-inorganic Hybrid Nanomaterials. Edited by E. Ruiz-Hitzky, K. Ariga, Y. M. Lvov Copyright Ó 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31718-9 278j 9 Biodegradable Polymer-based Nanocomposites

1 polymer/layered filler nanocomposites (PLFNCs), a Nylon 6/layered silicate hybrid 2 [4] reported by Toyota Central Research & Development Co. Inc. (TCRD), was 3 successfully prepared by in situ polymerization of e-caprolactam in a dispersion of 4 montomorrillonite (MMT). The silicate can be dispersed in liquid monomer or a 5 solution of monomer. It has also been possible to melt-mix polymers with layered 6 silicates, thus avoiding the use of organic solvents. The latter method permits the use 7 of conventional processing techniques such as injection molding and extrusion. The 8 extensive literature on nanocomposite research is covered in recent reviews [5–7]. 9 The study of nanocomposites has gained greater momentum. This new class of 10 materials is now being introduced in structural applications, such as a gas barrier 11 film, a flame retardant product, and other load-bearing applications [8]. 12 This chapter is intended to highlight the biodegradable polymer-based nanocom- 13 posites with the primary focus on nanostructure control and the foam processing 14 operation with the aim of producing nano-cellular plastics. Development of nano- 15 composite foams is one of the latest evolutionary technologies of polymeric foam 16 through a pioneering effort by Okamoto and his colleagues [9,10]. 17 18 19 9.2 20 Nano-structure Development 21 22 9.2.1 23 Melt Intercalation 24 25 Since the possibility of direct melt intercalation was first demonstrated [11], melt 26 intercalation has become a method of preparation of the intercalated polymer/ 27 layered silicate nanocomposites (PLSNCs). This process involves annealing, stati- 28 cally or under shear, a mixture of the polymer and organically modified layered 29 fillers (OMLFs) above the softening point of the polymer. During annealing, the 30 polymer chains diffuse from the bulk polymer melt into the nano-galleries between 31 the layered fillers. 32 In order to understand the thermodynamic issues associated with the nanocom- 33 posite formation, Vaia et al. have applied the mean-field statistical lattice model and 34 found that conclusions based on the mean field theory agreed nicely with the 35 experimental results [12,13]. The entropy loss associated with confinement of a 36 polymer melt is not prohibited to nanocomposite formation because an entropy gain 37 associated with the layer separation balances the entropy loss of polymer intercala- 38 tion, resulting in a net entropy change near to zero. Thus, from the theoretical model, 39 the outcome of nanocomposite formation via polymer melt intercalation depends on 40 energetic factors, which may be determined from the surface energies of the polymer 41 and OMLF. 42 Nevertheless, we have often faced the problem that the nanocomposite shows fine 43 and homogeneous distribution of the nanoparticles in the polymer matrix (e.g., poly 44 (L-lactide)) without a clear peak shift of the mean interlayer spacing of the (001) plane, 45 as revealed by wide-angle X-ray diffraction (WAXD) analysis [14]. Furthermore we 9.2 Nano-structure Development j279

1 sometimes encounter a decrease in the interlayer spacing compared with that of 2 pristine OMLF, despite very fine dispersion of the silicate particles. For this reason, 3 information on the structure of the surfactant (intercalant)–polymer interface is 4 necessary to understand the intercalation kinetics that can predict the final nano- 5 composite morphology and overall materials properties. 6 7 9.2.2 8 Interlayer Structure of OMLFs and Intercalation 9 10 9.2.2.1 Nano-fillers 11 In characterizing layered silicate, including layered titanate (HTO), the surface 12 charge density is particularly important because it determines the interlayer structure 13 of the intercalants as well as the cation exchange capacity (CEC). Lagaly proposed a 14 method of calculation consisting of total elemental analysis and the dimensions of 15 the unit cell [15]: 16 Surface charge : e =nm2 ¼ x=ab ð1Þ 17 18 where x is the layer charge (1.07 for HTO, 0.66 for synthetic fluorine hectrite (syn-FH) 19 and 0.33 for MMT. a and b are cell parameters of HTO (a ¼ 3.782 Å, b ¼ 2.978 Å [16]), 20 syn-FH (a ¼ 5.24 Å, b ¼ 9.08 Å [17]), and MMT (a ¼ 5.18 Å, b ¼ 9.00 Å [18]). For syn- + 21 FH, however, about 30 % of the interlayer Na ions are not replaced quantitatively by 22 intercalants since they do not take part in ion-exchange reactions [17]. For HTO, only + + 23 27 % of the interlayer H (H3O ) is active for ion-exchange reactions [14]. Thus the 24 incomplete replacement of the interlayer ions is ascribed to the intrinsic chemical 25 reactivity. The characteristic parameters of two nano-fillers are also summarized in 2 26 Table 9.1 [18]. HTO has a high surface charge density of 1.26 e /nm compared with 2 2 27 those of syn-FH (0.971 e /nm ) and MMT (0.780 e /nm ). From these results, we 28 can estimate the average distance between exchange sites as 0.888 nm for HTO, 29 1.014 nm for syn-FH and 1.188 nm for MMT. This estimation assumes that the 30 31 Tab. 9.1 Characteristic parameters of nano-fillers. 32 33 Parameters HTO syn-FH MMT 34 35 chemical H1.07Ti1.73O3.95 0.5H2ONa0.66Mg2.6Si4O10(F)2 Na0.33(Al1.67Mg0.33) 36 formula Si4O10(OH)2 – – – 37 particle size/nm 100 200 100 200 100 200 2 38 BET area/m /g 2400 800 700 CEC/meq/100 g 200 (660) 120 (170) 90 (90) 39 e /charge/nm2 1.26 0.971 0.708 40 density/g/cm3 2.40 2.50 2.50 41 20 refractive index (nD ) 2.3 1.55 1.55 42 pH 4–69–11 7.5–10 43 a Methylene blue adsorption method. The values in parentheses are calculated from the chemical 44 formulae of the nano-fillers. 45 Source: Reprinted with permission from [18] Ó 2006, Wiley-VCH. 280j 9 Biodegradable Polymer-based Nanocomposites

1 cations are evenly distributed in a cubic array over the silicate surface and that half of 2 the cations are located on the one side of the platelet and the other half reside on the 3 other side. 4 5 9.2.2.2 Molecular Dimensions and Interlayer Structure 6 The calculated models of the intercalant structures are presented in Figure 9.1. For + 7 octadecylammonium (C18H3N ), the molecular length, thickness and width are 8 2.466, 0.301 and 0.301 nm, respectively. Since the length of all alkyl units is more than 9 2 nm, these spacings (distance between exchange sites) of 0.888–1.188 nm do not 10 allow parallel layer arrangement like flat-lying chains [15] in each gallery space of the 11 nano-fillers. All the intercalants are oriented with some inclination to the host layer in 12 the interlayer space to form an interdigitated layer. This is suggested by the paraffin- 13 type layer structure proposed by Lagaly, especially in the case of clay minerals with 14 high surface charge [15]. 15 WAXD patterns for three OMLF powders are presented in Figure 9.2. The mean

16 interlayer spacing of the (001) plane (d(001)) for the HTO intercalated with qC14(OH) 17 (HTO-qC14(OH)) obtained by WAXD measurements is 2.264 nm (diffraction angle,   18 2Y ffi 3.90 ). The appearance of small peaks observed at 2Y ffi 7.78, 11.78 and 15.74 19 confirmed that these reflections were due to the (002) to (004) planes of HTO-

20 qC14(OH). HTO-qC14(OH) has a surprisingly well-ordered suprastructure, as proved 21 by WAXD, with diffraction maxima up to the fourth order due to the high surface 22 charge density of the HTO layers. On the other hand, syn-FH and MMT, which have 23 low surface charge density compared with that of HTO, show a less-ordered 24 interlayer structure, that is, the coherent order of the silicate layers is much lower 25 for syn-FH and MMT intercalated with surfactants. 26 From the WAXD results the interlayer opening is estimated after subtraction of the 27 layer thickness value of 0.374 nm for HTO [16], 0.98 nm for syn-FH [17] and 0.96 nm 28 for MMT [15]. This is an important point for the following discussion on the 29 interlayer structure. The illustration of a model of the interlayer structure of the fi 30 qC14(OH) in the gallery space of the HTO is shown in Figure 9.3. For nano- llers with 31 a high surface charge density, the intercalants can adopt a configuration with an 32 orientation where the alkyl chains are tilted under the effect of the van der Waals 33 forces, which decreases the chain–chain distance. For this reason, the angle a should 34 be directly related to the packing density of the alkyl chains. The value of a decreases 35 until close contact between the chains is attained, giving an increase in the degree of 36 crystallinity of the intercalants in the nano-galleries. To estimate the tilt angle a,we 37 combined the molecular dimensions, interlayer spacing and loading amount of the 38 intercalant in the layers, which was calculated from thermogravimetry analysis 39 (TGA). The characteristic parameters are summarized in Tables 9.2 and 9.3. Note 40 that HTO exhibits a large value of layer opening accompanied by large values of a 41 and endothermic heat flow DH, due to the melting of the intercalants in the 42 galleries, when compared with those of syn-FH and MMT. This indicates that HTO 43 leads to a highly interdigitated layer structure and the interlayer opening becomes 44 more uniform compared with MMT and syn-FH (possessing lower surface charge 45 density). 9.2 Nano-structure Development j281

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Fig. 9.1 Molecular dimensions of intercalants: (A) octadecy- + lammonium (C18H3N ), (B) octadecyltrimethylammonium 37 + (C18(CH3)3N ), (C) dioctadecyldimethylammonium 38 + (2C18(CH3)2N ), and (D) N-(cocoalkyl)-N,N-[bis(2-hydro- 39 xyethyl)]-N-methyl ammonium cations (qC14(OH)). Reprinted 40 from [18], Ó 2006, Wiley-VCH. 41 42 This explains why the well-defined diffraction peaks up to the (004) plane are 43 observed (see Figure 9.2). The entropic contribution of the intercalants, which leads 44 to the entropy gain associated with the layer expansion after intercalation of the 45 polymer chains, may not be significant because of the interdigitated layer structure. 282j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Fig. 9.2 WAXD patterns of HTO, syn-FH and MMT intercalated 24 + Ó with qC14(OH) . Reprinted from [18], 2006, Wiley-VCH. 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Fig. 9.3 Illustration of a model of interlayer thickness and width are 2.09, 0.881 and 42 structure of intercalant N-(cocoalkyl)-N,N-[bis 0.374 nm, respectively (see upper panel). The Q5 43 (2-hydroxyethyl)]-N-methyl ammonium cation tilt angle a of the intercalants can be estimated 44 (qC14(OH)) in the gallery space of layered tita- by combination of the interlayer spacing, mo- 45 nate (HTO). The average distance between lecular dimensions and loading amount of in- exchange sites is 0.888 nm, calculated from the tercalants when the alkyl chains adopted an all- surface charge density of 1.26 e /nm2. For trans comformation. Reprinted from [18], qC14(OH), the obtained molecular length, Ó 2006, Wiley-VCH. 9.2 Nano-structure Development j283

1 Tab. 9.2 Comparison of characteristic parameters of HTO, syn-FH 2 and MMT prepared with qC14(OH). 3 Parameter HTO-qC (OH) syn-FH-qC (OH) MMT-qC (OH) 4 14 14 14 5 layer opening/nm 1.889 1.083 0.895 6 tilt angle a/ 64.4 31.1 25.3 7 organic content/wt% 39.6 30.4 32.5  8 Tm/ C 108.3 111.3 97.7 D 9 H/J/g 214.5 141.2 138.6 a fl +  10 The melting and heat ow of qC14(OH) Cl are 35.8 C and 69.8 J/g, respectively. 11 Source: Reprinted with permission from [18] Ó 2006, Wiley-VCH. 12 13 14 9.2.2.3 Correlation of Intercalant Structure and Interlayer Opening 15 For the interdigitated layer structure in MMT, the alkyl chain length (i.e. C18H37,CH3 16 and (CH2)2OH in the amine structure) changes the interlayer opening. That is, when 17 + we compare different intercalants having the same long alkyl chain (i.e., C18H3N 18 + and C18(CH3)3N ), three methyl (CH3) substituents instead of hydrogen (H) disturb 19 the contact with the silicate surfaces. The value of a decreases until close contact 20 between the ammonium cations and silicate surfaces is attained, giving a decreasing 21 in the interlayer opening (d(001)) (see Table 9.3 and Figure 9.2). 22 + In the case of an intercalant having two long alkyl chains (i.e., 2C18(CH3)2N ), the 23 packing density of the alkyl chains is reduced and sterically limited to the nano- 24 + galleries. Consequently, MMT-2C18(CH3)2N exhibits a large interlayer opening 25 accompanied by low crystallinity of the intercalant (DH 130 J/g) compared with 26 + + MMT-C18H3N and MMT-C18(CH3)3N . Accordingly, the disordered diffraction peak 27 + of the (001) plane of MMT-2C18(CH3)2N is observed in the WAXD analysis (see 28 Figure 9.1 in Ref. [19]). We have to pay attention to the molecular size of the 29 substituents instead of H attached to the nitrogen to get a better understanding of the 30 interdigitated layer structure and direct polymer melt intercalation. This feature has 31 32 33 34 Tab. 9.3 Comparison of characteristic parameters of MMT-based + + + 35 OMLF prepared with C18H3N ,C18(CH3)3N and 2C18(CH3)2N . 36 + + + 37 Parameter C18H3N C18(CH3)3N 2C18(CH3)2N 38 layer opening/nm 1.350 1.011 1.540 39 tilt angle a/ 33.2 22.9 40.1 40 organic content/wt% 35.5 29.5 39.8 41  Tm/ C 69.9 69.5 44.0 42 DH/J/g 177.7 189.6 129.7 43 a fl + + +  The melting and heat ow of C18H3N ,C18(CH3)3N and 2C18(CH3)2N are 83.8 C and 95.6 J/g; 44 103.5 C and 161.2 J/g; and 37.0 C and 54.6 J/g, respectively. 45 Source: Reprinted with permission from [18] Ó 2006, Wiley-VCH. 284j 9 Biodegradable Polymer-based Nanocomposites

1 been observed in the results of OMLFs intercalated with various intercalants (such 2 as octadecyl di-methylbenzylammonium, n-hexadecyl tri-n-butyl phosphonium, 3 n-hexadecyl tri-phenyl phosphonium cations) [20]. 4 5 9.2.2.4 Nanocomposite Structure 6 Figure 9.4 shows the results of transmission electron microscopy (TEM) bright field 7 images of PLA-based nanocomposites, in which the dark entities are the cross section 8 of the intercalated MMT layers. The organically modified MMT content in all 9 nanocomposites was 4 wt%. From the TEM images, it becomes clear that there are 10 some intercalated-and-stacked silicate layers in the nanocomposites. Yoshida et al. 11 estimate the form factors obtained from the TEM images, that is the average value of 12 the particle length (L) of the dispersed particles and the correlation length (x) between 13 them [21]. From the WAXD patterns, the crystallite size (D) of the intercalated stacked 14 silicate layers of each nanocomposite is calculated by using the Scherrer equation. 15 The calculated value of D (ffi thickness of the dispersed particles) and other 16 parameters for each nanocomposite are presented in Table 9.4. + 17 For PLA/MMT-C18(CH3)3N , L and D are in the range (200 + 25) nm and 10.7 nm. + 18 On the other hand, PLA/MMT-C18H3N exhibits a large value of L (450 + 200 nm) 19 with a large level of stacking of the silicate layers (D21 nm). The x value of the PLA/ + + 20 MMT-C18(CH3)3N (80 + 20 nm) is lower than that of PLA/MMT-C18H3N 21 (260 + 140 nm), suggesting that the intercalated layers are more homogeneously fi + 22 and nely dispersed in the case of PLA/MMT-C18(CH3)3N . The number of stacked  þ + 23 individual silicate layers ( D/d(001) 1) is 5 for PLA/MMT-C18(CH3)3N and the x 24 value for this nanocomposite is one order of magnitude lower than those of PLA/ + + 25 MMT-C18H3N and PLA/MMT-2C18(CH3)2N , suggesting that the intercalated sili- 26 cate layers are more homogeneously and finely dispersed. + 27 Althoughthe (initial)interlayer openingofMMT-C18(CH3)3N at1.011 nmissmaller + + 28 thanthatofMMT-C18H3N at1.350 nmandthatofMMT-2C18(CH3)2N at1.540 nm,the 29 intercalation of the PLA in these different OMLFs gives almost the same basal spacing 30 afterpreparationofthenanocomposites.NotethattheexistenceofasharpBraggpeakin 31 the PLA-based nanocomposites after melt extrusion clearly indicates that the dispersed 32 silicate layers still retain an ordered structure after melt extrusion. 33 In Table 9.4 are summarized the layer expansions after preparation (¼D opening) 34 of three nanocomposites, or after subtraction of the initial layer opening. For the + 35 same MMTwith different intercalants (e.g., comparison between MMT-C18(CH3)3N + 36 and MMT-2C18(CH3)2N ), the layer expansion of the former (0.879 nm) exhibits a 37 large value compared with that of the latter (0.45 nm) in PLA-based nanocomposites. 38 In other words, the smaller interlayer opening caused by the configuration with small ¼  + 39 tilt angle (a 22.9 for C18(CH3)3N ) promotes a large amount of intercalation of the + fi 40 polymer chains. Accordingly, PLA/MMT-C18(CH3)3N exhibits ner dispersion of fi + + 41 the nano- llers than do PLA/MMT-2C18(CH3)2N and PLA/MMT-C18H3N ,as 42 discussed before (see Figure 9.4). 43 One more interesting feature is the absolute value of D opening. According to the 44 molecular modeling, the width and thickness of the PLA are 0.76 and 0.58 nm (see 45 Figure 9.5). This may suggest that the polymer chains could not penetrate into the 9.2 Nano-structure Development j285

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Fig. 9.4 Bright field TEM images of PLA-based nanocomposites + + 36 prepared with (A) MMT-C18H3N , (B) MMT-C18(CH3)3N and (C) + 37 MMT-2C18(CH3)2N . The dark entities are the cross section and/ 38 or face of the intercalated-and-stacked silicate layers and the bright areas are the matrix. Reprinted from [18], Ó 2006, Wiley-VCH. 39 40 + 41 galleries in the case of MMT-2C18(CH3)2N when we compare the apparent interlayer 42 expansion (¼D opening). 43 Now it is necessary to understand the meaning of the interlayer expansion in the 44 intercalated nanocomposites. As discussed before, we have to take the interdigitated 45 layer structure into consideration. This structure may suggest that a different 286j 9 Biodegradable Polymer-based Nanocomposites

1 Tab. 9.4 Form factors of three nanocomposites obtained from 2 WAXD and TEM observations. 3 Form factor PLA/MMT-C H N+ PLA/MMT-C (CH ) N+ PLA/MMT-2C (CH ) N+ 4 18 3 18 3 3 18 3 2 5 d001/nm 3.03 2.85 2.95 6 D opening/nm 0.72 0.879 0.45 7 final layer 2.07 1.89 1.99 8 opening/nm 9 D/nm 20.9 10.73 14.71 (D/d ) þ 1 7.9 4.8 6.0 10 001 L/nm 450 + 200 200 + 25 655 + 121 11 x/nm 260 + 140 80 + 20 300 + 52 12 Ó 13 Source: Reprinted with permission from [18] 2006, Wiley-VCH. 14 15 orientation angle could be adopted when the polymer chains penetrate into the 16 galleries, giving a decrease in the basal spacing after intercalation. At the same time, 17 this structure apparently provides a balance between polymer penetration and the 18 different orientation angles of the intercalants. That is, we have to pay attention to the 19 polymer chains intercalation into the galleries from the change in the basal spacing, 20 as revealed by WAXD. Presumably the penetration of the polymer chain is prevented 21 or reduced by the steric-limitation of the configuration with a large value of a (e.g., 22 a ¼ 40.1 for MMT-2C (CH ) N+. Accordingly, we sometimes observe small inter- 23 18 3 2 layer expansion and encounter a decrease in the interlayer spacing after melt 24 intercalation. As seen in Table 9.4, the initial interlayer opening depends on the 25 interlayer expansion (¼D opening) after melt intercalation. The smaller initial 26 opening leads to the larger interlayer expansion, and gives almost the same final 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Fig. 9.5 Molecular dimensions of the PLA-backbone using the molecular dynamics program (MM2 in Quantum CAChe) with 43 van der Waals radii taken into consideration. Optimization of 44 structure is based on minimization of the total energy of the 45 molecular system. 9.2 Nano-structure Development j287

1 interlayer opening. This feature has been observed in the results of other nano- 2 composites prepared from different OMLFs intercalated with different surfactants 3 [22]. From this result, the entropic contribution of the intercalants, which leads to the 4 entropy gain associated with the layer expansion after intercalation of the small 5 molecules and/or polymer chains, may not be significant due to the interdigitated 6 layer structure. Presumably the penetration takes place by pressure drop within the 7 nano-galleries, nano-capillary action, generated by the two platelets. 8 As reported in the literature [20], the pressure drop (Dp) into the nano-galleries, 9 which makes the polymer penetration more difficult, should be discussed. The 10 estimated pressure difference (24 MPa) is much larger than the shear stress 11 (0.1 MPa) during melt compounding [20]. This suggests that the shear stress has 12 little effect on the delamination (exfoliation) of the layer. This reasoning is consistent 13 with the intercalated structure reported by so many nanocomposite researchers, 14 who can prepare only intercalated (not exfoliated) nanocomposites via a simple 15 melt extrusion technique [5]. A novel compounding process is currently in progress. 16 Solid-state shear processing may be an innovative technique to delaminate the 17 layered fillers [23]. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Fig. 9.6 Three-component model used for basal spacing simu- þ 42 lations, consisting of two layers of MMT with K cations (stick model), four molecules of trimethylammonium cation (A) or 43 dimethylstearylammonium cation (B) (stick and ball model), and 44 one molecule of maleated PP (PP-MA) (ball model). Reprinted 45 from [24], Ó 2006, Elsevier Science. 288j 9 Biodegradable Polymer-based Nanocomposites

1 Compared to OMLFs, nanocomposite structure is difficult to model using atomic 2 scale molecular dynamics (MD) because the intercalated polymer chain conforma- 3 tion is complex and can hardly be the equilibrium state. However, Pricl et al. [24] 4 explored and characterized the atomic scale structure to predict binding energies and 5 basal spacing of PLFNCs based on polypropylene (PP) and maleated (MA) PP (PP- 6 MA), MMT, and different alkylammonium ions as intercalants (see Figure 9.6). From 7 a global interpretation of all these MD simulation results, they concluded that 8 intercalants with smaller volume are more effective for clay modification as they 9 improve the thermodynamics of the system by increasing the binding energy. On the 10 other hand, intercalants with longer tails are more effective for intercalation and 11 exfoliation processes, as they lead to higher basal spacing. Additional information is 12 necessary to predict a more reasonable nanostructure of PLFNCs; some studies using 13 coarse-grained MD simulation of the confined polymer chains within the silicate 14 galleries have been reported [25–28]. 15 16 17 9.3 18 Control of Nanostructure Properties 19 20 9.3.1 21 Flocculation Control and Modulus Enhancement 22 23 Most nanocomposite researchers obdurately believe that the preparation of a 24 completely exfoliated structure is the ultimate target for better overall properties. 25 However, these significant improvements are not observed in every nanocomposite 26 system, including systems where the silicate layers are near to exfoliated [29]. While, 27 from the barrier property standpoint, the development of exfoliated nanocomposites 28 is always preferred, Nylon 6-based nanocomposite systems are completely different 29 from other nanocomposite systems, as discussed [3,8]. 30 In Figure 9.7, Okamoto summarized the clay content dependence of the dynamic 0 31 storage modulus (G ) of various types of nanocomposites obtained at well below the fi 32 Tg of the matrices. The Einstein coef cient (kE), derived by using Halpin and Tai’s 33 theoretical expression modified by Nielsen, is shown in the figure, and represents the

34 aspect ratio (Lfiller/dfiller) of dispersed MMT particles without intercalation. From 35 Figure 9.7, it is clearly observed that poly(butylene succinate)(PBS)-based nanocom- 0 36 posites (PBSNCs) show a very high increment in G compared to other nanocom- 37 posites having the same content of clay in the matrix. N6NCs are well-established 38 exfoliated Nylon 6-based nanocomposites. 39 PLANCs are becoming established intercalated-and-flocculated PLA-based nano- 40 composites, while PBSNCs are intercalated-and-extended flocculated nanocompo- 41 site systems [30,31]. Due to the strong interaction between hydroxylated edge–edge 42 groups, the MMT particles are sometimes flocculated in the polymer matrix. As a 43 result of this flocculation the length of the MMT particles increases enormously and 44 hence the increase in the overall aspect ratio. For the preparation of high molecular

45 weight PBS, di-isocyanate [OCN-(C6H12)-NCO] type end-groups are generally used as 9.3 Control of Nanostructure Properties j289

1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 0 15 Fig. 9.7 Plots of G nanocomposite=G matrix vs. vol.% of MMT for 16 various nanocomposites. The Einstein coefficient kE is shown by the number in the box. The lines show the calculated results from 17 Halpin and Tai’s theory with various kE. 18 19 a chain extender. These isocyanate end-group chain extenders form urethane bonds 20 with hydroxy-terminated low molecular weight PBS, and each high molecular weight 21 PBS chain contains two such bonds. These urethane-type bonds lead to strong 22 interaction with the silicate surface by forming hydrogen bonds and hence to strong 23 flocculation (see Figure 9.8). For this reason, the aspect ratio of dispersed clay 24 particles is much higher in the case of PBSNCs than for all nanocomposites, and 25 hence the high enhancement of the modulus. 26 This behavior can be explained with the help of the classical rheological 27 theory of suspension of conventional filler reinforced systems. According to this 28 Q1 theory [32], rotation of the filler is possible when the volume fraction of clay 29 < ffi 1 ffiller fcritical (aspect ratio) . All PBSCNs studied here follow this relation except 30 ¼  1 PBSCN4 (MMT 3.6 wt%), in which ffiller (aspect ratio) . For this reason, in 31 32 33 34 35 36 37 38 39 40 41 42 43 Fig. 9.8 Formation of hydrogen bonds between PBS and MMT, 44 which leads to the flocculation of the dispersed silicate layers. 45 Reprinted from [30], Ó 2003 American Chemical Society. 290j 9 Biodegradable Polymer-based Nanocomposites

1 PBSNC4, rotation of dispersed intercalated with flocculated stacked silicate layers is 2 completely hindered and only translational motion is available. Hence PBSNC4 3 exhibits a very high modulus. This behavior is clearly observed in dynamic 4 storage modulus measurements in the molten state [30]. In the case of N6NC3.7 0 5 (MMT ¼ 3.7 wt%) we can see the same high increment in G as for the PBSCNs. The 6 development of the flocculated structure occurs even though N6NCs are well- 7 established exfoliated nanocomposite systems. 8 9 10 9.3.2 11 Linear Viscoelastic Properties 12 13 The measurement of rheological properties of the PLFNCs in the molten state is 14 crucial in order to gain a fundamental understanding of the nature of the processabil- 15 ity and the structure–property relationships for these materials. 16 Dynamic oscillatory shear measurements of polymeric materials are generally ¼ 17 performed by applying a time dependent strain of g(t) gosin(ot) and the resultant ¼ 0 þ 00 0 00 18 shear stress is s(t) go[G sin(ot) G cos(ot)], with G and G being the storage and 19 loss modulus, respectively. 20 Generally, the rheology of polymer melts depends strongly on the temperature at 21 which the measurement is carried out. It is well known that for thermorheological 0 00 22 simplicity, isotherms of storage modulus (G (o)), loss modulus (G (o)) and complex à 23 viscosity (Z (o)) can be superimposed by horizontal shifts along the frequency axis: 24 0 0 bTG ðaTo; Tref Þ¼bTG ðo; TÞ; ð2Þ 25 26 00 00 bTG ðaTo; Tref Þ¼bTG ðo; TÞð3Þ 27 28 à à jZ jðaTo; T ref Þ¼jZ jðo; TÞð4Þ 29 30 where aT and bT are the frequency and vertical shift factors, and Tref is the reference 31 temperature. All isotherms measured for pure polymer and for various nanocom- 32 posites can be superimposed along the frequency axis. 33 In the case of polymer samples, it is expected that, at the temperatures and 34 frequencies at which the rheological measurements were carried out, the polymer 35 chains should be fully relaxed and exhibit characteristic homo-polymer-like terminal 0 2 00 36 flow behavior (i.e., the curves can be expressed by a power-law of G / o and G / o). 37 The rheological properties of in-situ polymerized nanocomposites with end- 38 tethered polymer chains were first described by Krisnamoorti and Giannelis [33]. 39 The flow behavior of PCL- and Nylon 6-based nanocomposites differed extremely 40 from that of the corresponding neat matrices, whereas the thermorheological 41 properties of the nanocomposites were entirely determined by the behavior of the 0 00 42 matrices [33]. The slope of G (o) and G (o) versus aTo is much smaller than 2 and 1, 43 respectively. Values of 2 and 1 are expected for linear mono-dispersed polymer melts, 44 and the large deviation, especially in the presence of a very small amount of layered 45 silicate loading, may be due to the formation of a network structure in the molten 9.3 Control of Nanostructure Properties j291

1 state. However, such nanocomposites based on the in-situ polymerization technique 2 exhibit fairly broad molar mass distribution of the polymer matrix, which hides the 3 structure-relevant information and impedes the interpretation of the results. 4 To date, the melt state linear dynamic oscillatory shear properties of various kinds 5 of nanocomposites have been examined for a wide range of polymer matrices 6 including Nylon 6 with various matrix molecular weights [34], polystyrene (PS) 7 [35], PS-polyisoprene (PI) block copolymers [36,37], poly(e-caprolactone) (PCL) [38], 8 PLA [39,40], PBS [30,41], and so on [42]. 9 The linear dynamic viscoelastic master curves for the neat PLA and various PLA- 10 based nanocomposites (PLANCs) are shown in Figure 9.9 [40]. The linear dynamic 11 viscoelastic master curves were generated by applying the time–temperature super-

12 position principle and shifted to a common temperature Tref using both the 13 frequency shift factor aT and the modulus shift factor bT. The moduli of the 14 nanocomposites increase with increasing clay loading at all frequencies o. At high 0 00 15 os, the qualitative behavior of G (o) and G (o) is essentially the same and unaffected 0 00 16 by frequency. However, at low frequencies G (o) and G (o) increase monotonically 17 with increasing clay content. In the low frequency region, the curves can be expressed 0 00 18 by the power-law of G (o) / o2 and G (o) / o for neat PLA, suggesting that this is

19 similar to those of the narrow Mw distribution homopolymer melts. On the other < 1 0 20 hand, for aT 5 rad s , the viscoelastic response [particularly G (o)] for all the 21 nanocomposites displays significantly diminished frequency dependence as com- 0 22 pared to the matrices. In fact, for all PLANCs, G (o) becomes nearly independent at 00 23 low aToand exceeds G (o), characteristic of materials exhibiting a pseudo-solid-like 24 behavior [33]. The values of the terminal zone slopes of both neat PLA and PLANCs < 1 25 are estimated in the lower aTo region ( 10 rad s ), and are presented in Table 9.5. 26 The lower slope values and the higher absolute values of the dynamic moduli 27 indicate the formation of a “spatially-linked” structure in the PLANCs in the molten 28 state [36]. Because of this structure or high geometric constraints, the individual 29 stacked silicate layers are incapable of rotating freely and hence, by imposing small

30 aTo, the relaxations of the structure are prevented almost completely. This type of 31 prevented relaxation due to the high geometric constraints of the stacked and 32 intercalated silicate layers leads to the presence of the pseudo-solid-like behavior, 33 as observed in PLANCs. This behavior probably corresponds to the shear-thinning < 1 Ã 34 tendency, which appears strongly in the viscosity curves (aTo 5 rad s )(Z vs aTo). 35 Such a feature depends strongly on the shear rate in the dynamic measurement 36 because of the formation of shear-induced alignment of the dispersed MMT 37 particles [43]. – – 38 The temperature dependence frequency shift factor (aT, Williams Landel Ferry 39 type [44]) used to generate the master curves shown in Figure 9.9 is shown in 40 Figure 9.10. The dependence of the frequency shift factors on the silicate loading 41 suggests that the temperature-dependent relaxation process observed in the visco- 42 elastic measurements is somehow affected by the presence of the silicate layers [33]. 43 In the case of the Nylon 6-based nanocomposite, where there is hydrogen bonding to 44 the silicate surface, the system exhibits a high flow activation energy nearly one order 45 of magnitude greater than that of neat Nylon 6 [45]. 292j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Fig. 9.9 Reduced frequency dependence of storage modulus, loss 44 modulus and complex viscosity of neat PLA and various nano- 45 composites (PLANCs). Reprinted from [40], Ó 2003, Elsevier Science. 9.3 Control of Nanostructure Properties j293 0 00 o 1 Tab. 9.5 Terminal slopes of G and G vs. aT for PLA and various PLACNs. 2 3 0 00 System G G 4 5 PLA 1.3 0.9 6 PLACN4 0.2 0.5 7 PLACN5 0.18 0.4 PLACN7 0.17 0.32 8 9 Source: Reprinted from [40], Ó 2003, Elsevier Science. 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 Fig. 9.10 (A) Frequency shift factors aT and (B) modulus shift factor b as a function of temperature. Reprinted from [40], 45 T Ó 2003, Elsevier Science. 294j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 Fig. 9.11 Flow activation energy of pure PLA and various PLA- 15 based nanocomposites as a function of MMT content. Reprinted from [47], Ó 2003 Wiley-VCH Verlag GmbH & Co. 16 17 The shift factor b shows a large deviation from a simple density effect, it would be 18 T expected that the values would not vary far from unity [44]. One possible explanation 19 is an internal structure development occurring in PLANCs during measurement 20 (shear process). The alignment of the silicate layers probably enables the PLFNC 21 melts to withstand the shear force, thus leading to the increase in the absolute values 22 of G0(o) and G00(o). 23 Galgali and his colleagues [46] have also shown that the typical rheological 24 response in nanocomposites arises from frictional interactions between the silicate 25 layers and not from the immobilization of confined polymer chains between the 26 silicate layers. They have also shown a dramatic decrease in the creep compliance for 27 the PP-based nanocomposite with 9 wt% MMT. They showed a dramatic three orders 28 of magnitude drop in the zero shear viscosity beyond the apparent yield stress, 29 suggesting that the solid-like behavior in the quiescent state is a result of the 30 percolated structure of the layered silicate. 31 Figure 9.11 represents the MMT content-dependent (wt%) flow activation energy 32 (E ) of pure PLA and various PLACNs obtained from an Arrhenius fit of the master 33 a curves [47]. It is clearly observed that the E value increases significantly for the 34 a nanocomposite containing 3 wt% MMTand then is almost unchanged with increas- 35 ing MMT content. This result indicates that, in the presence of MMT, it is very 36 difficult for the materials to flow. This behavior is also ascribed to the formation of a 37 spatially linked structure in the nanocomposite in the molten state. 38 39 40 9.3.3 41 Elongational Flow and Strain-induced Hardening 42 43 Okamoto et al. [48] first conducted an elongation test on PP-based nanocomposites in _ 44 the molten state at constant Hencky strain rate, e0 using elongation flow optorheo- 45 metry [49]. They also attempted to control the alignment of the dispersed silicate 9.3 Control of Nanostructure Properties j295

1 layers with nanometer dimensions of an intercalated PP-based nanocomposite under 2 uniaxial elongational flow.

3 Figure 9.12 shows double-logarithmic plots of transient elongational viscosity (ZE) 4 against time (t) observed for the PLA-based nanocomposite (PLACN) (MMT ¼ 4 wt%)  _ 1 5 at 170 Cwith different Hencky strain rates (e0) ranging from 0.01 to 1 s [47]. A very 6 strong tendency for strain-induced hardening in the nanocomposite melt was ob-

7 served. In the early stage, ZE gradually increases with t but is almost independent of 8 _ e0; this is called the linear region of the viscosity curve. After a certain time, tZE which is 9 called the up-rising time (marked with the upward arrows in Figure 9.12(A)), strongly _ 10 dependent on e0, rapid upward deviation of ZE from the curves of the linear region is 11 seen. The very low shear viscosity of pure PLA was given as the main reason for this, 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Fig. 9.12 (A) Time variation of elongational viscosity for PLA- 43 based nanocomposite (MMT ¼ 4 wt%) melt at 170 C; (B) Strain 44 rate dependence of up-rising Hencky strain. Reprinted from [47], 45 Ó 2003 Wiley-VCH Verlag GmbH & Co. 296j 9 Biodegradable Polymer-based Nanocomposites

1 because the minimum viscosity range of their instrument was greater than 104 Pa s. 2 However, they confirmed that neither strain-induced hardening in elongation nor 3 rheopexy in shear flow took place with pure PLA having the same molecular weight 4 and polydispersity as the nanocomposite. _ ffi 5 As in PP-based nanocomposite systems, the extended Trouton rule, 3Z0 (g;t) ZE _ 6 (e0;t), also does not hold for PLANC melts, in contrast to the melt of pure polymers. 7 These results indicate that in the case of PLANC, the flow induced internal structural 8 changes also occur in elongation flow [48], but the changes are quite different in shear 9 flow. The strong rheopexy observed in the shear measurements for the PLA-based 10 nanocomposite at very slow shear rate reflects the fact that the shear-induced 11 structural change involved a process with an extremely long relaxation time. 12 Regarding elongation-induced structure development, Figure 9.12(B) shows the 13 ð Þ¼_ Â Hencky strain rate dependence of the up-rising Hencky strain eZE e0 tZE  14 _ recorded for PLANC at 170 C. The eZE increases systematically with the e0. The 15 _ lower the value of e0, the smaller the value of eZE . This tendency probably corresponds 16 to the rheopexy of PLANC under slow shear flow. 17 18 19 9.4 20 Physicochemical Phenomena 21 22 9.4.1 23 Biodegradability 24 25 Another most interesting and exciting aspect of nanocomposite technology is the 26 significant improvements in biodegradability of biodegradable polymers after 27 nanocomposite preparation with OMLS. Aliphatic polyesters are among the most 28 promising materials for the production of environmentally friendly biodegradable 29 plastics. Biodegradation of aliphatic polyester is well known, in that some bacteria 30 degrade it by producing enzymes, which attack the polymer. Tetto and his colleagues 31 [50] first reported some results on the biodegradability of nanocomposites based on 32 PCL, where they found that the PCL-based nanocomposites showed improved 33 biodegradability compared to pure PCL. They suggest that the improved biodegrad- 34 ability of PCL after nanocomposite formation may be due to the catalytic role of the 35 OMLS in the biodegradation mechanism. But it is still unclear how the clay increases 36 the biodegradation rate of PCL. 37 In 2002, Lee et al. [51] reported the biodegradation of aliphatic polyester-based 38 nanocomposites under compost. Figure 9.13(A, B) represent the clay content 39 dependence of biodegradation of APES-based nanocomposites prepared with two 40 different types of MMT clays. They assumed that the retardation of biodegradation 41 was due to the improvement of the barrier properties of the aliphatic APSE after 42 nanocomposite preparation with clay. However, there are no data about permeability. 43 Recently, Yamada and Okamoto et al. [52–54] first reported the biodegradability of 44 neat PLA and PLA-based nanocomposites prepared with trimethyl octadecylammo- fi + 45 nium-modi ed MMT (MMT-C18(CH3)3N ) with a detailed mechanism. The compost 9.4 Physicochemical Phenomena j297

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Fig. 9.13 Biodegradability of APES nanocomposites with: (A) 29 Closite 30B and (B) Closite 10A. Reprinted from [51], Ó 2002, 30 Elsevier Science. 31 32 33 used was prepared from food waste and tests were carried out at temperature of  34 58 + 2 C. Figure 9.14(A) shows photographs of the samples of neat PLA and 35 nanocomposite (MMT ¼ 4 wt%) recovered from the compost at various times. The

36 decreased molecular weight Mw and residual weight percentage Rw of the initial test 37 samples with time is also reported in Figure 9.14(B). The biodegradability of neat PLA 38 is significantly enhanced after nanocomposite preparation. Within one month, both

39 the Mw and the weight loss are almost the same for both PLA and the nanocomposite. 40 However, after one month, a sharp change occurs in the weight loss of nanocompo- 41 site, and within two months it is completely degraded into compost. The degradation 42 of PLA in compost is a complex process involving four main phenomena, namely: 43 water absorption, ester cleavage and formation of oligomer fragments, solubilization 44 of oligomer fragments, and finally diffusion of soluble oligomers by bacteria [55]. 45 Therefore, the factor which increases the hydrolysis tendency of PLA ultimately 298j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Fig. 9.14 (A) Photographs of biodegradability of neat PLA and 32 PLA-based nanocomposite recovered from compost with time. 33 Initial size of the crystallized samples was 3 Â 10 Â 0.1 cm3. 34 (B) Time dependence of residual weight, Rw and of matrix, Mw of PLA and PLACN4 under compost at 58 + 2 C. Reprinted from 35 [52], Ó 2002 American Chemical Society. 36 37 38 39 controls the degradation of PLA. The authors expect that the presence of terminal 40 hydroxylated edge groups on the silicate layers may be one of the factors responsible 41 for this behavior. In the case of PLA-based nanocomposite, the stacked (4 layers) 42 and intercalated silicate layers are homogeneously dispersed in the PLA matrix (as 43 seen from the TEM image [54]) and these hydroxy groups start heterogeneous 44 hydrolysis of the PLA matrix after absorbing water from the compost. This process 45 takes some time to start which explains why the weight loss and degree of hydrolysis 9.4 Physicochemical Phenomena j299

1 of PLA and nanocomposite is almost the same in the first month (see Figure 9.14(B)). 2 However, after one month there is a sharp weight loss with the nanocomposite 3 compared to that with PLA. This means that one month is the critical time after which 4 heterogeneous hydrolysis starts and the matrix breaks into very small fragments and 5 disappears as compost. This assumption was confirmed by conducting same type of + fi 6 experiment with PLANC prepared by using 2C18(CH3)2N modi ed syn-FH which 7 has no terminal hydroxylated edge group, the degradation pattern was almost the 8 same as with neat PLA [53]. 9 They also conducted a respirometric test to study the degradation of the PLA matrix  10 in a compost environment at 58 + 2 C. For this test the compost used was made 11 from bean-curd refuse, food waste, and cattle feces. Unlike weight loss, which reflects

12 the structural changes in the test sample, CO2 evolution provides an indicator of the 13 ultimate biodegradability of PLA in a PLA-based nanocomposite (prepared with fi 14 qC14(OH) modi ed syn-FH,), that is mineralization of the samples. Figure 9.15 15 shows the time dependence of the degree of biodegradation of neat PLA and 16 nanocomposite, indicating that the biodegradability of PLA in the nanocomposite 17 is enhanced significantly. The presence of OMLF may thus cause a different mode of 18 attack on the PLA component, which might be due to the presence of hydroxy groups. 19 Details of the mechanism of biodegradability are presented in the relevant literature 20 [53,54]. 21 K. Okamoto and M. Okamoto also investigated the biodegradability of neat PBS 22 before and after nanocomposite preparation with three different types of OMLF. 23 They used alkylammonium or alkylphosphonium salts for the modification of 24 pristine layered silicates, and these surfactants are toxic for microorganisms [56]. 25 Figure 9.16(A) shows photographs of samples of neat PBS and various nanocom- 26 posites recovered from the compost after 35 days. It can clearly be seen that many 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Fig. 9.15 (A) Degree of biodegradation (i.e., CO evolution), and 43 2 (B) time-dependent change of matrix Mw of neat PLA and PLA- 44 based nanocomposite (syn-FH ¼ 4 wt%) under compost at 45 58 + 2 C. Reprinted from [53], Ó 2004 WILEY-VCH. 300j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Fig. 9.16 Biodegradability of neat PBS and various nanocompo- 16 site sheets (A) under compost, and (B) under soil field. Reprinted 17 from [56], Ó 2003 John Wiley & Sons, Inc. 18 19 20 cracks have appeared in the nanocomposite samples compared to that of neat PBS. 21 This indicates the improved degradability of nanocomposites in compost. This kind 22 of fracture has an advantage for biodegradation because it is easy to mix with the 23 compost and create much more surface area for further attack by microorganisms. It 24 should be noted that the extent of fragmentation is directly related to the nature of 25 OMLF used in the nanocomposite preparation. Gel permeation chromatography 26 (GPC) was performed on the samples recovered from the compost, and it was found 27 that the extent of molecular weight loss was almost the same for all samples. This 28 indicates that the extent of hydrolysis of PBS in the pure state or in OMLF filled 29 systems is the same in compost.

30 Except for the PBS/SAP-qC16 (n-hexadecyl tri-n-butyl phosphonium cation modi- 31 fied saponite) system, the degree of degradation is the same for other samples. This 32 indicates that MMT or alkylammonium cations, and at the same time other proper- 33 ties, have no effect on the biodegradability of PBS. The accelerated degradation of

34 PBS matrix in the presence of SAP-qC16 may be due to the presence of alkylpho- 35 sphonium surfactant. This kind of behavior is also observed in the case of PLA/MMT- 36 based nanocomposite systems. 37 The nature of the degradation of PBS and various nanocomposites under a soil 38 field was also observed. This experiment was conducted for one, two, and six months. 39 After one and two months, there was no change in the nature of the surface of the 40 samples, but after six months black or red spots appeared on the surface of the 41 nanocomposite samples. Figure 9.16(B) represents the results of degradation of neat 42 PBS and various nanocomposite sheets recovered from a soil field after six months. 43 The spots on the sample surface were reported as being due to fungus attack because 44 when they were put into a slurry clear growth of fungus was observed. These results 45 9.4 Physicochemical Phenomena j301

1 also indicate that nanocomposites exhibit the same or a higher level of biodegrad- 2 ability compared with the PBS matrix. 3 4 9.4.2 5 Photodegradation 6 7 Hiroi and Okamoto et al. [57] first reported the photodegradability of neat PLA and 8 the corresponding PLA-based nanocomposite prepared by using organically modi- 9 fied layered titanate (HTO) as a new nano-filler. One of the features of this material

10 is its photocatalytic reactivity, like that of titania (TiO2). The photocatalytic reactions 11 of anatase-TiO2, such as the evolution of hydrogen gas from water or the oxidative 12 degradation of organic compounds, have attracted intense research interest be- 13 cause of their possible application to the conversion of solar energy into chemical 14 energy [58]. 15 Figure 9.17 shows the UV/VIS transmission spectra of pure PLA and nanocom- 16 posite (PLANC1.7: HTO ¼ 1.7 wt%). The spectra show change in the VIS region 17 (>400 nm) with increasing absorption in the presence of titanate layers compared 18 with neat PLA. For UV wavelengths, there is strong absorption up to 320 nm, 19 resulting in 0 % transmittance. This significant change in the spectra may indicate 20 photodegradation of the PLA matrix. To confirm this, some preliminary experiments 21 on the photodegradation of PLA-based nanocomposites under a sunshine weath-  22 ermeter at 60 C were carried out. After 300 h, there was no change in the nature of 23 the sample surfaces of neat PLA, however, the surface color of the nanocomposite 24 samples became yellow and/or light brown. Table 9.6 shows the GPC measurement

25 of samples recovered from the test. The drop in Mw accompanied by broadening of fi 26 Mw/Mn indicates that enhancement of degradation of PLA in the titanate- lled 27 system has occurred. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Fig. 9.17 UV/VIS transmission spectra of neat PLA and 43 nanocomposite (PLANC1.7). Reprinted from [57], Ó 2004 44 WILEY-VCH. 45 302j 9 Biodegradable Polymer-based Nanocomposites

1 Tab. 9.6 GPCresults of sample recovered from weathermeter after 2 300 h.

3 Samples M Â 10 3 (g/mol) M /M M =M0 a 4 w w n w w 5 PLA 198 1.53 0.94 6 PLANC1.7 93.7 1.89 0.68 7 PLANC3.9 86.3 1.86 0.76

8 a 0 Mw is molecular weight before test. 9 Source: Reprinted from [57], Ó 2004 WILEY-VCH. 10 11 9.5

12 Foam Processing using Supercritical CO2 13 14 9.5.1 15 PLA-based Nanocomposite 16 17 PLA-based nanocomposites have already been shown to exhibit a tendency toward 18 strong strain-induced hardening. This strain-induced hardening behavior is an 19 indispensable characteristic for foam processing due to its capacity to withstand 20 the stretching force experienced during the latter stages of bubble growth. On the 21 basis of this result, the first successful nanocomposite foam, processed by using

22 supercritical (sc)-CO2 as a physical foaming agent, appeared through a pioneering 23 effort by Okamoto et al. [9,10]. A small amount of nano-fillers in the polymer matrix 24 serve as nucleation sites to facilitate bubble nucleation during foaming. Novel fi 25 nanocomposite foams based on the combination of new nano- llers and sc-CO2 26 have led to a new class of materials. The process consists of four stages: (i) saturation

27 of CO2 in the sample at the desired temperature, (ii) cell nucleation when the release 28 of CO2 pressure starts (supersaturated CO2), (iii) cell growth to an equilibrium size 29 during the release of CO2, and (iv) stabilization of the cell via cooling of the foamed 30 sample. The set-up of the autoclave used in the study is shown in Figure 9.18. 31 Figure 9.19 shows typical scanning electron microscope (SEM) images of the 32 fracture surfaces of the PLA-based nanocomposite (PLA/MMT-ODA) and neat PLA  33 without MMT foamed in the temperature range 100–140 C under different isobaric 34 saturation conditions (14, 21 and 28 MPa) [59]. All foams exhibit nicely the closed-cell 35 structure. We note here that homogeneous cells were formed in the case of nano- 36 composite foams, while neat PLA foams show a rather non-uniform cell structure 37 having large cell size. The nanocomposite foams show smaller cell size (d) and larger

38 cell density (Nc) compared to the neat PLA foam, suggesting that the dispersed 39 silicate particles act as nucleating sites for cell formation [9]. 40 For both foam systems, the calculated distribution functions of the cell size from 41 the SEM images are presented in Figure 9.20. The nanocomposite foams nicely  42 obeyed a Gaussian distribution. In the case of PLA/ODA foamed at 150 C under a 43 high pressure of 24 MPa, we can see that the width of the distribution peaks, which 44 indicates the dispersity for cell size, became narrow, accompanied by a finer 45 dispersion of silicate particles. j 9.5 Foam Processing using Supercritical CO2 303

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fig. 9.18 Schematic representation of autoclave set-up. Reprinted 20 with permission from [10], Ó 2002, Society of Plastic Engineers. 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Fig. 9.19 Typical results of SEM images of the fracture surfaces of 43 PLA/MMT-ODA and neat PLA foamed in the temperature range 44 100–140 C under different isobaric saturation conditions (14, 21 45 and 28 MPa). Reprinted from [59], Ó 2006, Elsevier Science. 304j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Fig. 9.20 Typical example of cell size distribution of foamed PLA/ 21 MMT-ODA and neat PLA in experiment at 150 C under 24 MPa. s2 2 22 Average values of d in mm and variances d in mm in the Gaussian 23 fit through the data are 24.2 and 19.1 for PLA/MMT-ODA foam and 58.3 and 171.0 for PLA foam. Reprinted from [59], Ó 2006, 24 Elsevier Science. 25 26  27 Obviously, with decreasing saturation pressure (140 C and 14 MPa), both foams

28 exhibit large cell size due to the low supply of CO2 molecules, which can subsequently 29 form a small population of cell nuclei upon depressurization. The incorporation of 30 MMT induces heterogeneous nucleation because of a lower activation energy barrier 31 compared with homogeneous nucleation [60]. However, the competition between 32 homogeneous and heterogeneous nucleation is no longer discernible. 33 34 9.5.2 35 Temperature Dependence of Cellular Structure 36

37 The dependence of the foam density (rf) at the Tf under different CO2 pressures is 38 shown in Figure 9.21. Throughout the whole CO2 pressure range, the mass density of 39 PLA/MMT-ODA foams remains constant at low foaming temperature (Tf) range, 40 abruptly decreases beyond a certain Tf, and then attains a minimum constant value  41 up to 150 C. It can be said that such behavior of mass density is due to the  42 competition between cell nucleation and cell growth. In the low Tf range ( 110 C), 43 where a large supply of CO2 molecules is provided, cell nucleation is dominant, while  44 at high Tf,( 140 C), the cell growth and coalescence of the cell predominate due to  45 the low viscosity of the systems compared with the low Tf range ( 110 C). This j 9.5 Foam Processing using Supercritical CO2 305

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Fig. 9.21 Foaming temperature dependence of mass density for 17 PLA/MMT-ODA foamed under different CO2 pressures. Rep- Ó 18 rinted from [59], 2006, Elsevier Science. 19 20 ffi 21 behavior is clearly seen in the plots of the cell size ( 2d), the cell density (Nc), and the 22 mean cell wall thickness (d) versus Tf under various pressure conditions. As seen in 23 Figure 9.22, with increasing Tf all nanocomposite foams show an increasing 24 tendency for 2d and/or d to attain a maximum. On the other hand, the temperature

25 dependence of Nc shows the opposite behavior compared with the tendency of 2d due 26 to cell growth and coalescence. Both 2d and Nc affect the mass density of the foams. 27 Using the glass transition temperature (Tg) depressions (corresponding to DTg), þ 28 plots of rf versus Tf DTg were constructed from the data of Figure 9.21. The results 29 are shown in Figure 9.23. All the data, including those for neat PLA and different 30 types of PLA-based nanocomposites (PLA/MMT-SBE), conform well to a reduced 3 þ <  31 curve with rf 1.0 + 0.1 g/cm at Tf DTg 140 + 4 C (nano-cellular region), 3 þ 32 whereas rf values approach 0.3 + 0.15 g/cm as the reduced temperature (Tf DTg)  33 is increased well above 150 C (microcellular region). The critical temperature is thus  34 140 + 4 C, above which cell growth prevails. Below the critical temperature, cell 35 nucleation dominates and cell growth is suppressed due to the high modulus and 0 36 viscosity, as revealed by the temperature dependence of the storage, G (o) and loss, 00 0 00  37 G (o) moduli (G ¼ 162 MPa and viscosity component; G /o ffi 2 MPa s at 140 C). þ 38 Figure 9.24 shows temperature-reduced plots of 2d, Nc and d versus Tf DTg. All 39 the data conform well to a reduced curve like Figure 9.23. Interestingly, when the

40 authors used both Tg and the melting temperature (Tm) [61] depressions to conduct 41 superposition, they recognized that the reduced curve is nicely constructed but there fi þ 42 is no signi cant difference compare with the case of Tf DTg. This indicates that Tg 43 depression is important in optimizing foam processing conditions but Tm depression fi 44 is not a signi cant factor for processing because the Tf range is still below Tm after 45 CO2 saturation. 306j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Fig. 9.22 Foaming temperature dependence of (A) cell size, (B) 27 cell density and (C) mean cell wall thickness under different CO2 28 pressures. Reprinted from [59], Ó 2006, Elsevier Science. 29 30 31 In Figure 9.25, the relations between 2d and N , and d and 2d are shown. The 32 c relation obeys well Equations (2) and (3). 33 ½ ðr =r ފ 34 4 3 1 f p Nc ¼ 10 ð2Þ 35 4pd3 36 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 37 ¼ ð = ð = Þ ÞðÞ d d 1 1 rf rp 1 3 38

39 where rp and rf are the mass density of pre-foamed and post-foamed samples, 40 respectively. 12 3 41 But the deviation occurs beyond the value of Nc 10 cell/cm for panel (A) and 42 below 2d 1 mm for panel (B). The downward and upward deviations indicate that the 43 heterogeneous cell distribution mechanism due to the rigid crystalline phases in the 44 PLA matrix is caused by the high degree of crystallinity (49 wt%) in the low foaming  45 temperature range (100 C). As seen in Figure 9.26, the PLANC foams exhibit j 9.5 Foam Processing using Supercritical CO2 307

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Fig. 9.23 Plot of mass density for PLA/MMT-ODA, PLA/MMT- SBE and neat PLA versus reduced foaming temperature 21  (Tf þ DTg). The critical temperature (140 + 4 C) is shaded. 22 Reprinted from [59], Ó 2006, Elsevier Science. 23 24 25 heterogeneous cell distribution. The PLA foam reduces the value of Nc accompanied 26 by a large value of d compared with that of PLANC foams. In the case of PLANC 27 ffi foams, the controlled structure is from microcellular (2d 30 mm and Nc 28  7 3 ffi ffi  13 3 3.0 10 cells/cm ) to nano-cellular (2d 200 nm and Nc 2.0 10 cells/cm ). 29 30 31 9.5.3

32 CO2 Pressure Dependence 33 34 At high pressure, both homogeneous and heterogeneous nucleation mechanisms fi 35 may be of comparable signi cance. All systems demonstrate that Nc increases –  36 systematically with increasing CO2 pressure in the low Tf region ( 100 120 C). 37 For PLA/MMT-ODA foams, it appears that heterogeneous nucleation is favored at 38 high pressure. Cell nucleation in a heterogeneous nucleation system such as PLA/ 39 MMT-ODA foams takes place in the boundary between the matrix and the dispersed 40 MMT (nano-filler) particles. Accordingly, the cell size decreases without individual 41 cell coalescence for PLA/MMT-ODA and neat PLA systems, as seen in Figure 9.26. To 42 investigate whether the addition of internal surfaces of the dispersed nano-clay may

43 hinder CO2 diffusion by creating a more tortuous diffusive pathway [21], characteri- 44 zation of the interfacial tension between bubble and matrix was carried out using the 45 modified classical nucleation theory [60]. 308j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Fig. 9.24 Temperature-reduced plots of (A) 2d, (B) Nc and (C) d 33 versus Tf þ DTg for PLA/MMT-ODA, PLA/MMT-SBE and neat 34 PLA. Reprinted from [59], Ó 2006, Elsevier Science. 35 36 37 According to the theory proposed by Suh and Colton, the rate of nucleation of cells 38 per unit volume (N_ ) can be written as 39 "# 40 3 ð Þ _ 16pg S y ð Þ 41 N Cf exp 2 4 3ðDPCO2Þ kBT 42

43 where C is the concentration of CO2 and/or the concentration of heterogeneous 44 nucleation sites, f is the collision frequency of CO2, g is the interfacial tension 45 between bubble and matrix, S(y) is the energy reduction factor for the heterogeneous j 9.5 Foam Processing using Supercritical CO2 309

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Fig. 9.25 Relation between (A) Cell size and cell density and 42 (B) cell wall thickness and cell size for all foams. Reprinted from [59], Ó 2006, Elsevier Science. 43 44 45 310j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Fig. 9.26 SEM images of the fracture surfaces of (A) neat PLA,  32 (B) PLA/MMT-ODA and (C) PLA/MMT-SBE foamed at 100 C under 28 MPa. Reprinted from [59], Ó 2006, Elsevier Science. 33 34 35

36 nucleation (i.e., PLA/MMT-ODA), DPCO2 is the magnitude of the pressure quench 37 during depressurization, kB is the Boltzmann constant, and T is the absolute 38 temperature. 39 The theoretical cell density is given by ð 40 t _ 41 Ntheor ¼ Ndt ð5Þ 42 0 43 where t is the foaming time, approximately 3 s. Assuming no effect of the coalescence

44 of the cell on the value of Nc, we estimate the interfacial tension of the systems 45 calculated using Equations (9.4) and (9.5), that is, the slope of the plots (Nc, versus j 9.5 Foam Processing using Supercritical CO2 311

1 Tab. 9.7 Characteristic interfacial parameters of two systems. 2 T  S 1/3 2 S  3 f/ C g (u) /mJ/m (u) u/ 4 PLA/CO2 110 7.43 5 PLA/MMT-ODA/CO2 6.65 0.717 107.3 6 PLA/CO2 120 7.08 7 PLA/MMT-ODA/CO2 5.38 0.439 85.3 8 Source: Reprinted from [59], Ó 2006, Elsevier Science. 9 10 = 11 1 DPCO2 ). The characteristic parameters of two systems are shown in Table 9.7. The 12 interfacial tension of PLA/MMT-ODA and neat PLA are 6.65 and 7.43 mJ/m2 at  13 110 C, respectively. These estimated g values are in good agreement with that of the 2 14 poly(methyl methacrylate) (PMMA)-CO2 system ( 10 mJ/m ) [62]. The PLA/MMT- 15 ODA system has a low value compared to that of neat PLA. This trend reflects the 16 relative importance of heterogeneous nucleation, which dominates over homoge-

17 neous nucleation when the amount of CO2 available for bubble nucleation is limited 18 because of a lower activation energy barrier, as mentioned before. That is, in 19 heterogeneous nucleation (PLA/MMT-ODA), we have to take the reduction of the 20 critical energy into consideration because of the inclusion of nucleants, which is a 21 function of the PLA–gas–nano-clay contact angle (y) and the relative curvature (W)of 22 the nucleant surface to the critical radius of the nucleated phase [63]. When W  10, 23 the energy reduction factor S(y) can be express by 24 SðyÞ¼ð1=4Þð2 þ cosyÞð1cosyÞ2 ð6Þ 25 26 In the case of homogeneous nucleation S(y) is unity (y ¼ 180). The obtained values 27 of the contact angle are 107.3 at 110 C and 85.3 at 120 C. The estimated reduction 28 factor (S(y) ¼ 0.4–0.7) was not so small when compared with the other nano-filler 29 (e.g., carbon nanofingers, S(y) ¼ 0.006) [64]. However, experimentally, nano-clay

30 particles lead to an increase in Nc. – 31 For PLA/MMT-SBE foams prepared under the condition with low Tf ( 100 32 110 C) and high pressure (28 MPa), the nanocomposite foams exhibit no signifi-

33 cant difference in Nc compared with PLA/MMT-ODA foams. This reasoning is 34 consistent with the large value of W in both systems. 35 36 9.5.4 37 TEM Observation 38 39 To confirm the heterogeneous nucleation and the nanocellular feature in the foam 40 processing, TEM observation of the cell wall in the PLA/MMT-ODA foam was 41 conducted.  42 Figure 9.27 shows the TEM micrograph of the cell wall foamed at 100 C under 43 28 MPa. Interestingly, the grown cells with a diameter of 200 nm are localized along 44 the dispersed nano-clay particles in the cell wall. In other words, the dispersed nano- 45 filler particles act as nucleating sites for cell formation and the cell growth occurs on 312j 9 Biodegradable Polymer-based Nanocomposites

1 the surfaces of the MMTs, that is, the cellular structure has an oval-faced morphology  2 rather than a spherical cellular structure for the high Tf condition ( 140 C). In 3 Figure 9.27, in addition to the nano-cellular structure formation, a lammellar pattern 4 beside the nano-clay (MMT) particles is observed. This behavior appears to arise from 5 the formation of the a-phase of the PLA crystal in the presence of nano-clay particles 6 [65]. This is a unique observation of the epitaxial crystallization of PLA grown from 7 MMT (clay) surfaces due to the nucleation effect of the dispersed nano-clays. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Fig. 9.27 TEM micrograph for the structure of PLA/MMT-ODA 44 cell wall foamed at 100 C under 28 MPa. Reprinted from [59], 45 Ó 2006, Elsevier Science. 9.6 Porous Ceramic Materials via Nanocomposites j313

1 9.5.5 2 Mechanical Properties of Nanocomposite Foams 3

4 Figure 9.28 shows the relation of the relative modulus (Kf/Kp) to the relative density 5 (rf/rp) of neat PLA and PLA-based nanocomposite foams, taken in the directions 6 parallel (A) and perpendicular (B) to the flow. 7 To clarify whether the modulus enhancement of the nanocomposite foams was 8 reasonable, Equation (9.7) proposed earlier by Kumar [66] was applied to estimate the 9 relative moduli with various foam densities: ! ! ! 10 4 2 Kf rf rf rf 11 ¼ þ ð7Þ Kp rp rp rp 12 13 where Kp and Kf are the modulus of pre-foamed and post-foamed samples, respec- fi fi 14 tively. The solid line in the gure represents the t to Equation (9.7). The neat PLA 15 foams do not show any difference between the two moduli. On the other hand, for 16 nanocomposite foams, the relative moduli exhibit a large value compared with the 17 theoretical one. The dispersed MMTparticles in the cell wall align along the thickness 18 direction of the sample. In other word, the MMT particles arrange due to the biaxial fl 19 ow of material during foaming. The MMTparticles seem to act as a secondary cloth 20 layer to protect the cells from being destroyed by external forces. In the direction fl 21 perpendicular to the ow, the relative moduli of PLA/MMT-ODA and PLA/MMT-SBE 22 foams appear higher than the predicted values even though at the same relative mass – 23 density, in the range 0.7 0.85 (see Figure 9.28(a)). This upward deviation suggests 24 that the small cell size with large cell density enhances the material property, as 25 predicted by Weaire [67]. This may create an improvement in the mechanical 26 properties of polymeric foams through polymeric nanocomposites. 27 28 29 9.6 30 Porous Ceramic Materials via Nanocomposites 31 32 A new route for the preparation of porous ceramic material from thermosetting 33 epoxy/MMT nanocomposite was first demonstrated by Brown et al. [68]. This route 34 offers attractive potential for diversification and application of the PLFNCs. Okamoto 35 and coworkers have reported the results on a novel porous ceramic material formed 36 via burning of the PLA-based nanocomposite [69]. In the nanocomposite containing 37 3.0 wt% inorganic MMT. The SEM image of the fracture surface of the porous Q2 38 ceramic material prepared from simple burning of the PLANC in a furnace at up to  39 950 C is shown in Figure 9.29. After complete burning the nanocomposite becomes 40 a white mass with a porous structure. The bright lines in the SEM image correspond 41 to the edge of the stacked silicate layers. In the porous ceramic material, the silicate 42 layers form a ‘house of cards’ structure, which consists of large plates of length 43 1000 nm and thickness 30–60 nm. This implies that the further stacked platelet 44 structure is formed during burning. The material exhibits an open-cell type structure 2 1 45 having a 100–1000 nm diameter void, a BET surface area of 31 m g and a low 314j 9 Biodegradable Polymer-based Nanocomposites

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

41 Fig. 9.28 The relation of the relative modulus (Kf/Kp) and the 42 relative density (rf/rp) of neat PLA and PLA-based nanocompo- 43 site foams, taken in the directions parallel (A) and perpendicular (B) to the flow. 44 45 9.7 Future Prospects j315

1 2 3 4 5 6 7 8 9 10 11 Fig. 9.29 SEM image of porous ceramic material after coating 12 with a platinum layer (10 nm thickness). Reprinted from [69], Ó 13 2002 American Chemical Society. 14 15 density of porous material of 0.187 g ml 1, estimated by the buoyancy method. The 16 BETsurface area value of MMT is 780 m2/g and that of the porous ceramic material is 17 31 m2/g, suggesting that there are about 25 MMT plates stacked together. When   18 MMT is heated above 700 C (but below 960 C) first all the OH groups are eliminated 19 from the structure and thus MMT is decomposed into a non-hydrated aluminosili- 20 cate. This transformation radically disturbs the crystalline network of the MMT, and 21 the resulting diffraction pattern is indeed often typical of an amorphous (or non- 22 crystalline) phase. The estimated value of the compression modulus (K) is of the 23 order of 1.2 MPa, which is five orders of magnitude lower than the bulk modulus of 24 MMT (102 GPa) [3]. In the stress–strain curve, the linear deformation behavior is 25 nicely described in the early stage of the deformation, that is, the deformation of the 26 material closely resembles that of ordinary polymeric foams [70]. This open-cell type 27 porous ceramic material consisting of the ‘house of cards’ structure is expected to 28 provide strain recovery and an excellent energy dissipation mechanism after un- 29 loading in the elastic region up to 8 % strain, probably each plate bends like a leaf 30 spring. This porous ceramic material is a new material possessing elasticity and is 31 very lightweight. This new route for the preparation of porous ceramic material via 32 burning of nanocomposites can be expected to pave the way for a much broader range 33 of applications of the PLSNCs. This porous ceramic material closely relates an Q3 34 excellent insulator property for flame reterdant of PLFNCs [3]. The flame behavior 35 must derive from the morphological control of the shielding properties of the 36 graphitic clay created during polymer ablation. 37 38 39 9.7 40 Future Prospects 41 42 Development of the PLFNCs is one of the latest evolutionary steps in polymer 43 technology. The PLFNCs offer attractive potential for diversification and application 44 of conventional polymeric materials. Some of the PLFNCs are already commercially 45 available and applied in industrial products. 316j 9 Biodegradable Polymer-based Nanocomposites

1 Biodegradable polymers based on nanocomposites and foams have great future 2 promise for potential applications as high-performance biodegradable materials. 3 They are entirely new types of materials based on materials from plants and nature.

4 When disposed of in compost, they are safely decomposed into CO2, water, and 5 humus through the activity of microorganisms. The CO2 and water will become corn 6 or sugarcane again through plant photosynthesis. Undoubtedly, their unique prop- 7 erties originating from their controlled nanostructure pave the way to a much broader 8 range of applications (some already commercially available through Unitika Ltd., 9 Japan), and open a new dimension for plastics and composites. The major impact is at 10 least a decade away. 11 12 13 References 14 15 1 Smith, R. (2005)Biodegradable Polymer Ohshima, M., Hasegawa, N. and 16 for Industrial Applications, CRC Press, Usuki, A. (2002) Polymer Engineering 42 17 New York. and Science, (9), 1907. 2 Johnson, R.M., Mwaikambo, L.Y. and 11 Vaia, R.A., Ishii, H. and Giannelis, E.P. 18 Tucker, N. (2003) “Biopolymer”, Rapra (1993) Chemistry of Materials, 5, 1694. 19 Review Report No 159, 148 pp Rapra 12 Vaia, R.A. and Giannelis, E.P. (1997) 20 Technology Ltd., London. Macromolecules, 30, 7990. 21 3 Okamoto, M. (2003) “Polymer/Layered 13 Vaia, R.A. and Giannelis, E.P. (1997) 22 Silicate Nanocomposites”, Rapra Macromolecules, 30, 8000. 23 Review Report No 163, 166 pp Rapra 14 Hiroi, R., Sinha Ray, S., Okamoto, M. 24 Technology Ltd., London. and Shiroi, T. (2004) Macromolecular 25 4 Usuki, A., Kojima, Y., Okada, A., Rapid Communications, 25, 1359. 26 Fukushima, Y., Kurauchi, T. and 15 Lagaly, G. (1970) Clay Minerals, 16,1. 27 Kamigaito, O. (1993) Journal of 16 Nakano, S., Sasaki, T., Takemura, K. 8 28 Materials Research, , 1174. and Watanabe, M. (1998) Chemistry of 5 Sinha Ray, S. and Okamoto, M. (2003) Materials, 10, 2044. 29 Progress in Polymer Science, 28, 1539. 17 Teyama, H., Nishimura, S., 30 6 Gao, F. (2004) Materials Today, 7, 50. Tsunematsu, K., Jinnai, K., Adachi, Y. 31 7 Okada, A. and Usuki, A. (2006) and Kimura, M. (1992) Clays and Clay 32 Macromolecular Materials and Minerals, 40, 180. 33 Engineering, 291, 1449. 18 Yoshida, O. and Okamoto, M. (2006) 34 8 Okamoto, M. (2006) Recent advances Macromolecular Rapid 35 in polymer/layered silicate Communications, 27, 751. 36 nanocomposites: an overview from 19 Sinha Ray, S., Yamada, K., Okamoto, 37 science to technology. Materials Science M. and Ueda, K. (2003) Journal of 22 – 3 38 and Technology, (7), 756 779. Nanoscience and Nanotechnology, , 503. 9 Okamoto, M., Nam, P.H., Maiti, M., 20 Yoshida, O. and Okamoto, M. (2006) 39 Q4 Kotaka, T., Nakayama, T., Takada, M., Journal of Polymer Engineering, 40 Ohshima, M., Usuki, A., Hasegawa, N. in press. 41 and Okamoto, H. (2001) Nano Letters, 21 Sinha Ray, S., Yamada, K., 42 1, 503. Okamoto,M.,Ogami,A.andUeda, 43 10 Nam, P.H., Okamoto, M., Maiti, P., K. (2003) Chemistry of Materials, 15, 44 Kotaka, T., Nakayama, T., Takada, M., 1456. 45 References j317

1 22 Saito, T., Okamoto, M., Hiroi, R., 37 Mitchell, C.A. and Krishnamoorti, R. 2 Yamamoto, M. and Shiroi, T. (2006) (2002) Journal of Polymer Science Part B: 40 3 Macromolecular Materials and Polymer Physics, , 1434. 291 4 Engineering, , 1367. 38 Lepoittevin, B., Devalckenaere, M., 23 Saito, T., Okamoto, M., Hiroi, R., Pantoustier, N., Alexandre, M., Kubies, 5 Yamamoto, M. and Shiroi, T. (2006) D., Calberg, C., Jerome, R. and Dubois, 6 Macromolecular Rapid P. (2002) Polymer, 43, 1111. 7 Communications, 27, 1472. 39 Sinha Ray, S., Maiti, P., Okamoto, M., 8 24 Totha, R., Coslanicha, A., Ferronea, M., Yamada, K. and Ueda, K. (2002) 9 Fermeglia, M., Pricl, S., Miertus, S. and Macromolecules, 35, 3104. 10 Chiellini, E. (2004) Polymer, 45, 8075. 40 Sinha Ray, S., Yamada, K., Okamoto, 11 25 Sinsawat, A., Anderson, K.L., Vaia, R. M. and Ueda, K. (2003) Polymer, 44, 12 A. and Farmer, B.L. (2003) Journal of 857. 13 Polymer Science Part B: Polymer Physics, 41 Okamoto, K., Sinha Ray, S. and 41 14 , 3272. Okamoto, M. (2003) Journal of Polymer 41B 15 26 Kuppa, V., Menakanit, S., Science Part B: Polymer Physics, , Krishnamoorti, R. and Manias, E. 3160. 16 (2003) Journal of Polymer Science Part B: 42 Krishnamoorti, R. and Yurekli, K. 17 Polymer Physics, 41, 3285. (2001) Current Opinion in Colloid and 18 27 Zeng, Q.H., Yu, A.B., Lu, G.Q. and Interface Science, 6, 464. 19 Standish, R.K. (2003) Chemistry of 43 Okamoto, M., Taguchi, H., Sato, H., 20 Materials, 15, 4732. Kotaka, T. and Tatayama, H. (2000) 21 28 Sheng, N., Boyce, M.C., Parks, D.M., Langmuir, 16, 4055. 22 Rutledge, G.C., Abes, J.I., and Cohen, 44 Williams, M.L., Landel, R.F. and Ferry, 23 R.E. (2004) Polymer, 45, 487. J.D. (1955) Journal of the American 24 29 Chen, J.S., Poliks, M.D., Ober, C.K., Chemical Society, 77, 3701. 25 Zhang, Y., Wiesner, U. and Giannelis, 45 Nam, P.H. (2001) Master Thesis, 43 26 E.P. (2002) Polymer, , 4895. Toyota Technological Institute. 30 Sinha Ray, S., Okamoto, K. and 46 Galgali, G., Ramesh, C. and Lele, A. 27 Okamoto, M. (2003) Macromolecules, (2001) Macromolecules, 34, 852. 28 36, 2355. 47 Sinha Ray, S. and Okamoto, M. (2003) 29 31 Okamoto, K., Sinha Ray, S. and Macromolecular Rapid 30 Okamoto, M. (2003) Journal of Communications, 24, 815. 31 Polymer Science Part B: Polymer Physics, 48 Okamoto, M., Nam, P.H., Maiti, P., 32 41, 3160. Kotaka, T.,Hasegawa, N. and Usuki, A. 33 32 Utracki, L.A. (1990) Polymer Alloys and (2001) Nano Letters, 1, 295. 34 Blends: Thermodynamics and Rheology, 49 Kotaka, T., Kojima, A. and Okamoto, 35 Hasser Publishers, New York. M. (1997) Rheologica Acta, 36, 646. 36 33 Krishnamoorti, R. and Giannelis, E.P. 50 Tetto, J.A., Steeves, D.M., Welsh, E.A. 30 37 (1997) Macromolecules, , 4097. and Powell, B.E. (1999) ANTEC’99, 34 Fornes, T.D., Yoon, P.J., Keskkula, H. 1628. 38 and Paul, D.R. (2001) Polymer, 42, 51 Lee, S.R., Park, H.M., Lim, H.L., Kang, 39 9929. T., Li, X., Cho, W.J. and Ha, C.S. (2002) 40 35 Hoffman, B., Dietrich, C., Thomann, Polymer, 43, 2495. 41 R., Friedrich, C. and Mulhaupt, R. 52 Sinha Ray, S., Okamoto, M., Yamada, 42 (2000) Macromolecular Rapid K. and Ueda, K. (2002) Nano Letters, 2, 43 Communications, 21, 57. 1093. 44 36 Ren, J., Silva, A.S. and Krishnamoorti, 53 Sinha Ray, S., Yamda, K., Ogami, A., 45 R. (2000) Macromolecules, 33, 3739. Okamoto, M. and Ueda, K. (2002) 318j 9 Biodegradable Polymer-based Nanocomposites

1 Macromolecular Rapid polymers via supercritical CO2,Kyoto 2 Communications, 23, 943. University. 3 54 Sinha Ray, S., Yamada, K., Okamoto, 62 Goel, S.K. and Beckman, E.J. (1994) 34 4 M. and Ueda, K. (2003) Macromolecular Polymer Engineering and Science, , Materials and Engineering, 288, 936. 1137. 5 55 Liu, J.W., Zhao, Q. and Wan, C.X. 63 Fletcher, N.H. (1958) Journal of 6 (2001) Space Medicine & Medical Chemical Physics, 29, 572. 7 Engineering, 14, 308. 64 Shen, J., Zeng, C. and Lee, L.J. (2005) 8 56 Okamoto, K., Sinha Ray, S. and Polymer, 46, 5218. 9 Okamoto, M. (2003) Journal of 65 Nam, J.Y., Sinha, S.R. and Okamoto, 10 Polymer Science Part B: Polymer Physics, M. (2003) Macromolecules, 36, 7126. 11 41, 3160. 66 Kumar, V. and Weller, J.E. (1991) 12 57 Hiroi, R., Sinha Ray, S., Okamoto, M. ANTEC,1401. 13 and Shiroi, T. (2004) Macromolecular 67 Weaire, D. and Fu, T.L.(1988) Journal of 25 32 14 Rapid Communications, , 1359. Rheology, , 271. 15 58 Fujishima, A. and Honda, K. (1972) 68 Brown, J.M., Curliss, D.B. and Vaia, R. Nature, 37, 238. A. (2000)Proceedings of PMSE Spring 16 59 Ema, Y., Ikeya, M. and Okamoto, M. Meeting, San Francisco, 17 (2006) Polymer, 47, 5350. California,278. 18 60 Colton, J.S. and Suh, N.P. (1987) 69 Sinha Ray, S., Okamoto, K., Yamada, K. 19 Polymer Engineering and Science, 27, and Okamoto, M. (2002) Nano Letters, 20 485. 2, 423. 21 61 Takada, M. (2004) PhD Thesis, 70 Gibson L.J. and Ashby M.F. (eds) 22 Crystallization control and foam (1988) Cellular Solids, Pergamon 23 processing of semi-crystalline Press, New York, p.8. 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 1 Author Query 2 3 Q1. You had abbreviated many nanocomposites as e.g. PBSCN. I have 4 assumed you mean-PBSNC etc and have corrected accordingly 5 Q2. This is not a complete sentence, what does it relate to? 6 Q3. I do not understand the sentence "This porous ceramic., please 7 rewrite. 8 Q4. Reference 20 update? 9 Q5. Caption to Fig. 9.3 " (see upper panel)" What upper panel? 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45