For hydrogen production, photocatalytic water-splitting (Ahmad et al., 2015; Ng et al., 2021; Villa et al., 2021) is a promising way to convert solar energy into clean energy (Fujishima and Honda, 1972). Numerous studies comparing photocatalytic hydrogen production using various photocatalyst types have been reported [for example, see Refs (Maeda, 2011; Sakata et al., 2015; Chen et al., 2019; Rafique et al., 2020)]. TiO₂-based photocatalysts are the most widely used and efficient photocatalysts (Do et al., 2020; Li et al., 2020; Dharma et al., 2022). Nowadays, employing heterojunction photocatalysts to produce hydrogen is alluring and very attractive; one such catalyst is NiO-TiO₂, which demonstrated intriguing results (Yu et al., 2015).

Photocatalytic hydrogen production can be affected not only by the type of photocatalyst and its properties but also by many other factors, such as the composition of the reaction mixture, reaction conditions, irradiation source (wavelength, intensity in reaction volume, irradiation distribution), and reactor type (Ahmed et al., 2011; Amakiri et al., 2021; Enesca, 2021). The irradiation source is a significant factor in photocatalysis, as photon transfer or low quantum efficiency are still significant problems in photocatalytic applications. To circumvent some of the current limitations, other reactor designs and operating conditions have been proposed (Deba et al., 2023).

In the case of heterogeneous photocatalysis, the required energy source depends on the band gap energy of the photocatalyst [e.g., for TiO₂ λ ≤ 387.5 nm (Schneider, 2015)]. The fundamental physical principle of all irradiation sources is the luminescence of excited atoms or molecules, where the electron is returned from the excited higher states to the ground state with simultaneous irradiation emissions. Based on operating principles, lamps can be divided into arc, incandescent, fluorescent, and lasers (Pareek, 2005). Ideally, using sunlight containing approximately 3%–5% ultraviolet irradiation would be desirable. Thus, mercury lamps and solar light are used in the research and application of photocatalysis. Mercury is a toxic, rare metal that burdens the environment. This limitation is associated with the end of the lamp’s life. As a mercury lamp only runs for approximately 1,500–20,000 h (Purpura et al., 2014), UV-LED lamps have inherently long lifetimes (Rasoulifard et al., 2011; Purpura et al., 2014), typically up to tens of thousands (50,000–80,000 h) of operating hours. The most significant advantage of using a UV-LED lamp is the substantial energy savings.

LEDs are also affordable and highly efficient irradiation sources, cheap, compact, lightweight, and have a lower operating temperature than conventional light sources. UV LEDs are also irradiation sources capable of producing monochromatic light with a narrow emission spectrum. In recent years, the electrical efficiency of UV-LED lamps has increased to values between 40% and 50%, which is significantly higher than traditional mercury vapor lamps and is rapidly approaching visible LEDs. Their efficiency is expected to increase further in the coming years if their development is comparable to that of visible light LEDs, greatly improving their photocatalytic water treatment efficiency (Tokode et al., 2014; Martín-Sómer et al., 2023). In addition, they can be turned on and off instantly (no warm-up) and do not contain harmful materials (Almquist et al., 2022).

Many research groups have already studied the influence of lamp intensity on photocatalytic efficiency (Pansamut et al., 2013; Kumar and Pandey, 2017; Wang et al., 2020; Amakiri et al., 2021). In most cases, the authors reported that hydrogen production increased with the increasing intensity of irradiation used (Baniasadi et al., 2013; Bell et al., 2013; Cheng et al., 2018). However, higher light intensities can cause much more significant energy losses instead of favoring the degradation of organic compounds. The key is determining the appropriate irradiation intensity to reduce the energy loss due to charge carrier recombination and ensure reproducibility in all measurements at different times (Amakiri et al., 2021). In general, in the case of heterogeneous reactions, it is observed that in some cases, the dependence of the reaction rate on the irradiation intensity becomes nonlinear, and the yields begin to decrease gradually (Bloh, 2019). However, more attention should be focused on comparing commercially available lamps and their illumination distribution and profile. The choice of an appropriate light source is also a crucial design factor for the efficient excitation of photogenerated species, which results in the creation of active radical species. Furthermore, the placement of the light source in the photocatalytic reactor impacts the effectiveness of the light received by the photocatalyst. Because of the high surface-to-volume ratios in the micro-photoreactor, the illuminated surface area is an important design parameter (Shukla et al., 2021). Photocatalytic water splitting is most often performed in two basic reactor configurations depending on the deployed state of the photocatalyst (powder form suspended in liquid and photocatalyst immobilized onto continuous inert carriers).

Most authors use batch reactors and powder photocatalysts in water-splitting reactions (Xing et al., 2013; Azam et al., 2019; Deng et al., 2019; Fajrina and Tahir, 2019; Guba et al., 2019; Visan et al., 2019; Toe et al., 2022; Sarkar et al., 2023). Its advantage is the possibility of using an external irradiation source during the reaction, excellent irradiation distribution in the entire volume of the reaction mixture, good mass transport, and a large specific area of the photocatalyst concerning the reactor volume. On the other hand, separating photocatalyst particles from the reaction mixture is very difficult, and there is a limited depth of irradiation transmission. The distribution of light intensity depends on whether the lamp is along a reactor (cylindrical shape) (Enesca, 2021) or is located at the top of the reactor (Sergejevs et al., 2017). Another disadvantage is that the reaction products remain in contact with the metal-doped photocatalyst particles for longer, promoting an undesired back reaction (Reilly et al., 2012). The unwanted back reaction is predominant in suspended photocatalyst suspensions because the H₂ and O₂ produced remain in contact with the high surface area photocatalyst particles for an extended period (Reilly et al., 2018).

Monolithic reactors (Lin and Valsaraj, 2006; Tahir and Amin, 2013; Maroto-Valer and Ola, 2015; Fajrina and Tahir, 2019), fixed bed fluidized bed reactors (Tahir and Amin, 2013; Vaiano et al., 2015; Fajrina and Tahir, 2019), optical fibers (Lin and Valsaraj, 2006; Tahir and Amin, 2013; Maroto-Valer and Ola, 2015), or reactors with photocatalysts immobilized in thin film (Oelgemoeller, 2012; Sambiagio and Noël, 2020) represent other attractive systems possessing easy separation and re-usability of photocatalysts. Immobilized thin-film photocatalysts are limited by poor light distribution, limited mass transfer, and a small photocatalyst surface area relative to reactor volume (Belver et al., 2019). However, they offer the convenient handling of the photocatalyst and reduce the risk of backreaction because the reaction products and the metal-doped photocatalyst are easily separated.

The significant advantage of immobilized photocatalysts is that they eliminate the need for separate catalysts while preventing photocatalyst agglomeration and deactivation (Mei et al., 2023). Micro-reactors using photocatalysts immobilized in thin film ensure an even light distribution due to their small size, short optical path, and large area-to-volume ratio. Due to the high density of photons in microreactors, it is clear that fast reaction times are required compared to a conventional batch reactor (Adamu et al., 2020). Properties unique to microreactors include laminar flow, short molecular diffusion distances, large specific interfacial areas, and excellent heat transfer characteristics. Particularly in photochemical reactions, microreactors show higher homogeneity of spatial illumination and better light penetration over the entire reactor depth than large-area reactors (Rashmi Pradhan et al., 2019; Zhan et al., 2020). Although external mass transport limitations are often considered for microreactors, internal (porous) diffusion limitations are only occasionally discussed (Zhan et al., 2020).

In this manuscript, we have focused our attention on comparing the efficiency of using six different lamps, which differed not only in the intensity of irradiation and their power but also in the profile and distribution of irradiation inside the batch photoreactor. In this part, NiO-TiO₂ powder photocatalyst is studied with very low NiO loading (0.2 wt% NiO), which is expected to improve the photocatalytic behavior of pure TiO₂. The second part is focused on a micro-photoreactor system–a batch reactor with a micro-reactor through which that reaction mixture is circulating. A micro-photoreactor uses a photocatalyst immobilized in a thin film on the glass. In this part, pure TiO₂ thin film is used as a photocatalyst. TiO₂ photocatalysts could be taken as standards to which the photocatalytic activity is usually compared, especially regarding a new design or type of process or reactor. In this manuscript, we focused mainly on the continuity of the entire process, reproducibility, stability, and evaluation of the contribution of the given micro-photoreactor.

Scientific teams use different types of commercially available lamps in photocatalytic reactions. This work compared six different lamps, i.e., lamps with different optics, and thus the intensity of irradiation, their power, and the irradiation distribution inside the conventional batch photoreactor.

The highest yield of hydrogen was achieved with the L3 lamp (UV-LED Solo P, Opsytec Dr. Gröbel, Germany), which, although it has lower radiation energy at 6 cm from the source than the other two lamps (L5 and L6), nevertheless it distributes the radiation better in space.

Due to the different properties of all the lamps, the best lamp for both the conventional batch reactor (L3) and the micro-photoreactor (L4) was evaluated and selected. The contribution of that article is the introduction of very different properties of individual but commercially available lamps, which could find broad use depending on the application.

The micro-photoreactor–a batch reactor with a micro-reactor with the circulating reaction mixture offered the possibility of long-term generation of hydrogen from the methanol-water solution due to continuous recirculation of the reaction mixture using the immobilized photocatalyst for easy photocatalyst separation. In the micro-photoreactor, emphasis was placed on the possibility of a long-term continuous hydrogen generation process with excellent reproducibility and stability. The TiO₂ thin film led to the production of 333.7 ± 21.1 µmol during 168 h (i.e., 252.8 ± 16.0 mmol.m⁻², the formation of hydrogen per area of the thin film). The micro-photoreactor also achieves stable/reproducible results in long-term experiments - after three consecutive cycles (three repeating cycles each of 168 h).