Small Angle Neutron Scattering Shows Nanoscale PMMA Distribution in Transparent Wood Biocomposites

Transparent wood biocomposites based on PMMA combine high optical transmittance with excellent mechanical properties. One hypothesis is that despite poor miscibility the polymer is distributed at the nanoscale inside the cell wall. Small-angle neutron scattering (SANS) experiments are performed to test this hypothesis, using biocomposites based on deuterated PMMA and “contrast-matched” PMMA. The wood cell wall nanostructure soaked in heavy water is quantified in terms of the correlation distance d between the center of elementary cellulose fibrils. For wood/deuterated PMMA, this distance d is very similar as for wood/heavy water (correlation peaks at q ≈ 0.1 Å–1). The peak disappears when contrast-matched PMMA is used, indeed proving nanoscale polymer distribution in the cell wall. The specific processing method used for transparent wood explains the nanocomposite nature of the wood cell wall and can serve as a nanotechnology for cell wall impregnation of polymers in large wood biocomposite structures.

. Electron density scattering length (EDSL) of wood components, MMA, PMMA, and 58 water, calculated through online Neutron/X-ray scattering Length Density Calculator 59 (https://sld-calculator.appspot.com), assuming a chemical formula of C 6 H 10 O 5 for cellulose,  Figure S1: Fitting of anisotropic SANS profiles (I(q)×q) at relative high q range (0.01-0.5 Å -1 ) 66 using two decay power law (-3 & -1) and one gaussian function for various samples. The fitted 67 parameters are summarized in Table S2 69 70 Figure S2: Fitting of the isotropic SANS profiles using two decay power law (-4 & -2) and one 71 spherical Bessel function for various samples. The fitted radiuses of gyration are 2, 1.5 and 2.8 72 from left to right.

77
This isotropic scattering profiles were fitted using two decay power law (-4 & -2) and one 78 spherical Bessel function and a constant as described: 79  Figure   112 113 is shown in Figure S5a, which shows an oscillation with a peak at around 5/r. Y represents the 114 normalized form factor.

115
The structure factor depends on the arrangement of cylinder in the cell wall but since the 116 cylinders are rigid, they cannot inter-penetrate. Thus the radial distribution function is 0 at 117 distance up to two times the radius, and has a sharp peak at about this distance 2r, followed by a 118 second shell at slightly beyond 4r and then converging to average number density, or unity at 119 large distance (Fig. S5b). The Fourier transform of g(r) is the structure factor (Fig. S5c), now 120 having a peak at around π and 2π. The peak position of the structure factor is thus at the

129
The intensity we observe are the product of curve a and c in Fig. S5, and thus the first peak 130 is actually observed at smaller q than the peak position of structure factor due to the descending 131 slope of |F(q)| 2 .

132
The effect of the slope of the form factor to the peak position can be estimated to the first 133 approximation as parabolic function and straight segment. We place the peak at x=0 as

162
The sharpness of the structure factor peak depends on the extent of oscillation of the pair 163 distribution function, which is expected to be attenuated faster when the packing density is lower.
164 Thus, there is no analytical solution directly relating the observed peak position and the structure 165 factor of the fibril arrangement and further analysis requires explicit model building.

166
167 Origin of the differences in peak position between X-ray and neutron scattering

168
In the above discussion we considered the scattering length density boundary to coincide 169 with the hard cylinder. At molecular level, there are however nuances depending on the probe.
170 For X-ray, the electron density centers of carbon and oxygen atoms are a few angstrom away 171 from the boundary. For cell wall immersed in heavy water, the hydroxyl groups of the outermost 172 chain will be exchanged to deuterium leading to higher scattering length density, whereas in non-173 aqueous system, this outermost chain is hydrogenated and shares the scattering length density of 174 the core. Thus the difference in peak position between X-ray and neutron for the three cases can 175 be explained as follows: