Janus nanostructures for heterogeneous photocatalysis

When Fujishima and Honda discovered the photo-assisted water splitting properties of titania anodes in 1972, it marked the beginning of a new era in the field of photo-catalysis.¹ This was quickly followed by Allen Bard's approach of a short-circuited photo-electrochemical cell on a particle.² Bard's concept laid the foundation for modern heterogeneous photocatalytic systems, and the field has since remained true to its original concept.³ 

Traditional photocatalysts are semiconducting materials with a band-gap of 2.5 eV to 3.4 eV (corresponding to visible to ultraviolet spectra).⁴ Photons which are incident on a material with an energy greater than, or equal to, its band gap are converted to electrons in the conduction band (cb) and holes in the valence band (vb), respectively. These photogenerated charge carriers can then migrate to the surface to participate in electrochemical reduction and oxidation (REDOX) reactions.⁴ A single phase semiconductor often suffers from several shortcomings including low quantum efficiency, charge-recombination, and chemical back-reactions.^5,6 It is generally accepted that no single material is able to overcome all of these limitations.^7–9 Hence, combinations of materials, using noble metals or co-catalysts, are often employed to increase the net photocatalytic efficiency and to provide favourable sites for REDOX reactions to occur.^8,10 These efforts can be broadly categorised as either a heterojunction (Schottky)^11,12 or a localised surface plasmon resonance (LSPR, plasmonic)¹³ photocatalysis. Furthermore, since heterogeneous catalysis is primarily a surface phenomenon, significant effort has been made to tailor the shape, size, composition, and functionality of catalyst particles.^14,15 Many of the reported techniques are focussed towards the fabrication of nanoparticles with isotropic surface properties; this is similar to Bard's original approach using short-circuited nano-sized photoelectrochemical cells.² However, many advanced applications demand specialized properties, which requires the engineering of particle shape, surface, and morphology to produce suitable reaction sites. Hence, a number of new and exotic nano-materials have been reported¹⁶ and, in this regard, Janus particles (JPs) are gaining increasing attention.

In 1988, Casagrande and Veyssie introduced Janus particles which exhibited distinct surface properties.¹⁷ Initially, these particles were intended to stabilize emulsions owing to their readily tailored amphiphilic surface properties.¹⁸ This was further advanced by Sir Pierre-Gilles de Gennes in his Nobel lecture (1992), which highlighted their anisotropic physical and chemical behaviour.¹⁹ Since then, Janus nanoparticles have attracted interest for a range of potential applications such as water-repellent textiles, sensors, nanomotors, optical sensing, stabilisation of emulsions, magnetic field imaging, and optical sensing devices.^20,21 Recently, heterogeneous photocatalysis have also considered Janus particles and initial studies have reported promising results, which will be discussed in the subsequent sections.

This paper will build upon available reviews concerning the fabrication, assembly, and application of Janus particles.^22–24 In addition, the field of heterogeneous photocatalysis has been reviewed in detail, including some recent articles.^4,6,7,25,26 Hence, it becomes important to distinguish this article from existing reviews and highlight its relevance and importance. Most reviews on Janus particles deal with their broader classification with only a brief overview of their various applications. In addition, reviews in the field of heterogeneous (photo)catalysis attempt to broadly cover recent advances in this field, which often limits a detailed discussion. This translates into grouping of catalyst species by composition rather than structure and the advantages of a particular morphology/arrangement remain undocumented. Given the promising results recently reported in this field of Janus particles for photocatalytic applications, a focussed and an updated review in this field is timely. This field is still in its infancy and carries significant potential for growth and development. Therefore, we aim here to provide a review of the recent achievements of Janus particles with regard to photocatalysis and environmental applications. This article is also intended to draw attention and highlight the capabilities of these relatively simple, yet intriguing nano-structures, so that this field may attract accelerated growth.

Photocatalysis implies photon-assisted generation of catalytically active species that are capable of performing reduction and oxidation reactions. In photocatalysis, when a photon with energy greater or equal to the band gap (E_g) of the semiconductor interacts with the material, it initiates non-equilibrium photophysical and photochemical processes.²⁷ The photon excites an electron from the highest occupied molecular orbital to lowest unoccupied molecular orbital. The process lasts for a few femtoseconds and is followed by the relaxation of the generated holes and electrons to the top and bottom of vb and cb, respectively, on a similar time-scale,^27,28

Depending on the band positioning and band bending configuration, these charge carriers can migrate to the surface of the catalyst and initiate REDOX reactions. The migration itself could be completed on a scale of microseconds, while the reactions themselves are much slower (millisecond timescale).²⁷ In an aqueous environment, positive holes can oxidize adsorbed water at the surface to produce hydroxyl radicals (OH^•), which are extremely powerful oxidants.²⁹ The hydroxyl radicals, more powerful than ozone, can subsequently oxidize organic species producing mineral salts, CO₂ (carbon dioxide) and H₂O (water)²⁹ 

Equations (2)–(8) summarise the main REDOX and the subsequent chemical reactions that take place at the surface of a photocatalyst. Electrons in the conduction band can be rapidly trapped by molecular oxygen absorbed on the surface, which is reduced to form a superoxide radical anion (⁠$O2•−$⁠). This superoxide anion can further react with (⁠$H+$⁠) to generate a hydroperoxide radical (⁠$OOH•$⁠). These reactive oxygen species (ROS) can also contribute to the oxidative pathways, such as the degradation of a pollutant, see Eq. (3). Thus, both the photocatalytic oxidation and the reduction reactions generate powerful oxidising agents for mineralization of pollutants that are prone to oxidation in an aqueous medium. These ROS can also be employed to produce a variety of desirable reactions such as oxidation of alcohols or organic synthesis, which can be powered through sunlight.

However, not all incident photons result in the generation of a charged pair. A majority of the generated charge pairs could either be trapped as ions or may recombine to generate heat (phonon) in the process. Further, the semiconductor-photon interaction is dependent on a number of factors including the average cross-section and surface area of the particle as well as the wavelength of incident light. The absorption coefficient $α(E)$ of an idealized semiconductor is given by²⁷ 

or

where $α0$ indicates the intrinsic absorption of the semiconductor and $E$ is the energy of the incident photon. Equations (9) and (10) indicate that photons with higher wavelength are absorbed closer to the surface. While visible light (400–800 nm) can penetrate much deeper into the material (up to a few microns) before generating a charge-pair. Hence, care must be taken while selecting the morphology of the designated catalyst. A lower particle size increases the surface area thereby increasing the number of incident photons and reducing the migration distance of electron and holes. However, a larger cross-section improves the photon to charge pair conversion ratio.

To design an efficient and stable photocatalyst for utilization of solar energy, several critical requirements must be satisfied. First, the semiconductor responsible for light harvesting must possess a band gap large enough to provide energetic electrons, so that E_g ≫ 1.23 eV, and typically E_g > 2.0 eV.^8,25 However, the band gap must also be sufficiently small to allow for efficient absorption overlap with the solar spectrum, so that E_g < 3.0 eV.^8,25 The selection of an appropriate band gap is a trade-off between the low band gap (lower REDOX potential, high absorption) and the higher band gap (higher REDOX potential, reduced absorption). For example, to efficiently generate •OH radicals an E_g of ≥2.7 eV is required. Furthermore, suitable band positions located above and below the hydrogen and oxygen evolution energy levels are also important to initiate and sustain several reactions, such as water splitting for hydrogen evolution.³⁰ Second, there must be a mechanism to efficiently drive charge separation and the transportation process. Third, the primary semiconductor catalyst should be closely integrated with a selected co-catalyst to allow for efficient utilization of the photogenerated charge carriers and initiate the photochemical reactions. Finally, there should be suitable provisions to protect the catalyst from unwanted galvanic reactions, such as dissolution of components into ions in the solution or plating from the ions in the solution to form surface layers, which can deteriorate or destabilize the material over prolonged periods of use.

Since a single material is unlikely to satisfy all these conditions, co-catalysts, sensitizers, or a combination of the two are often used.⁴ The majority of these heterogeneous catalysts can be broadly classified into the following categories: (a) semiconductor-semiconductor; (b) molecule-semiconductor; (c) nano-hybrid; and (d) metal-semiconductor. Figure 1 shows the classification of these heterogeneous catalysts, which will now be described.

As the name implies, a semiconductor-semiconductor (S-S) catalyst is composed of two different semiconducting materials with different band-gaps but with an overlap of either the cb or vb positions, see Fig. 1(a). This allows the lower band-gap semiconductor to absorb in the visible region and pump energetic charge carriers into its partner. Furthermore, the resulting difference in Fermi energy creates a Schottky junction at the interface which effectively resists charge recombination.^11,15 Finally, depending upon the nature of the REDOX reactions and pollutants, a co-catalytic effect can also be observed. An example of such an arrangement includes titania and nickel/cobalt compounds,³² TiO₂-CdS,³³ TiO₂-ZnO,³⁴ and CdS-CdSe³⁵ among others. Hence, this combination offers advantages which can overcome some of the associated problems described above. However, the lower band gap semiconductors are generally less stable in a corrosive environment and prone to galvanic reactions at its surface, which reduces catalyst lifetime.³⁶ Surface passivation can be used to improve chemical stability but often results in reduced performance;³⁷ hence, the issue of chemical stability still needs to be addressed.³⁸ 

Another approach to acquire surface passivation, photosensitization, and a co-catalyst effect involves binding a co-catalytic molecule to the host semiconductor [see Fig. 1(b)].³⁹ Figure 1(b) shows how the molecular catalyst helps to sensitize the semiconductor towards visible light absorption and allows rapid charge transfer between the surrounding dielectric medium and the depleted charge carriers on the catalyst surface. The molecular catalyst, which can be organic or inorganic in nature, acts in a similar manner to that in a dye-sensitized solar cell.⁴⁰ A good example is chlorophyll, which is responsible for the green pigmentation in plants and aids photosynthesis.⁴⁰ Catalytic molecules are often used to suppress the recombination of photogenerated charge carriers and facilitate rapid charge transfer. Furthermore, this step can also enable protection of the photocatalyst from oxidation, thereby prolonging the operational lifetime.⁴¹ Molecular co-catalysts can be further subdivided into two individual classes depending upon the nature of the interaction between the molecule and the host semiconductor. Type 1 catalysts include an arrangement where the molecule is simply anchored to the semiconductor surface.^42,43 In contrast, Type 2 catalysts are created when there exists a distinct chemical bonding between the molecule and semiconducting phase.⁴⁴ Molecular catalytic systems possess several advantages over traditional heterogeneous materials; these systems are highly selective towards the adsorption/desorption of reactant species and products. This characteristic enables the desired REDOX reaction to proceed at a much faster rate. Furthermore, a better surface distribution requires lower loading percentages as opposed to other catalyst classes.⁴⁵ Finally, these systems have highly reduced REDOX overpotentials which allows them to operate at quantum efficiencies that are superior to bulk or nano-scale systems.^45,46

Molecular systems are often employed as dye sensitizers to enable large bandgap semiconductors to absorb and utilize visible light.⁴⁷ These systems are dependent on rapid electron transfer between the anchored dye molecules and the semiconductor. The dye reacts to visible light by producing electrons which are then injected into the conduction band of the semiconductor [see Fig. 1(b)]. In return, the oxidized dye can be returned to its original state by accepting electrons from the donor species in the reaction. However, this mechanism places a limitation on the speed of REDOX reactions, and the dye is exposed to potentially unwanted galvanic reactions. In some instances, these molecules also act as co-catalysts.^44,48 For example, hydrogenases, including [Ni-Fe] or [Fe-Fe] hydrogenases, have been widely studied as co-catalysts for photogeneration of H₂.^49–51 However, there are limits in terms of the long-term stability of the catalyst or cost-effective production techniques.⁵¹ 

As the name suggests, this category includes multi-component and nano-scale architectures. However, their distinguishing feature is that each component serves a specific purpose in the overall structure. Since a designated shape, geometry, and chemical composition are obtained, several benefits can be achieved simultaneously. These include desired REDOX reactions, efficient photogenerated charge transfer, enhanced catalytic activity, and prolonged stability. Photocatalyst systems based on BiOCl-Au-CdS,³¹ CdS–Au–TiO₂,⁵² and CdS–Pt–TiO₂⁵³ are examples of such heterostructures. The photocatalytic performance of such materials far exceeds that of TiO₂ alone. Integration of multiple semiconductors or co-catalysts with an appropriate architecture allows functionality unattainable within a single material. For example, Fig. 1(c) shows the BiOCl-Au-CdS system, where BiOCl is a large band-gap catalyst that is sensitized by the lower band-gap CdS through an Au junction which facilitates rapid charge transfer and prevents the photodegradation of CdS in the process. This methodology has been used to address some critical and fundamental challenges in the field of photocatalysis. However, since the fabrication of nano-structures is often a multi-step and bottom-up approach, the difficulty and cost of fabrication scale with complexity and number of participating heterostructures. If the economics of fabrication can be resolved, this technique is potentially lucrative for wide-scale applications.

Noble or rare-earth metals, which are bonded with traditional semiconductors, have been reported to provide several benefits including band-bending, surface plasmonic resonance, Schottky junction, and enhanced scattering and absorption.^49,54 As described by Zhang et al., the metal can be in solo, embedded, encapsulated, or isolated form, as in Fig. 1(d).¹³ Noble metals (Au, Ag, and Pt) are often used due to their inertness in chemically aggressive environments. This is important as the sole purpose of these photocatalysts is to retain the metallic characteristic, whereas oxidation of the element will give rise to semiconductor-semiconductor type catalyst, as in Fig. 1(a).

When the metal clusters form a close-contact with the host semiconductor, it gives rise to a Schottky junction.⁵⁵ However, unlike traditional S-S bonding, this junction allows for the separation of photogenerated charge carriers by acting as an electron reservoir.^55,56 This lowers the probability of recombination, thereby allowing more charged species to migrate to the surface and initiate catalysis. Second, these metals often have their own co-catalytic properties.^57,58 Hence, their incorporation also attempts to reduce the overpotential for surface electrochemical reactions. Finally, the abundant availability of freely conducting electrons at the metal/water dielectric interface allows these metal clusters to display surface plasmon resonances (SPR).⁵⁹ Localized SPR, or LSPR, is observed when the interacting light has a wavelength comparable to the cluster size.¹³ This generates a resonance in the surface electrons of the metal which results in a very high energy transfer. These “hot-electrons” can overcome the forbidden zone of the Schottky junction and can be used for surface reduction reactions. LSPR and plasmonic catalysis are a growing field and the reader is directed to these articles for further information.^13,59,60 In brief, the most commonly employed nanoclusters are based on Ag, Au, and Pt, while the semiconductors used for evaluation have included TiO₂, CdS, and Fe₂O₃.^13,59 Several morphologies can be obtained, and a classification involves either asymmetric, axis-symmetric, or core-shell structures. However, the use of noble metals and wet-chemistry techniques, often employed to fabricate these catalysts, makes them expensive and places them at a commercial disadvantage. This can be overcome by simplifying the fabrication process and shifting the focus to employ more earth-abundant elements, such as copper.

Barring some exceptional cases, the above categories can be used to classify the majority of heterogeneous photocatalyst materials. This classification will also be applied to various Janus particles used for photocatalytic applications, which will now be described in Sec. III.

As previously mentioned, the term Janus was first used by Casagrande and co-workers to describe the amphiphilic properties of small (50 μm to 90 μm) spherical silica glass beads.¹⁸ Since this initial work, the term Janus has been used to broadly define any bi-phasic particulate system with a clear anisotropic distinction of each phase. However, the approach itself came into the limelight when Sir Pierre-Gilles de Gennes, in his Nobel lecture, highlighted their anisotropic physical and chemical properties.¹⁹ The Janus morphology has, so far, been reported in eight different configurations as shown in Fig. 2.^21,22,61

Over the last two decades, the Janus morphology has been employed to create nano-materials in a number of different fields,²⁴ and one emerging application for Janus particles is photocatalysis. In a heterogeneous catalyst with two distinct phases, the secondary phase is responsible for improving either charge-separation or providing co-catalysis, or both. When such phases are arranged in a clear an-isotropic structure, as in Fig. 2, they can be described as Janus particles. Recently, published literature indicates that a Janus morphology can be advantageous for some photocatalytic applications, and Janus particles have been reported to be almost an order of magnitude superior than a conventional nanostructure for the similar amount of material and experimental conditions.^62,63 However, a common pattern between studies and observations dealing with different Janus morphologies and compositions has not been reported. This can be credited to the lack of an exhaustive comparative analysis of these studies. Therefore, the main objective of this review is to highlight their potential and areas of future research.

Since Janus photocatalytic materials form a sub-category of the heterogeneous photocatalysts, they can be categorized using the same classifications as previously described in Sec. II. An overview of the published literature in this field indicates that many are based on a noble metal/semiconductor junction. This is followed, in terms of popularity, by semiconductor/semiconductor type particles and others which cannot be placed under either category. Hence, the order of discussion follows the same pattern as in Sec. II.

The metal part in the metal/semiconductor (M-S) junction is composed of either Au, Ag, or Pt owing to their excellent chemical stability and resistance to oxide formation in aqueous media. However, the field of Janus catalysts is primarily dominated by Au-TiO₂ nanostructures. In fact, of all the articles using the term “Janus” to describe their catalyst morphology, a large number are based on spherical Au and titania structures.

Pradhan et al. reported the first Au-TiO₂ Janus catalyst in 2009.⁶⁴ The Au-TiO₂ heterodimers were fabricated by partially passivating the surface of Janus Au nanoparticles (∼2 nm) using hexanethiolates. The hydrophilic surface was then utilized to grow TiO₂ particles (∼6 nm) using a sol-gel process with titanium butoxide as a precursor. The results, when compared to pristine TiO₂ particles, revealed that the presence of Au nanoparticles facilitated charge separation by acting as an electron reservoir. These particles were then employed for catalytic oxidation of methanol. In the presence of Janus catalyst, almost linear conversion of alcohol could be attained with time, while pristine TiO₂ did not display any activity.

Fu et al. also examined the photocatalysis of Au-TiO₂ Janus particles, prepared by using a UV-treated copolymer template.⁶⁵ In this study, the catalysts were fabricated by a two-step process involving a polystyrene-block-poly (ethylene oxide) (PS-b-PEO) copolymer template. The precursor for spin-coating was obtained by preparing a (1%) mixture of PS-b-PEO and HAuCl₄ in toluene. Titanium isopropoxide was employed as a source for the TiO₂ and the mixture was spin coated on a silicon wafer to yield a monolayer of thickness ∼10 nm. This film was then treated with deep UV (48 h) to slowly etch away the block-copolymer while simultaneously reducing chloroauric acid to Au nanoparticles. The Janus particles were found to be marginally better at dye degradation (∼10%) when compared to bare TiO₂ and conventional Au-TiO₂. The authors suggested that the improved performance of the Janus particles could be attributed to the nature of the heterointerface between Au and TiO₂; this resulted in an increased lifetime of charge carriers.

The incorporation of Au nanoparticles can also impart excellent co-catalytic properties and surface plasmon effects. This topic was investigated by Seh et al.,⁶² who fabricated Au-TiO₂ Janus particles with Au nanoparticles of size (diameter) 30–70 nm. Fabrication of the Janus catalysts required Au nanoparticles (30–70 nm) to be suspended in an aqueous solution of pH 9–10. A known quantity of this solution was used for controlled hydrolysis of titanium diisopropoxide bis(acetylacetonate) in isopropyl alcohol. Figure 3(a) shows a transmission electron microscopy (TEM) image of Au_50 nm-TiO₂ Janus particles. Other morphologies including bare gold particles, pristine TiO₂, a core@shell structure, and a 5 nm Au-TiO₂ composite were also prepared. It is to be noted that the TiO₂ particles in this study were amorphous in nature. The optical extinction spectra indicated a strong absorption in the vicinity of ∼550 nm, Fig. 3(b). Figures 3(c)–3(e) correspond to the plasmonic near field maps for Au₅₀ Janus, core-shell, and bare nanoparticles, respectively.

When employed for hydrogen production using a sacrificial electron donor (isopropanol, IPA), it was observed that the Janus particles provided the best catalytic activity, and their effectiveness increased with an increase in Au nanoparticle (np) diameter. The enhanced performance of the Janus particles was attributed to the stronger localization of the plasmonic near-fields which allowed the Janus morphology to absorb 1.75 times more energy in the visible spectrum.

Ding et al. examined the benefits of M-S heterojunction in their study of Au/TiO₂ heterodimers, for visible light photocatalysis.⁶⁶ The particles were prepared using commercial TiO₂ (P25 Degussa) and colloidal Au nanoparticles. The use of a ligand exchange process allowed easy control of particle size owing to low processing temperatures, which facilitated a close metal-semiconductor contact, giving rise to a Schottky junction. 3-Mercaptopropionic acid was utilized as an intermediate ligand to facilitate bonding. The thiol group possesses a large affinity for Au, while the carboxylic group aids bonding with TiO₂. The process was followed by vacuum evaporation and processing at 200 °C to remove the ligand and aid in junction formation. The fabricated particles were tested for dye degradation and hydrogen evolution experiments. Results indicated that only the Janus heterodimers displayed significant catalytic activity in visible light. The high performance of ligand-prepared Janus particles over other catalysts was attributed to several factors including: (a) the formation of a heterojunction, (b) strong LSPR at λ ∼ 500 nm, and (c) improved absorption through amplification of near electric field. The authors also reported that conforming to a Janus morphology also prevented a back-reaction due to clear separation of sites for REDOX reactions.

The above studies highlight the multiple benefits of incorporating Au nanoparticles onto traditional TiO₂. However, TiO₂ itself can be modified for efficient catalysis by doping or band engineering.⁶⁷ In this regard, Liu et al. prepared Janus particles based on Au nanocage (AuNC) and C–TiO₂ (Carbon-TiO₂) using a microemulsion process, as shown in Fig. 4.⁶⁸ The asymmetric Janus particles were prepared by first fabricating Ag nanocubes through a template assisted sol-gel technique, which was then treated with HAuCl₄. The as-prepared Au nanocages were bonded to C-TiO₂, P25 Degussa, and a core@shell structure by heating a microemulsion based on cyclohexane, ethanol, and sodium dodecylsulfate, containing the necessary catalyst and Au nanocage. TEM images of the prepared catalysts are shown in Figs. 4(a)–4(d). The Janus particles were observed to be twice as efficient at aerobic oxidation of isopropyl alcohol as the core@shell structure and almost 300% more efficient than pristine C-TiO₂. The superior catalytic activity of the Janus structure was attributed to the preferential localization of the plasmonic near-fields close to the M-S junction. A finite-difference time-domain (FDTD) method allowed assessment of the electromagnetic field distribution over the Janus AuNC/(C–TiO₂), as shown in Fig. 4(e). The presence of higher internal scattering in the Au nanocage allowed it to absorb more power, thereby generating a larger number of “hot-spots” and increased activity.

Similar conclusions were also drawn by Cao et al.⁶⁹ who prepared a variety of Au/TiO₂ nanocomposites, including Janus particles, to evaluate their visible light activity. Particles were prepared using an inverse miniemulsion technique with titanium ethoxide acting as a precursor for TiO₂ and HAuCl₄ as a source of Au. The catalytic activity of different titania phases was compared for a similar Au concentration. The improved performance of Janus particles was credited to localization of plasmonic near fields. Among the Janus structures, the amorphous Au/TiO₂ particles displayed the highest visible light activity followed by anatase and rutile phase, respectively. The higher activity of the amorphous phase was credited to the presence of localized electronic states and a larger surface area, where both attributes allowed the material to absorb more power in the visible spectrum.

Particle geometry also plays an important role in dictating the energy absorption characteristics. This has been elucidated by Zhang et al. who employed Whispering Gallery Mode (WGM) resonances in Au/TiO₂ composites.⁶³ A finite element simulation was performed using an isolated (60 nm) Au particle on the surface of a much larger TiO₂ particle. The analysis revealed that TiO₂ particles below 200 nm and continuous substrates were unable to display WGM resonance. However, plasmonic absorption was greatly enhanced for TiO₂ in the size range of 200–600 nm, with absorption increasing with particle diameter. This was associated with the enhanced number of sub-surface resonance modes available with increasing particle size. The results indicate that a Janus morphology with an Au sphere half-buried in the TiO₂ matrix would exhibit the highest activity, and this was confirmed by experimental analysis. While the Au nanoparticle size did not vary (60 nm), it was observed that by varying TiO₂ particle size the absorption could be expanded over the whole visible spectrum (400–800 nm). It was also reported that owing to significant enhancement in the local electric field, the Janus configuration particularly benefits from a LSPR effect, with plasmonic absorbance increasing by a factor of 40.

Another attempt in this regard was made by Sharma et al. who investigated the effect of clustered Au-TiO₂ Janus nano-assemblies.⁷⁰ Two configurations, namely, nanospheres (∼8 nm dia.) and nanorods (20 nm length, aspect ratio 3.6 nm) were fabricated, which were then bonded to TiO₂ particles to form snowman-like Janus particles. Nanospheres were fabricated using a seed solution by reducing 20 ml (25 mM) HAuCl₄·3H₂O and trisodium citrate (25 mM) with 0.6 ml (0.1M) NaBH₄. The growth solution was prepared by dissolving 25 mM of HAuCl₄ and 6 gm cetyltrimethylammonium bromide in 200 ml water. 9 ml of growth solution was mixed with 0.05 ml (0.1M) ascorbic acid to which 1 ml of seed solution was added resulting in the formation of nanospheres. The nanorods were prepared in a similar manner in the presence of cetyltrimethyl ammonium bromide (8.33 ml, 180 mM), NaCl (225 μl, 0.1M), and AgNO₃ (180 μl, 0.01M), without a seed solution. Thereafter, Janus particles were prepared by mixing 25 ml aqueous solution of required nanostructure with 1 ml hydroxypropyl cellulose (1 wt. %). The resulting solution was suspended in isopropanol (IPA) and homogenised using ammonia. Finally, 10 ml (10 mM in IPA) titanium diisopropoxide bis(acetylacetonate) was added to form the Janus structures. These catalysts were tested for degradation of methylene blue (MB) and carbendazim in the presence of UV and visible light. In both cases, Janus particles performed better under visible illumination than UV irradiation, with Au nanorods performing slightly better than nanospheres. The authors reasoned that the presence of UV limits the function of Au in acting as an electron reservoir, with TiO₂ being responsible for majority charge carrier generation. However, this role is reversed in visible light where Au nanoparticles act as light harvesting antennas though LSPR and injecting hot electrons into TiO₂.

Deviating from conventional wet-synthesis, Byeon and Kim reported an ambient spark discharge technique for producing ultrafine Au-TiO₂ heterodimers.⁷¹ The particles were fabricated by generating a spark between a Ti anode and Au cathode (3 mm diameter, 100 mm length). The high gas temperature inside the spark channel induced sublimation and subsequent rapid cooling of the particles. The setup specifications were as follows: resistance, 0.5 MΩ; capacitance, 1.0 nF; loading current, 2.6 mA; applied voltage, 3.4 kV; and frequency, 1020 Hz with a gas flow rate of 3.2 l min⁻¹. The resulting particles (38 nm across with an embedded Au particle of 2 nm) were tested for visible light assisted aerobic oxidation of alcohol and photothermal decomposition of CO. The enhanced catalytic activity of the Janus particles was attributed to LSPR and efficient charge separation at the Au-TiO₂ junction.

Another method to fabricate a large array of Au-TiO₂ Janus particles was explored by Wen et al. using a template-guided approach.⁷² Initially, the templates were created on polished Al foil using a Ni imprinted mould (400 nm pitch). This was followed by atomic layer deposition of TiO₂ using TiCl₄ and deionized (DI) water as precursors, with a growth rate of approximately 0.05 nm per cycle (300 °C). The barrier layer and any over growth of TiO₂ nanotubes were removed using Ar ion milling followed by the deposition of a thin supporting layer of Ti/Au (10/30 nm) using physical vapor deposition. This was followed by electrodeposition of a thick Ni supporting layer, consequent over-etching of the secondary pores, and ultimately electrodeposition of Au nanorods. Another sample containing a thin Pt interface was also prepared. These particles were employed for photocatalytic hydrogen production and dye degradation. Janus particles were reported to be three time more effective at hydrogen production than control TiO₂ particles under solar illumination, and the activity could be further improved by 161% through incorporation of a Pt co-catalyst. The high catalytic performance of Janus particles was attributed to the strong plasmon resonance energy transfer and efficient charge separation at the metal-semiconductor interface.

Among the M-S type Janus particles, there have been limited attempts towards employing Ag as a co-catalyst. The first study in this regard was reported by Jiang et al. who produced Ag–Ag₂S Janus particles using a one-pot thermal decomposition method.⁷³ Silver diethyldithiocarbamate was injected into hot oleylamine amine solution. Thermal decomposition of the resulting mixture was carried out at 180 °C for 3 h under a N₂ atmosphere. The initial Ag nuclei were formed at 90 °C over which the sulphide phase could grow. The effect of temperature and processing parameters on particle shape and formation was also reported. The morphology of the as-formed Ag–Ag₂S was characterized using scanning tunnelling electron microscopy (STEM), XRD, and energy dispersive X-ray analysis (EDX) as shown in Figs. 5(a)–5(c), respectively. The resulting Janus particles (15 nm–25 nm) with an eggplant shape were coupled with commercial P25 Degussa and tested for dye degradation. The higher catalytic activity of Janus particles was attributed to rapid charge transfer between the titania and Ag containing phases.

Chen et al. also reported Ag-TiO₂ based Janus particles.⁷⁴ The particles were prepared using an emulsion swelling assisted protrusion synthesis. First, Janus composite particles of titania–poly(vinylbenzyl chloride-divinylbenzene) (VBC-DVB) were produced, where the lobe size of VBC-DVB was used to tune the TiO₂ particle size (∼260 nm). These particles were then reacted with a silver ammonia solution at 70 °C to obtain the desired Janus particles. The article describes in detail the reactions involved, including the properties of the intermediate products, and presents a thorough characterization of the process. However, it does not report on the catalytic activity. Given the fact that TiO₂ has already been established as a good semiconductor photocatalyst and the excellent plasmonic and pesticidal nature of Ag nanoparticles,^75–77 it would be of interest to explore potential applications.

Recently, Mohanta and co-workers have reported a biogenic synthesis route for preparing Ag-SnO₂ Janus particles using plant (Iresine herbstii leaf) extract.⁷⁸ The particles were prepared through sol-gel synthesis route. The required amounts of SnCl₄ 5H₂O (3.0 g m) and AgNO₃ (0.17 g m) were dissolved in water and added dropwise to the plant extract prepared by boiling fresh Iresine herbstii leaf (2.5 g m) in 50 ml water. The solution was maintained at 70 °C for 2 h and later irradiated with thirty-ten second microwave blasts (1400 W). The resulting residue was washed, recovered, and calcined at 400 °C for 2 h to produce Janus particles. The authors reported that ethanol extract of Iresine herbstii contains many polar phytochemicals which would result in a chelation effect on Sn⁴⁺ and Ag²⁺ ions, thereby regulating the growth of particles. However, no direct evidence or characterization of the plant extract was reported. The as synthesized particles were tested for degradation of Doripenem dye.

Moving away from TiO₂, Xu et al. recently reported on Janus silica nanosheets that support Au particles for the conversion of organic compounds (p-nitrophenol and p-nitroanisole).⁷⁹ Initially, Janus silica sheets were prepared using a sol-gel route. The oil mixture was prepared by melting together tetraethylorthosilicate, amino-propyltrimethoxysilane, and paraffin at 70 °C. This mixture was then dispersed in a solution containing hydrolysed styrene-maleic anhydride, with pH adjusted to 2.5 through addition of HCl. The mixture was held at 70 °C for 12 h after which the Janus hollow spheres were crushed using a colloidal mill to obtain Janus silica nanosheets. These sheets were then re-dispersed in ethanol and HauCl₄, which was reduced into Au nanoparticles using NaBH₄. As a control experiment, biphasic particles [see Fig. 2(e)] were also prepared. It was observed that Janus silica nanosheets were 16 times better than the biphasic interface. The amphiphilic properties, large surface area, and high mass transfer of Janus nanosheet were reported to be responsible for its superior performance and this advantage is apparently lost with isotropic heterodimers.

Yu et al. reported on a combination of CeO₂, an oxide semiconductor, with Au to form Janus nanoparticles.⁸⁰ The particles were prepared using a polymer template assisted growth technique. In brief, individual solutions of polystyrene-block-poly (ethylene oxide), Ce(NO₃)₃, and HauCl₄ were prepared in toluene (0.03 wt. %) and ethanol (0.25M each), respectively. These were mixed together to form a micelles solution, to which NaOH (dissolved in ethanol) was added dropwise and stirred for 1 h. The obtained Janus precipitates were washed and recovered. It was observed that the average diameter of Au and CeO₂ particles were 9.4 nm and 18.2 nm, respectively. Alternately, pure CeO₂ and Au-CeO₂ nanocomposites were also prepared by precipitating in the absence of a polymer template. The as prepared catalysts were calcined at 400 °C for 2 h in a N₂ atmosphere. Contrary to the conventional approach, these catalysts were analysed for dye degradation using ultrasonic waves in the dark. The applied technique is termed sonocatalysis where ultrasonic waves led to cavitation bubbles, thereby generating hot-spots and hydroxyl ions in the process.⁸¹ The high catalytic activity of Janus particles was credited to the formation of a heterointerface between the Au and CeO₂, thereby enhancing surface electron transfer and promoting the generation of free hydroxyl radicals.

Transition metal-oxides are often employed as lower band gap sensitizers, where TiO₂ is the primary catalyst. Compared to (noble) metal-based Janus catalysts, these tend to be chemically stable in a range of acidic and basic media while being economic to produce. Depending upon the transition metal element, additional benefits can also be obtained such as magnetic separation and gas-phase synthesis.

An example was demonstrated by Mou et al.⁸² where the authors employed asymmetric shrinkage between two different precursors for γ-Fe₂O₃ (iron citrate) and TiO₂ (tetra-n-butyl titanate) to create hollow “bowl-like” Janus particles, see Fig. 6. The precursors were atomized through an oppositely charged twin-headed electrospray. A bias voltage of 15 kV was applied to the nozzles under a fixed feed rate (0.1 ml h⁻¹) and an aluminium board, 10 cm away was used as a collector. The resulting particles were calcined at 450 °C for 2 h. The resulting particles were characterized using scanning electron microscopy (SEM), STEM, and XRD, as seen in Fig. 6. The hollow concave bowl-in-bowl structure was attributed to the different drying and shrinkage rate of the precursors during gel formation. EDX analysis revealed the outer and inner faces to be TiO₂ and γ-Fe₂O₃, respectively. The photocatalytic activity of the as-prepared Janus bowls was evaluated against traditional nano-composite and pure γ-Fe₂O₃ for degradation of Rhodamine B. Given the magnetic nature of the catalyst, the powder could be easily collected using magnetic separation. The Janus catalyst was observed to have a degradation rate almost twice as large as single phase (γ-Fe₂O₃) particles, and the presence of a charge separating heterojunction between the TiO₂ and γ-Fe₂O₃ was thought to explain the higher activity of the Janus catalyst.

The same group (Mou et al.) later reported on the effect of operating parameters for different Janus particles, obtained using a “twin-head electrospray” technique.⁸³ Tetrabutyl titanate, cerium nitrate, and cadmium nitrate with SC(NH₂)₂ were used as TiO₂, CeO₂, and CdS sources, respectively. Depending upon the precursors and the post-heat-treatment, different morphologies could be produced. These included “solid snowman,” “hollow bowl,” and “pot-like” structures of CeO₂-TiO₂ and CdS-TiO₂ compositions. It was also proposed that the technique can be easily adapted for large scale production of Janus particles. The resulting powders were evaluated for their hydrogen evolution capacity using sacrificial agents. Compared to traditional CeO₂-TiO₂ composite, the Janus particles displayed an order of magnitude higher hydrogen evolution.

An unconventional S-S type Janus catalyst was reported by Hu et al. with comparable performance to P25 Degussa.⁸⁴ The particles were composed of ZnO/ZnAl₂O₄, with and without the presence of chemisorbed Cr (IV) ions. While the authors employ the term “Janus,” a clear anisotropic distribution of individual oxide phases is not demonstrated. The precursor was prepared by precipitating salts of Zn and Al in a molar ratio of 4:1 and calcination at 500 °C. These precipitates were then used for removal of dissolved Cr(IV) species and recalcined at 900 °C. Calcination allowed bonding of the adsorbed Cr ion to the lattice, thereby, rendering it immobile. The catalyst was characterized as having a band-gap of 3.17 eV indicating its UV active nature, and an agglomerated particle size of ∼1 μm. When irradiated with UV-light, the catalyst showed a similar activity to P25 Degussa while also having an application as a Cr recycling agent. The Cr (IV) containing ZnO/ZnAl₂O₄ can be photo-reduced to Cr (III) after recycling.

As demonstrated, lower band-gap semiconductors are often employed to sensitize TiO₂ to visible light. This can be partially associated with the excellent stability and photoactivity of anatase TiO₂ which is only limited by its high band-gap and surface area. However, a study by Panwar and co-workers demonstrated that through careful design, a high band-gap Janus catalyst of TiO₂-SiO₂ could perform better than commercial P25 Degussa.⁸⁵ The binary oxide Janus particles were prepared through a Pickering emulsion method. Amine functionalized SiO₂ sub-micro spheres (0.3 g m) were dispersed in paraffin wax (3 g m, 75 °C) followed by addition of 30 ml water. After stirring for 1 h, the colloidosomes were separated and dispersed in 20 ml ethanol. This was followed by addition of water (0.202 g m), hydroxyethyl cellulose (0.075 g m), and tetrabutyl orthotitanate (dropwise, in 5 ml ethanol). This mixture was incubated in a shaker at 200 rpm for 3 h at 40 °C. The resulting particles were recovered, calcined at 450 °C and 750 °C, and tested for UV-assisted degradation of Solophenyl green dye. Janus particles calcined at 450 °C were found to be significantly (>6 times) better than P25 Degussa. The higher activity of the Janus particles was credited to the enhanced charge carrier lifetime obtained by transfer of photogenerated holes into the type-II interface at the Ti-O-Si junction.

The same group (Panwar et al.) has recently reported on similar Janus particles for a self-cleaning fabric application.⁸⁶ A piece of cotton fabric was treated with P25 Degussa and Janus (450 °C) TiO₂-SiO₂ particles through padding (pH 2) and exhaustion (pH 7) methods, respectively. The functionalized cotton fabric was stained with Solophenyl green dye and exposed to UV. It was observed that Janus particles demonstrated comparable performance to P25 under lower titania loading, as well as better adhesion and mechanical stability. It was reasoned that the presence of SiO₂ allows better adhesion to the cotton fabric under neutral pH conditions which enables the particles to resist being removed during washing.

The Janus morphology is not limited to simple particulate systems, as demonstrated by Li et al. in their study of a Janus bilayer junction.⁸⁷ The authors constructed 1 T MoS₂ monolayers on Bi₁₂O₁₇ face-ends of a Bi₁₂O₁₇Cl₂ monolayer to create a 2-D Janus (Cl₂)–(Bi₁₂O₁₇)–(MoS₂) junction. Initially, nanosheets of Bi₁₂O₁₇Cl₂ and MoS₂ were prepared using a hydrothermal precipitation. From the bulk nano-sheets, individual monolayers were extracted using lithium intercalation based liquid exfoliation. Thereafter, an oxygen deficient Bi₁₂O₁₇Cl₂ monolayer suspension in water was used as bonding sites for MoS₂ monolayers, thereby yielding the Janus catalyst. In this heterojunction, Bi₁₂O₁₇Cl₂ acts as the primary catalyst intercepting visible light. Given the internal electric field of the monolayer, the photogenerated electrons and holes quickly migrate to the Bi₁₂O₁₇ and Cl₂ ends, respectively, resulting in an extended carrier lifetime of 3446 ns. The Janus bilayer demonstrated an excellent visible light photocatalytic hydrogen evolution corresponding to a quantum efficiency of 36% (420 nm) over a continuous operation (>100 h), indicating the superior performance and stability of the bi-layers.

It is also possible to obtain a Janus-like structure within the domain of a z-scheme through careful fabrication. In this regard, Yuan et al. reported a Janus-like z-scheme for photocatalytic hydrogen evolution.⁸⁸ Their particle was based on roxbyite Cu₇S₄ and γ-MnS. Hexagonal bipyramid roxbyite was fabricated using a modified hot-injection method and was selected as template for the growth of a Janus structure using a cation-exchange process. As-prepared roxbyite crystals were dispersed in toluene and mixed with tri-n-octylphosphine followed by 90 min of stirring. During this process, a mixture of MnCl₂•4H₂O and oleylamine was prepared through sequential introduction under a constant Ar flow. The temperature of the mixture was then raised to 250 °C for 30 min during which a colour change of the solution confirmed the reaction; the temperature was then lowered to 100 °C. To this mixture, the suspension of roxbyite crystals was added and by controlling the reaction time differing ratios of Mn/Cu could be obtained. The high density of Cu vacancies in Cu₇S₄ facilitated ion exchange and LSPR. In addition, incorporation of Mn²⁺ into the Cu₇S₄ crystals allowed the Janus catalyst to absorb over a broad wavelength (λ > 240 nm). This catalyst was then tested for visible light assisted hydrogen production in the presence of sacrificial agent (quantum efficiency η = 18.8%) and pure water (η < 2.5%). The poor performance of the Janus catalyst for overall water splitting was associated with lack of sites for oxygen evolution. After depositing with a MnO_x co-catalyst, this Janus-MnO_x hybrid structure achieved a quantum efficiency of η = 5.5%. This is one of the few studies which has discussed the overall water-splitting ability of the Janus catalyst, and more work is warranted in this area.

Different compounds of the same metal can also be used to create Janus structures as demonstrated by Jo et al.⁸⁹ through their Ag₂O-Ag₂CO₃ particles. The authors prepared these Janus nanocomposites through an ion exchange process followed by in-situ phase transformation. To synthesize the particles, silver nitrate (0.0425 g m) was dissolved in a 50 ml solution of water/ethanol mixture and the solution is heated to 60 °C. To this, an equimolar solution (50 ml) of sodium bicarbonate was added and the temperature was increased to 80 °C. A colour change of the solution indicated reaction completion after which the particles (∼10 nm) were washed and recovered. As a control, phase pure Ag₂O and Ag₂CO₃ particles were also prepared. The synthesized catalyst was then tested for mineralization of Rhodamine B (Rh B) and 4-chlorophenol in visible light. It was observed that Janus particles were several times more effective at dye degradation than the phase pure particles and almost 20 times better than P25 Degussa. An in-depth analysis revealed that suppressed charge recombination at the heterointerface was responsible for the superior catalytic performance.

The earliest record that could be found for nanohybrid Janus particles is reported by Sotiriou and co-workers who fabricated silica-coated Fe₂O₃-Ag Janus particles.⁹⁰ A single step flame spray pyrolysis technique was used for the synthesis of Janus particles with different Ag compositions. Iron(III) acetylacetonate (fixed at 0.5M) and silver acetate, dissolved in 2-ethylhexanoic acid and acetonitrile, were used as precursors. The solution, fed at a rate of 5 ml min⁻¹, was dispersed with O₂ gas (5 l min⁻¹) and ignited by a ring shaped pre-mixed O₂/CH₄ flame. The obtained Janus particles were coated in-flight with SiO₂ (23 wt. % w.r.t. Fe₂O₃) using hexamethyldisiloxane vapor using N₂ as a carrier gas (15 l min⁻¹). The resultant Janus particles (∼50 nm) displayed both excellent magnetic susceptibility and strong LSPR effects. Diffuse reflectance study revealed a broad absorbance curve with a maximum at ∼400 nm. The absorptivity increased with increasing Ag content and reached a maximum with 50 wt. % Ag. Coating with SiO₂ also removed the cytotoxicity and improved the dispersibility of the Janus particles. The authors then employed the Janus particles as bio-markers for imaging Raji and HeLa cells. The study also reported on the manoeuvrability of these Janus particles using an external magnetic field. Given the excellent stability and plasmonic absorption of the Janus particles in the visible region, it would be of interest to explore their application for photocatalysis.

Lv et al.⁹¹ created TiO₂/CNT (carbon nanotube) based Janus particles for potential application as a catalyst. Electrohydrodynamic co-jetting followed by calcination in air (500 °C) was used to obtain the primary bi-compartmental particles. Two separate polymer solutions of anhydrous poly(d,l-lactide-co-glycolide) and chloroform dimethylformamide were prepared by mixing with titanium butoxide. Ferrite nanoparticles (10 nm) were added to one of the solutions, while the processing rate for both the syringes was fixed at 0.3 ml h⁻¹ with a DC voltage of 7.5 to 9.0 kV. These primary Janus particles had an average diameter of ∼800 nm, with Fe₂O₃ particles confined to one half. After calcination, the diameter was reduced to ∼600 nm. The calcined particles were placed in a quartz tube furnace at 775 °C which reduced iron oxide to iron particles. These metallic iron particles acted as sites for inception and growth of CNTs using C₂H₂ as the carbon source. The possibility of CNT functionalization was also speculated. Depending upon the rolling vector (n, m values), CNTs can exhibit both metallic and semiconducting nature.⁹² It would be intriguing to see how these characteristics can be applied for the benefit of catalytic reactions.

Similarly, Singh et al. reported on a Si-Ag based Janus catalyst.⁹³ The particles were fabricated through gas-aggregated co-sputtering of Si and Ag targets, which were located besides each other on an integrated magnetron sputtering head. The deposition system consisted of a nanocluster sputtering source, a quadrupole mass filter, and a deposition chamber. The chamber was flushed with a mixture of Ar and He gases with flow rates of 70 and 5 sccm, respectively. The measured base pressure in the aggregation zone and main chamber was 2.5 × 10⁻¹ and 7 × 10⁻⁸ mbar, respectively. The magnetron power could be tuned for individual targets allowing volume fraction control of sputtered Si and Ag, leading to different particle morphologies. The Si target input power was fixed at 90 W. At a power ratio (Si:Ag) of 1.8, Janus particles were produced which changed to a core@shell morphology when this ratio was increased to 2.25. This article focussed on fabrication methods, while no direct evidence of catalytic activity was reported. However, the authors refer to an earlier study by Chen et al.⁹⁴ which describes the photocatalytic property of Ag-decorated SiO₂ nanospheres, thereby indicating their catalytic potential.

A direct comparison of catalytic activity is often advantageous as it provides proof for any enhanced performance imparted by a Janus morphology. Huang et al. compared the catalytic activity of their Janus TiO₂/C catalyst against pristine TiO₂.⁹⁵ A polystyrene (PS)-titania core-shell structure was dispersed in water using sodium dodecyl benzene sulfonate. The particles were polymerized at 70 °C by adding azobisisobutyronitrile and divinylbenzene to obtain TiO₂/MPS (3-(trimethoxysilyl) propylmethacrylate). Phenolic resin was coated on the lobe of TiO₂/MPS and calcined at 800 °C for 4 h in a nitrogen atmosphere to achieve a porous TiO₂/C Janus composite. The “snowman-like” Janus particles had an average TiO₂ and C diameter of 226 nm and 178 nm, respectively. The photocatalytic activity of the prepared catalyst was evaluated against hollow TiO₂ and P25 Degussa particles.

The majority of the reported Janus particles are solid “bare” formulations with a morphology tending to spherical. However, this hinders their potential as spheres have the lowest specific surface area. In this regard, Kirillova and co-workers have made an attempt to fabricate a “hairy” Janus particle with enhanced surface area, see Fig. 7(a).⁹⁶ 

This catalyst has a silica core (200 nm) with two distinct hydrophilic (poly acrylic acid) and hydrophobic (polystyrene) polymer shells on each half of its surface. The hydrophilic surface can be loaded with active catalyst species, such as Au or Ag particles. The core silica particles (200 nm) were prepared by hydrolysing tetraethylorthosilicate in an ammonia and ethanol solution. “Grafting from” and “grafting to” approaches were used to coat the particles with acrylic acid and polystyrene, respectively. Finally, metallic nanoparticle loading was achieved by allowing these hairy Janus particles to swell in deionized water (48 h) before using triethylamine to reduce the dissolved metal salt(s) onto the surface. With their ability to swell and de-swell in an appropriate medium, the particles are better suited to stabilize emulsions as well as possessing a higher activity owing to the larger surface area. Figure 7(a) illustrates the role of Ag NP-modified PAA/PS (polyacrylic/polystyrene) Janus particle for catalytic degradation of methylene blue with NaBH₄ and the corresponding UV–vis spectra of the methylene blue reduction are shown in Fig. 7(b). Figures 7(c)–7(e) are logarithmic plots of the absorbance as a function of time for methylene blue, eosin Y, and 4-nitrophenol, respectively, in the presence of Ag NP-modified PAA/PS-JP. This approach clearly demonstrates the multifaceted benefits achieved in a hybrid Janus particle for catalytic applications.

Graphitic carbon nitride (GCN) is a 2D carbon based material which is recently gaining popularity as a high surface area photocatalyst.⁹⁷ However, owing to the poor band-positions, functionalization of GCN with co-catalysts is often attempted to improve performance.⁹⁸ In the same direction, Zheng and co-workers fabricated bi-functional hollow CN structure for water splitting.⁹⁹ The hollow bi-functional Janus spheres were synthesized using a sacrificial SiO₂ template. A dense SiO₂ core (∼200 nm dia.) was modified with a monolayer of 3-aminopropyl triethoxysilane onto which Pt nanoparticles (∼3 nm) were reduced. Thereafter, a thin mesoporous silica shell was obtained by coating with a mixture of tetraethoxysilane and n-octadecyltrimethoxysilane and calcining at 550 °C. The resulting particles were loaded with cyanamide and the silica core removed using NH₄HF₂. Finally, the hollow sphere was deposited with Co₃O₄ nanoparticles (∼2 nm) to obtain the Janus catalyst. The spatially separated REDOX sites were then tested for hydrogen evolution, oxygen evolution, and overall water splitting. Compared to unfunctionalized and homogeneously deposited samples, Janus catalysts were observed to be an order of magnitude more effective. The higher activity was primarily credited to the spatially separated sites for evolution of H₂ and O₂ and the rapid electron transfer at the CN interface, thereby preventing any recombination or back reactions.

Another emerging method to fabricate high surface area catalyst is through bi-functionalized graphene/graphene oxide (GO) sheets.^100,101 The article describing bi-functionalized graphene sheets was reported by Yu et al. who deposited prepared Au and ZnO particles on the opposing faces of the nanosheet.¹⁰¹ In their study, the authors first coated the individual graphene sheets onto a silicon wafer substrate which was later functionalized by carboxylic groups through acetic acid plasma treatment. This allowed ZnO spheres (∼10 nm) to be attached to the functionalized and exposed surface of the graphene. The sheets were subsequently coated with poly(methyl methacrylate) (PMMA) and extracted using HF acid solution. For functionalization with Au particles (cubic, 60 nm edge), the recovered graphene sheets were grafted with 1-pyrenemethylamine and PMMA was removed using CHCl₃. The authors employed these Janus graphene sheets as UV activated photodiode. Given the semiconducting nature of ZnO and plasmonic capacity of Au nanoparticles,¹⁰² it would be of interest to investigate these bifunctional sheets for catalysis applications.

A similar study by Holm et al. regarding fabrication of TiO₂-GO-Pt Janus nanosheets was recently reported.¹⁰⁰ In this study, functionalization was accomplished in-situ using molecular precursors of the respective phases. GO sheets were first deposited onto a SiO₂/Si surface by Langmuir−Blodgett (LB) method. The exposed surface was then coated with a titanium butoxide solution to produce amorphous titania. By controlling the exposure time, different particle sizes (2–6 and 5–15 nm) could be obtained. Titania was crystallized using steam treatment method and later spin-coated with PMMA (280 nm) and extracted using concentrated HF. The obtained GO sheets were finally functionalized with Pt by photodeposition from an aqueous solution of K₂PtCl₄. Finally, PMMA coating was removed by dissolving in N-methyl-2-pyrrolidone to obtain the Janus particles. In-depth characterization revealed that the Janus catalyst displayed excellent transfer of photogenerated electrons from TiO₂ to Pt surface. However, no direct evidence of any photocatalytic activity was provided, which needs to be explored. Further, given the generic approach of the method and its excellent performance, it would be worthwhile to explore it for the preparation of Janus compounds which could impart higher photostability to some lower-band gap, but unstable catalysts, including sulphides and phosphides.

In addition to traditional catalysis, the Janus morphology may also possess some unconventional applications including nano/micromotors. Due to the anisotropic placement of REDOX sites in Janus particles, these structures can exhibit directional motion under appropriate conditions. In fact, the application of Janus particles as micromotors is an emerging field with several articles being published in the last two years alone. Depending upon the fabrication technique, its surrounding medium, and incident energy, several methods are responsible for imparting motion to such Janus motors. These include, but are not limited to, electrophoretic, diffusophoretic, and thermophoretic effects. However, an in-depth discussion on the propulsion mechanisms and types of Janus motors is beyond the scope of this review. Hence, the standard classification of heterogeneous photocatalysts will also be employed to classify Janus motors.

One of the first Au-TiO₂ based Janus micromotor was reported by Li et al.⁵⁷ The micromotors were prepared by first synthesizing amorphous TiO₂ microspheres via a microemulsion method. These were then drop-coated onto a glass slide to create a monolayer, which was subsequently coated with Au using ion-sputtering. The motion of Janus micromotors could be readily controlled by switching on/off and varying the intensity of UV-light, as shown in Figs. 8(a)–8(d). The resulting particles used H₂O₂ as a fuel to produce oxygen bubbles at the exposed TiO₂ surface, as illustrated in Figs. 8(e)–8(j). A quantum efficiency of 28% was reported for peroxide degradation imparting a maximum propulsion speed of 135 μm s⁻¹.

The same group (Mou et al.) later published another article dealing with Pt/TiO₂ Janus micromotors.¹⁰³ In this study, the host TiO₂ particles were crystalline with a size restriction of ∼800 nm. These Janus particles, when exposed to UV radiation, produce hydrogen bubbles on the Pt side to encourage forward motion and eliminate the need for an external fuel such as H₂O₂. This study also incorporated photocatalytic degradation of Rhodamine B in the absence of external stirring. Nevertheless, UV radiation is still necessary to drive Janus motors and hence the catalytic reaction.

In this regard, Li et al. reported on self-propelled Janus micromotors as a mobile catalyst.¹⁰⁴ The particles consisted of a Mg core partially covered by a shell of Au-containing TiO₂. The Mg microspheres of a required size were ion-sputtered with Au before being immobilized on a glass slide and coated with TiO₂ using atomic layer deposition. A small opening served as a site for Mg to react with water to produce hydrogen bubbles, while Au containing TiO₂ acted as the primary catalyst for pollutant treatment. The Mg core is depleted in the process and eventually the motors cease moving when the core is consumed. This is a possible limitation since once mobilized these micromotors cannot be switched off. In addition, the presented approach makes it difficult for catalyst retrieval, although an inherent advantage in this approach is that a larger TiO₂ area is available for photocatalysis.

BiOI supported Au nanoparticles is another active catalyst to be explored as a Janus micromotor.¹⁰⁵ Dong and co-workers fabricated BiOI microspheres (2.0 μm to 4.0 μm) by dissolving Bi(NO₃)₃•5H₂O and KI in absolute ethanol and water, respectively. The iodide solution was added to the ethanol dropwise and the pH was adjusted to 7.0 using ammonia. The solution was maintained at 80 °C for 3 h after which the particles were recovered by centrifuging. The microspheres were suspended in ethanol and drop cast onto a glass slide for immobilization after which they were coated with different metals (Au, Pt, and Al; ∼40 nm) using a sputtering technique. As a control, silica microspheres (∼2.0 μm) were also prepared using the same technique. It was observed that under illumination by green light, the Janus micromotors exhibited active directional motion with an average speed of 1.62 μm s⁻¹, propagating with the BiOI side forward. Contrastingly, with green light absent, only Brownian motion was observed. Primary motion was credited to catalytically derived self-electrophoresis, as confirmed by control experiments. The propulsion speed of the micromotors was observed to improve with decreasing wavelength and increasing glucose concentration (10 mM). UV irradiation increased photocatalytic activity in BiOI, while glucose acted as a fuel for propulsion. Directional motion was only reported for particles coated with either Au or Pt. The illumination conditions of the experiment (∼2500–5500 lx) also correspond well to a typical solar overcast, indicating practical applicability.

Pollutant degradation using (Au-WO₃@C) Janus micromotors was reported by Zhang et al. for aqueous sodium-2,6-dichloroindophenol (DCIP) and Rh B.¹⁰⁶ The micromotors were prepared by first synthesizing WO₃@C microspheres through hydrothermal synthesis. D+ glucose and Na₂WO₄·2H₂O were dissolved in distilled water and treated at 180 °C for 12 h. The particles were recovered, dried, calcined in air (200 °C, 30 min), and dispersed in ethanol which was then dropcast onto a glass slide. The particles were then sputter coated with Au and Al (30 nm) resulting in an effective particle diameter of ∼1 μm. Under UV illumination (40 mW m⁻²), a propulsion speed of 16 μm s⁻¹ could be obtained, resulting from the build-up of an osmotic pressure of reactants on the uncoated site (diffusophoretic). This speed could be further improved to 26 μm s⁻¹ and 29 μm s⁻¹ for low concentration solutions of DCIP and Rh B, respectively. Finally, upon coating with a thin layer of nickel, directional control over particle motion could be achieved using an external magnetic field.

From the studies reported to date, we can observe that common limitations of micromotors include: (i) a requirement for a chemical fuel, (ii) a need for UV irradiation, and (iii) monochromatic sensitivity. In this regard, Jang et al. fabricated broad spectrum powered Au/B-TiO₂ Janus micromotors.¹⁰⁷ The (Ti³⁺ doped) black- or B-TiO₂ was prepared by hydrothermal treatment of an ethanol solution containing titanium isopropoxide and formic acid, for 6 h at 150 °C.¹⁰⁷ The prepared particles were then heat-treated in air with a ramp rate of 75 °C min⁻¹ to obtain B-TiO₂ microspheres (diameter ∼3.5 μm). An aqueous suspension of these microspheres (5 μg ml⁻¹) was coated onto a clean silicon wafer through dip coating and sonicated before drying in air. Afterwards, a 40 nm thick layer of Au was deposited on one half of the surface through an evaporation process (0.1 Å s⁻¹). It was observed that the propulsion speed increased linearly with fuel concentration, reaching a maximum of ∼30 μm s⁻¹ with 3 wt. % H₂O₂ (λ > 400 nm; 120 W Hg lamp). This was 30 times faster than the control particles. Speed was also observed to be linearly dependent on the photon flux and wavelength of the incident light: UV (1.59 μm s⁻¹), blue and cyan (0.23 μm s⁻¹), green (0.16 μm s⁻¹), red (0.08 μm s⁻¹), and visible (white) (0.54 μm s⁻¹) wavelengths in pure water.

Jurado-Sanchez and co-workers synthesized C and Pt based Janus micromotors for “on-the-fly” decontamination of organic and inorganic pollutants.¹⁰⁸ Activated C microspheres (60 μm across) were suspended in isopropanol and spread on a glass slide to create a monolayer. These particles were then coated with (90 nm) Pt using a magnetron sputtering system. The microporous Pt surface acted as a site for bubble evolution (H₂O₂ dissociation), which propelled the Janus particles forward. These Janus motors can attain speeds up to 500 μm s⁻¹ in distilled water and 350 μm s⁻¹ in 70% seawater solution, with accelerated pollutant removal through activated C surface. Figure 9(a) shows the propulsion of these activated carbon/Pt based Janus motors over a period of 4 s. Control experiments were performed by employing polystyrene (PS)/Pt Janus micromotors. Figure 9(b) shows the propulsion of PS/Pt micromotors in ultrapure water and 70% seawater containing 10% H₂O₂ and 2% sodium cholate. These micromotors can be operated in dark, as they do not rely on photocatalysis. However, Pt has been reported to produce LSPR and hot electrons, which in itself could be used for degradation of some dissolved pollutants and is worthy of future investigation.

Recently, Qin and co-workers¹⁰⁹ reported the fabrication of bare Au based Janus nanomotors through selected activation of the surface. Initially, citrate protected Au nanoparticles were modified with a Bis-p-(sulfonatophenyl)phenyl phosphine (BSPP) stabilizer through ligand exchange. After stirring for 10 h at room temperature, the nanoparticles were precipitated through addition of solid NaCl. The particles were then re-dissolved in an aqueous solution containing (3-aminopropyl)triethoxysilane coated silica beads. This allowed the nanoparticles to be absorbed on the surface of larger silica spheres owing to electrostatic forces. This mixture was later treated by a short thiolated oligonucleotide (T5) and the particles were recovered by etching with hydrofluoric acid. This allowed the surface coated by T5 to be catalytically active, while the other half coated with BSPP to be rendered inactive. These Janus nanomotors acquire motion through catalytically induced self-thermophoresis. The active surface provides a site for REDOX reactions (in the presence of Fe(CN)₆³⁻ and S₂O₃²⁻) which heats up in the process. This asymmetrical heating allows solvent molecules to be unevenly pressurized and hence, drives the particle with cooler surface forward. The advantage of such nanomotors is their independence of an external power source (UV light). Furthermore, unlike Mg core micromotors,¹⁰⁴ these Janus motors remain in perpetual motion until the reaction is complete. This makes them an ideal catalyst for larger water bodies and poor illumination conditions.

Particles with a Janus morphology have been widely applied as a heterogeneous photocatalyst or to support and mobilize a primary catalyst, for example, their use in micromotors. In either case, before such catalysts can be deployed for field testing, their activity level and effectiveness need to be evaluated. This has been attempted through standard dye mineralization (organic or inorganic), water splitting reactions, or additional tests. Section IV provides a summary of the reported activity and experimental conditions for the Janus structures described above:

This is one of the oldest and most frequently employed tests for quantifying catalyst performance. Dilute solutions are prepared by dissolving a known quantity of certain dyes in water, which are then degraded in the presence of a catalyst and proper irradiation. Not only is it relatively straightforward to perform but also the dyes can be visually inspected for change in intensity as the colour fades. Hence, it remains popular among the scientific community. Nevertheless, performance evaluation through dye degradation also has its shortcomings. The dye properties (light absorption, structure, and intermediates) when combined with different catalysts can yield inconsistent results. Furthermore, decomposition of pure/single dye solutions has limited practical applicability outside laboratory trials. Therefore, dye degradation is not a standardized benchmark for performance evaluation, but rather an indicator of catalytic activity. Some of the commonly employed dyes, as used for Janus particles, are:

Fu et al. tested their Au-TiO₂ Janus nanoparticles by monitoring MB degradation as a function of time.⁶⁵ For the test, similar sized glass slides containing the Janus nanoparticles, and a traditional catalyst, were immersed in a 5 ml MB aqueous solution (concentration of 10⁻⁵M) in a quartz flask. The setup was irradiated with UV light with a wavelength of 254 nm. It was observed that after an irradiation time of 390 min, 37% degradation of the dye was observed for the control titania particles. However, for Janus nanoparticles the degradation was observed to be 48%, indicating the superior performance of the Janus based nanostructure.

Another study regarding the degradation of MB using Janus nanoparticles was presented by Kirillova et al. for their hybrid hairy Janus particles,⁹⁶ as described in Sec. III C. A volume of 0.25 ml (0.33 mg) of the as-prepared catalyst was added to a mixture of 40 ml dye solution (10⁻⁵M) and 2 ml freshly prepared aqueous solution of sodium borohydride (0.1M). Immediately after mixing, samples were taken each minute and analysed using UV-visible spectroscopy. A k value (degradation rate constant) of 0.017 s⁻¹ was obtained and complete degradation could be achieved after 180 s. Similar experimental conditions confirmed a k value of 0.42 s⁻¹ for eosin.⁹⁶ 

Sharma et al. also reported on UV and visible light assisted degradation of MB solution in the presence of their TiO₂-Au nanocluster Janus particles.⁷⁰ 10 mg of required photocatalyst was added to 15 ml (5 × 10⁻⁵M) dye solution and kept in the dark for 15 min to achieve absorption-desorption equilibrium. UV illumination was achieved using a Luzchem LZC 4V UV irradiation chamber (365 nm; incident power unspecified), while visible light was produced using a solar simulator equipped with a >420 nm cut-off filter calibrated for 1 sun. For a nanosphere configuration, it was observed that complete degradation of dye could be achieved in 240 min under UV (k = 0.0074 min⁻¹) and 180 min under visible illumination (k = 0.009 min⁻¹). Similarly, for Au nanorods, the corresponding rate constants were calculated to be k = 0.0065 min⁻¹ and k = 0.0081 min⁻¹ under UV and visible lights, respectively.

MB degradation was also employed by Wen et al. to highlight the enhanced catalytic activity of their Au-TiO₂ Janus particles.⁷² 0.1 ml (0.6 mM) aqueous MB solution was injected into a 300 ml cylinder shaped quartz reactor containing 20 vol. % methanol solution, under magnetic stirring. The samples were submerged in this solution and kept in the dark for 90 min to achieve desorption equilibrium. The setup was then irradiated by visible light (100 mW cm⁻²) for 50 min. It was observed that Janus particles could degrade 80% and 96% dye for simple and Pt interface catalysts, respectively. In contrast, control TiO₂ nanotubes could provide less than 3% degradation. The higher activity of the Pt interface containing Janus catalyst was attributed to the improved collection of hot electrons through the close contact between Pt and TiO₂.

γ-Fe₂O₃-TiO₂ magnetic Janus hollow bowl particles, as seen in Fig. 7, fabricated by Mou et al. were tested for Rh B degradation using visible light.⁸² For the experiment, 0.05 g m of the Janus catalyst was dispersed in 20 ml aqueous solution (10⁻⁵M) of Rh B. The setup was sonicated for 5 min followed by stirring in the dark for 30 min. The mixture was then irradiated with a 300 W Xenon (Xe) lamp with UV cut-off filters (λ > 400 nm). After a predetermined time, the catalyst was separated from the solution using a permanent magnet and the Rh B concentration was analysed (absorbance λ = 553 nm). Figure 10(a) shows the original absorbance of Rh B and subsequent degradation in the presence of Janus hollow bowls, as a function of time. Figure 10(b) shows the ratio of concentration C_t to C₀ (initial concentration) as a function of time “t” in the presence of γ-Fe₂O₃/TiO₂ Janus hollow bowls and single-component γ-Fe₂O₃ particles. It was observed that the absorbance peak steadily decreased with exposure time for the Janus nanoparticles, while little change was observed for simple γ-Fe₂O₃ particles. The degradation rate was almost twice as fast when compared to single component γ-Fe₂O₃. Furthermore, given the magnetic nature of the γ-Fe₂O₃ phase, magnetic separation of the Janus particles could be achieved. Figure 10(c) shows the separation of Janus hollow bowls after photocatalytic degradation of Rh B was accomplished, after intervals of 0, 5, 15, and 30 s.

Cao and co-workers used Rh B to evaluate the activity of their TiO₂-Au heterodimers.⁶⁹ 30 mg of photocatalyst was dispersed in 30 ml solution of Rh B (0.2 μmol l⁻¹) and the mixture was stored in the dark for 40 min to achieve absorption-desorption equilibrium. The setup was irradiated with a 500 W metal-halogen lamp with a UV filter (λ > 400 nm). The amorphous TiO₂-Au particles displayed the highest activity, reaching a degradation of ∼90% in 60 min followed by anatase (∼60%) and rutile (negligible) TiO₂-Au Janus particles. The higher activity of amorphous catalyst was attributed to a larger surface area and the presence of localized electronic states for efficient charge trapping.

Degradation of Rh B was also reported by Mou et al. for their TiO₂/Pt Janus submicromotors.¹⁰³ 100 μl solution containing 2 × 10⁶ pcs ml⁻¹ TiO₂/Pt particles and 20 μmol Rh B was added into two wells of a 96 well plate. The setup was irradiated by continuous and pulsed UV light (1 W cm⁻²) for 80 min, although the authors did not indicate the wavelength of the illumination source. It was observed that the rate constant was an order of magnitude higher with pulsed UV and catalyst (0.0384 min⁻¹) than pure UV (0.0034 min⁻¹). This was also better than continuous irradiation (0.023 min⁻¹), which resulted in particle agglomeration. However, no control experiment with pristine TiO₂ was reported.

Huang and co-workers reported Rh B degradation for their TiO₂/C Janus catalyst.⁹⁵ The degradation profile of hollow TiO₂ and P25 Degussa was also reported for comparison. 10 mg of catalyst was dispersed in a 100 ml Rh B solution (0.1 μmol) and kept in dark for 50 min. The mixture was then irradiated using a high-pressure Hg lamp (500 W). The Janus particles were observed to be more efficient (60% degradation) than hollow TiO₂ (55% degradation) but performed poorly than P25 Degussa (100% degradation). The poor performance of Janus particles compared to P25 Degussa was attributed to the difference in particle size/surface area. However, no direct comparison for Brunauer-Emmett-Teller (BET) surface area of the two catalysts was reported. Similarly, the higher catalytic activity of the Janus particles compared to hollow TiO₂ was attributed to a higher porosity and crystallinity.

Zhang et al. also reported degradation of Rh B in ultra-low concentrations through the use of Au-WO₃@C Janus micromotors.¹⁰⁶ The experimental setup consisted of a 400 W Xe lamp irradiating a 15 ml solution of either DCIP (25 μM) or Rh B (6.25 μM), consisting of 250 μl (500 μg) Au-WO₃@C Janus particles. It was observed that under UV illumination (40 mW cm⁻²) the Janus micromotors could degrade ∼80% and 40% DCIP and Rh B in 5 min, respectively. However, the authors note that a higher than critical concentration of pollutants (DCIP/Rh B) impedes the motion of micromotors, thereby limiting their applicability. Since dilution of pollutants is not always possible, the ability to operate at higher pollutant concentrations is worthy of investigation.

Jo et al. also investigated the catalytic performance of their Ag₂O-Ag₂CO₃ Janus particles through monitoring visible light assisted degradation of Rh B.⁸⁹ The photocatalytic setup was illuminated using a 300 W halogen lamp (incident power not reported) providing visible light. 50 mg of required catalyst was dispersed in 200 ml of aqueous dye solution (5 mg l⁻¹) and kept in the dark for 30 min before being illuminated (60 min). Janus particles were observed to degrade ∼99% of the dye as opposed to 57%, 42%, and 28% achieved by pure Ag₂O, Ag₂CO₃, and P25, respectively. The rate constant k for Janus catalyst (0.0555 min⁻¹) was observed to be 4 (0.0142 min⁻¹), 6 (0.0089 min⁻¹), and 10 (0.0056 min⁻¹) times higher than corresponding phase pure catalysts. Stability tests were also performed by recycling the catalyst for four consecutive cycles. The Janus Ag₂O-Ag₂CO₃ particles could retain 90% of their performance as opposed to a steady drop to nearly 0% observed in the pure Ag₂O catalyst. It was reported that the presence of heterojunction not only improves catalytic performance but also alleviates photoreduction of Ag₂O phase, thereby improving stability.

Jiang et al. reported enhanced photodecomposition of commercial MO dye using their novel Ag-Ag₂S Janus particles coupled with P25 TiO₂.⁷³ 40 mg of the composite catalyst was added to 40 ml of MO aqueous solution (10 mg l⁻¹) in a beaker. The setup was illuminated with a 500 W Xe lamp. Photocatalytic activity indicated that the Janus catalyst achieved 100% degradation in 40 min, while this number was limited to 80% and 13% for traditional Ag₂S-TiO₂ and bare P25, respectively. The enhanced activity of Janus particles was attributed to rapid electron transfer achieved between TiO₂ and Ag–Ag₂S phases. This indicates the apparent advantage of the Janus morphology over a conventional catalyst.

MO degradation was also used by Hu and co-workers for their Cr containing Zn/Al oxide biphasic catalyst.⁸⁴ In this case, 40 ml of 20 mg l⁻¹MO solution was mixed with 20 mg photocatalyst in a 50 ml quartz vessel. After agitating the mixture in the dark for 60 min, the setup was illuminated using a 300 W Hg arc lamp with a visible light filter (λ > 380 nm). The Cr containing catalyst performed comparable to P25 Degussa and 90% degradation was achieved in 1 h. The stability of the catalyst was also reported, with only a 4% reduction in activity observed after three consecutive cycles. However, a larger number of trials are required to confirm the long-term usability of the catalyst.

Enhanced MO degradation under visible light was also observed by Ding et al. for their Au-TiO₂ based Janus particles.⁶⁶ 10 mg of catalyst was dispersed in 30 ml of MO (5 mg l⁻¹) in a quartz tube, which was then kept in the dark (time unspecified). The mixture was then irradiated with a Xe light source (34 mW cm⁻²; λ > 410 nm). It was reported that ligand-exchange based Au-TiO₂ Janus particles achieved 100% degradation in 240 min. Salt-deposited sample achieved 30% degradation, while P25 Degussa displayed negligible activity. The higher activity of Janus catalyst was attributed to the heterojunction and monodispersity of Au particles. This behaviour was absent in salt deposited samples, resulting in a poor absorptivity and power transfer. The Janus catalysts were also tested for long term stability and exhibited a steady performance over multiple cycles.

Recently, Yu and co-workers reported on the sonocatalytic degradation of MO using Au-CeO₂ Janus nanoparticles.⁸⁰ In a standard experiment, 6 mg of catalyst was added to an MO solution (40 ml; 10 mg l⁻¹) and kept in the dark for 40 min to achieve absorption-desorption equilibrium. Subsequently, the sample was irradiated with ultrasonic waves using an ultrasonic generator (40 kHz; 150 W). Janus nanoparticles could achieve 100% degradation of MO dye in 40 min. Under similar conditions, only 72% and 76% of the dye could be degraded with pristine CeO₂ and Au-CeO₂ nanocomposites, respectively. The results indicate that conforming to a Janus morphology could increase the catalytic performance by ∼30%, and this improvement was credited to efficient charge separation at the heterojunction of Janus particles.

Degradation of Solophenyl green under UV irradiation was reported by Panwar et al.⁸⁵ For the experiment, 50 ml of 28.2 mM dye solution was mixed with 50 mg P25 Degussa or equivalent TiO₂ mass containing TiO₂-SiO₂ catalyst (173 mg core-shell, 230 mg Janus) in a 150 ml beaker. The mixture was sonicated (15 min) and then stirred (45 min) in dark to achieve stable absorption of dye molecules on the catalyst surface. The setup was then illuminated by two 8 W UV-A lights (365 nm) from above. The k values reported for different catalyst geometries were 0.0092 min⁻¹ (Janus-450), 0.0014 min⁻¹ (P25), 0.0028 min⁻¹ (Janus-750), and 0.0008 min⁻¹ (core-shell-450). The higher activity between the two Janus morphologies and other catalysts was attributed to the presence of a type-II heterojunction at the Ti-O-Si interface. High temperature processing led to an increase in crystallite size of the TiO₂ (Janus-750), thereby reducing the band-gap to 2.5 eV. This prevented the photogenerated holes from migrating into the heterojunction, thus causing recombination. In addition, photogeneration was reduced due to the majority of UV being blocked by the SiO₂ surface in core-shell arrangement. Hence, the Janus particles processed at 450 °C could outperform the other catalysts.

In a later experiment, the Janus-450 and P25 particles were attached to the surface of a cotton fabric to analyse their self-cleaning ability.⁸⁶ Treated cotton fabrics (4 × 3 cm²) were stained with (0.03 wt. %) solophenyl green dye solution and exposed to 7 UVA lamps for 12 h. The change in reflectance percentage over time was used to determine catalytic self-cleaning ability. It was reported that a 78% decrease in reflectance was observed with Janus particles, which is comparable to that observed for P25 (76%). However, it is to be noted that Janus particles contained only half (5.9 wt. %) the active component (TiO₂) of P25 (13.8 wt. %). Furthermore, after a single wash at neutral pH, the TiO₂ content for P25 functionalized fabric was reduced to 3.9 wt. %, while only a minor decrease (4.9 wt. %) was observed for the Janus catalyst. A second test run resulted in a 64% reduction for the Janus particles, while only a 58% was reported for P25. These results indicate the higher stability and effectiveness of Janus particles for practical applications and under varying pH conditions.

Janus particles can also be used for dehydrogenation of linear or branched organic molecules. Depending on the original compound and rate of reaction, either partial conversion or complete degradation is possible. Reported applications are listed below:

Pradhan et al. employed catalytic oxidation of methanol as a test for their Au-TiO₂ heterodimers.⁶⁴ 1.0 mg of Janus nanoparticles were dissolved in a mixture of tetrahydrofuran and water (1:1 volume) amounting to 20 ml. To this mixture, 0.01M acetylacetone, 0.33M ammonium acetate, and 0.62 mM methanol were added. The mixture was subjected to vigorous stirring and saturated with ultrapure oxygen. The resulting solution was exposed to UV irradiation (365 nm, 6 W) under a nitrogen atmosphere.

With the assistance of UV irradiation, methanol was readily converted into diacetyldihydrolutidine (absorbance ∼400 nm). After the initial delay of 50 min, the concentration of diacetyldihydrolutidine increased linearly with time. A conversion rate of 1.12 × 10⁻³mM min⁻¹ was observed indicating that 0.14 mM (out of 0.62 mM) methanol had been converted after 150 min. A control experiment with commercial TiO₂ displayed a lack of activity even after 180 min of run time. This higher catalytic activity could be attributed to the presence of the Au nanoparticles acting as an electron reservoir, thereby increasing the lifetime of the photogenerated charge carriers.

Liu et al. reported visible light assisted isopropanol degradation.⁶⁸ Several different morphologies were tested. In a typical setup, 0.18 g of catalyst was spread uniformly in a quartz vessel (9.0 cm²). The air inside was replaced by artificial air (O₂/N₂ = 1/9) at 1 atm. A predetermined amount of gaseous isopropanol was then injected into the vessel. The setup was kept in dark for 1.5 h after which it was illuminated with a 300 W Xe lamp using Y50 + R900 + water filters. It was observed that commercial TiO₂ displayed negligible catalytic activity, while C-TiO₂ produced 1.8 μmol of acetone in 5 h. However, with the Au nano-cage and C-TiO₂ Janus nanoparticles, a total of 5.7 μmol of acetone could be produced, which is 3.2 times higher than C-TiO₂ catalyst. Control experiments under dark conditions also evolved 0.51 μmol of acetone, indicating the thermocatalytic effect of Au. Subsequent analysis under monochromatic irradiation confirmed the presence of strong plasmonic resonance and superior performance compared to bare C-TiO₂ alone.

Kirillova et al. also reported catalytic reduction of 4-nitrophenol in a similar experimental condition as previously described (Sec. III C) for dye degradation (MB and eosin).⁹⁶ The difference being this experiment was carried out at a higher temperature (313 K). An effective k value of 0.101 s⁻¹ was observed, which was reported to be slightly lesser than similar systems. However, the reduction could be obtained with a smaller amount of catalyst.

In an experiment by Xu et al.,⁷⁹ authors reported on the catalytic reduction of p-nitrophenol. Typically, 4.5 ml of aqueous solution (0.1 mM) was sonicated with 2.5 mg Janus nanosheets. To this mixture, 0.5 ml of 0.2M NaBH₄ solution was added, after which the samples were taken at regular intervals and analysed using UV-vis spectroscopy. It was observed that p-nitrophenol was completely transformed into p-aminophenol in approximately 10 min. The authors also reported on the excellent stability of the particles with no appreciable change in catalytic activity even after five cycles.

In a following experiment, Janus silica nanosheets were used to create a stabilized emulsion with an oil to water ratio of 9:10.⁷⁹ Toluene containing p-nitroanisole and NaBH₄ aqueous solution was chosen as the reactant owing to mutual immiscibility of the two liquids. It was observed that 92.3% of p-nitroanisole was converted into p-anisidine in 2 h, which was improved to 94.47% in 3 h. In a controlled experiment, snowman-like Janus particle was also used. The conversion efficiency of nanosheets was found to be 16 times better than biphasic particles.

Carbendazim is a broad-spectrum fungicide which is known to cause infertility in high doses. Sharma et al. tested their TiO₂-Au nanocluster Janus particles for UV and visible light assisted degradation of aqueous Carbendazim.⁷⁰ The experimental setup is as described in Sec. IV A 1. For carbendazim, the degradation rate constant k was observed to be 0.006 min⁻¹ and 0.008 min⁻¹ for nanosphere and 0.0078 min⁻¹ and 0.0075 min⁻¹ for the nanorod configuration, under UV and visible illumination, respectively.

Doripenem is a widely used anti-biotic pharmaceutical compound and its presence in the ecosystem is detrimental to the natural microbial community. Mohanta et al. reported on the UV and solar-assisted degradation of aqueous Doripenem solution with their Ag-SnO₂ Janus particles.⁷⁸ 2 mg of photocatalyst was dispersed in a 50 ml (10⁻⁴M, pH 6 at 28 °C) solution of the compound. The mixture was irradiated with either a UV lamp (11 W, 254 nm) with incident power of 1.27 mW cm⁻² or solar irradiation (unspecified) with illumination intensity of 55 mW cm⁻². After an irradiation time of 180 min, it was observed that Janus particles demonstrated 80% degradation efficiency in the presence of UV, while it decreased to 49% in solar radiation. However, compared to pristine Ag (∼30%) and SnO₂ (∼25%) particles in solar light, Janus particles appeared to have a superior performance. The higher catalytic activity was attributed to the presence of heterojunction which improved charge separation and reduced the effective band gap of SnO₂ enabling it to harvest visible light. However, the performance steadily declined with each consecutive use indicating poor stability of the catalyst.

Jo et al. investigated the visible light degradation of 4-chlorophenol in the presence of their Ag₂O-Ag₂CO₃ Janus particles.⁸⁹ The experimental setup is as described in Sec. IV A 2 with an irradiation time of 180 min. It was reported that the Janus catalyst could achieve 90% degradation as opposed to 43%, 35%, and 10% observed with phase pure Ag₂CO₃, Ag₂O, and P25 samples, respectively. Hence, Janus particles were observed to be 4, 6, and 20 times more effective than corresponding pristine catalysts.

Another important, but less explored, application of heterogeneous catalysts is for air purification by photodegradation of harmful gaseous molecules. The easiest way to accomplish this would be to use this catalyst as a coating or paint, which is made easier when the base semiconductor in such catalysts is TiO₂. Owing to the potential benefits of this application, it becomes vital that such catalysts are not only easy to fabricate at a large scale but also readily available and low cost.

Byeon and Kim demonstrated this application with their Au-TiO₂ Janus nanoparticles, for enhanced photocatalytic oxidation of CO (carbon monoxide).⁷¹ It was observed that the Janus catalyst was able to assist with 47% oxidation of CO at 293 K, in the presence of visible light (λ > 450 nm), while the conversion was reduced to approximately 29% in the dark. Given the short lifetime of photogenerated charge carriers, conversion observed in the dark was attributed to the thermocatalytic activity of Au. The conversion efficiency was again increased to 36% when exposed to visible light (>600 nm). In a controlled experiment, it was revealed that under similar experimental conditions (λ > 450 nm), pure TiO₂ could only account for a total of <4% conversion.⁷¹ Hence, the Janus particles were observed to be 8 (λ > 600 nm) to 12 (λ > 450 nm) times more active than pristine TiO₂ particles.

The TiO₂ assisted photoelectrocatalytic water splitting experiment by Fujishima and Honda is one the seminal articles in the field of photocatalysis. Achieving efficient visible light water splitting through heterogeneous photocatalysis is still being investigated at a significant level. The biggest attraction of this scheme is the free availability of the driving force (solar energy) which makes the prospect very lucrative. It is speculated that any significant progress in the efficiency of such catalysis could make a prominent contribution towards mitigating the global energy crisis. Some of the recently reported Janus particles have demonstrated significant potential in this direction.

Mou et al. reported photo-assisted water splitting using hollow bowl snowman-like CeO₂-TiO₂ and CdS-TiO₂ Janus heterostructures.⁸³ The proposed catalysts were characterised using SEM, TEM, and EDX analysis [Figs. 11(a) and 11(b)]. The illumination source was a 300 W Xe lamp kept at a distance of 30 cm. The as prepared powder (10 mg) was dispersed into 40 ml aqueous solution containing Na₂S (0.1M) and Na₂SO₃ (0.02M). It was observed that the as-prepared “snowman-like” CeO₂-TiO₂ Janus nanoparticles could produce 1185 μmol·h⁻¹ g⁻¹ of hydrogen gas (effective surface area of 44.34 m² g⁻¹) as shown in Fig. 11(c), while the control experiment containing traditional CeO₂-TiO₂ composite could only produce 28–102 μmol·h⁻¹ g⁻¹. For the Janus nanoparticles, the rate of hydrogen evolution was an order of magnitude larger than that of conventional catalyst, which indicates the superior performance of the Janus morphology. The higher activity of Janus particles was credited to the presence of abundant heterojunction, enabling efficient separation of photogenerated electron and holes. However, more in-depth analysis is necessary to fully examine this claim.

Ding et al. also reported their Au-TiO₂ Janus catalyst for hydrogen production through oxidation of methanol (33 vol. %) solution.⁶⁶ For the experiment, 15 mg of catalyst was suspended in 10 ml alcohol solution followed by degassing with nitrogen. The setup was illuminated for 6 h with an incident power of 10 mW cm⁻². When a cut-off filter of λ > 390 nm was employed, a net hydrogen evolution of 350 μmol g⁻¹ was observed for ligand-processed catalyst, followed closely by salt-deposition catalyst (300 μmol g⁻¹). Negligible activity (∼5 μmol g⁻¹) was observed for P25 Degussa. For λ > 515 nm, Janus particles showed an evolution rate of ∼6 μmol g⁻¹, which was five times higher than salt-derived composites (∼1 μmol g⁻¹).

Another study relating to the hydrogen production was presented by Seh et al., for Au-TiO₂ Janus nanoparticles.⁶² The Janus catalysts were tested under visible light using iso-propyl alcohol as a sacrificial hole trap. The prepared Janus particles (45 mg), along with other morphologies (in equal quantities), were incorporated into a solution containing 1:2 ratio of iso-propanol and water (22.5 ml). The setup was illuminated using a tungsten halogen lamp (500 W) having a cut-off filter λ > 400 nm. It was observed that bare gold nanoparticles and amorphous titania did not show any appreciable activity. However, the rate of hydrogen evolution was 0.9 ml min⁻¹ for core-shell structure and 1.5 ml min⁻¹ for Janus particles. This was 1.7 times higher than the iso-symmetric structure. The higher rate of activity was attributed to the large plasmon fields generated in the local structure of the asymmetric Janus morphology. The activity increased with increasing Au particle size due to the stronger localization of electric near fields.

Similarly, Zhang et al. also reported hydrogen evolution for their Au-TiO₂ based particles.⁶³ Three configurations, namely: TiO₂ supported Au, Janus, and core@shell structures were investigated under visible light illumination (λ > 420 nm). 40 mg of catalyst was dispersed in a solution of methanol (30 ml) and water (70 ml) and degassed in vacuum for 30 min. The system was then irradiated for 8 h using a 300 W Xe arc lamp maintained at a distance of 10 cm. The highest hydrogen production (120 ml mg⁻¹) was reported for the Janus particles which was 2–4 times higher than other morphologies. The outstanding performance of half-embedded Janus particles was credited to the high energy transfer through WGM resonance. Another contributing factor is the visible light used for illumination, which favours strong LSPR absorbance and a stronger localization of near electric fields (E/E₀ > 3700 at 650 nm), which is significantly (∼28 times) higher than conventional particles.

Photocatalytic hydrogen evolution was also reported by Li et al. for their 2D Janus motif of (Cl₂)-(Bi₁₂O₁₇)-(MoS₂) bi-layer junction.⁸⁷ 10 mg of catalyst was dispersed in 80 ml aqueous solution containing ascorbic acid (0.3 mol l⁻¹) as the scavenging agent. The solution was transferred into a Pyrex flask (120 ml) and bubbled with nitrogen for 30 min. The setup was then irradiated with a 300 W Xe arc lamp. A maximum hydrogen evolution rate of 33 mmol·h⁻¹ g⁻¹ was reported, corresponding to a quantum yield of 36% (λ = 420 nm). The catalytic activity could be sustained for over 100 h. Given the excellent catalytic properties and hydrogen evolution rate of the Janus bi-layers, it would be interesting to consider and forecast its application for overall water splitting, which was not reported for these bi-layers.

Another attempt at hydrogen evolution and water splitting under simulated sunlight was reported by Zheng et al. for their hollow Janus CN nanospheres (CNS).⁹⁹ The photocatalytic setup consisted of a water-cooled Pyrex-top reaction vessel connected to closed gas circulation system and illuminated by a 300 W Xe lamp. The evolved gases were analysed using a thermal conductivity detector column with Ar as a carrier gas. For hydrogen production, 20 mg of required catalyst was dispersed in 100 ml (10 vol. %) triethanolamine solution. Similarly, for oxygen evolution, 100 ml (0.01M) AgNO₃ solution was used with 20 mg catalyst and 0.2 g m La₂O₃ (pH buffer). For water splitting, only pure water (100 ml with 20 mg catalyst) was employed.

Pure hollow CNS displayed a hydrogen evolution rate of 1.2 μmol h⁻¹, whereas an optimized composition of Janus CNS (1 wt. % Pt, 3 wt. % Co₂O₃) achieved an impressive rate of 96.3 μmol h⁻¹. When the functionalization was reversed (Pt outside), this activity could be further improved to 124.2 μmol h⁻¹, possibly due to limiting diffusion rate of reactants through the CN membrane. Similarly, the highest oxygen evolution rate of 15.4 μmol h⁻¹ was exhibited by the Janus catalyst. However, with overall water splitting experiments, these values dropped to about 0.3 μmol h⁻¹ and 0.1 μmol h⁻¹ for H₂ and O₂ gases, respectively. These numbers further improved to 12.2 μmol h⁻¹ and 6.3 μmol h⁻¹, respectively, when 200 mg catalyst was employed. Nevertheless, the Janus catalyst could achieve significant activity even in pure water, while none was observed for symmetric and phase pure samples. The stability of the Janus catalyst was also confirmed through repeated cycling and consequent characterization of the sample, indicating no appreciable change in functional or morphological attributes.

Yuan et al. reported on the efficient visible light driven hydrogen production and overall water splitting using their γ-MnS/Cu₇S₄ Janus-like z-scheme.⁸⁸ In a typical experiment, 1 mg of catalyst was suspended in a 6 ml aqueous solution of 0.35M Na₂S and 0.25M Na₂SO₃. Afterwards, the system was irradiated with a 300 W Xe lamp having a cut-off filter (λ > 420 nm). The illumination intensity was reported to be 100 mW cm⁻² and the experiment was run for 4 h. It was observed that the sample designated as NC3 (Mn:Cu molar ratio of 3) could exhibit a hydrogen evolution rate of 718 μmol·g⁻¹ h⁻¹. This corresponded to a quantum efficiency of 18.8%. However, this performance dropped to 113 μmol·g⁻¹ h⁻¹ for pure water, due to lack of oxygen evolution sites. This number could be further improved to 209 μmol·g⁻¹ h⁻¹ when the Janus NC3 particles were alloyed with MO_x co-catalyst, which aided in oxygen evolution. The results could be reproduced for five consecutive cycles, indicating high stability. These Janus particles offer efficient overall water splitting under visible illumination, with a high efficiency without the need of noble metal co-catalysts. This is a significant advancement in the direction of practical photocatalytic application.

Hydrogen evolution was also reported by Wen et al. for their Au-TiO₂/Pt Janus heterostructures.⁷² The experimental setup is as descried in Sec. IV A 1, without the dye. The methanol solution was purged with Ar gas for 1 h to achieve complete removal of dissolved O₂. The Janus nanostructures produced 12.8 μmol h⁻¹ cm⁻² hydrogen, which was three times larger than simple titania nanotubes (4.8 μmol h⁻¹ cm⁻²). This number could be further improved to 17.5 μmol h⁻¹ cm⁻² with the incorporation of (∼2 nm) Pt co-catalyst. These results, achieved with visible light illumination, indicate the benefit imparted by a Janus morphology. Further tests confirmed the presence of strong LSPR, which coupled with clear separation of reduction and oxidation sites led to a large improvement in hydrogen production capacity.

In addition to the above information, Table I provides condensed information on different Janus particles, their preparation technique, application and mode of irradiation.

Beginning from the simple Au-TiO₂ heterodimer in 2009,⁶⁴ to recently reported self-propagating nanomotors,¹⁰⁹ the field of Janus particles for catalytic applications has seen tremendous progress. Nevertheless, Janus catalysts form a relatively small fraction of the total volume of literature published in photocatalysis. This review has highlighted that Janus based catalysts often outperform their traditional/isotropic counterparts.

The majority of studies published regarding metal-semiconductor type Janus catalysts focus only on one particular composition, namely, Au-TiO₂. There exists ample literature to suggest that the incorporation of Au nanoparticles offers several advantages,¹¹¹ with Localised Surface Plasmon Resonance (LSPR) being the primary contributor.¹¹² The fabrication of Au nanoparticles is well documented, which makes it easier to control its size and shape,¹¹³ and this allows precise tailoring of the absorption characteristics, since LSPR is a function of particle size. This could be a reason for the popularity of Au based Janus nanoparticles. Another major advantage is that Au has significant co-catalytic properties of its own; an example would be the thermocatalytic ability of Au to oxidise organic compounds.¹¹⁴ Other benefits, which are common in M/S type catalysts, are the creation of a Schottky junction for efficient charge separation and metal nanoparticles acting as an electron reservoir. In practice, not all advantages can be realised simultaneously, for example, hot electron injection from LSPR is only observed when the primary semiconductor (TiO₂) is unable to produce sufficient charge carriers. The same catalyst will utilise Au as an electron reservoir when illuminated with white light (UV-visible).^66,70 However, since Au is a precious element, the high cost of the catalyst raises concerns about its potential wide scale applicability.

The incorporation of Pt and Ag also offers similar merits and, irrespective of the metal used, conforming to Janus morphology does have its benefits. Compared to an isotropic/homogeneous composite, Janus particles demonstrate strong localization of near electric fields. As demonstrated in the case of Au-TiO₂, this enhancement could be several orders of magnitude⁶³ and, consequently, the energy absorption, charge transfer, and charge separation characteristics are considerably improved. Monodisperse Janus particles are also sensitive to the direction of incident light and normally display a narrow, but strong, absorption near the wavelength corresponding to their LSPR.⁶⁶ Depending upon the placement scheme and size of the individual phases, other modes of enhanced energy transfer can also be obtained, e.g., Whispering Gallery Mode.⁶³ Anisotropic placement of metal and semiconductor phases also helps to separate the sites for oxidation and reduction reactions, thereby reducing the probability for back-reactions.^87,99 Their bi-phasic morphology also provides the unique ability for un-assisted unidirectional propulsion, under suitable environment. This has given rise to their application as micro/nanomotors, which cannot be achieved with conventional catalysts.

In the S-S category, however, such advantages are not clearly observable. A close contact/heterojunction between two desirable semiconductors can provide several benefits, including band-gap engineering and a co-catalyst effect. However, such advantages are common among all particle morphologies. In order to isolate and highlight the benefits of conforming to Janus morphology, suitable control experiments are required. Unfortunately, the current articles on S-S type Janus composites are lacking in such detail. Hence, it becomes difficult to do a comparative analysis and highlight the advantages of Janus arrangement, and therefore work in this area is necessary. Another major issue which needs to be properly addressed is the long-term stability, recovery, and reusability of the catalyst. These points are crucial for promoting commercialisation and large-scale applicability of Janus catalysts.

Photocatalytic nano or micro-motors is another important emerging application of Janus particles with several good studies being published in the years 2016–2017.^{57,103,105–107,109,115} Anisotropic/asymmetric placement of reduction and oxidation sites creates an imbalance of energy and reaction products. This incentivises unidirectional motion in unrestrained particles. Hence, this is one application which is exclusive to Janus morphology and carries its own unique benefits. Based on the type of Janus particle, incident energy, and surrounding medium, different mechanisms for propulsion can be observed including electrophoretic,¹¹⁵ diffusophoretic,¹⁰⁶ and thermophoretic.¹⁰⁹ These Janus motors carry several important advantages over conventional catalysts. First, their ability to exhibit un-assisted motion makes them a desirable candidate for treatment of large water-bodies where achieving external stirring is difficult. Second, through careful selection of participating materials, the need for an external fuel could be completely eliminated.^104,105,107 Third, by substituting traditional catalysts with activated carbon, pollutant removal can be achieved through adsorption in place of photocatalysis.¹⁰⁸ Finally, some degree of directional control and ease of collection through techniques such as modulation of UV light⁵⁷ and magnetic separation has also been demonstrated.¹⁰⁶ 

Moreover, unconventional Janus architectures (nanohybrid) promise to offer some novel and intriguing applications.^87,96,100 However, this field needs to be properly explored in order to tap into the huge potential of such unique Janus morphologies. Finally, for the perspective of photocatalysis based on Janus nanostructures, molecular-scale or atomic-scale characterization and mechanism analysis are required. This is essential to gain new insights into the photocatalytic performance of these materials and understanding of the enhancement of the electron-hole separation.

As this review highlights, there are clearly promising results for the application of Janus particles as heterogeneous photocatalysts. However, only a limited number of configurations have been explored. Analysis of the published literature in this field reveals that, when compared to traditional morphologies such as core@shell or satellite configurations, Janus particles exhibit a number of advantages. However, there is a lack of studies where a suitable combination of semiconductor and abundant metal(s) has been demonstrated, such as Ag and Cu. Silver has been widely reported for its excellent catalytic^42,75 and pesticidal⁷⁷ properties while being able to demonstrate strong LSPR effects.¹¹⁶ Furthermore, since it is relatively inexpensive, a larger volume of the catalyst can be employed and recovered, if necessary. Cu¹¹⁷ and Pd¹¹⁸ are other candidate metals in this field and studies should be performed to examine the beneficial effects and trade-offs of replacing Au and Pt with these elements. Additionally, most studies have been limited to using TiO₂ as the parent semiconducting phase. Recently, several other excellent and stable semiconducting materials have been reported including ferroelectrics¹¹⁹ and oxy-halide/iodide compounds.¹²⁰ It would also be of interest to explore how these materials can benefit from being employed in a Janus-like configuration.

Similarly, research should also be extended to include other low band-gap semiconducting materials such as ZnS, Cu₂O, and CdS. There have been limited attempts to fabricate and test a S-S type Janus catalyst, even though the conventional S-S configuration has been extensively explored. Owing to the lack of a systematic study in this direction, their potential advantages are yet to be proven. Nevertheless, the use of an anisotropic arrangement of oxide phases could certainly benefit some specialized applications and an example in this regard could be water splitting at a neutral pH.^88,99 In a traditional configuration, the close proximity of reduction and oxidation sites gives rise to a back reaction of the unstable radicles to form water. Hence, sacrificial agents are often employed to promote hydrogen evolution. A Janus structure could help to mitigate this effect by providing a clear separation of reduction and oxidation sites on the respective phases. Hence, this area needs to be properly examined for its potential benefits.

Furthermore, non-conventional Janus structures and their applications should also be explored. Janus particles as nano/micromotors for “on-the-fly” catalysis are an emerging field in this regard. Other potential benefits could include magnetic separation,¹²¹ Stern-layer formation (for ferroelectrics),¹²² or catalytic conversion of gaseous pollutants (carbon capture, solar fuel).¹²³ Efforts should also be made to develop economic methods for fabrication and size control of Janus particles. The majority of the heterogeneous catalysts reported to date are fabricated using a bottom-up approach employing wet-chemistry. However, techniques such as gas-phase synthesis and ambient spark discharge have also been reported.^71,93 This can help to reduce the fabrication cost of such nano-materials, which would be a significant step towards developing commercially competitive products.

The field of heterogeneous photocatalysis has achieved limited success in moving beyond laboratory-trials despite being investigated for almost fifty years. With the fossil fuel supply being rapidly depleted and increasing carbon emissions, effort needs to be made to switch to alternate energy sources. Photocatalysis should be no exception and technological/commercial device prototyping needs to be initiated. This should ideally be pursued with a modular approach, where the core catalyst can be “upgraded” when necessary. Such a configuration would allow low cost replacement of the photocatalyst as and when a better material becomes available. Such a step would break ground in the field of environmental remediation and would allow much needed technological development of this field.

Aditya Chauhan would like to acknowledge the support of SERB, India, in the form of Cambridge India Ramanujan fellowship. Paul Scheier and Monisha Rastogi would like to acknowledge the support received from the Austrian Science Fund (FWF) through Grant Nos. P26635 and W1259.

There are no conflicts to declare.