Cellulose-TiO₂ nanocomposite with enhanced UV–Vis light absorption

Microcrystalline cellulose (CAS:9004-34-6), potassium hydroxide, acetic acid as well as all the solvents were purchased from commercial suppliers (Sigma-Aldrich, Fluka) and used without further purification. (Cyclohexyl)hexyldimethylammonium bromide was synthesized in Pernak’s group at Poznan University of Technology.

(Cyclohexyl)hexyldimethylammonium bromide (0.5 mol) was dissolved in methanol and a stoichiometric amount of saturated methanolic solution of KOH was added. The solution was stirred at room temperature for 5 min, after which the partially precipitated KBr was filtrated. Then a stoichiometric amount of acetic acid was added. The solution was stirred again at room temperature for 15 min and after evaporation of the solvent, the product was dissolved in anhydrous acetone. The inorganic salt was separated and acetone was evaporated to give the product, which was finally dried under reduced pressure at 80 °C for 24 h.

A flask containing solution of 45.0 g of dried (cyclohexyl)hexyldimethylammonium acetate and 45.0 g of dimethyl sulfoxide (DMSO) was immersed in an oil bath at temperature 100 °C with a deviation of temperature from the set temperature of ±1 °C. Microcrystalline cellulose was added in the amount of 7.9412 g and stirred until the solution became optically clear under a microscope (Bresser Biolux LCD Microscope). Afterwards the solution was cooled down, and 150 mL of water was added. Precipitated cellulose was separated and dried under reduced pressure. Finally it was powdered in a mortar grinder Retsch RM 200.

A commercial TiO₂ supplied by the Chemical Factory “POLICE” S.A. (Poland) pre-dried at 100 °C was used as a starting material in preparation process of nitrogen modified TiO₂. Nitridation of the lattice and the surface of titanium dioxide was carried out using HEL Ltd. “Autolab” E746 installation. 350 mL of ammonia water (2.5 %) and 600 g of starting nanocristalline material were placed in a chamber of pressure reactor (BHL-800, Berghof, Germany). Both, titania and NH₄OH substrates were blended using magnetic stirrer to obtain homogeneous slurry. The suspension was heated up to 100 °C for 4 h afterwards. The next step on TiO₂/N preparation process was final drying in air atmosphere at 100 °C for 4 h. The obtained material was ground with a mortar to form a fine powder (Bubacz et al. 2010). The structural and textural parameters of N-doped TiO₂ photocatalyst were listed in Table 1. The BET SSA of 235 m²/g is formed by particles average size of 168 nm and average anatase crystallites size of 6.1 nm by TEM and 11 nm by XRD.

A flask containing solution of 45.0 g of dried (cyclohexyl)hexyldimethylammonium acetate and 45.0 g of DMSO was immersed in an oil bath at temperature 100 °C with a deviation from the set temperature of ±1 °C. Microcrystalline cellulose was added in the amount of 7.9412 g, and stirred until the solution became optically clear under a microscope. Afterwards 0.8823 g of titanium dioxide was added to the solution, stirred for 5 min and cooled down. 150 mL of water was added to the solution, and cellulose-TiO₂/N composite was precipitated. The composite was separated and dried under reduced pressure. Finally, it was powdered in a mortar grinder (Retsch RM 200). Obtained composite contained 10 wt% of N-doped titanium dioxide.

The photocatalysts and cellulose were characterized by UV–Vis/DR using V-630 spectrophotometer (Jasco, Japan) equipped with an integrating sphere accessory for diffuse reflectance spectra (BaSO₄ was used as a reference material). FTIR/DRS spectra of tested photocatalyst was recorded using FT/IR 4200 spectrometer (Jasco, Japan) equipped with DR accessory of Harrick Company (USA). N₂ adsorption measurements at 77 K for calculation of N₂-BET SSA were performed using Quadrasorb SI (Quantachrome, the USA) instrument. The total organic carbon (TOC) measurements were conducted using Multi N/C 2000 analyzer (Analytik Jena, Germany). Thermogravimetric analysis on STA-449C (Netzsch Company, Germany) was managed to check the character of cellulose precipitated from (cyclohexyl)hexyldimethylammonium acetate IL-DMSO solution and composite of cellulose and 10 wt% of TiO₂/N catalyst. The sample was heated from 25 to 950 °C (10°/min) under nitrogen atmosphere (gas flow 30 mL/min). X-ray diffraction patterns were obtained by using X’Pert PRO diffractometer with Cu Kα radiation (λ = 1.54056Å). Dynamic light scattering DLS was used to measure the particles size distribution of prepared TiO₂/N photocatalyst (Zetasizer Nano series ZS, Malvern Instruments, Great Britain). Emission spectrum of UV–Vis irradiation source (6 lamps × 20 W each, Philips), presented in Fig. 1, was measured using USB4000 equipment (Ocean Optics, Netherlands). The radiation intensity was measured using radiation intensity meter LB 901 (Lab-El, Poland) equipped with CM3 and PD204AB Cos sensors.

0.1 g of cellulose-TiO₂ nanocomposite with 10 wt% of TiO₂/N catalyst was stirred (500 rpm) in the dark and under UV light radiation in 500 mL of demineralized water to check the purity of cellulosic material (measurements of total organic carbon TOC). It was noted that demi water contained 0.26 mg/L of organic carbon and 0.09 mg/L of inorganic carbon. Aqueous solution of model organic compound—mono azo dye Reactive Red 198 (RR 198) with the concentration of 20 mg/L was also stirred in the dark to check the adsorption abilities of tested material. Secondly, UV–Vis lamp radiated high UV intensity was switched on to check the photocatalytic activity of above-mentioned materials.

Synthesis route, full characteristic of (cyclohexyl)hexyldimethylammonium acetate and dissolution of cellulose was described earlier (Pernak et al. 2012d).

In Fig. 2 the UV–vis/DR spectra of microcrystalline cellulose, precipitated cellulose (reference material), TiO₂/N photocatalyst and TiO₂/N-cellulose composite are presented. Crystalline cellulose absorbs radiation in both, UV range with the efficiency of ca. 50 % and visible range with efficiency of ca. 15 %. In the case of reference material the increase of light absorption (ca. 35 % for UV range and ca. 30 % for visible range) was caused by yellowish color of prepared sample. Addition of even 10 wt% of TiO₂/N leads to change of absorption range. As it can be seen cellulose mixed with 10 wt% of TiO₂/N catalyst absorbs light stronger than crystalline cellulose, increases absorption under UV to complete 100 % as well as under visible range (400–800 nm) up to 50 %. This could be assigned to the presence of nitrogen in the lattice of TiO₂ (Wawrzyniak and Morawski 2006; Janus et al. 2009) and also because of the yellow colour of new composite. The colour of obtained nanocomposite was noted to be more intensive than precipitated cellulose. The UV–vis/DR spectra was plotted in Kubelka–Munk function, that allows to calculate the reflectance from the materials which both scatter and absorb light. Equation 1 represents the modified form of Kubelka–Munk functionwhere k—excitation coefficient, h—Planck’s constant [Js], υ—light frequency [s⁻¹], A—absorption constant, E _g —band gap energy [eV].

Thus the E _g can be calculated by extrapolating to zero a linear fit to a plot of (khν)^1/2 against hν. The E _g value estimated for initial TiO₂/N and cellulose mixed with 10 wt% of TiO₂/N material was 3.31 and 3.09 eV, respectively. This means that the presence of nitrogen in a titania structural lattice results with decreasing of band gap energy and extends the light absorption to the visible region what was also confirmed e.g. by Wawrzyniak and Morawski (2006). Another factor that might contribute to the decrease in band-gap value is the smaller crystalline size of tested photocatalyst due to a quantum size effect.

In Fig. 3 FTIR/DRS spectra of used TiO₂/N photocatalytic nanomaterial, crystalline cellulose, precipitated cellulose and composite of cellulose with 10 wt% of TiO₂/N catalyst are presented. Strong hydrogen-bonded OH stretching at around 3,800–2,995 cm⁻¹ is commonly observed in all cellulosic materials’ spectra. For material prepared by mixing of cellulose and TiO₂/N, the maximum absorbance of hydrogen-bonded OH stretching is shifted to higher wavenumber (from 3,455 to 3,470 cm⁻¹). Nevertheless, this change was insignificant and it was caused by the presence of ionic liquid in tested samples. The intensity of abovementioned OH stretching and CH stretching bonds at 2,901 cm⁻¹ decreased after mixing with ionic liquid and TiO₂/N and the CH bond was shifted to 2,924 cm⁻¹. This TiO₂/N modification leads to intensity decrease of OH bending of adsorbed water at 1,640 cm⁻¹, the HCH and OCH on-plane bending vibrations at 1,371–1,378 cm⁻¹, the COC, CCO and CCH deformation modes and stretching vibrations at 897 cm⁻¹ and also the C–OH out-of-plane bending mode at 673–675 cm⁻¹. All of these changes could prove that interaction between cellulose and TiO₂/N photocatalyst lead to immobilization of modified titania nanoparticles in the cellulose matrix. On the TiO₂/N spectrum it was possible to observe bonds associated with OH vibrations (3,300–3,500 cm⁻¹) assigned to the dissociated and molecularly adsorbed water and OH groups at 1,630–1,640 cm⁻¹ assigned for molecular water. The narrow bond at 1,428 cm⁻¹ in the synthesized catalysts was attributed to the ammonium groups (NH₄ ⁺) (Randorn et al. 2004), which may suggest that nitrogen was not only successfully incorporated into the structural lattice but also bounded with the surface of TiO₂. A weak bond at 1,536 cm⁻¹ in TiO₂/N sample (named as 4) could be assigned to either NH₂ or NO₂ and NO groups as it was described by Janus et al. (2009). As it can be seen after mixing of cellulose with TiO₂/N material the bond assigned to Ti⁴⁺ at 955 cm⁻¹ disappeared. This means that during preparation of composite there was a reaction between cellulose and TiO₂/N nanocatalyst. In general, on the surface of TiO₂/N-cellulose composite the bonds of cellulose dominates whereas bonds of TiO₂ is practically disappeared, in spite of presence 10 wt%.

Figure 4 shows the X-ray diffraction patterns of tested materials. In starting cellulose only crystalline phase of this cellulose was found (JCPDS card no 00-056-1718) with BET SSA of 4.06 m²/g. Peaks at 2θ = 14.5°, 16.6°, 22.7° and 34.4° are corresponding to (\( 1\overline{1} 0 \)), (110), (200) and (004) planes of crystalline cellulose which peaks have been separated by Wada et al. (1993). It is generally accepted that peak broadening is due to the amorphous cellulose. In the case of nitrogen modified TiO₂ photocatalyst several characteristic peaks attributed to anatase at 25.4°, 38.0°, 48.3°, 53.8° and 62.8° are observed. A weak diffraction line of rutile at 27.5° could be also possible to observe what was associated with the addition of rutile nuclei during the production process (sulfate technology). It was noted that additive of TiO₂/N material to crystalline cellulose leads to disappearance of peaks corresponding to cellulose. The XRD pattern indicates that anatase TiO₂ phase dominates.

In spite of high BET SSA of TiO₂/N (235 m²/g), the total BET SSA of cellulose-TiO₂/N nanocomposite was small and practically the same as microcrystalline cellulose, e.g. 4.64 m²/g. The BET SSA measured for reference material-precipitated cellulose was 3.7 m²/g.

In Fig. 5a the results of studies on water resistivity carried out without light radiation (in dark) for both the precipitated cellulose and cellulose-TiO₂/N composite dissolved in water are presented. It was found that for precipitated cellulose the TOC values of filtered water are very high (up to 38 mg/L of TOC) in comparison to cellulose-TiO₂/N composite. This means that on the surface of prepared cellulosic material some impurities are adsorbed (residues of components used in preparation process). The TOC concentration calculated for tested nanocomposite with titania rises insignificantly with stirring time from 3.09 to 3.82 mg/L (after 24 h of stirring). It was proved that immobilization of nitrogen modified titania in cellulose matrix stabilizes the surface of new complex and suppresses the releasing of organic residues to aqueous solution.

The experiments described above for both samples were repeated in aqueous solution under UV light irradiation. Results of these tests are presented in Fig. 5b. After 24 h of radiation the concentration of organic carbon in water was higher than after stirring in the dark (44 vs. 37.7 mg/L of TOC). This was related to the loss of solution volume because dissolved TOC did not evaporated with a steam as well as photodecomposition the precipitated cellulose because for composite cellulose with titania the reached TOC level was almost the same as without UV radiation.

Photostability measurements were also carried out in azo dye solution under UV lamp (Fig. 6). Adsorption of RR 198 was conducted in the dark during 24 h of stirring. It was noted that adsorption abilities of crystalline cellulose and precipitated cellulose were insignificant what is related with small specific surface area of tested materials. Addition of TiO₂/N catalyst cellulose matrix leads to stabilization of the surface of composite. It could be deduced that particles of TiO₂/N photocatalyst are loaded on the surface of precipitated cellulose and these nitrogen modified titania particles effectively block the releasing of ionic liquid residues to aqueous solution. It is worth mentioning that minor photoactivity of tested composite was observed. No photodegradation of used azo dye was possible to observe. This was also confirmed by UV–Vis spectroscopy measurements and no changes in RR 198 concentration was observed.

In Fig. 7 thermal stability of crystalline cellulose, ionic liquid ((cyclohexyl)hexyldimethylammonium acetate), precipitated cellulose and TiO₂/N immobilized in cellulose matrix are shown. A small mass loss observed at 70–130 °C is related to the release of moisture from the samples. In TiO₂/N-cellulose nanocomposite sample mass loss in a range of 150–280 °C is possible to observe. This could be assigned to the presence of ionic liquid residues. The thermal decomposition of pure (cyclohexyl)hexyldimethylammonium acetate could be observe in a temperature range from 150 to 250 °C with the maximum rate at 163 °C. It was noted that additive of TiO₂/N photocatalyst decreases the decomposition temperature of new composite. Some nanoparticles of cellulose could be present in water and that results in the increase of TOC value and that fact could not be excluded during the efficiency estimation of separation method.