Printer Friendly

Analysis of the structure and mass transport properties of nanocomposite polyurethane.

INTRODUCTION

Polymer composites reinforced with inorganic clay minerals of dimensions in the nanometer range, known as nanocomposites, have attracted a great interest of researchers, because of unexpected synergistically properties derived from the two components. This new class of materials is characterized by improved thermal, mechanical, and barrier properties compared with either the matrix either to conventional composites (called also particulate microcomposites) because of their unique phase morphology deriving by layer intercalation or exfoliation, that maximizes interfacial contact between the organic and inorganic phases and enhances bulk properties. The most studied nanocomposites are composed of thermoplastic or thermosetting matrix and an organically modified montmorillonite (OMM) (1-5).

Natural montmorillonite consisted of layers made up of two silicate tetrahedron fused to an edge-shared octahedral sheet of either aluminum or magnesium hydroxide. The physical dimensions of these disc-like shaped silicate layers were typically of about 100 nm in diameter and 1 nm in thickness. Isomorphic substitution within the layers generates negative charges that are normally counter-balanced by cations ([Na.sup.+], [Ca.sup.2+], [K.sup.+]) residing in the interlayer galleries space. Since momtmorillonite is hydrophilic and it is characterized by a poor affinity with hydrophobic organic polymers, organic cations such as alkylammonium cations are used to change the originally hydrophilic silicate interlayer spacing into a hydrophobic surface. The organic cations lower the surface energy of silicate layers and enhance the miscibility between the silicate layers and the polymer matrix (6-10).

Polyurethanes (PUs) are unique polymeric materials with a wide range of physical and chemical properties with applications as coatings, adhesives, fiber, foams, and thermoplastic elastomers (11). They are also used as adhesives to produce multilayer laminates for food packaging. However, their function in laminated films is limited to bond the films and their contribution to the overall barrier performance of the laminate is usually neglected. If the adhesive could contribute to the barrier performance of the laminate, besides adding a new value to the adhesive component, this would lead to a reduction in laminate thickness. A polyurethane gas-barrier coating can also be used. The permeation-barrier properties of the PUs could be achieved by incorporating nano-clay, because the penetrant molecules have to wiggle around them in a random walk, thus diffusing through a tortuous pathway (12). The addition of nanofiller could enhance the gas and vapor barrier properties of the adhesive and hence of a laminated multilayer of packaging film. Since polyurethane/clay nanocomposite was introduced by Pinnavia (13), many polyurethane/clay nanocomposite studies have been carried out (14-19). In contrast to microcomposites, impressive improvements in performance were achieved with a small amount of filler. This was ascribed to the high aspect ratio of the exfoliated montmorillonite layers.

In this work, nanocomposite adhesives for flexible packaging obtained using an OMM in a polyurethane matrix were synthesized and characterized. The effect of different preparation procedures was compared in order to correlate exfoliation, rheological, and physical properties. The microstructure of the composites was investigated by X-ray diffraction. The viscosity of polyols-OM systems was studied as function of shear rate in a cone--plate rheometer in order to either check the possible degradation of the polyol oligomer or to correlate the viscosity with the aggregation state of OMM. In particular, the rheological behavior of the polyol changes from that of Newtonian fluid to a strongly non-Newtonian fluid.

The rheology of multiphase systems, and more specifically of solid-liquid suspensions, has been the object of large number of investigations (20-24), whereas a more limited attention has been paid to the case of nanometric fillers suspended in a non-Newtonian liquid (25). In this article, a simple model accounting for an apparent increase of rheological units size associated with the intercalation of macromolecules into OMM galleries is proposed. In particular, model results are compared with experimental data for different contents of OMM in the polyol.

The viscosity of the polyol nanocomposite was reduced adding ethyl acetate and then the isocyanate curing agent. Curing was performed at room temperature for 1 week. The basal distances of crosslinked PU nanocomposites were obtained by X-ray diffraction.

The glass transition temperature [T.sub.g] of PU nanocomposites, measured using differential scanning calorimeter, increases with increasing volume fraction of OMM.

Finally, the permeability to oxygen and water vapor of polyurethane clay-nanocomposites was measured. The gas permeation through the composites was correlated to the volume fraction of the impermeable inorganic part of the OMM.

EXPERIMENTAL

Materials

The organoclay supplied by Laviosa (Livorno, Italy) are commercially available as Dellite HPS and Dellite 43B. Dellite HPS is a purified unmodified natural montmorillonite, while Dellite 43 B is an OMM derived from natural montmorillonite especially purified and modified with a high content of dimethyl benzylhydrogenated tallow ammonium salt.

Two-component aromatic polyester-based polyurethane adhesive consisting of a prepolymer and 60% solution of an aromatic polyester polyol in ethyl acetate were used. The cross-linker, Polurene FP 75, consisted of an aromatic isocyanate TDI/TMP (toluene diisocyanate--trime-tylpropane). The polyol Polurene FP28A is a copolymer of isophthalic acid and di-ethylene glycol characterized by an average molecular weight of 6600. Polurene FP 28A and FP 75 are the components of a solvent-based adhesive currently used for flexible food packaging and technical laminates.

Nanocomposite Preparation

Nanocomposites were obtained through in situ intercalative polymerization method (26): after mixing of organoclay and polyol Polurene FP 28 A for 1 h at 30[degrees]C in a Haake reomix 600/610, with a mixing velocity = 60 rpm, the polyol was dried by ethyl-acetate solvent evaporation. Ethyl-acetate was, then, added to reduce viscosity as in the industrial practice. The polyols-organoclay systems were dried for 24 h at 80[degrees]C and then mixed with isocyanate Polurene FP 75. A glass slab with a uniform (0.4-0.6 mm) layer of the polyol-organoclay-isocyanate solution was coated and, finally, the nanocomposites were cured for 1 week at room temperature (about 25[degrees]C).

FTIR spectra indicated that these curing conditions lead to complete disappearance of peaks associated with the isocyanate group indicating that PU is fully cured.

Characterization Techniques

Wide Angle X-ray diffraction (WAXD) were collected on a PW 1729 Philips, using Cu K[alpha] radiation in reflection mode ([lambda] = 0.154 nm). The samples were step-scanned at room temperature from 1.3[degrees] to 10[degrees] 2[theta] in order to determine the d-spacing of organoclay, Polyol-OMM systems, and nanocomposites. The samples were held in the diffractometer using a socket glass sample holder.

The mass fraction of the organic modifier in each formulation was obtained from thermogravimetric analysis (TGA Netzsch STA 406). Samples were heated from room temperature up to 1000[degrees]C at 10[degrees]C/min.

The viscosity of polyols-OMM systems ([eta]), as a function of the volumetric amount of filler loaded ([x.sub.vol]), was measured at 50[degrees]C using a strain controlled rheometer (Ares Rheometric Scientific) with a cone and plate geometry operating in steady shear between 0.1 [s.sup.-1] and 10 [s.sup.-1].

Glass transition temperature ([T.sub.g]) of PU nanocomposites was measured using differential scanning calorimeter (DSC METTLER 622 Toledo). Dynamic scans have been run between - 100[degrees]C and 50[degrees]C at 10[degrees]C/min.

The water vapor transmission rate trough PU unfilled films and nanocomposites was measured using a standard test method (dish method), ASTM E96-00. The transmission rate was normalized with respect to the films thickness ranging between 40 and 60 [micro]m. Water vapor permeability P [g/Pa s m] is obtained from Eq. 1:

P = [[G/[1 * A]]/[[S * ([R.sub.1] - [R.sub.2])]/100]].[delta] (1)

where G is the weight change (g), t is the time during which G occurred (h), A is the test area [[m.sup.2]], S is the saturation vapor pressure at test temperature [Pa], [R.sub.1] is the relative humidity at in the test chamber. [R.sub.2] is the relative humidity at the dry side of the film and finally [delta] is the thickness of the film.

Oxygen permeation tests were performed in gas-membrane-gas configuration apparatus based on the measurement of the pressure increase of the downstream side of the nanocomposite film or meat film, while maintaining a constant upstream side driving pressure. The apparatus and experimental procedure were similar to those reported elsewhere (27). In each experiment sufficient time was allowed to ensure attainment of steady state permeation. The permeability was computed from the slope of the linear, steady state part of the curve representing the permeated gas volume as a function of time. The "time lag", [theta], i.e. the intercept of the steady state permeation curve on the time abscissa, was also determined. From its value oxygen diffusivity was estimated based on the assumption of ideal Fickian behavior. In fact, in such a case, diffusivity, D, is related to [theta], by the following expression (28):

D = [[L.sup.2]/60] (2)

where L is the film thickness.

All the tests were performed at a vanishingly small downstream pressure which was assumed to be equal to zero. The use of a very accurate transducer on the downstream side (MKS Baratron 121 pressure transducer with a full scale of 10 Torr, a sensitivity of 0.001 Torr and an accuracy of [+ or -]0.5% of the reading) allowed reliable measurement of permeability even though the downstream side pressure never exceeded 0.25% of the upstream side value. The experiments were performed at upstream pressures of pure oxygen equal to 760 Torr and a temperature equal to 25[degrees]C.

RESULTS AND DISCUSSION

Organofiller Characterization

The X-ray diffraction patterns of unmodified and modified montmorillonite samples are reported in Fig. 1. The higher d-spacing of Dellite 43B is attributed to the exchange of the small metallic cations with the large organic cations. The larger interlayer spacing can promote OMM intercalation during mixing with the polymer (6). However the rather large d-spacing of montmorillonite layers is not the only factor which can affect intercalation during mixing with the polymer. The alkylammonium cations can provide functional groups that can react with the polymer matrix (26). In this case only dispersion forces and intermolecular interaction between the polyol segments and the alkylammonium tail can occur.

[FIGURE 1 OMITTED]

The mass fraction of the organic modifier in each formulation obtained from thermogravimetric analysis is reported in Table 1. The weight loss of Dellite HPS is only due to the loss of absorbed and bonding water. Two distinct weight loss mechanisms are observed for Dellite HPS in Fig. 2. The first is observed between 30 and 100[degrees]C, and is attributed to water desorption. The second, observed between 500 and 750[degrees]C, is attributed to the loss of chemically bonded water. Two main degradation mechanisms are also observed in the thermogravimetric analysis (TGA) of Dellite 43B. The first weight loss starting at 200[degrees]C is attributed to the decomposition of the organic modifier. The second is located at roughly the same temperatures of the second degradation step observed for Dellite HPS. As a consequence, the second degradation step for both clays is attributed to chemically bonded water loss. Therefore, the organic weight fraction of the OMM ([wt.sub.org]) was obtained by subtracting the final weight loss of HPS ([wt.sub.HPS]) from the final weight loss of each sample ([wt.sub.OMMT]), as reported in Eq. 3.

[FIGURE 2 OMITTED]
TABLE 1. Physical properties of  Dellite 43B nnd Dellite HPS.

Sample  Volume fraction of organic content  2 [theta]  d-spacing (nm)
                  [x.sub.org] (%)

 HPS                    0                      6.8          1.3
 43B                   28.9                    4.7          1.9


w[t.sub.org] = w[t.sub.OMMT] - w[t.sub.HPS] (3)

The volumetric fraction of organic modifier was calculated according to Eq. 4.

w[t.sub.org] = [[m.sub.org]/[m.sub.[n - cl]]] = [[[x.sub.org][d.sub.org]]/[d.sub.[n - cl]]] (4)

where [wt.sub.org] is the weight fraction of organic modifier, obtained from TGA results, [m.sub.org] is the weight of organic modifier, [m.sub.n-cl] is the weight of nano-clay, [x.sub.org] is the volumetric fraction of organic modifier, [d.sub.org] is the density of organic modifier (0.83 g/[cm.sup.3]) and [d.sub.n-cl] is the density of nano-clay (1.6 g/[cm.sup.3]).

Polyol-OMM Nanocomposites

XRD. A comparison between the X-ray diffraction patterns of Dellite 43B and the nanofilled polyol obtained by mixing intercalation with Dellite 43B is reported in Fig. 3. Similar results were obtained by Ke et al. (29).

[FIGURE 3 OMITTED]

XRD patterns of samples filled with 2.1% vol, 4.2% vol, and 5.7% vol, clearly show that the peak associated to the d-spacing of Dellite 43B is shifted to angle smaller than 20 = 1.3[degrees] (corresponding to d-spacing >6.7 nm). This suggests that the organoclay, during mixing with the polyol, was at least intercalated with a lamellar spacing higher than 6.7 nm or it was exfoliated.

The sample filled with 7.1% vol of OMM shows a peak centered at about 2[theta] = 2.4[degrees], corresponding to a d-spacing of 3.68 nm. This indicates that the OMM was intercalated during mixing, but intercalation is less efficient than in the case of samples filled with a lower content of OMM. Nevertheless, it cannot be excluded the presence of exfoliated OMM even in the sample with 7.1% of OMM.

Rheological Characterization and Modeling

The viscosity of neat polyol as a function of the shear rate is compared with that of polyol filled with Dellite 43 B in Fig. 4. A clear increase with the volume fraction of OMM is observed. Furthermore, the unfilled polyol and polyol with 2.1% vol of OMM show a quasi-Newtonian behavior, whereas the samples with higher content of OMM behave like a pseudo-plastic fluid. A number of equations of varying complexity are available to predict non-Newtonian rheological behavior. A simple expression which used successfully despite its simple structure is the empirical "power law equation" (30):

[FIGURE 4 OMITTED]

[eta] = K[[gamma].sup.[(n - 1)]] (5)

which is able to phenomenologically describe the fluid behavior in steady shear rheological experiments. K and n are model parameters, generally referred to as consistency and flow index, respectively; [gamma] is the shear rate. Equation 5 is applicable for [gamma] values comprised between upper and lower Newtonian plateau regions. Obviously, a Newtonian fluid is a special case for which n = 1 and K becomes equal to the viscosity. The power law index, n, of polyols filled with Dellite 43 B, obtained by power law fitting of Fig. 4 data, are reported in Table 2. In particular the rheological behavior of the polyol abruptly changes from that of a quasi-Newtonian fluid to a strongly non-Newtonian fluid (pseudo-plastic behavior, i.e. n < 1) when the OMM content is higher than 2.1% v/v. Above this value, i.e. for volume fractions of OMM ([x.sub.vol]) ranging between 4.2% v/v and 7.1% v/v, filled polyol displays a pseudoplastic behavior with a flow index which gradually decreases as the amount of filler increases. This behavior indicates that OMM is responsible of the aggregation of polyol chains which behaves as larger and entangled polymer molecules. It can be assumed that clay lamellar crystals bonding several polyol chains among galleries and at their free surface, when their concentration is higher than 2.1% v/v, act as weak physical cross linking points playing a similar role of that played by entanglements in high molecular weight polymers. During steady shear experiments, normal stress build up was also observed, confirming that nanocomposites with an OMM content above a threshold value present the typical rheological behavior of a pseudoplastic fluid.
TABLE 2. Power law index of polyol filled with Dellite 43B.

Polyol with 43 B (% vol)  Power index (n)

          0                    0.94
          2.1                  0.95
          4.2                  0.52
          5.7                  0.48
          7.1                  0.41


To correlate a rheological parameter with the OMM content a more complex model is fitted to data reported in Fig. 4, which accounts for the presence of Newtonian plateaus in the limits [gamma] [right arrow] 0 and [gamma] [right arrow] [infinity]. In particular Cross proposed a model that in the two limits gives, respectively, the Newtonian viscosity at zero shear rate, i.e., [[eta].sub.0] and the Newtonian viscosity for extremely high values of shear rate. i.e., [[eta].sub.[infinity]]. After simplification introduced by Ellis (28), based on the assumption that [[eta].sub.0] [dmt] [[eta].sub.[infinity]]. Cross equation reads:

[eta] = [[[eta].sub.0]/[1 + [([lambda][gamma]).sup.m]]] (6)

where [[eta].sub.0] represents the viscosity at zero shear rate, [lambda] and m are model parameters, [gamma] is the shear rate. In particular, [lambda] is the reciprocal of the shear rate at which the calculated value of [eta] becomes equal to [[eta].sub.0]. At low shear rates, the viscosity approaches [[eta].sub.0] which in turn is a function of the average molecular weight of the polymer (30), while at high shear rates ([lambda] [gamma] [much greater than] 1) power law behavior is predicted, with m related to the power low index, n, as follows:

m = 1 - n (7)

then, shear rate dependence of polyol/dellite systems, [eta], was modeled using Eq. 6. The values of the index m in Eq. 5 are obtained from Eq. 6, using the n values, reported in Table 2. Once n is determined the parameters of Eq. 6, [[eta].sub.0] and [lambda], were calculated by a nonlinear fitting procedure. The comparison between Eq. 6 prediction and experimental data is shown in Fig. 4. As shown in Table 3 the value of [[eta].sub.0] is a function of the OMM content.
TABLE 3, Values of parameters fitting of rheological data with Ellis
equation.

Polyol with      [lambda] (S)       m           [[eta].sub.o] (Pa s)
43 B (% vol)

     0                0            0                56.2 [+ or -] 0.003
     2.1              0            0.05  5.74 [10.sup.2] [+ or -] 1.40
     4.2      1.77 [+ or -] 0.042  0.48  5.23 [10.sup.3] [+ or -] 58.4
     5.7      1.83 [+ or -] 0.059  0.52  5.36 [10.sup.3] [+ or -] 145
     7.1      2.49 [+ or -] 0.051  0.59  1.23 [10.sup.4] [+ or -] 137


The dependence of the viscosity of polyol/dellite systems from the volume fraction of OMM was also modeled. The rheology of multiphase systems, and more specifically of solid-liquid suspensions, was the object of numerous investigations, both theoretical and experimental, starting from the work of Einstein (31), (32).

The Einstein equation can be applied to very dilute suspensions (solid volume fraction [PHI] < 0.02) of rigid spheres in a Newtonian field (30):

[[eta].sub.r] = 1 + 2.5 * [PHI] (8)

where [[eta].sub.r], is the relative viscosity of the suspension, calculated as the ratio between the viscosity of the filled suspensions, [eta], and the viscosity of the suspending medium, [[eta].sub.s].

The classical equation of Einstein was generalized in order to include (a) the effect of viscoelesticity of the particles (33): (b) the nonzero Reynolds number corrections (34); (c) the deformability of the particles (35); (d) the nonspherical shape of particles (36); (e) the sedimentation process of the filler (37). Starting from Einstein equation, a large number of correlations between the relative viscosity. [[eta].sub.r] and the volume fraction of the solid particles. [empty set] were published in the literature (20). These theoretical models derive from three assumptions: (1) the diameter of rigid particles is large compared with the suspending medium molecules, but small compared with the smallest dimension that can be measured by the rheometer; (2) flow is at steady state without inertial, concentration gradient or wall slip effects; and (3) the medium liquid adheres perfectly to particles (30). There may be also a fourth assumption taking into account the inter-particle interaction, depending on the model chosen.

Among these equations, we found that a good fit of our experimental data by using Batchelor Equation:

[[eta].sub.r] = 1 + [k.sub.1] * [PHI] + [k.sub.2] * [[PHI].sup.2] (9)

Equation 9 is obtained from the classical Einstein model, Eq. 7, operating a correction that take into account the effect of more concentrated suspensions of particles. In particular, referring to the physical significance of the parameters introduced in Eq. 8, [k.sub.1] takes into account the shape of the particles in suspension and equals 2.5 for spherical particles, as in Einstein equation. Guth found that, in the case of nonspherical particles. [k.sub.1] depends on the aspect ratio (36), p, according to:

[k.sub.1] = [p/[2ln(2p) - 3]] + 2 (10)

[k.sub.2] was first introduced by Batchelor in order to consider increase in viscosity taking place in more concentrated suspensions (solid volume fraction [PHI] > 0.02) (37). Other authors report for [k.sub.2] values ranging from 7 to 14 (38). On the other hand, the shape of the OMM particles is not spherical and according to Fornes and Paul (39), as measured by transmission electron microscopy, the aspect ratio of a single layer of OMM is about 120. The results of mechanical characterization obtained in a former article (40) on isocyanate crosslinked polyol led to an aspect ratio of 49, indicating that stacks of two MMT lamellas are present in the average. By substituting this last value in Eq. 10, a value of 8.5 can be calculated for [k.sub.1]. Steady shear viscosity (see Fig. 5) was fitted with Batchelor equation, for shear rate ranging from 0 [s.sup.1] to 10 [s.sup.1] and assuming [k.sub.1] = 8.5 Pa s. The values of the parameter [k.sub.2], obtained by nonlinear fitting are reported in Table 4 as a function of the shear rate. These values, very high in comparison with those reported in the literature, indicate a strong nonlinear behavior of Eq. 9 even at very low of filler loading, probably related with the nanoscale dispersion. The decrease of [k.sub.2] with the shear rate can be considered a direct consequence of the non-Newtonian behavior of these fluids. According to Fig. 4 the viscosity of filled polyols approaches that of unfilled ones increasing the shear rate, leading to a more limited variation of [eta] with [PHI].

[FIGURE 5 OMITTED]
TABLE 4. Values of [k.sub.2] parameter of the non linear fitting of
rheolocical data with Batchelor equation

[gamma]([S.sup.-1])    [k.sub.2] (KPas)

        0             39.9 [+ or -] 3.69
        1             15.7 [+ or -] 3.03
        2             12.0 [+ or -] 2.13
        3             9.614 [+ or -] 0.185
        4             8.57 [+ or -] 0.171
        5             8.03 [+ or -] 0.162
        8             6.93 [+ or -] 0.146
       10             6.41 [+ or -] 0.140


The dependence of the polyol/dellite systems viscosity on shear rate, can be interpreted assuming the presence of large rheological units coordinated by several OMM lamellae. OMM is therefore responsible of the aggregation of polyol chains which behaves as larger and entangled polymer molecules with a higher average molecular weight, with respect to the unfilled polyol. Therefore a fictive equivalent molecular weight can be associated with the presence of OMM lamellae in the polyol. To estimate this apparent or equivalent fictive average molecular weight, [M.sub.wf], a relationship of zero shear rate viscosity ([[eta].sub.0] in Eq. 6) and the volume fraction of the Dellite 43B, [PHI], must be adopted. The relationship between [[eta].sub.0] and the average molecular weight for highly entangled linear polymers, is (30):

[[eta].sub.0] = k[M.sub.w.sup.a] (11)

where [[eta].sub.0] is the zero shear rate viscosity of the polyol filled with Dellite 43B, [M.sub.w] is the apparent or equivalent average molecular weight of the polyol filled with Dellite 43B, k is a model parameter depending on the nature of the polymer and on the temperature and finally a is a model parameter, assuming equal to 3.4 for many polymers.

The rheological units made of large assembly of OMM lamellae and polyols closer resemble a branched molecule. Then this equation should be to adapted branched polymers introducing a correction factor accounting for the difference between the gyration radius of a branched and a linear molecule of the same molecular weight (30). However the correction factor does not alter the mathematics of Eq. 11 and can be included into the parameter k.

A relationship between the zero shear rate viscosity of the polyol/Desllite system [[eta].sub.0] and the volume fraction of the nanofiller [PHI], can be obtained starting from Batchelor equation (Eq. 9) applied at zero shear rate:

[[eta].sub.0] = [[eta].sub.0s] * (1 + [k.sub.1] * [PHI] + [k.sub.2] * [[PHI].sup.2]) (12)

where [[eta].sub.0] is the viscosity of the filled suspensions at zero shear rate and [[eta].sub.0s] is the viscosity of the suspending medium at zero shear rate.

Referring to Batchelor parameters, [k.sub.1] and [k.sub.2], the first was calculated from Eq. 10, as previously detailed ([k.sub.1] = 8.5), the second, [k.sub.2], was assumed equal to 3.99 [10.sup.4] which is the value of the parameter [k.sub.2] at lowest measured shear rate reported in Table 4.

Substituting Eq. 12 into Eq. 11, a relationship between a fictive equivalents average molecular weight of the polyol/Dellite systems, [M.sub.wf], and the volume fraction of the Dellite 43B, [PHI], is obtained:

[M.sub.wf] = [k.sub.3] * [[(1 + [k.sub.1] * [PHI] + [k.sub.2] * [[PHI].sup.2])].sup.[1/a]] (13)

Setting [PHI] = 0, [k.sub.3] = [([[[eta].sub.0s]/k]).sup.(1/a)] is the average molecular weight of unfilled polyol equal to 6600.

Assuming from the literature a = 3.4 (30) and that k and a are not dependent on [PHI], Eq. 13 can be used to calculate the value of the apparent fictive average molecular weight of the polyol/Dellite systems as a function of the volume fraction of Dellite 43B. The results reported in Table 5 and in Fig. 6 indicate that very large rheological units, corresponding to an equivalent increase of the molecular weight up 4.7 fold, are formed.

[FIGURE 6 OMITTED]
TABLE 5. Values of the apparent fictive molecular weight [M.sub.wf] as
function of the Dellite 43B content.

[PHI]  [M.sub.wf] (g/mol)

0       6.60 [10.sup.3]
0.021   1.56 [10.sup.4]
0.042   2.32 [10.sup.4]
0.057   2.77 [10.sup.4]
0.071   3.14 [10.sup.4]


PU-OMM NANOCOMPOSITES

Mass Transport Properties of Crosslinked Polyurethanes

The viscosity of the polyol nanocomposite was reduced adding ethyl acetate and then the isocyanate curing agent. Curing was performed at room temperature for 1 week. The basal distances of crosslinked PU nanocomposites, obtained by X-ray diffraction are reported in Fig. 7.

[FIGURE 7 OMITTED]

The comparison between XRD spectra of Fig. 3 and Fig. 7 obtained on the polyol and on the crosslinked poly-urethane respectively shows a reduction of basal distance, even if it is still higher than that of neat OMM. Diffraction peaks at higher angles indicate that the addition of the solvent and isocyanate decreases the lamellar spacing. The higher affinity of polyol toward ethyl acetate with respect to ammonium salts, determines its extraction from the OMM interlamellar galleries, resulting in a d-spacing decrease.

The glass transition temperature [T.sub.g] of PU nanocompo-sites were measured using differential scanning calorimeter. The values of [T.sub.g] as function of the Dellite 43B content in the polyurethane matrix are reported in Table 6. [T.sub.g] increases from -35[degrees]C for the unfilled polyol to -32[degrees]C for the crosslinked polyurethane. Increasing the OMM content, [T.sub.g] reaches -18, 8[degrees]C for the sample with 5.7% vol of Dellite 43B. On the other hand, the change at [T.sub.g] of heat capacity, [[DELTA]C.sub.p], normalized to the weight of the organic fraction, given by PU and organic modifier, decreases from 1.02 J/gK for the unfilled polyol to 0.74 J/gK for the sample with 5.7% vol of Dellite 43B. Therefore, a rigid amorphous fraction [x.sub.ra] (41), (42) can be calculated as the ratio between the changes of heat capacities of filled PU [[DELTA]C.sub.P]([Florin]) to that of unfilled one, [DELTA][C.sub.p](u):
TABLE 6. Values of [T.sub.g], [[DELTA]C.sub.p], and [X.sub.ra] as
function of the Dellite 43B content.

vol% Dellite    [T.sub.g]    [[DELTA]C.sub.p]        [X.sub.ra]
     43B      ([degrees]C)  (J/gK) PU matrix
                               and organic
                             (normalized to
                             modifier weight)

    0            -32.0      1.02 [+ or -] 0.09           0
    1.8          -24.4      0.88 [+ or -] 0.081  0.14 [+ or -] 0.01
    3.5          -21.1      0.80 [+ or -] 0.09   0.22 [+ or -] 0.02
    4.6          -20.1      0.79 [+ or -] 0.053  0.23 [+ or -] 0.009
    5.7          -18.8      0.74 [+ or -] 0.09   0.28 [+ or -] 0.06


[x.sub.ra] = 1 - [[[DELTA][C.sub.p]([Florin])]/[[DELTA][C.sub.p](u)]] (14)

The values of [x.sub.ra] calculated with Eq. 14 are also reported in Table 6 and in Fig. 8. The rigid amorphous fraction of the PU nanocomposites increases with increasing volume fraction of Dellite 43B. As shown in Fig. 4 the segmental mobility is significantly reduced as OMM increases indicating that chains immobilization occur when they are intercalated between OMM lamellae. This behavior can be compared with those of semi crystalline polymer where a rigid amorphous fraction is typically observed (43-46). Similarly, OMM lamellae can act as lamellar chain folded crystals presenting polymer molecules partially immobilized and even bridging two OMM stacks. This possible interpretation indicates that even in the polyol this is likely to occur: large rheological units made of OMM lamellae and polymer chains partially or totally intercalated are generated. OMM are able to bridge a larger number of OMM lamellae as the filler weight fraction increases, leading to a corresponding high viscosity of the OMM/polyol nano dispersion. These large units behaving as an entangled polymer of very high molecular weight are responsible of the strong non-Newtonian behavior of the filled polyols, as reported in a previous work (21).

[FIGURE 8 OMITTED]

Finally, the permeability to oxygen and water vapor of polyurethane clay-nanocomposites (P) was measured. According with the dish method, water vapor permeability is measured from the weight change of the desiccant as shown in Fig. 9, for different OMM content. All curves show a linear behavior indicating that the steady state conditions are immediately achieved. Increasing the nanofiller content the permeability decreases, as expected according to the experimental and theoretical literature data (47). The raw data of Fig. 9 are then used to calculate the water vapor permeability reported in Table 7 and Fig. 10 as a function of the OMM content. The non linear dependence of P on [phi], characterized by an upper curvature in Fig. 10, indicates that larger lamellar stacks are present when a larger amount of OMM is used. A comparison between OMM/PU nanocomposite and a micro-composite obtained dispersing [Na.sup.+] montmorillonite (MMT) into the polyurethane is also shown in Table 7. Nanocomposites with similar filler content show a lower permeability than the microcomposite obtained with unmodified MMT (Dellite HPS). This observation confirms that a key factor for an improvement of barrier properties of a film is given by the size of the dispersed particles.

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]
TABLE 7. Water vapor permeability reported as a function of the OMM
content.

      vol% Dellite 43B        Permeability (g/Pa s m)  Permeability
                                                       decrease (%)

0                                    3.55E-13               0
2.1                                  2.64E-13             -26
4.2                                  1.46E-I3             -59
5.7                                  1.25E-13             -65
7.1                                  1.06E-13              70
Microcomposite with 3.5% HPS         2.79E-13             -21


A strong increase of barrier properties to oxygen is also observed. Permeability to oxygen of PU nanocomposites were obtained using 5.7% of 43 B. As shown in Fig. 11a and in Table 8 the PU nanocomposites with 5.7% vol of 43B displays a reduction of permeability and diffusivity to oxygen of quite an order of magnitude. In Fig. 11b and c are reported an enlarged plots of the regions used for the evaluation of time lag. As can be noticed, the time lag for pure PU is quite short, close to the lowest detectable value. As a consequence, the diffusivity reported for the case of pure PU has to be taken as a lower-bound value.

[FIGURE 11 OMITTED]
TABLE 8. Permeability and diffusivity to oxygen of the PU nanocomposite
with 5.7% vol of Dellite 43 B.

vol% Dellite 43B  P ([cm.sup.3] (STP) * cm)/  D ([cm.sup.2]/s])
                    [cm.sup.2] * atm * s])

       0                1.54 [E.sup.-8]        6.56 [E.sup.-7]
       5.7              3.29 [E.sup.-9]        5.31 [E.sup.-8]


These experimental results can be rationalized on the basis of the available models for mass transport in lamellar nanocomposirtes. In general, nanocomposites display permeabilities and diffusivities of low molecular weight compounds which are lower, at varying degree, than the starting neat polymer matrices. In fact, decrease of permeability and diffusivity of nanocomposites as compared with neat polymer matrix can range from tenth percent to more than two order of magnitude, depending on processing and mixing procedures, on the nature of matrix and nanofiller. and on the amount of nanofiller. Incidentally, we note that there are notable exceptions to this behavior in the cases of glassy polymers characterized by a high glass transition temperatures and a high excess free volume: in these cases nanofillers, like nanometric silica fume, affect the packing ability of glassy polymers determining an increase of free volume and, in turn, an increase of permeability (48).

The generally observed increase of barrier properties can be attributed to geometrical factors (49-61) as well as to indirect effects on structure and morphology of the polymer matrix (62-64) including change of free volume, development of interphases, qualitative and quantitative change of crystalline domains. From the geometrical point of view, the presence of nanolayers determines both an increase of tortuosity of the diffusive path as well as an effective decrease of the area of cross section exposed to the diffuse flux. More in details, geometrical models proposed in the literature accounts for the different effect that, at a varying degree, have been assumed to play a role in decreasing permeability and diffusivity, i.e. (a) "wiggling" effect related to the mentioned increase of diffusive path and reduction of area exposed to the flux; (b) effect related to the slits between contiguous lamellas, (c) necking effect related to the passage among the lamellas close to the upper and lower surfaces of the film. Improvements of these models account for polydispoersity of nanofiller size, degree of orientation order and aggregation of nanofiller.

It is believed that, in the case at hand, these geometrical effects play the major role in determining the observed significant decrease of permeabilities and diffusivities.

CONCLUSION

Nanocomposite adhesives obtained using a montmorillonite, modified with organic cations, in a polyurethane matrix were synthesized and characterized. Rheological analysis on OMM-filled polyol indicated that as the clay content increases, the clay lamellar crystals acts as weak and labile physical cross linking points inducing either an increase of viscosity of more than two order of magnitudes either a strong non-Newtonian behavior. A significant increase of [T.sub.g] of PU nanocomposites with OMM content was also observed, confirming that the nanoscale dispersion of the OMM is capable to significantly limit the molecular mobility of polymer chain segments. The presence of a large amount of rigid amorphous fraction can contribute to the overall barrier properties. The use of these PU nanocomposites as adhesives in multilayer packaging films with improved gas barrier properties is envisaged.

ACKNOWLEDGMENTS

Mr. D. Cannoletta is kindly acknowledged for the X-ray measurements.

REFERENCES

(1.) S.J. Komarneni, Mater. Chem., 2, 1219 (1992).

(2.) H. Gleiter, Adv. Mater., 4, 474 (1992).

(3.) B.M. Novak. Adv. Mater., 5, 422 (1993).

(4.) R.F. Ziolo, E.P. Giannelis, B.A. Weinstein, M.P. O'Horo, B.N. Granguly, V. Mehrota, M.W. Russel, and D.R. Huffman, Science, 257, 219 (1992).

(5.) F. Bauera, H.J. Glasela, E. Hartmanna, H. Langgatha. and R. Hinterwaldner, Int, J. Adhes. Adhes., 24, 519 (2004).

(6.) K. Yano, A. Usuki, A. Okada, T. Kurauchi, and O. Kamigaito, J. Polym. Sci. Part A: Polym. Chem., 31, 2493 (1993).

(7.) A.S. Moet and A. Akelah, Mater. Lett., 18, 97 (1993).

(8.) L.P. Meier, R.A. Shelden, W.R. Caseri, and U.W. Suter, Macromolecules, 27, 1673 (1994).

(9.) M.W. Noh and D.C. Lee, Polym. Bull., 42. 619 (1999).

(10.) F. Dietsche and R. Mullaupt, Polym. Bull., 43, 395 (1999).

(11.) B.K. Kim, J.W. Seo, and H.M. Jeong, Eur. Polym. J., 39, 85 (2003).

(12.) A.M. Osman, V. Mittal, M. Morbidell, and W.U. Suter. Macromolecules, 36, 9851 (2003).

(13.) Z. Wang and T.J. Pinnavia, Chem. Mater., 10(12), 3769 (1998).

(14.) C. Zilg, R. Thomann, R. Mulhaupt, and J. Finter, Adv Mater., 11(1), 49 (1999).

(15.) K.J. Yao, M. Song, D.J. Hourston, and D.Z. Luo, Polymer, 43(3). 1017 (2002).

(16.) T.K. Chen, Y.I. Tien, and K.H. Wei, J. Polym. Sci. Part A: Polym, Chem., 37(13), 2225 (1999)

(17.) T.K. Chen, Y.I. Tien, and K.H. Wei, Polymer. 41(4), 1345 (2000).

(18.) Y.I. Tien and K.H. Wei, Polymer, 42(7), 3213 (2001).

(19.) Y.I. Tien and K.H. Wei, Macromolecules, 34(26), 9045 (2001).

(20.) L. Nicolais and G. Astarita, Ing. Chim. Itla., 9, 123 (1973).

(21.) L. Nicodemo, L. Nicolais, and F. Landel, Chem. Eng. Sci., 29, 729 (1974).

(22.) L. Nicodemo and L. Nicolais, Polymer, 15, 589 (1974).

(23.) L. Nicodemo and L. Nicolais, Chem. Eng. Sci., 8, 155 (1974).

(24.) L. Nicodemo and L.Nicolais, J. Appl. Polym. Sci., 18, 2809 (1974).

(25.) M.E. Mackay, T.T. Dao, A. Tuteja, D.L. Ho, B. Van Horn, H. Kim, and C.J. Hawker, Nat. Mater., 2, 762 (2003).

(26.) S.S. Ray and M. Okamoto, Prog, Polym. Sci., 28, 1539 (2003).

(27.) G. Mensitieri, M.A. Del Nobile, T. Monetta, L. Nicodemo, and F. Bellucei, J. Membr. Sci., 89, 131 (1994).

(28.) J. Crank, The Mathematics of Diffusion, 2nd ed., Clarendon Press, Oxford (1975).

(29.) Y.C. Ke and P. Stroeve, Polymer-Layered Silicate and Silica Naocomposites, Elsevier, Amsterdam (2005).

(30.) J.M. Dealy, K.F. Wissbrun, and V. Reinhold, Melt Rheology and Its Role in Plastics Processing Theory and Applications, Van Nostrand Reinhold, New York (1990).

(31.) A. Einstein, Ann. Physik, 19, 289 (1906).

(32.) A. Einstein, Ann. Physik, 34, 591 (1911).

(33.) R. Roscoe, J. Fluid Mech., 28, 273 (1967).

(34.) L.D. Landau and E.M. Lifshitz, Fluid Mechanics, Pergamon Press, London (1959).

(35.) W.R. Scowalter, C.E. Chaffey, and M. Brenner, J. Coll. Interf. Sci., 26, 152 (1968).

(36.) H. Giesekus, Rheol, Acta, 2, 50 (1962).

(37.) J.V. Robinson, Trans, Soc. Rheol., 1, 15 (1957).

(38.) L.A. Utracki, Polymer Alloys and Blends, Thermodynamics and Rheology. Hanser Plublishers, New York (1989).

(39.) T.D. Fornes and D.R. Paul, Polymer, 44, 4993 (2003).

(40.) C. Esposito Corcione, P. Prinari, D. Cannoletta, A. Maffezzoli, and G. Mensitieri, Int. J. Adhesion Adhesives, 28, 91 (2008).

(41.) C. Schick and E. Donth, Phys. Scr., 43, 423 (1991).

(42.) B. Hahn and J. Wendorff, Macromolecules, 18, 718 (1985).

(43.) X. Lu and P. Cebe, Polymer, 37, 4857 (1996).

(44.) S. Iannace, A. Maffezzoli, G. Leo, and L. Nicolais, Polymer, 42, 3799 (2001).

(45.) H. Xu and P. Cebe, Macromolecules, 37, 2797 (2004).

(46.) A. Minakov, D. Mordvintsev, R. Tol, and C. Schick, Thermochim Acta, 25, 442 (2006).

(47.) R.K. Bharadwaj, Macromolecules, 34, 9189 (2001).

(48.) T.C. Merkel, B.D. Freeman, R.J. Spomtak, Z. He, I. Pannai, P. Meakin, and A.J. Hill, Science, 296, 519 (2002).

(49.) L.E. Nielsen, J. Macromol. Sci. (Chem.), A1(5), 929 (1967).

(50.) K. Yano, A. Usuky, A. Okada, T. Kuruachi, and O. Kamigaito, J. Polym. Sci., Part A, 31, 2493 (1993).

(51.) D.M. Eitzman, R.R. Melkote, and E.L. Cussler, AIChE J., 42, 1 (1996).

(52.) W.R. Falla, M. Mulski, and E.L. Cussler, J. Membr. Sci., 119, 129 (1996).

(53.) C. Yang, E.E. Nuxoll, and E.L. Cussler, AIChE J., 47(2), 295 (2001).

(54.) G.D. Moggridge, N.K. Lape, C. Yang, and E.L. Cussler, Proa. Organ. Coat., 46, 231 (2003).

(55.) C. Yang, W.H. Smyrl, and E.L. Cussler, J. Membr. Sci., 231, 1 (2004).

(56.) N.K. Lape, E.E. Nuxoll, and E.L. Cussler, J. Membr. Sci., 236, 29 (2004).

(57.) W.A. Wakeham and E.A. Mason, Ind. Eng. Chem. Fundam., 18(4), 301 (1979).

(58.) G.H. Fredrickson and J. Bicerano, J. Chem. Phys., 110(4), 2181 (1999).

(59.) A.A. Gusev and H.R. Lusti, Adv. Mater., 13, 21, 1641 (2001).

(60.) Y.P. Ly and Y. Cheng, J. Membr. Sci., 133, 207 (1997).

(61.) R.K. Bharadwaj, Macromolecules, 34, 9189 (2001).

(62.) P. Winberg, M. Eldrup, N.J. Pedersen, M.A. van Es, and F.H.J. Maurer, Polymer, 46, 8239 (2005).

(63.) Z.F. Wang, B. Wang, N. Qi, H.F. Zhang, and L.Q. Zhang, Polymer, 46, 719 (2005).

(64.) Y. Liang, J. Quian, J.W. Cho, V. Psihogios, and T. Lan, Applications of Plastics Nanocomposites, Technical Papers (2002).

C. Esposito Corcione, (1) G. Mensitieri, (2) A. Maffezzoli (1)

(1) Dipartimento di Ingegneria dell'Innovazione, Universita del Salento, Via Monteroni 73100, Lecce, Italia

(2) Dipartimento di Ingegneria dei Materiali e della Produzione, Universita di Napoli Federico II, Piazzale Tecchio 80, 80125 Napoli, Italy

Correspondence to: Dr. Carola Esposito Corcione; e-mail: carola.corcione@unile.it

Contract grant sponsor: Italian government (project Prin 04). DOI 10.1002/pen.21410
COPYRIGHT 2009 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Corcione, C.Esposito; Mensitieri, G.; Maffezzoli, A.
Publication:Polymer Engineering and Science
Article Type:Report
Geographic Code:1USA
Date:Sep 1, 2009
Words:7035
Previous Article:Hydro-entangled bi-component microfiber nonwovens.
Next Article:Physical aging of polycarbonate block copolymers: ductility rejuvenation below the glass transition temperature.
Topics:


Related Articles
Polymerization compounding: epoxy-montmorillonite nanocomposites.
Effects of clay on the morphology of poly(acrylonitrile-butadiene-styrene) and polypropylene nanocomposites.
Effect of intercalating agents on clay dispersion and thermal properties in polyethylene/montmorillonite nanocomposites.
Properties of new nanocomposite triblock copolymer gels based on expandable graphite.
Microstructure and water vapor transport properties of functionalized carbon nanotube-reinforced dense-segmented polyurethane composite membranes.
Nylon 66/clay nanocomposite structure development in a twin screw extruder.
Effects of direct melt compounding and masterbatch dilution on the structure and properties of nanoclay-filled polyolefins.
In situ polymerization of polyamide 66 nanocomposites utilizing interfacial polycondensation. II. Sodium montmorillonite nanocomposites.
Butyl nanocomposites with reduced air permeability: theory and practice.
Mechanical properties and water vapor permeability of starch/montmorillonite nanocomposites.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters