The materials used in this research were 4′-pentyl-4-biphenylcarbonitrile (5CB) (Sigma-Aldrich), which is a type of liquid crystal, tetra-n-butyl orthotitanate (TBOT) (Sigma-Aldrich) as the TiO₂ precursor, 2-propanol, and distilled water.

In a typical experiment, 5CB (9.8 mg), 2-propanol (2.057 ml), and distilled water (0.016 ml) were mixed well in a 5-ml sample bottle. The TBOT (0.1 ml) was then added dropwise into the mixture, with stirring. Then, the solution was transferred into a petri dish and covered with a perforated aluminum foil. The petri dish containing TBOT, 5CB, 2-propanol, and distilled water was then placed under a magnetic field (0.3 T) produced by neodymium block magnet bars as shown in Supplementary Data 3. The sample was allowed to dry at room temperature until a constant weight was reached. The TiO₂–5CB sample that was dried under the neodymium block magnet bars took 14 days to reach constant weight. The TiO₂–5CB sample that was dried outside the neodymium block magnet bars took 12 days to reach constant weight. The magnetic strength applied was measured by a handheld Gauss Meter teslameter. The relative humidity was 60%.

The scanning electron microscopy (SEM) images of the TiO₂ composites were taken using a JEOL JSM-6390LV instrument operating with an accelerating voltage of 15 kV. The specific surface area of the composites was obtained using a multipoint Brunauer–Emmett–Teller (BET) analysis via the measurement of nitrogen adsorption–desorption as a function of relative pressure. The samples were analyzed using a Thermo Fischer Scientific Sorptomatic 1990 surface analyzer utilizing a customary process at −196°C. For preparation, the samples were pretreated and degassed at 150°C for 12 h. Thermogravimetric analysis (TGA) was carried out using a Mettler Toledo TGA/SDTA851 thermogravimetric/differential thermal analyzer. The scanning was performed up to the temperature of 500°C, with a heating rate and air flow rate of 10°C min⁻¹ and 20 mL min⁻¹, respectively. Direct current (DC) electrical conductivity was tested in the range of 0 to 25 V at room temperature using a PHYWE Model with the main voltage of 230 V and frequency of 50/60 Hz. Hall effect studies were set up with the components of Hall probe (Ge Crystal–n-type), Hall probe wooden stand, Hall effect set up model (DHE)−21, electromagnet model (EMU)−50 V, constant power supply (DPS)−50, and digital gaussmeter model (DGM)−201. The current and Hall voltage values were recorded. Diffuse reflectance UV–visible (DR UV–Vis) spectroscopy was used to determine the chromophore–chromophore interactions in the samples. For the DR UV–Vis study, the spectra were recorded in the range of 200–500 nm using a Perkin Elmer UV–visible Spectrometer Lambda 900 with barium sulfate (BaSO₄) as the standard reference. For Fourier-transform infrared spectroscopy (FTIR), the spectra were recorded in the range of 500–4,000 cm⁻¹ using a Perkin Elmer Spectrum One Spectrometer. For X-ray photoelectron spectroscopy (XPS) analysis, the samples were analyzed using a PHI-5500 spectrometer coupled with monochromatic Mg K_α (1253.6 eV). The sample was scanned at a binding energy of 0 to 1200 eV with an X-ray source generated at 14 kV and 300 W. The crystallinity and phase content of the solid materials were determined using a Bruker AXS Advance D8 X-ray diffractometer (XRD) using a diffracted monochromatic beam of Cu K_α (λ = 1.5406 Å) radiation produced at 40 kV and 40 mA. The diffraction pattern was scanned in the 2θ range of 20–80° with a gradual increment of 0.05° and a step time of 1 s. Photoluminescence (PL) analysis was done to study the electrons–holes recombinations of the photocatalysts. The measurement was performed at room temperature using a JASCO spectrofluorometer (FP-8500), with a 150 W Xe lamp as the excitation source. The excitation wavelength was 279 nm.

The photocatalytic activity of the TiO2 composites was evaluated through the photocatalytic oxidation of styrene in the presence of 30% aqueous hydrogen peroxide (H₂O₂). The reaction solution was prepared using 0.575 mL of styrene, 0.817 mL of aqueous H₂O₂, and 5 mL of acetonitrile. About 50 mg of photocatalyst was then added to the solution. The final concentration of styrene, hydrogen peroxide, and acetonitrile were 0.78, 1.25, and 15.02 M, respectively. The reaction was carried out for 24 h at room temperature with the solution being photoirradiated using a 2 × 15 W UV lamp (Vilber Lourmat, VL-215-C, 254 nm) and magnetically stirred continuously. The setup of experiment is attached as Supplementary Data 1. After the reaction, the mixture was centrifuged, and the catalyst was removed. The solution obtained was analyzed by a gas chromatograph equipped with a capillary column and a flame ionization detector. Nitrogen gas was used as the carrier gas with a column flow of 3 mL min⁻¹.

One-dimensional-like TiO₂ composite was successfully prepared by the sol–gel method under a magnetic field (0.3 T) using TBOT as the TiO₂ precursor, 5CB liquid crystal as the structure-aligning agent, and a mixture of 2-propanol and distilled water as the solvent. Figure 1A shows the SEM image of TiO₂ composite synthesized without a magnetic field. The sample formed was not aligned and possesses irregular shape. Figure 1B shows the TiO₂–5CB composite synthesized under a magnetic field. This composite was whiskerlike in shape. The 1-D-like TiO₂–5CB synthesized under a magnetic field has a very high length-to-diameter ratio with a diameter and length in the range of 100 to 500 nm and 500 to 1,000 μm, respectively. The SEM images strongly confirm that this method can produce the 1-D-like TiO₂ composites. In addition, the magnetic field does affect the shape of the TiO₂ composite.

The surface properties of TiO₂–5CB composites were evaluated using multipoint the BET analysis. As shown in Figure 2, both types of TiO₂ composites have similar isotherms, which is type IV isotherm with hysteresis loops of H4 (Ertl et al., 1997; Lakhi et al., 2017). The Barrett-Joyner-Halenda (BJH) plot analysis confirmed that TiO₂–5CB synthesized under and without a magnetic field consisted of both micropore and mesopore with a uniform pore size distribution. The plausible location of pores in the TiO₂ composites is illustrated in Figure 3. It was suggested that the pores were formed between the composites. From the sorption isotherm of the samples, the BET surface area was obtained and has been tabulated in Table 1. Both samples have similar surface areas. Physically, the surface area and pore structure of TiO₂–5CB synthesized under and without the magnetic field are similar, and the only distinct difference is the shape of particles that could have a direct influence on the photocatalytic performance as discussed later.

The TiO₂ composites were analyzed using the TGA. The TGA of TiO₂ anatase standard is attached as Supplementary Data 2. Generally, the weight percentages of composites decreased with increasing temperature until 550°C as shown in Figure 4. The decomposition of 5CB occurred in the range of 130–550°C. The composition of 5CB in the TiO₂–5CB prepared under and without a magnetic field is about 13.2 and 13.0 wt%, respectively. The amount of 5CB obtained from TG analysis was compared with the amount of 5CB added during the synthesis of TiO₂–5CB. As shown in Table 2, it was confirmed that the content of 5CB in both samples did not vary significantly. The weight lost detected below 130°C corresponds to the evaporation of the solvent in the composites. It was found that the TiO₂–5CB synthesized without the magnetic field contained a higher amount of solvent. One-dimensional-like TiO₂–5CB synthesized under the magnetic field consisted of only 2.1% of solvent, which is 14.1% lower. This phenomenon explains the differences in the particle packing of each composite. The TiO₂–5CB composite synthesized without the magnetic field is believed to consist of a large number of voids, which are formed by the loosely packed irregular-shaped particles. The voids formed are able to trap a higher amount of solvent. On the contrary, 1-D-like TiO₂–5CB composites synthesized under a magnetic field had a more compact particle packing since the particles were uniform in shape. These particles were arranged tightly to each other, leaving negligible voids that trapped less solvent. The findings here are in line with our concept about the formation of pores in each composite as shown in Figure 3.

The formation of 1-D-like TiO₂ has a strong relationship with the existence of liquid crystal in a magnetic field. Nevertheless, the condition during the hydrolysis process is very crucial for the formation of 1-D-like TiO₂ too. It was found that a slow hydrolysis rate is the key factor for the formation of 1-D-like TiO₂–5CB under a magnetic field. Figure 5 shows the formation of 1-D-like TiO₂–5CB synthesized under a magnetic field by a slow hydrolysis process. During the slow hydrolysis, it was expected that the orientation of TBOT was templated by the anisotropic medium of 5CB; hence, the 1-D-like TiO₂ was obtained. Interestingly, the impact of the anisotropic 5CB on the orientation of TiO₂ molecules can be proven by the XRD study, where an anatase phase was observed in the 1-D-like TiO₂–5CB composite. As shown in Figure 6, TiO₂–5CB synthesized without a magnetic field was in an amorphous phase only. Whereas, the peaks representative of the anatase phase were observed for the TiO₂–5CB synthesized with a magnetic field. This composite was not fully amorphous. It is well established that diamagnetic assemblies having magnetic anisotropy will become oriented in a steady magnetic field to achieve the minimum-energy state (Collings and Hird, 1997). Therefore, the structure of organized molecular assemblies of 5CB liquid crystal can be controlled by a magnetic field. When 5CB molecules receive torque in a magnetic field, they will orientate in the direction where its magnetic energy is minimum. This orientated mesophase acts as the template for subsequence crystallization. This phenomenon made it possible to align a TBOT skeleton through the magnetic orientation of a nematic mesophase of the 5CB used as a template. During the hydrolysis process, the TBOT molecules polymerize slowly in the aligned mesophase. Hence, the ordered structure of TBOT is formed by the arrangement of anisotropic molecules, and the resultant crystallinity increases (Yamaguchi and Tanimoto, 2006). Several studies also reported this similar phenomenon, where the magnetic field affects the crystallinity of synthesized materials such as BiFeO₃/LDPE (Song et al., 2017), Fe₃O₄ (Hong et al., 2007), and CaCO₃ (Tai et al., 2008). Based on the data obtained, we suggest that the crystallization process is dependent on the orientation of TBOT molecules, and this concept is illustrated in Supplementary Data 4.

The FTIR spectroscopy is used to further resolve the interfacial interaction in TiO₂–5CB composites. The FTIR spectra of TiO₂, 5CB, TiO₂–5CB synthesized under or without magnetic field are depicted in Figure 7. The broad peaks at around 523–783 cm⁻¹ for TiO₂–5CB synthesized under and without magnetic field corresponded to the Ti–O bond in TiO₂ (Pavia et al., 2014). Besides that, the FTIR spectra of TiO₂–5CB synthesized under and without magnetic field also exhibited a broad peak at 3,300–3,400 cm⁻¹, which corresponded to the stretching vibration of hydroxyl (O–H) group since the solvent used during the synthesis process was 2-propanol. In addition, the characteristic peaks of 5CB liquid crystal, which were clearly observed, were assigned to as follows: C=N nitriles vibration at 2,226 cm⁻¹, = C–H aromatic stretching vibration at 3,026 cm⁻¹, C–H stretching vibration at 2,856–2,956 cm⁻¹, C = C aromatic stretching vibration at 1,494 cm⁻¹ and 1,606 cm⁻¹, and C–H aromatic out of plane bend at 553–815 cm⁻¹ (Pavia et al., 2014). The peaks of C–H aromatic out of the plane bend at 553–815 cm⁻¹ clearly disappeared after TiO₂ was mixed with the liquid crystals and the peaks of C = N nitriles appeared at 1,645 cm⁻¹. This is because the Ti–O bond in TiO₂ overlapped with this peak since the amount of TiO₂ precursor used in the synthesis is higher than that of liquid crystals and also due to the occurrence of the interaction between TiO₂ and liquid crystals. At the same time, the characteristic peaks of 5CB liquid crystal, which are C = N nitriles, C–H stretching vibration, and C = C aromatic stretching, and also Ti–O bond, exhibited a shifting when a comparison was done between the TiO₂–5CB synthesized under and without magnetic field. The peaks shifted marginally to a higher wave number, suggesting the occurrence of a strong interaction between TiO₂ and the 5CB liquid crystal when a magnetic field is applied.

Characterization using XPS was also carried out to strengthen the interfacial interaction discussion on TiO₂–5CB synthesized under and without magnetic field since the Ti–O bond shifted marginally to a higher wave number in the FTIR spectrum. The XPS analysis was done to validate the interaction that occurs in the Ti–O bond. The high-resolution spectra for Ti 2p and O 1s of TiO₂–5CB synthesized under and without magnetic field are shown in Figure 8. The binding energy of XPS is calibrated for interpretation. The peak at 456 eV represents Ti 2p^3/2 and the peak at 462 eV represents Ti 2p^1/2, while the peak at 530 eV represents O 1s (Moulder et al., 1992). The Ti and O peaks of TiO₂–5CB synthesized under magnetic field shifted 1.0 eV to a higher binding energy compared with that of TiO₂–5CB synthesized without magnetic field. This validated the FTIR results that there is the occurrence of a strong interaction between TiO₂ and 5CB.

The interaction of 5CB liquid crystal in the TiO₂ composite samples was further confirmed using DR UV–Vis spectrophotometer. One of the possible interactions that can be evaluated by DR UV–Vis spectroscopy is the chromophore–chromophore interaction and the ð^* stacking of the aromatic rings. Figure 9 shows the DR UV–Vis spectra of TiO₂ synthesized under and without magnetic field. Both spectra show a similar pattern. There are two peaks appearing in the DR UV–Vis spectra, i.e., the peak at 210 to 230 nm and 300 to 330 nm, which belong to 5CB and TiO₂, respectively. A reduction in intensity for the peak of 5CB at 210 to 230 nm can clearly be seen for the spectrum of TiO₂–5CB synthesized under magnetic field. This is caused by the quenching phenomenon that occurred in the sample. The 5CB liquid crystal has aromatic rings, which is a stack of aromatic chromophores that experience reduced relative intensity. Thus, when the aromatic rings of 5CB liquid crystal are stacked, interactions between those chromophores occur, causing a reduction in the intensity, known as ð^* stacking of the chromophores (Bauer et al., 1976). The chromophores may interact with one another and perturb their electronic transitions. The electronic transition is sensitive to the interactions between two chromophores that approach one another. These interactions can show up in electronic transitions as shifts in wavelength as well as in intensity changes (Bauer et al., 1976). As shown in the SEM image of the TiO₂ synthesized under magnetic field (see Figure 1B), the TiO₂ is well-aligned and closely stacked together. This arrangement allows the interaction between the chromophores, which is proven by the DR UV–Vis spectrum of the sample. But, for the case of TiO₂–5CB synthesized without a magnetic field, the random arrangements did not allow interactions between the chromophores; thus, no quenching can be seen in its DR UV–Vis spectrum. This shows that the alignment of the TiO₂ influences the interaction of its chromophores. As a result, the plausible chromophore–chromophore interaction of 5CB liquid crystal, which was embedded inside TiO₂, was illustrated in Figure 10.

The photocatalytic activity of TiO₂ synthesized with liquid crystal as the structure-aligning agent under magnetic field (0.3 T) was evaluated through the photocatalytic oxidation of styrene in the presence of the hydrogen peroxide (H₂O₂) as an oxidizing source. The stability of composites under irradiation of UV light was checked for prior to the experiment. The composites were irradiated under UV irradiation produced by a UV lamp (Vilber Lourmat, VL-215-C) with radiation intensity of 2 × 15 W and wavelength of 254 nm for 12 h. The morphology of 1-D-like TiO₂ composites remained intact after the photoirradiation process. This is due to the photostability properties of the 5CB liquid crystal (Wen et al., 2005) wherein it cannot be photodegraded easily under UV irradiation. The FTIR spectrum for TiO₂–5CB synthesized under magnetic field after the photoirradiation process shows the functional group of nitriles, -C=N, and aromatic rings, C = C, similar to those detected in the spectrum of TiO₂–5CB synthesized under magnetic field prior to photoirradiation. It is proven that the same functional groups were retained even after the photoirradiation process although the intensities of the peaks are slightly reduced. This indicates that the structure of the 1-D-like TiO₂–5CB synthesized under a magnetic field is quite stable and could not be destroyed easily even after the photoirradiation process. It suggests that the 1-D-like TiO₂–5CB synthesized under magnetic field could be suitable as a photocatalyst.

In order to evaluate the photocatalytic performance, three different photocatalysts were chosen; (B) TiO₂ anatase standard (Sigma Aldrich, 637254), conducted as the reference for this experiment, (C) TiO₂–5CB synthesized under magnetic field, and (D) TiO₂–5CB synthesized without magnetic field. A blank sample (A) consisted of only hydrogen peroxide, styrene, and acetonitrile was used as control experiment. Figure 11 shows the conversion of styrene to benzaldehyde in the photocatalytic reaction. The highest concentration (130 μmol) was obtained for the TiO₂–5CB composite synthesized under magnetic field (0.3 T). This is strong evidence that the 1-D-like structure can improve the photocatalytic power of TiO₂. The possible reason is that the electronic properties of TiO₂–5CB have been modified. For another reason, the 1-D-like structure has a limited diffusion direction throughout the structure, which resulted in the unidirectional traveling path of the electron (Thakur et al., 2017). As a result, the recombination of electron and hole is inhibited. Other than that, it can be assumed that the TiO₂–5CB synthesized under a magnetic field has better interfacial interaction compared with the TiO₂–5CB synthesized without magnetic field. This good interfacial interaction allows electron transfer between the 5CB liquid crystal and the TiO₂ during photocatalytic oxidation of styrene, which can enhance the electron transport and delay the recombination of the electron with the hole.

In order to prove our statement, the PL studies were carried out to investigate the efficiency of the charge carrier trapping, migration, and transfer and detecting the behavior of the electron–hole (e⁻/h⁺) pairs in the semiconductor particles. Figure 12 shows the PL spectra of TiO₂–5CB synthesized under and without a magnetic field, and the bulk TiO₂ and 5CB as references. The TiO₂ exhibited three emission peaks, which were at 402, 482, and 556 nm (Cong et al., 2007). The emission peaks at 392 and 363 nm are represented as 5CB (Bezrodna et al., 2017). The emission peak of TiO₂–5CB synthesized without magnetic field occurred at 394 nm, which was located between the emission peaks of 5CB and TiO₂. Since the PL emission peak is the result of the recombination of electron–hole, it was indicated that the TiO₂–5CB synthesized without a magnetic field has a higher recombination rate due to the increasing intensity of the emission peak. Meanwhile, the emission peak of TiO₂–5CB synthesized under a magnetic field occurred at 351 nm, and it was shifted to a lower wavelength with decreased intensity. It was shown that the decreased intensity of the emission peak leads to the efficient quenching of PL, which can be attributed to the excited electron transfer from the valence band to new levels that exist upper to the conduction band (Cong et al., 2007). For the shifted wavelengths of TiO₂–5CB synthesized under a magnetic field, this might have been caused by the presence of interactions between TiO₂ and π-π stacked 5CB (Bezrodna et al., 2017). Thus, it was concluded that electron charge transfer occurred in the well-aligned 1-D-like TiO₂–5CB synthesized under a magnetic field during UV irradiation in PL analysis.

The Hall effect studies can provide useful information on the motional behavior of the electron–hole pairs. From here, the correlation between the 1-D-like structure with the electron–hole pairs can be determined. Prior to the testing, the DC electrical conductivity of TiO₂–5CB composites was first determined. Figure 13 shows that when a higher voltage was applied to the TiO₂ composites, the amount of current generated was also higher. These results proved that the TiO₂ composites allow the flow of electricity. It can be seen that the plotted graph of TiO₂–5CB, synthesized under magnetic field, shows the highest current due to the difference in shape. The TiO₂–5CB synthesized under a magnetic field is well-aligned, while the TiO₂–5CB synthesized without a magnetic field was in an irregular aggregated shape. Thus, it can be concluded that, as the particles of TiO₂–5CB are arranged neatly next to each other, the flow of electricity has been enhanced and the diffusion of electrons is expected to be improved as well (Liu et al., 2015). Figure 14 shows that the TiO₂–5CB synthesized under a magnetic field displayed a higher Hall voltage when the current was higher. Once again, this is most probably caused by the arrangement of the TiO₂–5CB synthesized under a magnetic field, where it is neatly arranged next to each other, because with this arrangement the current can flow smoothly, and, therefore, the Hall voltage becomes higher (Ivanov and Nikolov, 2016). Our results show that the motional behavior of the electron in TiO₂–5CB of a different shape is quite distinct. For the TiO₂–5CB composite with 1-D-like structure, the separation of the electron from the hole is expected to be improved due to the better mobility of electron in the unidirectional pathway.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.