Self-Assembly Regulated Photocatalysis of Porphyrin-TiO2 Nanocomposites

Photoactive artificial nanocatalysts that mimic natural photoenergy systems can yield clean and renewable energy. However, their poor photoabsorption capability and disfavored photogenic electron–hole recombination hinder their production. Herein, we designed two nanocatalysts with various microstructures by combining the tailored self-assembly of the meso-tetra(p-hydroxyphenyl) porphine photosensitizer with the growth of titanium dioxide (TiO2). The porphyrin photoabsorption antenna efficiently extended the absorption range of TiO2 in the visible region, while anatase TiO2 promoted the efficient electron–hole separation of porphyrin. The photo-induced electrons were transferred to the surface of the Pt co-catalyst for the generation of hydrogen via water splitting, and the hole was utilized for the decomposition of methyl orange dye. The hybrid structure showed greatly increased photocatalytic performance compared to the core@shell structure due to massive active sites and increased photo-generated electron output. This controlled assembly regulation provides a new approach for the fabrication of advanced, structure-dependent photocatalysts.


Introduction
Energy sustainability is crucial for anthropogenic activity and the development of societies [1].Recent technological advancements, economic development, and population growth have exerted increasing pressure on the global energy supply.However, with the decline of fossil fuel reserves and climate change, the development of green sustainable alternative energy has become a major challenge [2,3].To address this energy shortage while reducing the dependency on fossil fuels, many new forms of energy, including wind energy, solar energy, geothermal energy, hydropower, hydrogen energy, and bioenergy have been explored [4][5][6][7].Renewable, sustainable, and green hydrogen energy is considered an attractive alternative to fossil fuels [8,9].Furthermore, the generation of hydrogen fuel from water and sunlight, two of the most abundant natural resources on Earth [10], offers a promising pathway for acquiring hydrogen.Solar water splitting transforms sunlight into hydrogen through photocatalytic, photoelectrochemical, and photovoltaic electrolysis routes [11][12][13].The photocatalytic approach works without a corrosive electrolyte or external power, providing a more competitive pathway to hydrogen production.
The catalyst is the key to photocatalysis, and among various photocatalysts [14][15][16][17], titanium dioxide (TiO 2 )-based nanomaterials are the most widely used.TiO 2 crystals were divided into anatase, rutile, and brookite phases [18][19][20], which have good photophysical properties and chemical stability, an appropriate band gap matching the redox level of water [21,22], and an adjustable microstructure regulated by hydrothermal, sol-gel, or redox methods [23,24].Specifically, TiO 2 -related catalysts have wide applications in photocatalytic water decomposition and pollutant degradation [25].The photocatalytic performance of TiO 2 is affected by its microstructural characteristics.N-doping, photosensitizer modification, and composite strategy are important approaches to increasing the light absorption, carrier concentration, and photocatalytic activity of TiO 2 crystals [26][27][28].The final catalytic efficiency of different structures suggests microstructure-dependent catalytic activity.However, the wide band gap of TiO 2 only responds to ultraviolet (UV) light (about 4% of the solar spectrum) instead of visible light (about 40% of the solar spectrum) [29][30][31].Moreover, the harmful crystal defects and low specific surface area of TiO 2 significantly hinder its photocatalytic performance [32,33].Therefore, it is urgent to develop a nanocatalyst with a unique structure to enhance the photocatalytic performance of TiO 2 in visible light.Although a variety of artificial TiO 2 microstructures that mimic natural light-capturing behavior have been studied over the past decades, the precise control of TiO 2 microstructures on the nanoscale in photocatalysis remains a major challenge.
Self-assembly is a common phenomenon in nature.For instance, self-assembly controls the chloroplast complex by programmatically regulating the size and aggregated spatial structure of functional motifs to drive natural photosynthesis.Moreover, self-assembly is a powerful strategy used to fabricate nanomaterials with unique microstructures and new functions [34][35][36][37].Among numerous photoactive motifs, porphyrins are superior assembly modules with a rigid planar structure and strong visible light absorption capability [38][39][40].However, they are hindered by the face recombination of photoelectrons and holes after irradiation and the risk of photobleaching without a protective external shell.The synergistic effect of porphyrin/TiO 2 with microstructural regulation offers a promising strategy to address these issues.
Herein, we developed two meso-tetra(4-hydroxyphenyl) porphyrin (THPP)/TiO 2 nanocomposites with various assembled microstructures (Figure 1), i.e., hybrid-type nanoparticles (THPP-TiO 2 NPs) and core@shell type (THPP@TiO 2 NPs), through the tailored self-assembly of porphyrin and the hydrolysis/condensation of the TiO 2 precursor, titanium diisopropoxide bis(acetylacetonate) (TAA).The controlled co-assembly can be regulated due to the stronger hydrogen bonding interactions between the hydroxyl group of the TiO 2 scaffold and THPP than those of meso-tetra(4-benzenesulfonic acid)porphyrin (TSPP) and meso-tetra(4-carboxyphenyl)porphyrin (TCPP).The sodium dodecyl sulfate (SDS) surfactant and hydrothermal treatment further limited the size of the obtained nanocomposite and increased the crystallization degree of TiO 2 , respectively.Acting as a light-capturing antenna, porphyrin assisted the THPP-TiO 2 NPs in harvesting and transferring visible light into a higher level of hydrogen using its large specific surface area, abundant active sites, superior light-capturing capability, and high crystallinity.Our advanced photocatalytic system provides a technically simple method for the microstructural regulation of catalysts to greatly improve photoenergy utilization and visible-light-responsive photocatalytic hydrogen evolution.

The Optimization of Light-Capturing Blocks
The THPP building blocks were self-assembled into 169.2nm smooth nanospheres in a 25 mL reaction system (Figure 2A-C).The smooth surface of the THPP NPs gradually became rough with the increasing ratio of TiO2 precursor (TAA).The proportion of TAA has a crucial role in the co-assembly process.The optimal ratio of TAA was determined according to the morphologies of the nanomaterials.The scanning electron microscopy (SEM) images of the THPP-TiO2 nanocomposites indicated that 5 µL of TAA could acquire a monodispersed and homogeneous distribution of TiO2 in nanoparticle than 10 µL or 100 µL (Figure 2D).The THPP in THPP-TiO2 NPs was replaced by TSPP (Figure 2E,F) or TCPP (Figure 2G,H) but both were unable to yield a homogeneous morphology similar to that of the THPP-TiO2 NPs.These results suggested that the benzene hydroxyl group in THPP had a stronger binding force with the hydroxyl group of TiO2 in the sol-gel preparation process, yielding a homogeneous incorporation.

The Optimization of Light-Capturing Blocks
The THPP building blocks were self-assembled into 169.2nm smooth nanospheres in a 25 mL reaction system (Figure 2A-C).The smooth surface of the THPP NPs gradually became rough with the increasing ratio of TiO 2 precursor (TAA).The proportion of TAA has a crucial role in the co-assembly process.The optimal ratio of TAA was determined according to the morphologies of the nanomaterials.The scanning electron microscopy (SEM) images of the THPP-TiO 2 nanocomposites indicated that 5 µL of TAA could acquire a monodispersed and homogeneous distribution of TiO 2 in nanoparticle than 10 µL or 100 µL (Figure 2D).The THPP in THPP-TiO 2 NPs was replaced by TSPP (Figure 2E,F) or TCPP (Figure 2G,H) but both were unable to yield a homogeneous morphology similar to that of the THPP-TiO 2 NPs.These results suggested that the benzene hydroxyl group in THPP had a stronger binding force with the hydroxyl group of TiO 2 in the sol-gel preparation process, yielding a homogeneous incorporation.

The Optimization of Light-Capturing Blocks
The THPP building blocks were self-assembled into 169.2nm smooth nanospheres in a 25 mL reaction system (Figure 2A-C).The smooth surface of the THPP NPs gradually became rough with the increasing ratio of TiO2 precursor (TAA).The proportion of TAA has a crucial role in the co-assembly process.The optimal ratio of TAA was determined according to the morphologies of the nanomaterials.The scanning electron microscopy (SEM) images of the THPP-TiO2 nanocomposites indicated that 5 µL of TAA could acquire a monodispersed and homogeneous distribution of TiO2 in nanoparticle than 10 µL or 100 µL (Figure 2D).The THPP in THPP-TiO2 NPs was replaced by TSPP (Figure 2E,F) or TCPP (Figure 2G,H) but both were unable to yield a homogeneous morphology similar to that of the THPP-TiO2 NPs.These results suggested that the benzene hydroxyl group in THPP had a stronger binding force with the hydroxyl group of TiO2 in the sol-gel preparation process, yielding a homogeneous incorporation.

The Regulation of Assembled Microstructure
The TEM images of THPP-TiO 2 NPs revealed a rough morphology (Figure 3A).The UV-vis spectra indicated successful THPP loading (Figure 3B), which disappeared after etching by sodium hydroxide (NaOH).The elemental mapping further suggested a homogeneous distribution of Ti, N, and C in the nanoparticles (Figure 3C).To further decrease the nanoparticle size to improve the catalytic performance, we used an interfacial self-assembly microemulsion process driven by SDS micelle.As a result, the THPP-TiO 2 NPs greatly decreased in size around the 50-100 nm range (Figure 3D).Moreover, the TEM images of the etched THPP-TiO 2 NPs revealed an obvious pore structure after etching (Figure 3E,F).The specific surface area is an important factor of the catalyst in the photocatalysis process.The N 2 adsorption-desorption isotherms were collected to detect the pore change.The results indicated that the specific surface area of the THPP-TiO 2 NPs increased from 76.83 to 94.96 m 2 /g after etching (Figure 3G).

The Regulation of Assembled Microstructure
The TEM images of THPP-TiO2 NPs revealed a rough morphology (Figure 3A).The UV-vis spectra indicated successful THPP loading (Figure 3B), which disappeared after etching by sodium hydroxide (NaOH).The elemental mapping further suggested a homogeneous distribution of Ti, N, and C in the nanoparticles (Figure 3C).To further decrease the nanoparticle size to improve the catalytic performance, we used an interfacial self-assembly microemulsion process driven by SDS micelle.As a result, the THPP-TiO2 NPs greatly decreased in size around the 50-100 nm range (Figure 3D).Moreover, the TEM images of the etched THPP-TiO2 NPs revealed an obvious pore structure after etching (Figure 3E,F).The specific surface area is an important factor of the catalyst in the photocatalysis process.The N2 adsorption-desorption isotherms were collected to detect the pore change.The results indicated that the specific surface area of the THPP-TiO2 NPs increased from 76.83 to 94.96 m 2 /g after etching (Figure 3G).

The Optimization for the Crystallinity of TiO2
Anatase generates a higher oxidation capacity due to a higher band gap (3.2 eV), oxygen vacancies to capture electrons, and catalytic activity compared to those of rutile.Therefore, the amorphous TiO2 in THPP-TiO2 NPs were further changed to anatase-type TiO2 by hydrothermal treatment.To expand their application, the volume of the reaction system was increased from 25 to 500 mL (Figure 4A).The SEM and TEM images indicated

The Optimization for the Crystallinity of TiO 2
Anatase generates a higher oxidation capacity due to a higher band gap (3.2 eV), oxygen vacancies to capture electrons, and catalytic activity compared to those of rutile.Therefore, the amorphous TiO 2 in THPP-TiO 2 NPs were further changed to anatase-type TiO 2 by hydrothermal treatment.To expand their application, the volume of the reaction system was increased from 25 to 500 mL (Figure 4A).The SEM and TEM images indicated that the products remained stable in size and morphology after scaled-up preparation (Figure 4B-D).The hydrolysis condensation product of TAA is amorphous TiO 2 in THPP-TiO 2 , exhibiting no prominent peaks in the X-ray diffraction (XRD) pattern (Figure 4E) and changed to the TiO 2 crystal of THPP-TiO 2 -H NPs after hydrothermal treatment.The characteristic peaks of THPP-TiO 2 -H NPs matched the JCPDS pattern (JCPDS 21-1272) of anatase.The SEM image of the THPP-TiO 2 -H NPs maintained a morphology similar to that before hydrothermal treatment (Figure 4F).The magnified TEM image indicated a 0.33 nm lattice spacing assigned to the (101) crystal face (Figure 4G, inset: lattice spacing) [41].These results indicated that the THPP-TiO 2 -H NPs had superior anatase crystal structure.
OR PEER REVIEW 5 of 10 (Figure 4B-D).The hydrolysis condensation product of TAA is amorphous TiO2 in THPP-TiO2, exhibiting no prominent peaks in the X-ray diffraction (XRD) pattern (Figure 4E) and changed to the TiO2 crystal of THPP-TiO2-H NPs after hydrothermal treatment.The characteristic peaks of THPP-TiO2-H NPs matched the JCPDS pattern (JCPDS 21-1272) of anatase.The SEM image of the THPP-TiO2-H NPs maintained a morphology similar to that before hydrothermal treatment (Figure 4F).The magnified TEM image indicated a 0.33 nm lattice spacing assigned to the (101) crystal face (Figure 4G, inset: lattice spacing) [41].These results indicated that the THPP-TiO2-H NPs had superior anatase crystal structure.

Structure-Dependent Photocatalysis
Lastly, we investigated the photocatalytic performance and verified the catalytic mechanism of the nanocomposites.The hybrid microstructure of the THPP-TiO2 NPs was adjusted to the core@shell control samples (THPP@TiO2 NPs) through a two-step method, i.e., the self-assembly of THPP combined with a TiO2 shell coating.The TEM image of the etched THPP@TiO2 NPs showed a hollow TiO2 shell structure (inset, Figure 5A).Methyl orange (MO) is a representative pollutant often used to investigate the photocatalytic performance of catalysts.Compared to the negligible photocatalytic activity of the commercial titanium dioxide photocatalyst P25, the THPP-TiO2-H NPs revealed a higher degradation output, exceeding 96.7% after 75 min (Figure 5B).The photocatalytic hydrogen production capabilities of both THPP-TiO2-H NPs and THPP-TiO2 NPs indicated irradiated time-dependent hydrogen generation, respectively reaching 4.80 and 1.03 mmol/g, which was much higher than that of P25, THPP NPs, THPP + TiO2, THPP@TiO2-H NPs, and THPP@TiO2 NPs (Figure 5C).These results indicated that the commercial P25 control sample was unable to efficiently respond to visible light (>388 nm) [42,43] due to the intrinsic wide band gap (band gap 3.2 eV) of TiO2.Moreover, hybrid structure and hydrothermal treatment were essential in the exhibited photocatalytic activity, suggesting that the assembly could greatly increase the microstructure-dependent photocatalytic activity.

Structure-Dependent Photocatalysis
Lastly, we investigated the photocatalytic performance and verified the catalytic mechanism of the nanocomposites.The hybrid microstructure of the THPP-TiO 2 NPs was adjusted to the core@shell control samples (THPP@TiO 2 NPs) through a two-step method, i.e., the self-assembly of THPP combined with a TiO 2 shell coating.The TEM image of the etched THPP@TiO 2 NPs showed a hollow TiO 2 shell structure (inset, Figure 5A).Methyl orange (MO) is a representative pollutant often used to investigate the photocatalytic performance of catalysts.Compared to the negligible photocatalytic activity of the commercial titanium dioxide photocatalyst P25, the THPP-TiO 2 -H NPs revealed a higher degradation output, exceeding 96.7% after 75 min (Figure 5B).The photocatalytic hydrogen production capabilities of both THPP-TiO 2 -H NPs and THPP-TiO 2 NPs indicated irradiated time-dependent hydrogen generation, respectively reaching 4.80 and 1.03 mmol/g, which was much higher than that of P25, THPP NPs, THPP + TiO 2 , THPP@TiO 2 -H NPs, and THPP@TiO 2 NPs (Figure 5C).These results indicated that the commercial P25 control sample was unable to efficiently respond to visible light (>388 nm) [42,43] due to the intrinsic wide band gap (band gap 3.2 eV) of TiO 2 .Moreover, hybrid structure and hydrothermal treatment were essential in the exhibited photocatalytic activity, suggesting that the assembly could greatly increase the microstructure-dependent photocatalytic activity.
ally, the holes of THPP and TiO2 NPs were used for MO degradation (Figure 5F).The hybrid structure in THPP-TiO2-H NPs is beneficial to the above process.However, the core@shell structure in THPP@TiO2-H NPs is unable to efficiently transfer the electron to TiO2 due to the harmful recombination of electrons and holes, inducing low catalytic efficiency.

Materials
THPP, TSPP, TCPP, sodium dodecyl sulfate (SDS), and titanium diisopropoxide bis(acetylacetonate) (TAA, 75 wt.% in isopropanol) were purchased from Sigma-Aldrich (St. Louis, MO, USA).Sodium hydroxide (NaOH, 1 M) standard solution was obtained from Acros Organics (Geel, Belgium).Tetrahydrofuran (THF) was obtained from Tianjin Kermel Chemical Reagent Co., Ltd.(Tianjin, China).Commercial titanium dioxide (P25) was purchased from Degussa Co., Essen, Germany.All chemicals were used as received without further purification.X-ray powder diffraction (XRD) patterns were obtained To further explore the photogenerated electron-transport process of catalysts in the photocatalytic process, the band energies of TiO 2 and THPP NPs were investigated.According to the UV-Vis diffuse reflectance spectra, the energy gaps (E g ) of the THPP assemblies and TiO 2 NPs were calculated to be 1.56 and 3.10 eV (Figure 5D).The Mott-Schottky curve showed that the conduction band (E CB ) positions of the THPP assemblies and TiO 2 NPs were −0.70 and −0.35 eV vs. NHE (Figure 5E).According to the equation E g = E VB − E CB , the E VB values of the THPP assemblies and TiO 2 NPs were calculated to be 0.86 and 2.75 eV.The staggered band structures between THPP and TiO 2 facilitate charge separation in the THPP-TiO 2 -H NPs.Under visible light irradiation, the electron-hole pairs are simultaneously generated at both THPP and TiO 2 .The photogenerated electrons at the conduction band of THPP would spontaneously transfer to the TiO 2 , with a lower conduction band potential.The transfer of the photogenerated carriers effectively accelerates the separation of photogenerated electron-hole pairs of THPP [44], yielding higher hydrogen production.Meanwhile, the surface Pt cocatalyst [38] facilitated photoelectron transport from TiO 2 to H 2 O by inhibiting the undesired recombination of photo-generated electron-hole pairs on TiO 2 , improving the utilization efficiency of visible light.Additionally, the holes of THPP and TiO 2 NPs were used for MO degradation (Figure 5F).The hybrid structure in THPP-TiO 2 -H NPs is beneficial to the above process.However, the core@shell structure in THPP@TiO 2 -H NPs is unable to efficiently transfer the electron to TiO 2 due to the harmful recombination of electrons and holes, inducing low catalytic efficiency.

Preparation of THPP Nanoparticles
The self-assembly method was used to prepare THPP NPs.Briefly, 1 mg of THPP powder was dissolved in 1 mL of tetrahydrofuran (THF) solvent under sonication for 5 min.Then, the mixture was injected into 25 mL of water and further grown at 25 • C for 24 h.

Preparation of THPP-TiO 2 Nanocomposites
To decrease the size of the nanocomposites, the SDS surfactant was introduced into the above system.Briefly, 50 mg of THPP powder and 250 µL of TAA were dissolved in 50 mL of THF and stirred for 0.5 h at 25 • C. Subsequently, the solution was quickly added to 500 mL of SDS aqueous solution (0.01 M) under magnetic stirring, followed by magnetic stirring for 24 h.The TiO 2 sample was prepared according to the same procedure but with the removal of THPP.The THPP-TiO 2 NPs with large sizes were prepared according to the same procedure but lack of SDS.The pure TiO 2 sample was prepared following the same procedure without adding THPP.

Preparation of THPP@TiO 2 Nanocomposites
The THPP@TiO 2 nanocomposites were prepared using a two-step method.Briefly, 1 mg of THPP power was dissolved in 1 mL of THF solvent under sonication for 5 min.The mixture was injected into 25 mL of SDS aqueous solution (0.01 M) and further grown at 25 • C for 24 h.Subsequently, 5 µL of TAA dispersed in 10 µL of THF solvent was added to the solution.The reaction system was allowed to further grow for 24 h and acquire the THPP@TiO 2 NPs after centrifugation at 8000 rpm for 10 min and washing two times with water.

Hydrothermal Treatment
To increase the crystallinity, the TiO 2 NPs, THPP@TiO 2 , and small-size THPP-TiO 2 nanocomposites required a hydrothermal process.First, the resulting NPs solution was centrifuged at 8000 rpm for 10 min to remove excess reactants.Then, the precipitate was redispersed in water and transferred to a reactor containing a Teflon liner.Lastly, the reactors were sent to the oven and carried a hydrothermal treatment at 180 • C for 16 h.The ultimate products were collected using a centrifugation procedure at 8000 rpm for 10 min.

Photocatalytic Methyl Orange (MO) Decomposition
To test the photocatalytic performance, 6 mg of various nanocomposites or P25 were added to 3 mL of methyl orange (MO, 20 mg/L) dye aqueous solution, which was further stirred for 30 min in a dark environment.Then, the mixture was sent to the irradiation of visible light (PLS-SXE300/300UV, Beijing Perfectlight Technology Co., Ltd., 300 W, wavelength: UV cutoff filter > 400 nm).Subsequently, 400 µL of solution was taken out every 15 min and performed five times.These tested solutions were centrifuged at 8000 rpm for 10 min.The absorption of the supernatant was measured.

Photocatalytic Hydrogen Generation
The photocatalytic hydrogen generation assay was administrated in 50 mL of aqueous solution (pH = 8.1-8.2 adjusted with HCl) containing 5 mL of triethanolamine (TEOA, 10 vol%) sacrificial reagent and 41 µL of potassium chloroplatinate (5 mM).Then, 2 mg of tested samples were respectively added to various parallel solutions to harvest the hydrogen generation under visible light (UV cutoff filter > 400 nm) irradiation and tested using gas chromatography.

Mott-Schottky Tests
The Mott-Schottky plots and photocurrent response were measured on an electrochemical workstation (Autolab, Utrecht, The Netherlands) equipped with a standard three-electrode system.Briefly, the electrolyte solution is Na 2 SO 4 (0.5 M) aqueous solution, which was purged with N 2 to remove air before utilization.The working electrode was prepared by coating the catalysts onto the Pt/C electrode.The reference and counter electrode respectively adopted the Ag/AgCl electrode and platinum plate.

Conclusions
In this paper, we controllably established two microstructural regulation modes to explore the structure-dependent photocatalytic performance.In the co-assembly process, the benzene hydroxyl group acquired optimal morphology and hybrid uniformity via the strong interaction between the hydroxy groups of THPP and TiO 2 .The introduction of the SDS template efficiently limited the growth of the resulting nanocomposites even in large-scale applications.Compared to the core@shell structure of the THPP@TiO 2 nanocomposites, the THPP-TiO 2 nanocomposites with hybrid microstructure exhibited the close contact of porphyrin with TiO 2 and abundant catalytic sites, which promoted photogenerated electron-hole separation.Anatase TiO 2 accelerated the transport efficiency of photoelectrons to Pt on the surface, leading to high MO decomposition efficiency and photocatalytic hydrogen production.The ingenious assembly of structure-regulated nanocatalysts provides insight into the fabrication of advanced photocatalysts and photocatalytic activity modulation.

Figure 1 .
Figure 1.Schematic representation of the co-assembly process and two nanocomposites (THPP-TiO2 NPs and THPP@TiO2 NPs) with various microstructures.The blue and purple represent TiO2 and THPP species.

Figure 1 .
Figure 1.Schematic representation of the co-assembly process and two nanocomposites (THPP-TiO 2 NPs and THPP@TiO 2 NPs) with various microstructures.The blue and purple represent TiO 2 and THPP species.

Molecules 2024 , 10 Figure 1 .
Figure 1.Schematic representation of the co-assembly process and two nanocomposites (THPP-TiO2 NPs and THPP@TiO2 NPs) with various microstructures.The blue and purple represent TiO2 and THPP species.

Figure 3 .
Figure 3. (A) Transmission electron microscopy (TEM) images of THPP-TiO2 NPs prepared without SDS at different magnifications; (B) UV-vis absorption spectra of the indicated samples; (C) element mapping of large-size THPP-TiO2 NPs; (D,E) TEM images of small-size THPP-TiO2 NPs before and after etching with NaOH (1 M), which was prepared assisted by SDS; (F) magnifying TEM images of THPP-TiO2 NPs before (top) and after (down) etching with NaOH; (G) N2 adsorption-desorption isotherms of small-size THPP-TiO2 NPs before and after etching.

Figure 3 .
Figure 3. (A) Transmission electron microscopy (TEM) images of THPP-TiO 2 NPs prepared without SDS at different magnifications; (B) UV-vis absorption spectra of the indicated samples; (C) element mapping of large-size THPP-TiO 2 NPs; (D,E) TEM images of small-size THPP-TiO 2 NPs before and after etching with NaOH (1 M), which was prepared assisted by SDS; (F) magnifying TEM images of THPP-TiO 2 NPs before (top) and after (down) etching with NaOH; (G) N 2 adsorption-desorption isotherms of small-size THPP-TiO 2 NPs before and after etching.

Figure 5 .
Figure 5. (A) TEM images of THPP@TiO2 nanospheres before and after (inset image) etching with NaOH (1 M); (B) relative degradation rate of MO by the indicated treatments.The mass of the catalysts was 6 mg in each catalytic system.The THPP + TiO2 is a mixture of THPP NPs (3.75 mg) of TiO2 NPs (2.25 mg).C0 is the initial concentration of MO; (C) photocatalytic hydrogen production of the indicated samples (2 mg) as a function of irradiation time.The THPP + TiO2 is the mixture of THPP NPs (1.25 mg) of TiO2 NPs (0.75 mg).The volume fraction of the triethanolamine sacrificial reagent (triethanolamine, TEOA) and the mass percent of the Pt co-catalyst were 10 vol % and 2%, respectively.(D) Tauc plots of (αhv) 2 vs. Eg; (E) Mott-Schottky plots of THPP NPs and TiO2 NPs; (F) description of photocatalytic mechanism in THPP-TiO2-H NPs.

Figure 5 .
Figure 5. (A) TEM images of THPP@TiO 2 nanospheres before and after (inset image) etching with NaOH (1 M); (B) relative degradation rate of MO by the indicated treatments.The mass of the catalysts was 6 mg in each catalytic system.The THPP + TiO 2 is a mixture of THPP NPs (3.75 mg) of TiO 2 NPs (2.25 mg).C 0 is the initial concentration of MO; (C) photocatalytic hydrogen production of the indicated samples (2 mg) as a function of irradiation time.The THPP + TiO 2 is the mixture of THPP NPs (1.25 mg) of TiO 2 NPs (0.75 mg).The volume fraction of the triethanolamine sacrificial reagent (triethanolamine, TEOA) and the mass percent of the Pt co-catalyst were 10 vol % and 2%, respectively.(D) Tauc plots of (αhv) 2 vs. E g ; (E) Mott-Schottky plots of THPP NPs and TiO 2 NPs; (F) description of photocatalytic mechanism in THPP-TiO 2 -H NPs.