Cellulose (CE) and starch biopolymers from various sources (corn or maize)

containing variable amylose and amylopectin content were chosen for the study (cf. Table 1). The biopolymers and ethanol (reagent grade, Commercial Alcohols Inc. Brampton, ON, Canada) were purchased from Sigma Aldrich (Oakville, ON, Canada) and used as received. Deionized and distilled water was used for the preparation of all aqueous solutions.

The adsorption isotherms were obtained using the Intelligent Gravimetric Analyzer system IGA-002 (Hiden Isochema, UK). 40 mg of the adsorbent sample was placed in a stainless steel chamber attached to the microbalance. The sample container was housed ina thermostat reactor that allowed for ultra-high vacuum conditions. The desired temperature inside the reactor was controlled by a water bath. Samples were dried at 70˚C in vacuo (≈10⁻⁸ mbar) for 6 h. Then, the sample temperature was held at 25˚C prior to sample analysis. The IGA system controls the input and output valves to achieve the desired partial pressure in the chamber. After reaching equilibrium of the sample mass at one partial pressure (mbar), the IGA system advanced to the next partial pressure for the isotherm at 25˚C for a range of pressures (5 mbar increments; 0 to 30 mbar).

The particle size distribution for starch granules (Table 2) and zeta potential (x) was measured using a Malvern Zetasizer Nano ZS (Malvern Instruments Ltd, Worcestershire, UK). The size distribution of the samples was obtained by measurement of light scattered (θ = 173˚) by particles (dynamic light scattering, DLS) illuminated with a laser beam. The CONTIN algorithm was used to analyze the decay rates that are a function of the translational diffusion coefficients of the particles, D. Size distribution values were derived from triplicate measurements, each consisted of a minimum of ten individual runs. x-values were estimated based on the laser Doppler electrophoresis and phase analysis light scattering [20]. The initial pH of the starch suspensions (40 mg sample in 7 mL)

of adsorbent (starches and cellulose) in water was adjusted over a pH range (3 to 12) using either HCl (0.01 M) or NaOH (0.01 M). After reaching an equilibrium pH, aliquots of the suspension were taken for estimation of the x-values, where the reported results are the average of triplicate measurements from a minimum of ten individual runs.

The biopolymer adsorption properties in aqueous media were studied by evaluating their uptake properties using an organic dye adsorbate (PNP) at pH 8.5, and is above its pK_a value (7.1) [21]. An aqueous solution of PNP was prepared using NaHCO₃ buffer solution (1.0 M) at pH 8.5. 10 mg of each powdered sample were added to 4 dram glass vials followed by the addition of dye solution (7 ml) at different concentration (0.1 - 50 mM). The vials were sealed with parafilm and allowed to equilibrate with shaking at 180 rpm on a horizontal mechanical shaker (Poly Science, Dual Action Shaker) for 24 h. The samples were left at 295 K for 48 h and centrifuged (Precision Micro-Semi Micro Centricone, Precision Scientific Co.) at 1600 rpm for 1 h. The upper region of the solution was gently removed to a clean vial using a pipette. The UV-vis absorbance of PNP in the supernatant was measured from 200 to 650 nm. The absorption values were obtained using a Varian CARY 100 double beam spectrophotometer at 295 ± 0.5 K. Each measurement was obtained in duplicate by measuring absorbance at 405 nm, along with a calibration curve using standard solutions of PNP at variable concentration. The molar absorptivity of PNP (ε = 1174 M⁻¹ cm⁻¹) was determined by the Beer-Lambert relation. The equilibrium concentration (C_e, mM) of PNP was determined and the polymer uptake (Q_e, mg/g) was obtained by Equation (1).

Q e = ( C 0 − C e ) m × W S o l u t i o n (1)

V is the volume of the solution (ml); C₀ is the initial stock concentration (mM); and m is the polymer weight (mg). The isotherms were obtained by plotting Q_e versus C_e.

EMI shielding measurements were performed using an E5071C Network Analyser (ENA series 300 kHz - 20 GHz) in the X-band frequency range of 8.2 to 12.4 GHz. Each sample was sandwiched between two waveguides of the network analyser. The network analyser transmits a signal down the waveguide incident to the sample. The scattering parameters (S-parameters) of each sample were recorded and used to measure the EMI Shielding Effectiveness (EMI SE). The EMI shielding measurements were performed using equilibrium solvent-swelled samples after imbibing in water, ethanol and a 50% w/w water-ethanol mixture for 6 h. Figure 3 shows a schematic diagram of the network analyser for measurement of EMI SE. Further information and experimental details on the

EMI shielding setup is provided elsewhere [22] [23].

ITC measurements were obtained using a Calorimetry Sciences Corp. (CSC) isothermally at 25˚C. The enthalpy of adsorption was derived for the binding interactions between the biopolymers and water vapour. The heat of adsorption was measured by passing the water vapour through a heat exchanger coil connected to a water bath at 70˚C into a stainless steel sample cell (1.430 mL) containing the sample (ca. 10 mg). The experimental data obtained from the ITC titration were analyzed using CSC commercial Bind Works 3.1 software. Also, deionized water was used as a standard sample. A single binding model with an independent set of multiple binding sites was used for the nonlinear regression. The surface area under the peak shows the enthalpy of adsorption (ΔH_ads) that was estimated using the Origin software.

Figure 4 shows the water vapour adsorption isotherms for the samples at 25˚C. The results reveal that the adsorption and desorption of water vapour for the samples follow a Type II isotherm profile [24]. In addition, the water vapour sorption process in cellulose and starch bear similarities that relate to the structural similarity of these biopolymers and the key role of -OH adsorption sites. In addition, the completion of the monolayer saturation profile takes place at a low relative partial pressure (p/p₀) near 0.2, while much of the vapour adsorption occurs at p/p₀ > 0.8. Parallel trends were observed for N₂ adsorption isotherms for the same materials. However, lower levels of N₂uptake were observed [25] relative to the water vapour uptake herein. The water uptake (w/w %) for the biopolymers is given in parentheses; AM (28.8), AP (29.4), CE (16.4) and AM50 (37.7), where the starch materials exceed the uptake capacity of CE. This may relate to the greater surface accessibility of -OH surface groups of starch over CE and the amorphous nature of the biopolymer. It is noted that the water uptake of AM50 exceeds that of AM and AP and is in agreement with the role of adsorption sites in the amorphous domains for this mixed starch system. The BET surface area (SA) of the biopolymers was calculated from water vapour isotherms using Equation (2) and the results are summarized in Table 3.

Q e = Q m K B E T C e ( C s − C e ) [ 1 + ( K B E T − 1 ) ( C e / C s ) ] (2)

The terms K_BET and C_s are the equilibrium adsorption constant and saturated concentration of the adsorbate, respectively.

The estimated SA of the starch biopolymers exceed that of cellulose, in agreement with the uptake capacity, however; the BET SA estimates obtained from N₂ isotherms are systematically lower (10⁰ to 10¹ m²/g) [25]. The offset in uptake values by these two different gas systems may relate to the role of dipolar versus van der Waals interactions for water vapour and nitrogen gas. The key role of dipolar interactions contributes to the structural changes known for biopolymer-water systems based on the swelling and dye adsorption results reported elsewhere [25] [26]. In addition, since diameter of water (0.20 nm) is less than the diameter of N₂ (0.43 nm). It can be inferred that water molecules can access micropore sites with sizes smaller than 0.43 nm, while N₂ has greater restrictions due to steric effects and weaker surface interactions. Thus, the sorption analysis provide complementary characterization of the material textural properties and surface adsorption sites, according to the nature of the adsorptive probe.

To provide a comparison of the surface area (SA) of different starch and cellulose biopolymers, selected examples of materials from the literature are listed in Table 4.

The pH value at the point of zero charge (pH_zpc) provides a means for understanding the electrostatic interactions between surfaces of the adsorbent and charged adsorbate species. Figure 5 shows pH_zpc values versus pH using both acidic and basic conditions for AM, AM50, AP and CE. All samples were observed to possess a positive charge (ZPC > 0) in an acidic environment with different pH_zpc values. As an example, the surface of cellulose has a negative x-value due to the abundant electron density of the polar groups and the role surface ionization effects of hydroxyl groups, especially when pH > pH_zpc. In turn, the cellulose biopolymer surface has excess positive charge at pH < pH_zpc, due to the adsorption of H⁺ ions at the Lewis base sites with high electron density.

The absorbance of PNP was used to estimate the uptake properties of each system, where greater decolorization of the supernatant relates to higher uptake of PNP at pH 8.5. The adsorption isotherms (cf. Figure 6) for the three types of starch provided estimates of the uptake capacity (Q_m) that adopted the following order: AM (9.82 mg/g) > AP (6.52 mg/g) > AM50 (3.82 mg/g). The relative ordering of the uptake for PNP differs from the water vapour results and can be related to the anion form of PNP and the role of biopolymer zeta-potential in aqueous media. The corresponding surface area (SA) of the adsorbent materials was estimated by use of the Q_m value and Equation (3).

S A = Q m × N × δ × Y (3)

Q_m is the monolayer adsorption capacity at equilibrium (mol/g), N is Avogadro’s number (6.02214 × 10²³ mol⁻¹), δ is the cross-sectional molecular area of the adsorbate (m²), and Y is the coverage factor (Y = 1 for PNP). The molecular area (δ) is 52.5 Ų for PNP when adsorbed in a co-planar orientation and 25.0 Ų when adsorbed in an orthogonal arrangement relative to a planar surface. The corresponding SA values (m²/g) for AM (17.2), AP (11.4), and AM50 (5.62) compare favourably with the SA of cellulose (9.83 m²/g) by Udoetok et al. [31] based on dye adsorption isotherm results with PNP. This trend in SA for high amylose starch and other biopolymers are related to the relative accessibility surface chemical groups of the materials. While the accessibility of the surface O-H groups of AM is higher, it is noted that the SA estimates are lower in comparison with values obtained by water vapour adsorption (Table 3). It is important to note that the adsorbate-adsorbent interactions relate to the cohesive interactions in the medium, as evidenced by the different trends in uptake of gas phase versus liquid adsorbates. The reduced SA estimates by dye adsorption is consistent with the reduced uptake (lower Q_m values) at these conditions due to

the role of hydration and electrostatic effects for the anion form of PNP and the negative x-value of the biopolymers.

The EMI shielding test was employed as a complementary technique to verify the level of sorbed polar solvent (W and/or E) onto the biopolymer materials. The trend in the EMI results (Figure 7) are in accordance with the swelling results reported elsewhere [25].

The EMI shielding technique was used previously [32] to characterize the amount of sorbed water in magnetic mesoporous photonic cellulose films. Water and ethanol have the ability to interact with the electromagnetic (EM) waves due to their dipolar nature. EMI SE, which is expressed in decibel (dB) units is defined by the logarithm of the ratio of the incident EM wave to the transmitted EM wave or log (incident EM wave/transmitted EM wave). EMI shielding in non-magnetic materials originates from the ohmic and polarization losses. The former arises from the dissipation of energy by free charge carriers/ions via conduction, hopping, and tunneling mechanisms, while the latter

arises from the energy derived from the momentum required to reorient electric dipoles in each half cycle of an alternating field [33] [34]. If one assumes that the key difference among the polymers relates to the ohmic losses, it can be inferred that closely bound water constitutes a minor fraction of the overall fraction of bound water. As such, these two mechanisms can be employed to obtain crude estimates of the amount of sorbed water and/or ethanol by the sorbents. It follows that higher water and ethanol sorption denotes more ohmic and polarization losses, and thus higher EMI SE values as reported herein.

Isothermal titration calorimetry (ITC) is useful for the direct determination of thermodynamic parameters (enthalpy, entropy, and stoichiometry) in macromolecular systems, especially for processes that undergo measureable changes between initial and final states that involve aggregation, absorption, and adsorption processes. In the case of adsorption processes that involve substantive changes in adsorbate-adsorbent interactions, ITC can aide in establishing structure-property relationships, especially for biopolymer hydration phenomena by analysis of the thermodynamic parameters. In this study, the thermal response of water vapour sorption for various biopolymers was estimated from the results in Figure 8. The use of ITC was surmised to provide further insight on the differences in sorption capacity of AM, AP, AM50, and CE.

The calculated values for the duration of the thermal response was estimated from the full-width-half-maximum (fwhm) of the exotherm events shown in the ITC profiles in Figure 8(A). The results are considered semi-quantitative due to a number of uncontrolled variables such as the sensitivity of the thermal response for such solid-vapour interactions. The trend in fwhm adopts the following order: AP > AM > AM50 > CE. The results generally follow the trends in solvent swelling and accessibility of the biopolymer -OH groups. However, the fwhm response also relate to the kinetics of heat and mass transfer during the water

vapour adsorption process, and should therefore not be over interpreted.

The area under the curve for each system can be related to the thermal behaviour of biopolymer-vapour adsorption process; however, the dosage level of vapour is unknown due to the continuous flow arrangement used herein. Notwithstanding, the conformational flexibility of the biopolymer backbone and presence of surface accessible biopolymer hydroxyl groups play a role. CE and AM50 show almost similar polar solvent interactions even though CE has fewer accessible surface -OH groups. The limited -OH group accessibility of CE is a consequence of its fibril structure and extensive inter/intra molecular hydrogen bonding [16]. Furthermore, the interaction of AM and AP starches with water vapour are influenced by tertiary structure (branching, conformation, and folding), since such factors govern the surface accessibility of the active adsorption sites.

The heat of adsorption was estimated from the area under the curve of the thermal profile under the curve. Some uncertainty regarding the biopolymer-water stoichiometry exists due to the nature of the experimental configuration for monitoring the thermal response of vapour adsorption among the samples. The comparable ITC response for AP and AM suggest that these biopolymers have similar hydrophilic nature due to their surface accessible -OH groups [35] [36]. The lower heat of adsorption for AM50 sample relates to the presence of linear and branched starch components which may reduce the accessibility of the adsorption sites and the water vapour adsorption properties. The exotherm response in Figure 8(B) indicates the release of heat (absolute values shown) upon water vapour adsorption for these materials. The heat of adsorption does not reach the value anticipated for the enthalpy of condensation at the saturation pressure of water vapour (−44 kJ/mol) [37]. Thus, the calorimetric response should be treated semi-quantitatively. The trend in ITC results provides complementary insight of the complex role of hydration for biopolymers with variable structure. However, a detailed analysis is precluded by the use of a continuum model for the analysis of the calorimetric response and the limited quantitative understanding of the biopolymer-water stoichiometry for the adsorption process. A detailed calorimetric investigation is underway to further characterize the thermodynamics of adsorption for such types of solid-vapour systems.