The synthesis of the TiO₂-SWCNTs started by loading 10 mg of SWCNTs into a 50 mL conical bottom polypropylene centrifuge tube. Next, 20 ml of 0.38 M titanium(III) sulfate solution was added. The centrifuge tube with the SWCNTs and solution was then placed in a test tube rack under a sonicator probe horn that is mounted on a ring stand. The sonicator used is a Qsonica ultrasonic cell disruptor, model Q125 by Misonix, equipped with a 1/8" diameter probe and operating at a frequency of 20 kHz. The probe tip was immersed into the solution and a power of 1 W/mL was utilized. Samples were exposed for set amount of times to achieve desired amount of TiO₂ deposition. After sonification was complete the tube containing the synthesized TiO₂-SWCNT in titanium(III) solution was centrifuged at 3500 RPM for 1 hour. The tube was then removed from the centrifuge and the liquid supernatant was decanted from the TiO₂- SWCNTs that were dropped to the bottom of the container. The TiO₂-SWCNTs were then washed by adding 20 mL of deionized water and exposing to ultrasound sonication for a duration of 3 minutes at a power of 1 W/mL to re-suspend the TiO₂-SWCNTs. The tube was then re-centrifuged at 3500 for 1 h, the liquid decanted and replaced with clean deionized water, and then resuspended by sonication for 3 minutes at a power of 1 W/mL. This wash procedure was repeated 2 more times. The TiO₂-SWCNTs were stored in water in the centrifuge tubes until ready for characterization.

It was determined by Raman analysis that an amorphous form of TiO₂ was deposited on the surface of the SWCNTs via this sonochemical process. It is of interest to convert amorphous (poorly crystallized) forms of deposited TiO₂ to their corresponding ordered anatase crystal structure. For this conversion, a hydrothermal calcination process was used. Here, 10 mg of the TiO₂-SWCNT reaction product and 10 mL of deionized water were placed in 23 mL Teflon^® lined, high pressure stainless steel digestion bomb, utilized as an autoclave. The autoclave was loaded into a furnace for 8 h at 250 degrees C to achieve hydrothermal calcination of the TiO₂-SWCNT product. Upon completion the finished autoclaved samples were cooled to ambient and stored in the deionized water.

The morphology of pristine SWCNTs and TiO₂-SWCNTs (2 h sonication) were characterized by scanning electron microscopy (S 4800 SEM, Hitachi Co., Tokyo, Japan). The pristine SWCNT sample for this analysis was prepared by ultrasonic dispersion in isopropanol followed by deposition onto a conductive porous silver membrane. The TiO₂-SWCNTs (stored in water from the synthesis step) were prepared for analysis by homogeneously dispersing in water followed by deposition onto a conductive porous silver membrane. A small sample was cut from this prepared composite and adhered via conductive tape to the microscope stage. Once the sample was staged it was then loaded into the microscope for analysis.

Sample spectra were obtained using a Thermo Scientific Nicolet Almega XR Dispersive Raman Spectrometer equipped with an Olympus BX-51 research microscope. Raman spectra analysis was conducted on samples from 100 to 4000 cm⁻¹ with the laser operating at 532 nm at 100% of 150 mW with the beam going through the microscope equipped with a MPlain 10X BD objective.

X-ray diffraction (XRD) data was collected by using a Rigaku Miniflex, with CuK_α radiation of 40 kV/4 mA, λ = 1.5406 Å. For pristine SWCNTs, the sample was prepared by deposition of SWCNTs on a silver membrane filter as described above for the SEM analysis. A small sample was cut from this prepared composite and the scan was ran from 2θ = 10˚ to 40˚. For the remaining samples (TiO₂-SWCNTs), measured as dry powders, preparation consisted filling the powder sample dish and loading into the diffractometer. The collected scans were obtained from a range of 2θ = 20˚ to 75˚.

The first point of reference is Figure 1 that depicts the SEM image of dispersed SWCNTs deposited on a silver membrane filter at 150 k magnification.

From the SEM image it is demonstrated that the SWCNTs show a superstructure that exhibits a high degree of entanglement between the bundles of tubes. This can be accounted for by the report that SWCNTs allow for remarkable Van der Walls interactions due to their smooth, uniform surfaces in close proximities. Thus, the majority of the tubes are bundled via direct van der Waals attractions along their entire lengths [41]. These SEMs are in good agreement with similar reports published in the literature [42-48].

The Raman spectrum for the SWCNTs is shown in Figure 2.

As reported by Eklund, five peaks centered at around 1340 cm⁻¹ (D mode), 1580 cm⁻¹ (G mode), 2450 cm⁻¹, 2680 cm⁻¹ (G’ mode), and 3180 cm⁻¹ are due to SWCNTs [49]. These five peaks are clearly visible and labeled in the spectrum. Here the three predominate modes include the D, G, and G’ modes. The D mode (1340 cm⁻¹) originates from structural defects and is present in all graphite-like carbons. The G mode (1580 cm⁻¹) corresponds to planar vibrations of carbon. The position of the G band confirms that the sample is mainly composed of single wall carbon nanotubes [49]. The G’ mode (2680 cm⁻¹) is the second overtone of the D mode, which makes the name misleading. It is given because it is usually the second strongest mode in graphite. It is reported that the diameter of the SWCNT influences the position of the G’ mode, thus a rough estimate of tube diameter can be deduced (this information is not reported in this study) [50]. Two additional modes worth noting include the radial breathing mode and the bundling mode. The radial breathing mode is the result of the expansion and contraction that occurs across the radial cross section in the nanotubes, typically occurring in the 100 to 350 cm⁻¹ range [51,52]. A special form of the radial breathing mode, the bundling mode, is reported to originate from the collective vibration in a bundle of SWCNTs [53]. It is apparent from Figure 2 that there is a percentage of the SWCNTs that are bundled.

XRD was performed for SWCNTs deposited on a silver membrane filter as shown in Figure 3.

A single peak centered at 2θ = 38.36˚ is identified as silver from the membrane filter support and agrees with published literature [54]. There are no peaks for SWCNTs because crystal plains are non-existent. It is worth noting, from the lack of MWCNT spectral lines, that the CNTs evaluated are confirmed to be of single wall configuration [55-57]. MWCNTs are reported to have a peak at 2θ = 26.06˚, corresponding to the interplanner spacing of 0.342 nm between the tubes [58]. This

spectrum is used for comparison against TiO₂-SWCNTs spectrums.

The Raman spectra for P25 (a commercially produced nanosized TiO₂ from Evonik) is shown in Figure 4.

The six Raman active modes, A_1g + 2B_1g + 3E_g, for anatase TiO₂ that have been reported in the literature are detected at 144 cm⁻¹ (E._g.), 197 cm⁻¹ (E._g.), 399 cm⁻¹ (B_1g), 513 cm⁻¹ (A_1g), 519 cm⁻¹ (B_1g), and 639 cm⁻¹ (E._g.),[59] and the peaks at 250, 450, and 620 cm⁻¹ are due to the Ti-O-Ti bond of rutile spectrum [60]. The characteristic bands for brookite (not found in this study) are reported at 128, 135, 153, 172, 195, 214, 247, 288, 322, 366, 396, 412, 454, 461, 502, 545, 585, and 636 cm⁻¹ [61]. In the spectra for P25, five anatase crystal peaks were detected at 144 cm⁻¹, 197 cm⁻¹, 399 cm⁻¹ (B_1g), and 639 cm⁻¹. One rutile peak can be seen at 450 cm⁻¹. The positions and intensities of the Raman active modes correspond well with the percent of anatase phase of TiO₂ relative to the percent of rutile. The spectrum agrees with the P25 spectrum published by Parussulo. [62] The full width at half maximum (FWHM) of the strongest peak (144 cm⁻¹) is 27.0 cm⁻¹ correlating to the manufactures reported 21 nm crystallite size.

XRD of P25 nanoparticle was ran from 2θ = 20˚ to 75˚. Anatase TiO₂ has diffraction peaks at 2θ = 25.18˚ (strongest peak), 37.78˚, 48.00˚, 53.89˚, 54.99˚, 62.57˚, 68.68˚, 70.15˚, and 75.01˚, corresponding to the reflections from 101, 004, 200, 105, 211, 204, 116, 220, and 215 crystal planes [63]. The strongest peaks for rutile and brookite are at 2θ = 27.5˚ for the 110 crystal plane and 2θ = 30.8˚ for the 121 crystal plane of rutile and brookite respectively. [64] The spectrum is shown in Figure 5.

Here, the spectrum shows predominately anatase TiO₂ with the necessary diffraction peaks at 2θ = 25.18˚ (strongest peak), 37.78˚, 48.00˚, 53.89˚, 54.99˚, 62.57˚, 68.68˚, 70.15˚, and 75.01˚. The remaining peeks are due to the rutile structure. The integrated intensities of the strongest peak of both anatase (I_a,101 at 25.4˚) and rutile (I_r,110 27.5˚) were used to calculate the percentage of rutile and anatase according to the formulas: W_r(%) = I_r/(0.8844xI_a + I_r) × 100 and W_a(%) = 100 − W_r(%) [65]. The values indicate a composition of 78.8 % anatase and 21.2% rutile. Additionally, the average crystal diameter was calculated from the largest peak by the Scherrer’s equation which is B = Kλ/βcosθ where B is the crystalline size, in nm; l is the wavelength for the radiation used, which is 1.54056A for Cu; β is the full width at half maximum intensity (FWHM); θ is the angle for the XRD maximum peak and the shape factor, K is 0.93 [66]. This value is 20.3 nm and agrees well with the published value of 21 nm.

This portion of the study was conducted in order to characterize and determine a profile for increasing sonication exposure times for the deposition of TiO₂ onto SWCNTs. Sonochemical deposition times of 10 minute, 20 minute, 30 minute, and 2 h were performed and investigated.

Figure 6 shows the SEM for TiO₂ deposited on SWCNT at a deposition time of 2 hours.

From the micrograph microscopy image it is clear that the entangled superstructure and, clearly demostrates the nanosized TiO₂ that has been successfully decorated on the SWCNTs. It is noted that homogeneous deposition of both agglomerates of small TiO₂ particles and isolated grains appear to be deposited on the SWCNTs. This is in agreement with findings reported by Yao [67].

Figures 7-10 depict the Raman spectrum of the manu-

factured TiO₂-SWCNTs from 90 cm⁻¹ to 4000 cm⁻¹, at deposition times of 10 minute, 20 minute, 30 minute, and 2 h exposures to sonication.

All the Raman spectra reveal the SWCNTs peaks but are absence of crystalline TiO₂, as noted by comparison with Figure 4. In the TiO₂ portion of the spectrums, below 650 cm⁻¹, the absence of the crystalline TiO₂ confirms the amorphous characteristic of the deposited material. It is apparent that amorphous TiO₂ was predominant in all the samples. This is further confirmed by Figure 11 where the XRD of the 2 h sample is presented.

Here the spectrum shows no discernible peaks that would be expected for a crystalline material, as noted by comparison with the XRD spectrum for P25 as seen in Figure 5.

Figures 12-15 depict the Raman spectrum of the synthesized TiO₂-SWCNTs from 90 cm⁻¹ to 4000 cm⁻¹, at

deposition times of 10 minute, 20 minute, 30 minute, and 2 h exposures to sonication, followed by hydrothermal

calcination treatment.

All the Raman spectra displayed the necessary peaks for anatase TiO₂ crystals and SWCNTs showing they are both present in the samples. This is illustrated by comparison with the spectrum obtained for P25 as shown in Figure 4. Figures 12-15 also allude to a trend where the peaks appear more robust and defined as the length of deposition time increases. It appears that exclusively anatase was made in all the samples with the exception of the 10 minute specimen. For the 10 minute sample rutile TiO₂ appears to also be present along with the anatase. A clearer illustration is portrayed in Figure 16 where is of interest to note the anatase peaks at 160.68 cm⁻¹ (E._g.), 212.51 cm⁻¹ (E._g.), 399.85 cm⁻¹ (B_1g), 513.27 cm⁻¹ (A_1g), and 631.58 cm⁻¹ (E._g.) and vaguely depicted rutile peaks at 260.56 cm⁻¹, 447.04 cm⁻¹, and 612.82 cm⁻¹. The strongest, well resolved 160.68 cm⁻¹ (E._g.) mode is due to the external vibrations of the anatase structure, indicating that long-range order was formed [68].

This is also the case for the other measured TiO₂-SWCNT

samples.

All the samples show spectral peaks that are broadened and/or shifted with respect to P25, with the 10 minute sample showing the largest shift. For the 10 minute sample, the largest shift differences are noted at the highest and lowest E._g. mode. The highest mode has blue shifted (peak at 631.58 cm⁻¹) while the lowest mode (E._g.) has red shifted by approximately Δ16 cm⁻¹ (from 147.20 cm⁻¹ for P25 anatase to 160.68 cm⁻¹). A similar shift was reported by Bersani et al. who attributed it to phonon confinement caused by the decrease in the crystal size of the anatase TiO₂ [69]. These shifts are apparent in nanoparticles that have a diameter smaller than 15 nm and are indicative of reduction of crystal sizes [70]. For comparison, Figure 17 show an overlay of the observed red shift, for the lowest mode (E._g.), for the samples evaluated.

Here the compiled Raman spectrum from 90 cm⁻¹ to 220 cm⁻¹ for the synthesized TiO₂-SWCNTs at deposition times of 10 minute, 20 minute, 30 minute, and 2 h exposures to sonication with hydrothermal calcination treatment is shown. The spectrum is normalized for the large TiO₂ (E._g.) peaks clustered around 147 cm⁻¹ to 161 cm⁻¹ with the spectrum of p25 added for reference. With the grain size of P25 equal to a measured 20.4 nm, it is clear that the samples have relatively smaller sized TiO₂ particles. The decrease in size trend corresponds to a decrease in sonication exposure time. In addition to the shift in peaks, an increase in band asymmetry and broadening are observed as the length of sonification time decreases. The broadening of the peaks is reported to result from strain gradients in systems where TiO₂ is being chemically anchored to CNTs. These strain effects can extend several nm into materials [71]. To illustrate the relative broadening of the peaks, a measure of full

width at half maximum (FWHM) for the samples at the lowest E._g. mode is: 10 minute = 49.1 cm⁻¹, 20 minute = 41.8 cm⁻¹, 30 minute = 41.2 cm⁻¹, 2 h = 39.2 cm⁻¹, P25 = 25.4 cm⁻¹.

The XRD spectra for the hydrothermally treated samples, including 10 minute, 20 minute, 30 minute, and 2 h are presented in Figures 18-21.

Here the scans were ran from 2θ = 20˚ to 75˚. The spectrums show the characteristics of predominately anatase TiO₂ with diffraction peaks at 2θ = 25.18˚ (strongest peak), 37.78˚, 48.00˚, 53.89˚, 54.99˚, 62.57˚, 68.68˚, 70.15˚, and 75.01˚. The average crystal diameter for each sample was calculated from the largest peak by the Scherrer’s equation as described previously. These values are reported in Table 1.

It is clear from the data that the size of TiO₂ decreases as the length of deposition time is reduced. The size of the deposited TiO₂ versus the observed Raman red opshift, for the lowest mode (E._g.), for the samples evaluated, is also depicted in Table 1. Here a significant correlation is realized where the decreasing crystal size of deposited TiO₂ is linked to the observed Raman red shift for the lowest E.g_. mode measured for the TiO₂.