Packed Bed Microreactors for Sustainable Chemistry and Process Development

Abstract

Microreactor technology is seen as a promising approach to achieve green and sustainable synthesis in chemical fields because of the significant process intensification and fine control over reaction parameters caused by the miniaturization of reactor scale. The incorporation of solid catalysts as a packed bed in microreactors opens numerous opportunities for the efficient heterogeneous catalysis that plays a pivotal role in many industrially relevant chemical processes. In this review, the recent development in the use of packed bed microreactors as a versatile research tool and intensified production unit will be highlighted in the application areas including the synthesis of valuable chemicals and fuels, high-throughput catalyst screening, and kinetic/chemistry investigation. Selected reaction examples involving different reactant phases and catalyst categories will be particularly discussed, with an emphasis on the reactor performance in relation to the fundamental chemistry and engineering principles under microflow. In the end, future challenges and the outlook of packed bed microreactors for sustainable chemistry and process development will be provided.

1. Introduction

With the growing emphasis on environmental protection and resource utilization over recent decades, there has been a significant push to upgrade the chemical industry to meet higher standards for green and sustainable production. This shift is guided by frameworks such as the 12 Principles of Green Chemistry and the 12 Principles of Green Engineering [1,2], which provide a foundation for developing more eco-friendly and efficient chemical processes while considering long-term environmental, economic, and social impacts. Along with this concept, heterogeneous catalytic reactions have emerged as a promising approach for achieving green and sustainable chemical transformations, largely due to the superior reusability and tunability of heterogeneous catalysts compared with their homogeneous counterparts [3]. However, in typical industrial reactors that utilize heterogeneous catalysts (e.g., trickle bed, fixed bed), flow non-idealities such as incomplete wetting, flow maldistribution, and uneven concentration/temperature gradients are common [4], leading to suboptimal reaction performance and the inefficient use of both chemicals and catalysts.

In response to these challenges, process intensification through microreactor technology—specifically, the miniaturization of reactor dimensions to less than 1 mm in diameter—was introduced in the early 1990s and has gained significant momentum recently [5]. There are also definitions/applications of microreactors that extend the reactor diameter to several millimeters, provided the merits of miniaturization is maintained compared with conventional reactors (typically on the centimeter to meter scale) [6]. Packed bed microreactors, which incorporate heterogeneous catalysts packed inside in powder or particle form, are increasingly recognized as an important tool for optimizing reaction conditions and achieving green and sustainable chemical synthesis [7].

To be specific, the confined dimension of both microchannels and packings significantly enhances the specific surface area of the reactor and then results in better mass/heat transfer ability. For example, the surface area-to-volume ratio of multiphase flow in microreactors is typically 10,000 m²/m³ (about 10 to 100 times greater than that of conventional reactors) [8]. Accordingly, relative mild conditions (i.e., low temperature/pressure) can be applied in microreactors to obtain the equivalent reaction performance (e.g., conversion, selectivity, yield) compared with conventional macroscale reactors [9]. Thus, packed bed microreactors represent an attractive choice for sustainable chemical production, where the precise control of reaction parameters and intensification towards maximized reaction performance are essential. As a part of the microfluidics technology, packed bed microreactors allow the manipulation of quite a small amount of fluids (e.g., 10⁻¹⁸ to 10⁻⁹ L) within microchannels [10]. The small inventory of reagents as well as catalysts in microreactors significantly reduces resource consumption and waste production. Additionally, microreactors enable the rapid and extensive investigation of reaction conditions (e.g., because of the short residence time and use of parallelized reactor units), allowing for reaction optimization, kinetic studies, and high-throughput catalyst screening with minimal catalyst usage (particularly relevant to expensive noble metal catalysts). The short running time, coupled with the more precise parameter control and the sensitive response of microreactors, also makes it a promising tool for the pharmaceutical industry as well as biological analysis [10,11,12]. For processes involving highly exothermic reactions, the rapid heat dissipation provided in microreactors helps maintain an isothermal environment. In explosive scenarios (e.g., reactions involving O₂ or H₂), the quenching distance (critical size below which flame propagation is inhibited) is normally larger than typical void sizes in microreactors [13]. As a result, packed bed microreactors can effectively prevent explosions, allowing the handling of these reactions safely. Furthermore, the scale-up process can be easily achieved through numbering-up, which involves increasing the number of parallel microreactors without altering the operating conditions of each individual reactor. This approach effectively avoids the usually time-consuming reaction optimization process associated with increasing reactor volumes. The decreased size of microreactor components is also preferable for field construction and transportation [14], making packed bed microreactors promising for the distributed production of chemicals. Moreover, the continuous flow mode achieved by packed bed microreactors provides a better control over the final product quality (imperative for fine chemicals and pharmaceutical industries) than the commonly applied (semi-)batch reactors (e.g., flasks and autoclave) [15].

The packed bed microreactor can be either capillary- or chip-based (Figure 1). The capillary-based microreactor utilizes readily available and inexpensive capillaries made from materials such as polymers, silica, or stainless steel. The chip-based microreactor features a packed bed typically housed within microchannels or microchambers that are fabricated onto a chip/plate-like substrate (often made from silica, glass, metal, or alloys). The capillary-based beds are simple to prepare but typically require an additional mixer to combine different fluids. In contrast, chip-based packed beds can seamlessly couple fluid handling, mixing, and reaction as well as separation/detection steps within a compact platform, which allow to realize fully integrated and functionalized microfluidic systems such as the so-called microscale total analysis systems or lab-on-a-chip devices [16,17,18]. In packed bed microreactors, the catalyst is held in place usually by packing constrictions (e.g., micro-pillars, inert particles, or filters) [19,20,21] at both ends of the catalyst bed. Commercial/homemade catalysts can be packed into microchannels in a straightforward manner, such as by manual filling or using a vacuum pump, without using complex assembly/coating processes [22]. As a result, packed bed microreactors facilitate easy catalyst incorporation and replacement while maintaining high catalyst loading, which broadens the range of their applications in heterogeneously catalyzed reactions. A detailed comparison of different types of microreactors coupled with heterogeneous catalysts is displayed in

Table 1. Comparison of different types of microreactors for heterogeneously catalyzed reactions [14,22,28,29,30].

Reactor Type	Reactor Sketch a	Configuration	Advantages	Disadvantages
Packed bed microreactor		Solid catalysts are packed in capillary- or chip-based microreactors in the form of powder/particles	Direct use of commercial off-the-shelf or self-prepared powder catalysts Easy fabrication Easy catalyst incorporation and replacement High catalyst loading	High pressure drop Complex hydrodynamics Possible reactor clogging
Wall-coated microreactor		A thin layer of catalyst is coated onto the microreactor inner wall surface	Low pressure drop Relatively simple hydrodynamics (similar to that of empty microreactor)	Complex fabrication Need for developing dedicated coating and catalyst removal methods Low catalyst loading
Slurry-based microreactors		Solid catalysts are suspended in the liquid phase (to form a slurry) and continuously flow through the microreactor	Large contact area of catalyst with fluid Direct use of commercial off-the-shelf or self-prepared powder catalysts Easy catalyst incorporation and replacement	Possible reactor clogging Complex catalyst motion pattern Possible catalyst sedimentation/aggregation Need for additional separation step for catalyst reuse

^a Illustrated for the idealized case of gas–liquid slug flow involving solid catalysts.

, where the merits of packed bed microreactors are clearly visible.

Up to now, packed bed microreactors have been increasingly adopted for the synthesis of valuable chemicals and fuels, accelerated catalyst screening, and kinetic study [28]. A comprehensive and timely review focusing on, particularly, their latest applications across various reaction schemes and catalyst systems, is thus conducive to the rapid development of packed bed microreactors in the relevant fields. In this work, the recent development of packed bed reactors in typical reaction examples involving different reactant phases and catalyst categories are summarized. The potential of packed bed microreactors is highlighted as a versatile research tool and intensified production unit. Finally, the future challenges and perspectives are discussed, emphasizing the role of packed bed microreactors in advancing sustainable chemistry and process development.

2. Application of Packed Bed Microreactors

Over recent years, packed bed microreactors have experienced a considerable expansion in their application scope. They are particularly advantageous to intensify reactions for the synthesis of valuable chemicals/fuels where heat or mass transfer is a limiting factor, such as gas–liquid reactions with low gas solubility (e.g., hydrogenation and oxidation), highly exothermic reactions requiring a tight temperature control. Additionally, packed bed microreactors are highly desirable for catalyst screening and reaction kinetic studies, as they generally mitigate external/internal mass transfer limitations under well-controlled process conditions with the use of a much reduced catalyst amount, enabling a fast and precise extraction of catalyst performance or intrinsic kinetic information. Typical reaction examples are summarized in the following several categories. The details of reaction systems and operational parameters for selected cases are also outlined in Table 2, Table 3 and Table 4.

2.1. Synthesis of Value-Added Chemicals

2.1.1. Selective Alcohol Oxidation to Aldehydes and Ketones

The production of aldehydes from alcohol oxidation is an important transformation in organic chemistry and pharmaceutical fields, given that aldehydes are highly versatile building blocks and key intermediates or precursors for constructing more complex chemicals, including, among others, carboxylic acids, amines, and aldols [31]. Bogdan and McQuade [23] applied a 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) catalyst, immobilized on AMBERZYME^® Oxirane (AO) resin, in the biphasic oxidation systems composed of an aqueous bleach solution and organic alcohol solution with dichloromethane (CH₂Cl₂) or ethyl acetate (EtOAc) as the solvent (Entry 1; Table 2). The TEMPO/AO catalyst was packed in a fluoroelastomeric capillary that was woven around metal bars and submerged in an ice bath (Figure 1a). The catalyst exhibited excellent recyclability while maintaining a high conversion and selectivity. This was evidenced by negligible catalytic activity loss over a 9 h continuous test for the oxidation of 4-chlorobenzyl alcohol to 4-chlorobenzaldehyde (Scheme 1), with >99% conversion and 93% yield under specific conditions. The good oxidation behavior indicated the effective interphasic mixing in the microreactor (visually supported by the emulsification of the upstream slug flow in the bed causing droplet coalescence at the microreactor outlet). The generality of the AO-TEMPO packed bed microreactor was demonstrated in the selective and effective oxidation of various substrates including primary (aromatic and aliphatic) and secondary alcohols. Moreover, the green solvent EtOAc demonstrated a performance almost comparable to CH₂Cl₂, as exemplified for benzyl alcohol oxidation to benzaldehyde.

Compared to toxic inorganic oxidants (e.g., CrO₃, KMnO_4, and MnO₂), oxidation utilizing molecular oxygen as a cheap and readily available oxidant exhibited a higher atom economy and greener reaction route (typically with water as the byproduct) [32]. A Teflon^® capillary microreactor packed with a silica-immobilized TEMPO catalyst was used by Aellig et al. [33] for the aerobic oxidation of alcohols in the presence of catalytic amounts of HNO₃ as a co-oxidant (Entry 1; Table 3). Solvent screening experiments revealed that 1,2-dichloroethane (DCE) rendered the best results in the oxidation of benzyl alcohol (Scheme 2), affording 98% alcohol conversion and 99% benzaldehyde yield under relatively mild conditions (55 °C and 5 bar O₂) within a short residence time of 30 s. Meanwhile, ca. one to three orders-of-magnitude higher space–time yield (STY; defined herein as the product amount per mole catalyst, per unit time, and per reactor volume) was obtained in the continuous microreactor system for this reaction compared with conventional batch systems employing different solid TEMPO catalysts and NOx sources as co-oxidants. The substrate scope was further explored in the same microreactor, where the oxidation of several other primary and secondary alcohols (to the corresponding aldehydes or ketones) resulted in high reaction rates and selectivities (close or above 90%) under similar reaction conditions. It is worth noting that, for reactions under relatively mild conditions, capillary tubings made of cheap and disposable polymers can be applied conveniently as the microreactor. This avoids possible wall oxidation when using metal microreactors while maintaining a low fabrication cost and easy observation of flow property in the bed [23].

The presence of solvents in the reaction systems summarized above complicates the downstream separation process, leading to increased consumption of both energy and resources. This issue could be mitigated by employing solvent-free oxidation, as shown by Al-Rifai et al. [31] for the aerobic oxidation of benzyl alcohol (Scheme 2) using 0.05 wt.% Au–0.95 wt.% Pd/TiO₂ catalyst packed within a silicon–glass microreactor chip (Entry 2; Table 3). The reaction performance (under 120 °C and 1 bar gauge pressure of pure oxygen) was found highly related with three-phase hydrodynamics in the microreactor. The increase in the gas flow rate (especially at a relatively high liquid flow rate) caused the flow regime to first change from a liquid-dominated slug flow to segregated flow. This was accompanied by an increase in benzyl alcohol conversion and benzaldehyde selectivity, because the improved interfacial area and decreased liquid film thickness facilitated the transfer of O₂ to the catalyst. At a relatively high gas flow rate, a fully or partially wetted gas continuous flow appeared. Despite the increased benzaldehyde selectivity in the partially wetted flow regime at very high gas–liquid flow ratios, the benzyl alcohol conversion dropped from 98% (under full wetting) to ca. 20% because of the incomplete utilization of a catalyst surface. Overall, the fully wetted gas continuous flow regime proved to be the most effective option, maintaining both high conversion (93%) and selectivity (80%). Recently, a similar packed bed microreactor was connected with two other micro packed beds in series to realize the multi-step synthesis of benzylacetone from benzyl alcohol [34]. That is, the telescoped microflow system consisted of three beds with the respective catalysts for the aerobic oxidation of benzyl alcohol (to benzaldehyde), the subsequent C–C coupling (with acetone to form benzalacetone), and, finally, hydrogenation of benzalacetone to benzylacetone. Each bed could be operated under its own optimization conditions, leading to a significant yield increase compared with the batch cascade.

It is also promising to apply packed bed microreactors in homogenously catalyzed biphasic reactions, that is, by packing inert particles to enhance a gas–liquid interaction and thus mass transfer. Zhang et al. [35] applied this idea using a homogenous Cu/TEMPO catalyst for the aerobic oxidation of alcohols in microreactors packed with glass beads (Entry 3 in Table 3; Figure 2d). An almost 100% yield of benzaldehyde was obtained from 0.4 M benzyl alcohol (in acetonitrile) at a mixed gas (9% O₂ in N₂) pressure of 35 bar and 45 °C within 30 s. The accelerated mass transfer and reaction rate in packed bed microreactors, compared with slug flow reactors without solid packing, was validated by the 10 times shorter residence time needed for a full conversion of benzyl alcohol and other primary alcohols.

2.1.2. Safe Synthesis of Hydrogen Peroxide and Phosgene

Hydrogen peroxide (H₂O₂) spans a wide range of applications, from industrial and environmental uses to medical, household, and agricultural purposes [36,37]. Its strong oxidizing properties make it a powerful tool for cleaning, disinfecting, and bleaching, while its environmental benefits (i.e., with water as the byproduct) position it as a sustainable oxidant choice in various processes. The direct synthesis of H₂O₂ from H₂ and O₂ (Scheme 3) is an attractive approach due to its simplified process, cost efficiency, and environmental benefits compared with the industrial multi-step anthraquinone process. However, the high pressure required for an adequate mass transfer and the inherently explosive nature of an H₂/O₂ mixture pose significant safety risks for the direct synthesis route. Thus, catalyst dilution is often needed to prevent over-reaction and the subsequent risk of runaway in conventional fixed-bed reactors, translating into an inevitable productivity decrease [38]. In this respect, microreactor technology is deemed capable of producing H₂O₂ efficiently under the full catalyst utilization, thanks to its enhanced heat and mass transfer capabilities. Additionally, the dimensions of flow passages between particles in a microreactor can be easily kept smaller than the quenching distance for hydrogen (ca. 500 µm at 1 atm), effectively suppressing explosions [36,37].

In the pioneer work of Inoue et al. [21], a silicon-Pyrex microreactor chip consisting of 10 parallel packed beds was tested for the direct synthesis of H₂O₂. In the microreactor design, inert silica particles were packed in the inlet section fed with parallel channels that contained the catalyst bed. This was used to distribute gases (H₂ and O₂) and a liquid phase (aqueous solution of dilute sulfuric acid, phosphoric acid, and sodium bromide) more uniformly into the reaction channels and, more importantly, to prevent explosions in the inlet void spaces. By using a commercial 5 wt.% Pd/C catalyst, 0.2 wt.% H₂O₂ with a peroxide selectivity of almost 100% was achieved from H₂–O₂ mixtures (ratios within the explosive region) in the packed bed microreactor that was conducted safely under 20 bar and 20 °C at a gas phase residence time of 9 s. The concentration of H₂O₂ in the reaction effluent more than doubled to 0.6 wt.% when a second packed bed microreactor was installed in series, due to the longer residence time and improved reactant–catalyst contact. By similarly coupling two microreactors (i.e., single-channel glass chips) in series, Inoue et al. [39] further showed that a product concentration of H₂O₂ as high as 11.3 wt.% with a yield of 42% was obtained over a Pd–Au/TiO₂ catalyst at 0.95 MPa and 20 °C (Entry 4; Table 3). The milder reaction conditions required for such high H₂O₂ concentration in microreactors compared to the conventional batch reactor was attributed to the inherent superior heat and mass transfer rates in the microflow. Among all catalyst compositions (Pd/Al₂O₃, Pd/TiO₂, Pd–Au/TiO₂) studied, Pd–Au/TiO₂ was found to be the most promising due to its suppression of byproduct pathways, resulting in higher peroxide selectivity. To further increase the throughput, Inoue and coworkers [26,38,40] fabricated the microreactor system with up to 32 parallel packed beds (of Pd–Au/TiO₂ or Pd/TiO₂ catalyst) in a chip (e.g., see Figure 1d and Entry 5 in Table 3). Via the parallel operation of several such microreactor modules, H₂O₂ could be produced at a concentration of ca. 4–10 wt.% and throughput of ca. 1–10 kg/day under similar conditions (1 MPa and 23–25 °C). Thus, the production capacity increase in microreactors by numbering-up was proven to be fast, reasonable, and promising for the future industrial production of H₂O₂ from H₂ and O₂.

Very recently, other solvent and catalyst systems were explored for the direct synthesis of H₂O₂ in packed bed microreactors [41,42]. A microreactor made of polymethylmethacrylate (PMMA) was loaded with a Pd/Al₂O₃ catalyst and demonstrated the safe and direct synthesis of H₂O₂ under ambient conditions (1 bar, 25 °C) [41] (Entry 6; Table 3). Out of the liquid media tested (methanol, ethanol, isopropanol, butanol, and water), the aqueous solution of methanol (20 vol. %) proved to be the most favorable solvent and gave an industrially acceptable H₂O₂ concentration of 0.388 wt.%. The DFT calculation revealed the beneficial effect of methanol presence in enhancing O₂ adsorption on the catalyst and facilitating the electron transfer to O₂, thus co-catalyzing the H₂O₂ formation reaction. The Pd−Sn/Al₂O₃ catalyst was synthesized by Yang et al. [42] and packed in a similar PMMA microreactor mentioned above for the direct H₂O₂ synthesis using aqueous methanol solution as the liquid medium (Entry 7; Table 3). The doping of Sn was found to facilitate the H₂O₂ release from the catalyst surface and inhibit the catalyst deactivation effectively. The catalyst with the Sn/Pd molar ratio of 0.5 could maintain a concentration of H₂O₂ at about 0.483 wt.% for at least 180 min of continuous microreactor operation at normal temperature and pressure, in contrast to a monotonous concentration decline (from 0.4 wt.%) over a Pd/Al₂O₃ catalyst.

Microreactor technology is also suitable for the safe synthesis of phosgene, a toxic and reactive chemical used in the production of isocyanates, polycarbonates, and other organic compounds applied in the polymer and pharmaceutical industries [43]. Industrial-scale phosgene production involves the reaction of carbon monoxide (CO) and chlorine gas (Cl₂) over an activated carbon catalyst (Scheme 4). The hazardous nature of the reactants and product necessitates the careful consideration of production, storage, and transportation processes [44]. Utilizing the precise temperature control nature of microreactors, Ajmera et al. [43] fabricated a silicon-based microreactor packed with an active carbon catalyst and achieved a production of phosgene at 9.3 kg/year from a stoichiometric feed of CO and Cl₂ under about 220 °C and 1.4 bar (Entry 1; Table 4). The external mass transfer limitation was proven negligible for the employed catalyst with a size of 63 μm, and thus the kinetic operation was attained. The microreactor system exhibited additional advantages for the on-site/on-demand production, as the tiny reactor volume can be easily placed in a small, ventilated enclosure (or fume hood) and the phosgene productivity can be substantially increased via a facile integration of a number of microreactors (each with multiple reaction channels inside).

2.1.3. Bio-Based Platform Chemical Production

The 5-hydroxymethylfurfural (HMF) is an organic compound typically derived from the dehydration of C6 sugars and is considered a key intermediate in the production of bio-based polymer precursors such as 2,5-diformylfuran (DFF), levulinic acid (LA) and 2,5-furandicarboxylic (FDCA), fuels, and additives such as 2,5-dimethylfuran and 5-ethoxymethylfurfural [45]. To achieve a low-cost and energy-saving downstream separation process (e.g., distillation), a heterogeneously catalyzed reaction system under continuous flow mode is preferable for HMF synthesis. Aellig and Hermans [46] successfully converted fructose to HMF using a capillary microreactor packed with a commercial Amberlyst-15 solid acid catalyst (Entry 2; Table 2). The reaction utilized a low-boiling solvent system, combining 1,4-dioxane and dimethyl sulfoxide (DMSO), which facilitated energy-efficient solvent recycling. In contrast to a maximum HMF yield of 75% in batch, the yield in the microreactor increased to 92% with an almost 100% conversion, exhibiting a 75-fold higher space–time yield under relatively mild conditions (110 °C and 1 bar). The catalyst demonstrated stability over 96 h, making this system a highly promising alternative for HMF production compared with the homogenous catalyst. Zhang et al. [47] studied the selective production of HMF and LA from inulin (dissolved in water or the DMSO–water mixture) using commercial HND-580 solid acid catalysts packed in the microreactor (Entry 3; Table 2). Smaller catalyst sizes were found to favor higher product yields, possibly due to the improved substrate–catalyst contact. The highest yields obtained were 81% for HMF and 84% for LA at a short residence time of ca. 2.4–7.3 min and 150 °C. The space–time yields of HMF and LA were 1–2 orders of magnitude higher than those in conventional batch reactors. However, the catalyst stability was not tested.

The continuous HMF synthesis from glucose (a cheaper but less reactive feedstock than fructose) was studied by Guo et al. [48] in a microreactor packed with self-prepared P-TiO₂ catalyst (Entry 4 in Table 2; Figure 2c). At 150 °C, a maximum HMF yield of 72% was obtained from 0.1 M glucose (via fructose as the sugar isomerization intermediate) in a water-2-methyltetrahydrofuran (mTHF) biphasic system (Scheme 5). The yield was comparable to batch results, though the presence of mass transfer limitations affected reaction performance at higher glucose concentrations. Despite the use of mTHF to in situ extract HMF produced in an aqueous solution, the byproduct humin was still formed (e.g., from side reactions involving HMF and sugars), which accumulated on the catalyst active sites causing deactivation after several hours of use. However, the catalyst could be regenerated by ex situ re-calcination without an appreciable activity loss. The study demonstrated significantly higher space–time yields of HMF in the microreactor (~2 orders of magnitude higher) than those in batch over similar catalysts and conditions and provided insights into the synthesis and performance of the in situ phosphated TiO₂ catalysts.

Like HMF, furfural was identified as one of the top value-added platform compounds from biomass [49]. Lu et al. [50] utilized a packed bed microreactor with a mixed solid acid catalyst (SO₄²⁻/γ-Al₂O₃ and HND 580) for the continuous conversion of xylose to furfural in water (Entry 8; Table 3). The process achieved 96.86% xylose conversion and 44.65% furfural yield at 170 °C and a short residence time of 1.6 min, significantly outperforming batch processes. The addition of CO₂ improved the generation of hydronium ions, thereby improving the yield of furfural (by ca. 10%) in a gas–liquid–solid system. The mixed catalyst underwent a gradual activity loss after 5 h on stream, out of which the SO₄²⁻/γ-Al₂O₃ particles could be regenerated (via calcination) and HND 580 had to be replaced due to its poor thermal stability. It is noteworthy that the furfural yield was still not high, indicating further room for reaction optimization (e.g., by improving the catalyst system or using a biphasic solvent system for in situ furfural extraction).

The catalytic oxidation of HMF towards valuable bio-based products such as DFF (important precursor for pharmaceuticals and polymers) and FDCA (key precursor for polyethylene furandicarboxylate (PEF)) was also studied in packed bed microreactors. Aellig et al. [33] investigated a Teflon^® capillary-based packed bed microreactor for HMF oxidation (0.41 M; dissolved in DCE) at 55 °C using 5 bar O₂ and HNO₃ as a co-oxidant (Entry 9; Table 3). The process achieved 98% DFF selectivity at 97% HMF conversion in just 2 min and a further oxidation towards FDCA occurred at an increased contact time. However, the crystallization of FDCA was noticed at longer contact times (>6 min) due to its low solubility in DCE. Therefore, an alternative solvent system is needed to mitigate this issue. Yang et al. [51] investigated the continuous production of FDCA from the aerobic oxidation of HMF in a microreactor packed with an Au/CeO₂ catalyst (Entry 10; Table 3). An FDCA selectivity (90%) was achieved at a complete HMF conversion within a short residence time (41 s) under 90 °C and 6 bar O₂, corresponding to a space–time yield of 3.08 kg_FDCA/(L_reactor ·h) that was 1–2 orders of magnitude greater than that of conventional stirred tank reactors. A simple kinetic analysis suggested the intermediate oxidation of 5-hydroxymethyl-2-furancarboxylic acid (HMFCA) to 5-formyl-2-furancarboxylic acid (FFCA) as the rate-controlling step in the reaction network. However, the HMF concentration used was low (0.025 M in water) with the presence of a base (NaOH added at a molar ratio of 4 to HMF), indicating further room for improvement when it comes to feasible industrial application.

Furfuryl alcohol (FA), derived from furfural hydrogenation, is a renewable chemical widely used in resins, coatings, construction materials, and specialty chemicals. Duan et al. [52] demonstrated efficient hydrogenation of furfural (dissolved in isopropanol) to FA in a packed bed microreactor using an Cu/SiO₂ catalyst, achieving 100% furfural conversion and 99.5% FA yield in 65 s under 80 °C and 0.1 MPa with H₂ as the hydrogen source (Entry 11; Table 3). The catalyst operated stably for 16 h. For the catalytic transfer hydrogenation with isopropanol as the hydrogen donor, N₂ was introduced to facilitate liquid–solid mass transfer and the packed bed microreactor exhibited a maximum FA yield of 96.3% at 100 °C and 3.1 MPa. Moreover, the experimental data could be fitted by the Langmuir–Hinshelwood model with non-competitive and dissociative adsorption of hydrogen.

Hommes et al. [25] studied LA hydrogenation to γ-valerolactone (GVL; a promising bio-based solvent and fuel additive) using PFA capillary microreactors filled with Ru/C catalysts (Entry 12 in Table 3; Figure 1c and Figure 2e). A complete conversion of LA with an 84% yield of GVL was achieved at 130 °C and 12 bar H₂. Despite these promising outcomes, the reaction rate was affected by mass transfer limitations, attributed to the relatively low gas–liquid mixture flow rates and relatively large catalyst particles (0.3 and 0.45 mm) in use.

Glycerol, one important byproduct of biodiesel production processes, can be valorized to important building blocks and intermediates such as glyceric acid, oxalic acid, and tartronic acid. Zope et al. [53] evaluated the oxidation of glycerol in water by molecular oxygen in both the packed bed microreactor and slurry-type batch reactor, applying a commercial 1.6 wt.% Au/TiO₂ catalyst (Entry 13; Table 3). It was concluded that, in the microreactor, the high O₂ flow rates applied possibly facilitated the direct gas–catalyst contact and reduced the film thickness around wetted catalyst particles. This rendered a higher oxygen availability and selectivity (e.g., >30%) towards secondary oxidation products like tartronic and oxalic acids even at low glycerol conversions under 60 °C and 10 bar O₂ compared with batch (selectivity typically < 5% even at glycerol conversions as high as 83%). Moreover, the microreactor maintained a stable pH level (because of the rapid product removal), making it more effective with lower base concentrations required.

2.1.4. Enzymatic Chemical Synthesis

Enzymatic reactions are becoming increasingly popular recently. This is because of the simplified purification process, relatively mild conditions (such as neutral pH and moderate temperatures), and their biodegradable and renewable characteristics. The combination of packed bed microreactors and enzymatic reactions contributes significantly to sustainable and eco-friendly synthesis towards pharmaceutical and chemical production, offering an alternative to traditional chemo-catalytic methods [54,55].

A benchmark study focusing on the engineering aspects (kinetics, operating parameter selection, etc.) of enzymatic reactions was carried out by Dencic et al. [54]. In this study, the transesterification reaction between ethyl butyrate and 1-butanol, catalyzed by Novozym 435 (Candida antarctica lipase B immobilized on a macroporous support), was employed as a model reaction. A complete conversion was achieved in about 4 min residence time in the packed bed microreactor because of good mixing and well-controlled reaction conditions. A 12 h continuous operation without apparent catalyst deactivation under mild condition (70 °C) was achieved, indicating the good stability of enzyme in continuous flow mode. Similarly, Strniša et al. [56] investigated the transesterification of vinyl butyrate and 1-butanol into butyl butyrate (an industrially relevant ester) in a microreactor chip packed with an immobilized Novozym 435 catalyst. Microreactors of various dimensions were tested and all performed almost equally at the same residence time, showing that favorable hydrodynamics and thus enzyme activity were maintained in the scale-up process. At 24 °C, the highest substrate conversion of approximately 95% could be achieved in less than 0.8 min. Brás et al. [55] developed microreactors packed with enzyme-functionalized silica beads for the synthesis of levodopa (L-DOPA) from tyrosine and of dopamine (a neurotransmitter playing a crucial role in the human body) in single or consecutive steps (from L-DOPA or tyrosine). Yields of ca. 30% for L-DOPA and 70% for dopamine could be achieved. A successful upscaling of L-DOPA production with a 780-fold increase was demonstrated in a milliliter reactor by using the same bead packing as the microfluidic system and thus preserving the same transport and reaction characteristics in microflow.

2.1.5. Organic Synthesis via Oxidation of Aromatics and Alkenes

Zhang et al. [57] proposed the approach of using microreactor technology to synthesize acetophenone (AP), a versatile organic compound applied as a fragrance ingredient and precursor of producing resins, coatings, and plasticizers (Entry 14; Table 3). They prepared the active carbon (AC)-supported N-hydroxyphthalimide (NHPI) catalyst combined with tert-butyl nitrite (TBN) as the co-catalyst for the aerobic oxidation of ethylbenzene (EB) to AP in acetonitrile (MeCN). By utilizing a packed bed microreactor, a quick exploration across a broad range of operation parameters was conducted. A high yield of AP (ca. 84%) was obtained at 90 °C and O₂ pressure of 5 bar within 2.4 min, while it took over 6 h in batch to reach such a yield under similar reaction conditions but over a homogenous NHPI catalyst. This significantly shortened reaction time in microflow was attributed to the remarkable enhancement of mass transfer in the micro packed bed.

Packed bed microreactors have also shown promising intensification potential in the ozonation of various substrates such as the synthesis of benzaldehyde from styrene [58] as well as the degradation of organic pollutants and antibiotics (e.g., when coupled with catalysts based on or supported by Al₂O₃) [59,60] The continuous microflow system exhibited a much improved reaction rate or product selectivity/yield when compared with other conventional (or large-scale) reactors thanks to the improved gas–liquid–solid mass transfer rates and the precise control over the reaction time in packed bed microreactors.

The idea of packing inert particles was also adopted in non-catalytic reaction systems. The solvent-free and metal catalyst-free oxidation of dodecenes to valuable ketone products under the strong and economical oxidizing agent N₂O (liquefied) was investigated in a microreactor packed with soda lime glass beads [61] (Entry 15; Table 3). Under the studied conditions of 300–375 °C and 62–103.4 bar, N₂O vaporized, forming a gas–liquid system. The yield of the desired C12 ketone product was found up to 45% and the conversion of dodecane isomers was up to 77%. The rate-limiting step was determined as the cycloaddition of N₂O with the alkene.

2.2. Liquid Fuel Production

The increasing demand for liquid fuels (e.g., diesel, gasoline), coupled with concerns over fossil fuel depletion and their significant environmental impact, has spurred the development of renewable alternatives, such as synthetic fuels from renewable natural gas (i.e., biogas), CO₂, biomass. or waste. Microreactors have gained increasing attention for the (distributed) production of renewable fuels and are envisaged to offer high reaction efficiency, better economic feasibility, and faster installation, making them a promising alternative to conventional reactors [62,63]. Typical examples were demonstrated for biodiesel synthesis [64], Fischer–Tropsch synthesis [65,66], and the synthesis of methanol [67,68] as well as DME [69,70].

2.2.1. Biodiesel Production

Biodiesel is seen as an environmentally friendly fuel alternative for diesel engines because of its renewable source and low exhaust gas emission compared to fossil fuels [71]. It is normally produced by transesterification between edible oils (e.g., vegetable oil and animal fats) and short-chain alcohols with the aid of a base catalyst (e.g., KOH) (Scheme 6). The use of homogeneous catalysts complicates product separation and catalyst reuse, leading to increased operational costs and high water consumption. One earlier work in using microreactors coupled with heterogeneous catalysts for biodiesel synthesis was demonstrated by Chueluecha et al. [64]. They achieved the efficient transesterification of refined palm oil with methanol over a CaO catalyst packed in a microchannel. The optimum operating conditions (methanol-to-oil ratio = 24:1, 65 °C, ambient pressure) were similar to those in conventional packed bed and batch reactors. However, high-quality biodiesel with a methyl ester content of 96.5% was synthesized under a much shorter residence time (8.9 min in the microreactor vs. 2–8 h in batch). This reaction was further optimized by incorporating isopropanol as a co-solvent with methanol in the same microreactor [19] (Entry 5; Table 2). This resulted in a significantly improved mass transfer rate between oil and methanol. A higher product quality (with a purity of methyl esters of 99%) and reduced methanol consumption (methanol-to-oil ratio = 20:1) were obtained within an even shorter residence time (6.5 min). However, a gradual pressure drop increase in the bed was observed, ascribed to the partial blockage of the microchannel by large particles resulting from coalescence of the formed calcium–glycerine complex during the reaction. To mitigate this blockage, a frequent solvent wash was shown to work (at least partially).

Recently, the immobilized lipase catalyst Lipozyme TL IM was applied in the synthesis of fatty acid ethyl ester via transesterification of acid oil (prepared from rice bran oil soapstock) and ethanol in a packed bed microreactor [72] (Entry 6; Table 2). The optimal reaction conditions were found to be a water content of 4% (essential to maintain enzyme activity), a temperature of 20 °C, and a molar ratio of oil to ethanol of 1:4, rendering a maximum biodiesel yield of ca. 92% at a mean residence time of 4 h. Removing the byproduct glycerol as a lipase inhibitor during the reaction helped maintain the lipase’s relative activity over 82% across 27 cycles (108 h).

Inedible microalgae are considered as third-generation biofuels because of their high lipid content, fast growth rate, and wide distribution in the ocean [73]. The hydrodeoxygenation process of microalgae oil is critical for hydrocarbon production from lipids (mainly triglycerides). Zhou and Lawal [74] evaluated the performance of the presulfided NiMo/γ-Al₂O₃, a conventional catalyst used in the hydrotreating process, for the hydrodeoxygenation of microalgae oil using packed bed microreactors. The microreactor performance (i.e., reactant conversion and product STY) was enhanced with the decreasing microchannel diameter due to the improved mass transfer. A nearly complete conversion of microalgae oil (98.7%) was obtained under ca. 35 bar H₂ and 360 °C with an H₂/oil ratio of 1000 NmL/mL within 1 s, giving a C13 to C20 hydrocarbon yield at 56.2%. The catalyst activity kept stable for at least 7 h because H₂ also protected the catalyst surface from accumulation of the oxygenated intermediates.

2.2.2. Fischer–Tropsch Synthesis

Fischer–Tropsch (FT) synthesis is a process converting synthesis gas, a mixture of CO and H₂, into liquid hydrocarbons over typically cobalt- or iron-based catalysts, providing an alternative to conventional crude oil-derived fuels (Scheme 7). The diverse feedstock (especially syngas derived from renewable sources such as biomass and biogas) and the resulting ultra-clean fuels contribute to green and sustainable energy systems and the carbon footprint reduction.

A good temperature control is crucial for obtaining the desired selectivity towards liquid hydrocarbons and extending the catalyst lifetime in highly exothermic FT reactions. The good heat management ability of packed bed microreactors holds promises for the intensified FT process, as reported by Cao et al. [75] (Entry 2; Table 4). The small catalyst particles (45 and 150 μm; 30 wt.% Co-4.5 wt.%Re/γ-Al₂O₃) used, in conjunction with the confined microreactor geometry, provided efficient heat dissipation between catalysts and the wall, which largely inhibited the methanation reaction pathway and maintained the catalyst activity. A productivity of 2.14 g C₂₊/(g_cat·h) and CO conversion of 90.2% were obtained with a low methanol selectivity (3.4%) from a stoichiometric ratio of H₂/CO at 2 and a high-gas hour–space velocity (GHSV) of 22,886 h⁻¹ under 224 °C and 35 bar. Similarly, Chambreya et al. [24] (Entry 3; Table 4) found that, due to the efficient radial heat transfer in their microreactor packed with a laboratory-made 25 wt.% Co-0.1 wt.%Pt/ γ-Al₂O₃ catalyst (Figure 1b), the temperature profile could be well controlled even with a fast temperature ramping rate of up to 15 K/min during the startup. A high initial CO conversion of ca. 62% was obtained in the microreactor with a 1.1–1.4 g C₅₊/(gc_at·h) productivity and low selectivity to methane of ca. 8% under 200 °C, 35 bar, and GHSV of 13.2 NL/(g_cat·h). In contrast, the centimetric fixed bed reactor exhibited a lower initial CO conversion (ca. 40%) under similar conditions. Moreover, the latter larger diameter reactor had to use a diluted catalyst bed (with SiC) and very slow temperature ramping (0.05 K/min) during the startup, in order to prevent an uncontrolled temperature rise and hot spots that could lead to major catalyst deactivation. The negligible heat transfer issue in microreactors was also proven by Potgieter et al. [76] and Myrstad et al. [65], where the packed bed microreactor was shown to work efficiently under severe conditions in FT synthesis. A reactor simulation considering the residence time distribution, vapor–liquid equilibrium of the product condensation, and a microkinetic model was conducted by Loewert et al. [77] for FT synthesis over commercial 20 wt.% Co-0.5 wt.%Re/Al₂O₃ catalysts packed in a microreactor. The modeled kinetic parameters obtained via comparison with the experimental data were on the same order of magnitude as the literature values. The simulation results were shown useful for the FT synthesis operation and control in microreactors under dynamic mode.

Complex hydrodynamics existed during FT reactions in packed bed microreactors because of the formed liquid hydrocarbons, though initially a gas-solid system is present. The liquid holdup in FT processes carried out in packed bed microreactors was estimated by Knobloch et al. [78] to be approximately 3% (based on the total reactor volume). The lower liquid holdup in microreactors is preferable compared with typical trickle bed reactors where the resulting liquid phase may cover the solid catalyst and influence the mass transfer between gas and FT catalysts.

Recently, the pilot-scale demonstration of FT synthesis in packed bed microreactors was reported by Na et al. [79]. They fabricated a compact reactor block composed of 528 parallel microchannels to perform FT synthesis over a packed 12 wt.% Co/γ-Al₂O₃ catalyst with a production capacity of 0.5 barrels of synthetic hydrocarbons per day. The modular microreactor exhibited an overall CO conversion of 83% (at around 220 °C and 20 bar) and a stable temperature control within 270 h, while a multi-tubular fixed bed reactor using centimeter-sized tubes experienced reaction runaway in parallel experiments. However, the methane selectivity was still high (50.13%) in the pilot plant operation, requiring further optimization in the catalyst preparation and microreactor operation. A further scale-up of FT synthesis with a Co-Re/γ-Al₂O₃ catalyst packed in microreactors at a production rate of liquid hydrocarbons of 500 barrels/day was shown as possible in the simulation by Knochen et al. [80], requiring a microreactor volume of ca. 20 m³.

2.2.3. Methanol Synthesis

Methanol is considered a promising fuel owing to its high energy density, clean combustion characteristics, and the potential for sustainable production from renewable feedstocks, making it an important candidate in the transition to greener energy solutions [81]. Methanol synthesis from (renewable) syngas is also a cornerstone of the chemical industry due to its wide range of applications in addition to fuels, such as solvent and feedstock for producing other chemicals (Scheme 8). Bakhtiary-Davijany et al. [82] investigated methanol synthesis over a Cu/ZnO/Al₂O₃ catalyst packed within a multi-slit-type microreactor integrated with a heat exchanger (Entry 4 in Table 4; Figure 2a). They found that the negligible heat transfer resistance in the microreactor facilitated methanol synthesis, which is typically limited by thermodynamic constraints. A near-equilibrium conversion of CO at ca. 60% with methanol selectivity exceeding 98.7% was achieved at 250 °C and 80 bar at a low contact time (approximately 470 ms·g_cat/mL). Meanwhile, they claimed that the employed microreactor could improve the mixing performance even under the low Reynolds numbers involved (about 1) [83]. The internal pore diffusion was experimentally shown as neglectable for catalyst particle diameters below 125 μm. The external mass transfer limitation was also excluded under varied operational conditions (235–255 °C and 50–90 bar).

The methanol production from CO₂ hydrogenation was also achieved in packed bed microreactors (Scheme 9) [84]. For instance, Jiang et al. [85] tested the performance of Pd/In₂O₃/SBA-15 powder catalysts in the microreactor (Entry 5; Table 4). The well-dispersed Pd and In₂O₃ nanoparticles on the SBA-15 support increased oxygen vacancies and enhanced hydrogen adsorption, improving its activity. A methanol selectivity of 83.9% and CO₂ conversion of 12.6% were obtained in the microreactor at 260 °C and 50 bar within a continuous operation of 120 h, indicating stable and promising catalyst behavior. Since the pressure increase shifts the equilibrium of CO₂ hydrogenation to methanol, Tidona et al. [86] conducted this reaction in stainless steel capillary microreactors at pressures up to 950 bar. Both the homemade Cu/Al₂O₃ and commercial Cu/ZnO/Al₂O₃ catalysts were tested in separate packed bed microreactors under 170–280 °C. The feed mixture (H₂/CO molar ratio at 4) was expected in the supercritical state and the outlet product stream was assumed in the single phase. The methanol space–time yields under such high-pressure conditions were found 3–15 times higher than those obtained when the reaction was conducted at 30 or 75 bar over the two catalysts. The results further showcase the benefits of microreactors in handling reactions safely at high pressures.

2.2.4. Dimethyl Ether Synthesis

Dimethyl ether (DME) is a clean and versatile fuel with applications as an alternative to diesel and liquefied petroleum gas and a potential hydrogen carrier [70]. DME is traditionally produced by methanol dehydration using acidic catalysts like HZSM-5 or γ-Al₂O₃. Hayer et al. [69,70,87] proposed the direct DME production from syngas (Scheme 10) in a packed bed microreactor that integrated methanol synthesis and its dehydration in a single unit using a dual-function catalyst (physical mixture of CuO/ZnO/Al₂O₃ and γ-Al₂O₃). This one-step synthesis approach shifts the reaction towards higher methanol conversion while avoiding the need for methanol separation and reducing complexity compared with conventional sequential reactor systems. At 50 bar and 250 °C, a DME yield up to ca. 50% at ca. 90% CO conversion was achieved in the microreactor, where the experiments were proven to be free from external/internal mass transfer limitations [87] (Entry 6; Table 4). Both experiments and simulations revealed an isothermal operation in the heat exchange-coupled microreactor. These findings highlight the packed bed microreactor’s potential for process intensification and its application in the sustainable and small-to-medium-scale production of DME.

2.3. Catalyst Screening and Kinetic Study

Catalyst screening is a systematic approach to identify and develop optimal catalysts for specific reactions, including the broad screening of reaction systems and parameters, catalyst preparation-structure-performance relationship, and kinetic study. For solid-catalyzed reactions involving liquids, the catalyst activity evaluation or reaction kinetic research is usually conducted in batch reactors, e.g., autoclaves [88]. In batch reactors, every independent experiment requires a series of procedures including, typically, the loading of reactants and catalysts, reaction testing, sampling, and cleaning. The time for temperature (and pressure) changes from ambient conditions also needs considering [61,89]. While in continuous flow, the reaction parameters can be adjusted flexibly without stopping the reaction, and the corresponding results can be analyzed directly. Moreover, the ability of a catalyst to maintain its performance (i.e., activity, selectivity, and structural integrity) over an extended period of use under industrial conditions needs to be assessed before its commercial application [90]. In batch reactors, the catalyst stability (reusability) is typically evaluated through consecutive experiments, requiring catalyst filtration and reloading between each run. In such a case, catalyst deactivation due to metal leaching (often in trace amounts) cannot be accurately identified because the leached metal may redeposit on the support [91]. This issue can be effectively addressed using flow reactors, which allow for the continuous monitoring of both the product and catalyst performance over time.

Potential limitations in external mass transfer and internal diffusion within catalyst particles could influence the overall reaction rate/efficiency, which need to be eliminated for the determination of intrinsic kinetics or catalyst performance [92]. For gas–solid catalyzed reactions, the external mass transfer resistance can be usually neglected, e.g., by working at sufficiently large gas flow rates [82,83]. For multiphase gas–liquid or liquid–liquid reactions, the external mass transfer resistance is highly dependent on the contacting pattern between fluids and tends to prevail in traditional large-scale reactors [35]. The plot of reaction parameters (e.g., reaction rate/conversion) versus the liquid velocity under the same residence time could be researched, and the limit point with unchanged reaction parameter values indicates the neglectable liquid–solid mass transfer resistance [3,92,93]. The internal diffusion resistance normally decreases for smaller particles due to the short diffusion distance inside. The contribution from intraparticle diffusion resistance can be estimated by calculating the catalyst effectiveness factor (with values close to 1 indicating the absence of diffusion limitation) [94]. Thus, a proper selection of intensified flow reactors and catalyst size as well as operational conditions is important.

With the increasing attention on process intensification via miniaturization, packed bed microreactors emerge as a promising option to facilitate the catalyst performance testing and reaction kinetic study in continuous flow. Microreactors with small volumes can be utilized for rapid catalyst evaluation, minimizing the consumption of (expensive) reactants/catalysts and generation of waste. Meanwhile, its precise manipulation of reaction conditions, fast response, and enhanced transport rates make it a reliable tool for the extraction of intrinsic catalyst activity and reaction kinetics [95]. Moreover, the integration of online (operando) monitoring instruments is easily attainable within microreactors, allowing to follow reaction events and catalyst working conditions, providing rich information on catalytic behavior [96].

2.3.1. Hydrogenation Reactions

The hydrogenation of functionalized nitroaromatics is a critical chemical transformation with significant importance across pharmaceutical, agrochemical, and polymer industries. This kind of selective hydrogenation reaction is industrially conducted in (semi-)batch slurry reactors using traditional Pt−Pb/CaCO₃ catalysts [97]. Nuzhdin et al. [93] tested the hydrogenation of functionalized nitroaromatics (Scheme 11) using a continuous three-phase microreactor packed with a nano-dispersed 1.9 wt.% Au/Al₂O₃ catalyst, where the substituent X included -Cl, -CH = CH_2, and -C(O)CH₃ (Entry 16; Table 3). The external mass transfer limitation was eliminated with an H₂ flow rate of 60 mL/min and liquid (nitroaromatic compound in toluene) flow rate of 0.5 mL/min. No considerable effect of the catalyst grain size increase (from 125–250 μm to 350–500 μm) was found on the conversion under specific tests, indicating little/no internal mass transfer limitation. The hydrogenation performance was influenced by the substitution type and position of the functional atom to the NO₂ group. Overall, a high selectivity to the target product (>92%) was obtained under relatively low temperatures (60–110 °C) compared to the batch reactor. Even though the inevitable catalyst deactivation existed due to the blockage of active sites by the carbonaceous deposit, the oxidative regeneration of catalysts could be achieved under an air flow at 400 °C to restore its activity. To improve the catalyst’s longevity and performance, future research is suggested to focus on modifying the support surface and optimizing reaction conditions.

Vile et al. [97] studied a similar reaction system using a packed bed microreactor with a ligand-capped Pt-HHDMA/C catalyst (HHDMA: hexadecyl(2-hydroxyethyl)-dimethylammonium dihydrogen phosphate), as shown in Scheme 11 (X including carbonyl, alkyl, halogen, and amino substitution) (Entry 17; Table 3). The microreactor system was equipped with six parallel packed beds and facilitated efficient heat/mass transfer for high-throughput catalytic testing. They claimed that Pt-HHDMA/C exhibited superior activity, chemo selectivity, and leaching resistance compared with the traditional and industrially relevant Pt−Pb/CaCO_3, because their catalyst owned improved H₂ activation and controlled nitroarene adsorption, which reduced unselective reaction pathways. This mechanistic insight underscores the potential of HHDMA-modified Pt catalysts in hydrogenation processes and guides future catalyst design.

The kinetics in the hydrogenation of m-dinitrobenzene (to m-phenylenediamine) and halogen nitroaromatics (to aromatic anilines), which find importance in the fine chemical industry, were studied by Duan et al. [98,99] in packed bed microreactors. In m-dinitrobenzene hydrogenation, the reaction network of the complex reduction in two nitro groups over Pt/C and Ni/SiO₂ catalysts was established and the kinetic parameters were obtained [98]. Even though Ni/SiO₂ catalysts are more effective at inhibiting the formation of byproducts (azoxy compounds), Pt/C catalysts exhibited a much higher hydrogenation activity. For the hydrogenation of halogenated nitroaromatics (p-chloronitrobenzene, p-bromonitrobenzene, and p-iodonitrobenzene) over Pt/C catalysts [99], the mechanism involves the simultaneous hydrogenation and hydrodehalogenation, yielding halogenated aniline and nitrobenzene, which can be subsequently converted to aniline. Based on this mechanism, the kinetic models were established from data collected from microreactors and validated with the literature values.

Tadepalli et al. [100] utilized a packed bed microreactor for the hydrogenation of o-nitroanisole to o-anisidine over a 2 wt.% Pd/zeolite catalyst. The study confirmed no significant internal mass transfer resistance (with effectiveness factors between 0.7 and 0.97) and minimal external mass transfer resistance under optimal conditions identified through experimental analysis. In addition, the radial heat transfer resistance was analyzed to be negligible. Consequently, the microreactor is suitable for conducting kinetic studies. The microreactor was operated under a differential mode by limiting the conversion below 10%. The reaction data could be fitted well by assuming Langmuir–Hinshelwood (L–H) rate expressions with the presence of intermediate 2-methoxynitrosobenzene. Yu et al. [3] further investigated the reduction in carboxyl-functionalized nitroaromatics (Entry 18; Table 3), where 2-(4-nitrophenyl) butanoic acid (NBA) was used as the substrate in the presence of a Pd/C catalyst to produce 2-(4-aminophenyl) butanoic acid (ABA; an essential intermediate in the synthesis of Indobufen). By applying the Langmuir–Hinshelwood–Hougen–Watson (LHHW) approach, a plausible kinetic model was developed based on the reaction mechanism of competitive adsorption dissociative adsorbed H₂, with the formation of 2-[4-(hydroxyamino)phenyl]butanoic acid as the intermediate involved in one of the two pathways to produce NBA (Scheme 12).

Many other hydrogenation reactions were also studied utilizing packed bed microreactors as a reliable platform to extract kinetics information, screen catalysts, and operation parameters. Van Herk et al. [88] compared the kinetic data of biphenyl hydrogenation over a Pt-Pd/Al₂O₃ catalyst (as a model fast reaction) obtained in a microreactor containing six parallel beds with those in a standard batch autoclave (Entry 19; Table 3). The intrinsic kinetic values in batch could be reproduced in the microreactor, showing the promising potential of flow approach. To obtain the same reaction results across each of the parallel beds, an even loading of the catalyst and diluent (SiC or glass) should be ensured to avoid the bed segregation. The bed segregation was affected by both the catalyst and diluent sizes and their filling procedure. The conversion level could vary significantly in the segregated bed due to the presence of a shorter bed, flow bypassing, and poor thermal management. Ashok et al. [101] verified the efficient application of packed bed microreactors in the early-stage testing of a nanoparticle immobilized Pd@UiO-66 MOF catalyst in semi-hydrogenation of phenylacetylene to styrene. Gas–liquid slug flow regime was maintained upstream of the bed that contained only 48 mg of catalyst to achieve reproducible reaction results. Out of 88 experiments under a wide range of conditions, the optimal styrene selectivity of 70 % at a phenylacetylene conversion of 98.9 % was attained under low H₂ pressure (0.28 bar) and mild temperature (45 °C). The catalyst remained stable over 20 h time on stream.

Joshi and Lawal [102] evaluated the hydrodeoxygenation of acetic acid (AA) in both vapor and liquid phases using the reduced sulfided NiMo/Al₂O₃ catalyst packed in microreactors. The reaction with vapor-phase AA showed a higher conversion (60%) under atmospheric pressure than that with liquid-phase AA (5% at 16.5 bar), attributed to the limited hydrogen diffusion rate in the latter gas–liquid–solid system. The reaction performance (e.g., reaction rate and conversion, space–time yield) improved with decreasing microreactor diameters due to the higher hydrogen concentration on the catalyst surface, resulting from enhanced gas–liquid mixing in smaller microchannels. The successful demonstration of AA hydrodeoxygenation at low pressures suggests the potential of this catalyst and microreactor system for treating pyrolysis oil, which typically requires high-pressure conditions. In their follow-up work, the hydrodeoxygenation of 4-propylguaiacol (a model compound of lignin derivatives in pyrolysis oil) was tested using the same catalyst in a packed bed microreactor [103]. The use of hexane as a solvent improved hydrogen dissolution, enhancing the reaction performance. A maximum conversion of 4-propylguaiacol (80%) was achieved when the hydrogen adsorbed on the catalyst surface reached saturation at 2.2 MPa and 400 °C. A rate expression based on the L–H reaction mechanism was proposed and, thus, the derived kinetic model could describe the experimental data, indicating the surface reaction as the rate-controlling step.

N-ethylcarbazole is considered a promising liquid organic hydrogen carrier due to its high hydrogen storage capacity and thermal stability. Fan et al. [104] studied the hydrogenation of N-ethylcarbazole (dissolved in methylcyclohexane) over a 1 wt.% Ru/Al₂O₃ catalyst prepared in-house, utilizing a packed bed microreactor. The experiments achieved a complete conversion with 99% selectivity towards dodecahydro-N-ethylcarbazole under relatively mild conditions (80 °C and 4 MPa). A kinetic model was developed based on the proposed stepwise hydrogenation reaction network and reaction tests in microflow that were believed free from mass transfer limitation. The model suggests the hydrogenation of the intermediate 9-ethyl-octahydrocarbazole into dodecahydro-N-ethylcarbazole as the rate-limiting step.

The complex hydrodynamics of gas–liquid–solid reactions are unavoidable when using a packed bed microreactor. To address this, Duan et al. [92] introduced a tube-in-tube contactor upstream of the micro packed bed, ensuring that the organic phase was saturated with H₂ for the subsequent hydrogenation reaction (Entry 7 in Table 2; Figure 2b). This approach simplified the three-phase system into a more manageable liquid–solid system for kinetic study. Moreover, utilizing the automatic flow rate ramping strategy combined with online analysis in continuous mode, over 10,000 data points of varied residence times, and the corresponding conversion data could be collected within one hour, with the additional advantage in the reduced reagent cost. The accuracy of this technique, attributed to the ample data availability and elimination of mass transfer resistance, was validated through the hydrogenation of α-methylstyrene and nitrobenzene as model reactions. The kinetic parameters derived in this platform were found consistent with the literature findings.

2.3.2. Oxidation Reactions

Catalyst screening and kinetic study were well demonstrated in packed bed microreactors for heterogeneously catalyzed oxidation reactions. A packed bed microreactor was utilized by Wu et al. [91] to analyze the deactivation characteristic of a 1 wt.% Au-Pd/TiO₂ catalyst in cinnamyl alcohol oxidation to cinnamaldehyde (with versatile applications in the food, pharmaceutical, and agricultural industries) (Entry 20; Table 3). Even though the catalyst was pre-reduced in pure hydrogen flow, the alcohol conversion experienced a significant drop from 62% to 10% after 22 h on stream with a minimal selectivity decrease (maintained at 62%) at 100 °C and 4 bar. The irreversible deactivation was correlated with Pd leaching, with approximately 24% of the total Pd loading lost within 7 h of operation. Although a higher cinnamaldehyde selectivity was achieved under elevated oxygen concentrations, this advantage was offset by a reduced conversion (caused by a more severe metal leaching and the strong adsorption of deposits formed). However, the same catalyst showed a stable activity in benzyl alcohol oxidation over 20 h, indicating different deactivation mechanisms for the two reactions.

The stability of Au–Pd/TiO₂ catalysts prepared by different impregnation methods was studied by Al-Rifai et al. [95] for a solvent-free selective oxidation of benzyl alcohol in a packed bed microreactor. The rate of catalyst deactivation was found significantly impacted by the concentrations of chloride and Au employed during catalyst synthesis. The agglomeration of Au nanoparticles was observed during the reaction, leading to a decline in catalyst performance. Concurrently, the presence of Cl^- may promote the formation of carbon deposits on the catalyst surface (revealed by online Raman spectroscopy measurements at different axial bed positions in the employed silicon-glass microreactor). Such deposition could obstruct active sites and diminish the catalytic efficiency. Thus, catalysts synthesized with relatively low Au loadings and low (or moderate) chloride contents (e.g., 0.05 wt.% Au loading and 2.22 × 10⁻³ M Cl⁻) appeared to be resistant to deactivation.

Sankar et al. [105] introduced the “excess-anion” modified impregnation method (MIm) to synthesize a Au-Pd/TiO₂ catalyst with controlled size, composition, and nanostructure. The stability of an equimolar 1 wt.% Au-Pd/TiO₂ catalyst by MIM was evaluated in the solvent-free aerobic oxidation of benzyl alcohol in a packed bed microreactor [20] (Entry 21; Table 3). The MIm-prepared catalyst achieved around 80% conversion at 120 °C in flow (similar to the activity found in a batch stirred reactor) and demonstrated a stability over 75 h. In contrast, the same catalyst prepared by the sol immobilization method resulted in a lower activity (ca. 55% conversion) when tested in the microreactor with an occurrence of catalyst deactivation in 10–12 h. The MIm catalyst’s stability was attributed to its minimal size change in gold-rich bimetallic particles under reaction conditions.

Liu et al. [27] screened the aerobic oxidation of 4-isopropylbenzaldehyde to cumic acid (an important intermediate towards producing pharmaceutical compounds) using a packed bed microreactor (Figure 1e). The efficient catalyst for this reaction was tested to be the commercial 5 wt.% Pd/Al₂O₃. A high selectivity (over 98%) to cumic acid was achieved under varying flow conditions at typically 90 °C and 2 bar O₂, ascribed to the efficient heat transfer and precise control over gas–liquid flow that avoided the secondary oxidation reaction. The reaction time in the microreactor (ca. a few seconds) appeared much shorter than hours required in semi-batch operations for reaction completion.

Very recently, Bawa et al. [106] developed an automated silicon-based packed bed microreactor platform, interfaced with computer software for automated control and analysis, to screen kinetic models for the gas-phase catalytic combustion of methane over a commercial 5 wt.% Pd/Al₂O₃ catalyst. The microreactor contained ca. only 10 mg catalyst, and the bed was proven to work isothermally. The platform allowed the online identification of candidate models by estimating kinetic parameters and applying model discrimination techniques (within 48 h). The two models of Langmuir–Hinshelwood and Mars–van Krevelen were both identified as probable for complete methane oxidation. The ability to conduct a large number of experiments with minimal user supervision of this automated platform makes it a valuable tool for refining catalytic reaction mechanisms and optimizing kinetic models.

2.3.3. Hydrolysis, Dehydration, and Chlorination Reactions

The utilization of packed bed microreactors for catalyst screening was shown in other types of reactions, though to a much less extent. One typical example is seen in the screening of solid acid catalysts (SO₄²⁻/γ-Al₂O₃ and HND-580) in the dehydration reaction of xylose to furfural [50]. A few kinetic studies of other reactions in the packed bed microreactor platform were also performed, such as in the synthesis of phosgene [43] and the production of HMF from inulin including hydrolysis and dehydration reactions in sequence [47]. More details are found in Section 2.1.

Figure 2. Application examples of packed bed microreactors in different phasic systems. (a) Gas–solid (G–S) system for methanol synthesis; the reactor module, arrangement of reaction slits, and oil heat-exchange channels are shown. Reproduced with permission from ref. [82]. Copyright 2011 Elsevier. (b) Liquid–solid (L–S) system using an automated flow platform for α-methylstyrene hydrogenation over Pd/Al₂O₃ with a tube-in-tube gas–liquid contactor before the reactor. Reproduced with permission from ref. [92]. Copyright 2020 Royal Society of Chemistry. (c) Liquid–liquid–solid (L–L–S) system for HMF synthesis from glucose in water-2-methyltetrahydrofuran biphasic solvents over P-TiO₂ catalyst. Reproduced with permission from ref. [48]. Copyright 2022 Elsevier. (d) Gas–liquid–solid (G–L–S) system for aerobic oxidation of benzyl alcohol over homogenous Cu/TEMPO catalyst with inert glass beads packed in the microreactor. Reproduced with permission from ref. [35]. Copyright 2022 Royal Society of Chemistry. (e) G–L–S system for levulinic acid hydrogenation to γ-valerolactone over Ru/C catalyst. Reproduced with permission from ref. [25]. Copyright 2020 Elsevier.

Figure 2. Application examples of packed bed microreactors in different phasic systems. (a) Gas–solid (G–S) system for methanol synthesis; the reactor module, arrangement of reaction slits, and oil heat-exchange channels are shown. Reproduced with permission from ref. [82]. Copyright 2011 Elsevier. (b) Liquid–solid (L–S) system using an automated flow platform for α-methylstyrene hydrogenation over Pd/Al₂O₃ with a tube-in-tube gas–liquid contactor before the reactor. Reproduced with permission from ref. [92]. Copyright 2020 Royal Society of Chemistry. (c) Liquid–liquid–solid (L–L–S) system for HMF synthesis from glucose in water-2-methyltetrahydrofuran biphasic solvents over P-TiO₂ catalyst. Reproduced with permission from ref. [48]. Copyright 2022 Elsevier. (d) Gas–liquid–solid (G–L–S) system for aerobic oxidation of benzyl alcohol over homogenous Cu/TEMPO catalyst with inert glass beads packed in the microreactor. Reproduced with permission from ref. [35]. Copyright 2022 Royal Society of Chemistry. (e) G–L–S system for levulinic acid hydrogenation to γ-valerolactone over Ru/C catalyst. Reproduced with permission from ref. [25]. Copyright 2020 Elsevier.

Table 2. An overview of selected liquid(–liquid)–solid reaction systems reported in packed bed microreactors.

Entry	Reactor a	Catalyst	Reaction	Operational Conditions b	Results	Reference
1	Fluoroelastomeric capillary, D = 1.6 mm, L = 60 cm	TEMPO/AO	Oxidation of primary and secondary alcohols	Aqueous phase: NaOCl and KBr, Organic phase: alcohol in CH2Cl2 or EtOAc. Qaq = 56 µL/min, Qorg = 44 μL/min, ~4.8 min residence time, T: 0 °C	>99% conversion of 4-chlorobenzyl alcohol and 93% yield of 4-chlorobenzaldehyde could be obtained, long-term catalyst stability within 9 h	[23]
2	Capillary, D = 1.65 mm, L = 15–30 cm	Amberlyst-15, 200 and 700 μm	Fructose dehydration to HMF	Organic phase: fructose in 1,4-dioxane and DMSO, Qaq = 0.1–0.35 mL/min, residence time: 3 min, T: 80–110 °C, atmospheric pressure	HMF yield of 92%, an STY of 9.2 × 10−5 mol/(mL·min) being 75 times higher than that in batch, good catalyst stability within 96 h	[46]
3	Stainless steel capillary, D = 4.35–5 mm, L = 10 cm	HND-580 260, 510, 770 μm	Production of fructose, HMF, and LA from inulin	Aqueous phase: inulin in (DMSO–)water solution, Qaq = 0.1–0.9 mL/min T: 70–190 °C, P: 2 MPa	Maximum yields of 89% for fructose, 81% for HMF, and 84% for LA obtained within a residence time of 7.26 min	[47]
4	PFA tubing, D = 4 mm, L = 2–8 cm	P-TiO2, 300–425 μm	Synthesis of HMF from glucose	Aqueous phase: glucose in NaCl solution, Organic phase: mTHF, Qorg/Qaq = 4:1, T: 150 °C, P: 10 bar	Glucose conversion of 96% and 72% HMF yield from 0.1 M glucose, STY at 145.64 mol/(m3·min)	[48]
5	Aluminum alloy plate-type microreactor, H × W × L = 0.5 mm × 1 mm × 60 mm	CaO	Biodiesel synthesis	Alcohol phase: methanol, Oil phase: refined palm oil (with or without iso-propanol as co-solvent), T = 65 °C, ambient pressure at outlet	>98% yield of methyl esters, with a purity of methyl esters of 99% at 6.5 min residence time	[19]
6	Stainless steel capillary, D = 4.8 mm, L = 5.1 cm	Lipozyme TL IM	Biodiesel synthesis	Alcohol: ethanol, Oil phase: fatty acid oil, water content: 0–6%, T: 0–40 °C, molar ratio of fatty acid oil to ethanol: 1:3–1:6	The highest biodiesel yield of 92% at a mean residence time of 4 h	[72]
7	Stainless steel capillary, D = 3 mm, L = 10 cm	Pd/Al2O3, 480 μm	Hydrogenation of α-methylstyrene and nitrobenzene	Organic phase: α-methylstyrene or nitrobenzene in methanol, a gas–liquid contactor for gas dissolution before the packed bed T: 26–46 °C, P: 1.0–1.3 Mpa	Accurate reaction kinetics was obtained in line with the literature findings	[92]

^a D and L are the microreactor diameter and (bed) length, respectively. H, W, and L are the height, width, and length of the rectangular microchannel, respectively. ^b Q_aq is the flow rate of aqueous phase, and Q_org is the flow rate of organic phase.

Table 2. An overview of selected liquid(–liquid)–solid reaction systems reported in packed bed microreactors.

Entry	Reactor a	Catalyst	Reaction	Operational Conditions b	Results	Reference
1	Fluoroelastomeric capillary, D = 1.6 mm, L = 60 cm	TEMPO/AO	Oxidation of primary and secondary alcohols	Aqueous phase: NaOCl and KBr, Organic phase: alcohol in CH2Cl2 or EtOAc. Qaq = 56 µL/min, Qorg = 44 μL/min, ~4.8 min residence time, T: 0 °C	>99% conversion of 4-chlorobenzyl alcohol and 93% yield of 4-chlorobenzaldehyde could be obtained, long-term catalyst stability within 9 h	[23]
2	Capillary, D = 1.65 mm, L = 15–30 cm	Amberlyst-15, 200 and 700 μm	Fructose dehydration to HMF	Organic phase: fructose in 1,4-dioxane and DMSO, Qaq = 0.1–0.35 mL/min, residence time: 3 min, T: 80–110 °C, atmospheric pressure	HMF yield of 92%, an STY of 9.2 × 10−5 mol/(mL·min) being 75 times higher than that in batch, good catalyst stability within 96 h	[46]
3	Stainless steel capillary, D = 4.35–5 mm, L = 10 cm	HND-580 260, 510, 770 μm	Production of fructose, HMF, and LA from inulin	Aqueous phase: inulin in (DMSO–)water solution, Qaq = 0.1–0.9 mL/min T: 70–190 °C, P: 2 MPa	Maximum yields of 89% for fructose, 81% for HMF, and 84% for LA obtained within a residence time of 7.26 min	[47]
4	PFA tubing, D = 4 mm, L = 2–8 cm	P-TiO2, 300–425 μm	Synthesis of HMF from glucose	Aqueous phase: glucose in NaCl solution, Organic phase: mTHF, Qorg/Qaq = 4:1, T: 150 °C, P: 10 bar	Glucose conversion of 96% and 72% HMF yield from 0.1 M glucose, STY at 145.64 mol/(m3·min)	[48]
5	Aluminum alloy plate-type microreactor, H × W × L = 0.5 mm × 1 mm × 60 mm	CaO	Biodiesel synthesis	Alcohol phase: methanol, Oil phase: refined palm oil (with or without iso-propanol as co-solvent), T = 65 °C, ambient pressure at outlet	>98% yield of methyl esters, with a purity of methyl esters of 99% at 6.5 min residence time	[19]
6	Stainless steel capillary, D = 4.8 mm, L = 5.1 cm	Lipozyme TL IM	Biodiesel synthesis	Alcohol: ethanol, Oil phase: fatty acid oil, water content: 0–6%, T: 0–40 °C, molar ratio of fatty acid oil to ethanol: 1:3–1:6	The highest biodiesel yield of 92% at a mean residence time of 4 h	[72]
7	Stainless steel capillary, D = 3 mm, L = 10 cm	Pd/Al2O3, 480 μm	Hydrogenation of α-methylstyrene and nitrobenzene	Organic phase: α-methylstyrene or nitrobenzene in methanol, a gas–liquid contactor for gas dissolution before the packed bed T: 26–46 °C, P: 1.0–1.3 Mpa	Accurate reaction kinetics was obtained in line with the literature findings	[92]

^a D and L are the microreactor diameter and (bed) length, respectively. H, W, and L are the height, width, and length of the rectangular microchannel, respectively. ^b Q_aq is the flow rate of aqueous phase, and Q_org is the flow rate of organic phase.

Table 3. An overview of selected gas–liquid–solid reaction systems reported in packed bed microreactor.

Entry	Reactor a	Catalyst	Reaction	Operational Conditions b	Results	Reference
1	Teflon capillary, D = 1.65 mm	TEMPO/silica, 160–240 μm	Oxidation of different alcohols	Liquid: alcohol and HNO3 in DCE, Gas: O2 QL = 0.1–0.4 mL/min, QG = 1.3–5 mL/min, T: 55–80 °C, P: 5 bar	98% benzyl alcohol conversion and 99% benzaldehyde yield within a residence time of 0.5 min, STY of 4.1 × 105 mol benzaldehyde/(molcat·L·h)	[33]
2	Silicon-glass chip, H × W × L = 0.3 mm × 0.6 mm × 3.2 mm	Au–Pd/TiO2, 64 ± 14 μm	Benzyl alcohol oxidation	Liquid: benzyl alcohol, Gas: O2, QL = 0–5 μL/min, QG = 0–0.6 NmL/min T: 120 °C, P: 2 bar	>90% selectivity to benzaldehyde in the partially wetted gas-continuous flow; 93% conversion and 80% selectivity at fully wetted gas continuous flow	[31]
3	Capillary, D = 4.35 mm	No solid catalyst, glass beads as packing, 300–350 μm	Benzyl alcohol oxidation	Liquid: benzyl alcohol and Cu/TEMPO in acetonitrile solution, Gas: 9% O2 in N2, molar ratio of O2 to alcohol = 0.65:1 T: 20–45 °C, P: 5–35 bar	Almost 100% yield of benzaldehyde within 30 s, STY = 7318.4 mol/(m3 h) at optimum condition, which is ca. 2 to 20 times higher than that in conventional flow reactors under similar conditions	[35]
4	Tempax glass chips, H × W × L = 0.9 mm × 0.6 mm × 40/120 mm	Pd–Au/TiO2, Pd/TiO2, Pd/Al2O3,	Direct synthesis of H2O2	Liquid: aqueous solution (of diluted sulfuric acid, phosphoric acid, sodium bromide), Gas: O2 and H2, QG = 5 NmL/min, QL = 0.1 mL/min, T: 20 °C, P: 0.95 MPa,	11.3 wt.% H2O2 with a yield of at 42%, H2O2 productivityat 17 × 10−4 mol/h	[39]
5	Glass chip, H × W × L = 0.9 mm × 0.6 mm × 4 0 mm	Pd–Au/TiO2, Pd/TiO2, 60 μm	Direct synthesis of H2O2	Liquid: aqueous solution (of diluted sulfuric acid, phosphoric acid, sodium bromide), Gas: O2 and H2, QG = 5 sccm, QL = 0.01–0.16 mL/min, T: 23–31 °C, P: 0.95 MPa	10.4 wt.% H2O2 and selectivity of 48% in 16-channel microreactor, ca. 1 kg/day H2O2 produced from parallel operation of 4 microreactors	[38]
6	PMMA chip, H × W = 0.8 mm × 0.8 mm	Pd/Al2O3, 180–250 μm	Direct synthesis of H2O2	Liquid: aqueous solution (of methanol, sulfuric acid, sodium bromide), Gas: O2 and H2, QG = 3–10 mL/min, QL = 0.01–0.05 mL/min, H2/O2 molar ratio: 0.2–1 T: 10–30 °C, P: atmospheric pressure (0.1 Mpa)	0.388 wt.% H2O2 produced under optimum conditions	[41]
7	PMMA chip, H × W = 0.8 mm × 0.8 mm	Pd-Sn/Al2O3, 180–250 μm	Direct synthesis of H2O2	Liquid: aqueous solution (of methanol, sulfuric acid, sodium bromide), Gas: O2 and H2, QG = 6 NmL/min, QL = 0.02 mL/min, H2/O2 molar ratio: 0.5, atmospheric T and P	0.4929 wt.% H2O2 produced under optimum conditions	[42]
8	Capillary, D = 4 mm L = 10 cm	SO42−/Al2O3 and HND-580, 300–580 μm	Dehydration of xylose	Liquid: aqueous xylose solution, Gas: N2/CO2, QL = 0.1–1 mL/min, QG = 0–50 mL/min, T: 140–170 °C, P: 1.5–3.5 MPa,	STY of 5 × 10−3 g/(mL·min), increase in the reaction rate at 1–2 and STY at 1–3 orders of magnitude higher than the batch reactor employing other acid catalysts	[50]
9	Teflon capillary, D = 1.65 mm	TEMPO/silica, 160–240 μm	Oxidation of HMF	Liquid: HMF and HNO3 in DCE, Gas: O2, QL = 0.1–0.4 mL/min, QG = 1.3–5 mL/min, T: 55 °C, P: 5 bar	97% conversion of HMF and 98% selectivity towards DFF within a residence time of 2 min	[33]
10	Stainless steel capillary, D = 2 mm, L = 2–6 cm	Au/CeO2	Oxidation of HMF	Liquid: HMF and NaOH (molar ratio at 1:4) in water, Gas: O2, QG/QL = 30–130, T: 90–130 °C, P: 0.1–2 MPa	100% HMF conversion and 90% FDCA selectivity were achieved within 41 s, an STY of 1–2 orders of magnitude higher than that of traditional reactors under similar conditions	[51]
11	Capillary, D = 3.87–5.35 mm L = 5.35–25 cm	Cu-/SiO2, Ni/SiO2, Pd/C, Pt/C, 425–600 μm	Hydrogenation of furfural	Liquid: furfural in isopropanol, Gas: H2 or N2, QL = 0.1–1 mL/min, QG = 10–50 mL/min, T: 60–120 °C, P: 0.1–3.1 MPa,	Furfural converted completely to FA under 80 °C and 0.6 MPa over Cu/SiO2 catalyst, an STY of 1–2 orders of magnitude higher than that of traditional reactors under similar conditions	[52]
12	PFA capillary, D = 1.6 mm, L = 40–80 cm	Ru/C 300 and 450 μm	Hydrogenation of levulinic acid	Liquid: levulinic acid in 1, 4-dioxane, Gas: H2 or N2, QL = 0.05–0.17 mL/min, QG = 0.16–0.33 mL/min, T: 70–130 °C, P: 9–13 bar	100% levulinic acid conversion and 84% γ-valerolactone yield were obtained	[25]
13	Stainless steel capillary, D = 4 mm, L = 3 mm	Au/TiO2, 90–180 μm	Oxidation of glycerol	Liquid: glycerol and NaOH in water, Gas: O2, QL = 0.2–4 mL/min, QG = 5–30 mL/min T: 30–70 °C, P: 5–15 bar	Selectivity towards secondary oxidation products like oxalic acid and tartronic acid was higher (>30%) than that in a batch slurry reactor (<5%) under similar conditions	[53]
14	Stainless steel capillary, D = 4.35 mm, L = 20 cm	NHPI/AC, 200–250 μm	Oxidation of ethylbenzene	Liquid: EB and TBN (molar ratio of TBN/EB = 0.6–1.5) in MeCN, Gas: O2, QG/QL = 25–125, T: 50–110 °C, P: 2–7 bar	93% EB conversion and 90.1% selectivity to AP within a reaction time of 2.4 min	[57]
15	Stainless steel capillary, D = 4.6 mm, L = 150 mm	No catalyst, soda lime glass beads as packing, 210–250 μm	Oxidation of alkenes with nitrous oxide	Liquid: dodecenes mixture, Gas: vaporized N2O, QL = 12.2 μL/min, QG = 6.1 μL/min, T: 300–375 °C, P: 6205–10,342 kPa	Conversions of dodecene mixture up to 77% and yields of primary product (ketones) up to 45%.	[61]
16	Stainless steel microreactors, D = 4 mm, L = 30 mm	Au/Al2O3, 125–500 μm	Hydrogenation of functionalized nitroaromatics	Liquid: substituted nitrobenzene in toluene, Gas: H2, QL = 0.5 mL/min, QG = 8–60 mL/min T: 60–110 °C and P: 10–20 bar	A high selectivity to the target product (>92%) and almost 100% conversion of substrate were obtained	[93]
17	6 parallel stainless steel microreactors, D = 3.5 mm	Pt-HHDMA/C, 200–400 μm	Hydrogenation of functionalized nitroaromatics	Liquid: functionalized nitroaromatics in THF, Gas: H2 QL = 0.3–3 mL/min, QG = 3–60 mL/min T: 30–90 °C, P: 1–60 bar,	Pt-HHDMA/C exhibited superior activity, chemo selectivity, and leaching resistance compared with Pt−Pb/CaCO3	[97]
18	Stainless steel capillary, D = 4.6 mm, L = 50 mm	Pd/C powder, 6–20 μm	Hydrogenation of 2-(4-nitrophenyl) butanoic acid	Liquid: NBA, Gas: H2, QL = 0.8–1.2 mL/min, QG = 11.3 NmL/min, T: 0–30 °C, P: 1–2.5 Mpa	Reaction kinetics were determined	[3]
19	Capillary, D = 2.2 mm, L = 200 mm	Pt-Pd/Al2O3 catalyst, 70–200 μm	Hydrogenation of biphenyl	Liquid: biphenyl in tetradecane, Gas: H2, QL = 0.5–5 g/h, QG = 5–50 mL/min T: 140 °C, P: 5 MPa	Intrisic kinetic values were obtained in microreactors, similar to those from the autoclave experiments	[88]
20	PTFE capillary, D = 0.8 mm, L = 30 cm	Au–Pd/TiO2, 53–63 μm	Oxidation of cinnamyl alcohol	Liquid: cinnamyl alcohol in toluene, Gas: O2 (in N2), T: 80–120 °C, P: 4 bar	Catalyst performance was tested and deactivation chracteristics were revealed	[91]
21	Silicon/glass chip, H × W × L = 0.3 mm × 0.6 mm × 190 mm	Au-Pd/TiO2, 53–63 μm	Aerobic oxidation of benzyl alcohol	Liquid: benzyl alcohol, Gas: O2 QL = 0.003 mL/min, QG = 0.6 mL/min, T: 120 °C	MIm-prepared catalyst demonstrated long-term stability for over 75 h and achieved around 80% conversion of benzyl alcohol with 60% selectivity to benzaldehyde	[20]

^a D and L are the microreactor diameter and (bed) length, respectively. H, W, and L are the height, width, and length of the rectangular microchannel, respectively. ^b Q_L is the flow rate of liquid phase, and Q_G is the flow rate of gas phase.

Table 3. An overview of selected gas–liquid–solid reaction systems reported in packed bed microreactor.

Entry	Reactor a	Catalyst	Reaction	Operational Conditions b	Results	Reference
1	Teflon capillary, D = 1.65 mm	TEMPO/silica, 160–240 μm	Oxidation of different alcohols	Liquid: alcohol and HNO3 in DCE, Gas: O2 QL = 0.1–0.4 mL/min, QG = 1.3–5 mL/min, T: 55–80 °C, P: 5 bar	98% benzyl alcohol conversion and 99% benzaldehyde yield within a residence time of 0.5 min, STY of 4.1 × 105 mol benzaldehyde/(molcat·L·h)	[33]
2	Silicon-glass chip, H × W × L = 0.3 mm × 0.6 mm × 3.2 mm	Au–Pd/TiO2, 64 ± 14 μm	Benzyl alcohol oxidation	Liquid: benzyl alcohol, Gas: O2, QL = 0–5 μL/min, QG = 0–0.6 NmL/min T: 120 °C, P: 2 bar	>90% selectivity to benzaldehyde in the partially wetted gas-continuous flow; 93% conversion and 80% selectivity at fully wetted gas continuous flow	[31]
3	Capillary, D = 4.35 mm	No solid catalyst, glass beads as packing, 300–350 μm	Benzyl alcohol oxidation	Liquid: benzyl alcohol and Cu/TEMPO in acetonitrile solution, Gas: 9% O2 in N2, molar ratio of O2 to alcohol = 0.65:1 T: 20–45 °C, P: 5–35 bar	Almost 100% yield of benzaldehyde within 30 s, STY = 7318.4 mol/(m3 h) at optimum condition, which is ca. 2 to 20 times higher than that in conventional flow reactors under similar conditions	[35]
4	Tempax glass chips, H × W × L = 0.9 mm × 0.6 mm × 40/120 mm	Pd–Au/TiO2, Pd/TiO2, Pd/Al2O3,	Direct synthesis of H2O2	Liquid: aqueous solution (of diluted sulfuric acid, phosphoric acid, sodium bromide), Gas: O2 and H2, QG = 5 NmL/min, QL = 0.1 mL/min, T: 20 °C, P: 0.95 MPa,	11.3 wt.% H2O2 with a yield of at 42%, H2O2 productivityat 17 × 10−4 mol/h	[39]
5	Glass chip, H × W × L = 0.9 mm × 0.6 mm × 4 0 mm	Pd–Au/TiO2, Pd/TiO2, 60 μm	Direct synthesis of H2O2	Liquid: aqueous solution (of diluted sulfuric acid, phosphoric acid, sodium bromide), Gas: O2 and H2, QG = 5 sccm, QL = 0.01–0.16 mL/min, T: 23–31 °C, P: 0.95 MPa	10.4 wt.% H2O2 and selectivity of 48% in 16-channel microreactor, ca. 1 kg/day H2O2 produced from parallel operation of 4 microreactors	[38]
6	PMMA chip, H × W = 0.8 mm × 0.8 mm	Pd/Al2O3, 180–250 μm	Direct synthesis of H2O2	Liquid: aqueous solution (of methanol, sulfuric acid, sodium bromide), Gas: O2 and H2, QG = 3–10 mL/min, QL = 0.01–0.05 mL/min, H2/O2 molar ratio: 0.2–1 T: 10–30 °C, P: atmospheric pressure (0.1 Mpa)	0.388 wt.% H2O2 produced under optimum conditions	[41]
7	PMMA chip, H × W = 0.8 mm × 0.8 mm	Pd-Sn/Al2O3, 180–250 μm	Direct synthesis of H2O2	Liquid: aqueous solution (of methanol, sulfuric acid, sodium bromide), Gas: O2 and H2, QG = 6 NmL/min, QL = 0.02 mL/min, H2/O2 molar ratio: 0.5, atmospheric T and P	0.4929 wt.% H2O2 produced under optimum conditions	[42]
8	Capillary, D = 4 mm L = 10 cm	SO42−/Al2O3 and HND-580, 300–580 μm	Dehydration of xylose	Liquid: aqueous xylose solution, Gas: N2/CO2, QL = 0.1–1 mL/min, QG = 0–50 mL/min, T: 140–170 °C, P: 1.5–3.5 MPa,	STY of 5 × 10−3 g/(mL·min), increase in the reaction rate at 1–2 and STY at 1–3 orders of magnitude higher than the batch reactor employing other acid catalysts	[50]
9	Teflon capillary, D = 1.65 mm	TEMPO/silica, 160–240 μm	Oxidation of HMF	Liquid: HMF and HNO3 in DCE, Gas: O2, QL = 0.1–0.4 mL/min, QG = 1.3–5 mL/min, T: 55 °C, P: 5 bar	97% conversion of HMF and 98% selectivity towards DFF within a residence time of 2 min	[33]
10	Stainless steel capillary, D = 2 mm, L = 2–6 cm	Au/CeO2	Oxidation of HMF	Liquid: HMF and NaOH (molar ratio at 1:4) in water, Gas: O2, QG/QL = 30–130, T: 90–130 °C, P: 0.1–2 MPa	100% HMF conversion and 90% FDCA selectivity were achieved within 41 s, an STY of 1–2 orders of magnitude higher than that of traditional reactors under similar conditions	[51]
11	Capillary, D = 3.87–5.35 mm L = 5.35–25 cm	Cu-/SiO2, Ni/SiO2, Pd/C, Pt/C, 425–600 μm	Hydrogenation of furfural	Liquid: furfural in isopropanol, Gas: H2 or N2, QL = 0.1–1 mL/min, QG = 10–50 mL/min, T: 60–120 °C, P: 0.1–3.1 MPa,	Furfural converted completely to FA under 80 °C and 0.6 MPa over Cu/SiO2 catalyst, an STY of 1–2 orders of magnitude higher than that of traditional reactors under similar conditions	[52]
12	PFA capillary, D = 1.6 mm, L = 40–80 cm	Ru/C 300 and 450 μm	Hydrogenation of levulinic acid	Liquid: levulinic acid in 1, 4-dioxane, Gas: H2 or N2, QL = 0.05–0.17 mL/min, QG = 0.16–0.33 mL/min, T: 70–130 °C, P: 9–13 bar	100% levulinic acid conversion and 84% γ-valerolactone yield were obtained	[25]
13	Stainless steel capillary, D = 4 mm, L = 3 mm	Au/TiO2, 90–180 μm	Oxidation of glycerol	Liquid: glycerol and NaOH in water, Gas: O2, QL = 0.2–4 mL/min, QG = 5–30 mL/min T: 30–70 °C, P: 5–15 bar	Selectivity towards secondary oxidation products like oxalic acid and tartronic acid was higher (>30%) than that in a batch slurry reactor (<5%) under similar conditions	[53]
14	Stainless steel capillary, D = 4.35 mm, L = 20 cm	NHPI/AC, 200–250 μm	Oxidation of ethylbenzene	Liquid: EB and TBN (molar ratio of TBN/EB = 0.6–1.5) in MeCN, Gas: O2, QG/QL = 25–125, T: 50–110 °C, P: 2–7 bar	93% EB conversion and 90.1% selectivity to AP within a reaction time of 2.4 min	[57]
15	Stainless steel capillary, D = 4.6 mm, L = 150 mm	No catalyst, soda lime glass beads as packing, 210–250 μm	Oxidation of alkenes with nitrous oxide	Liquid: dodecenes mixture, Gas: vaporized N2O, QL = 12.2 μL/min, QG = 6.1 μL/min, T: 300–375 °C, P: 6205–10,342 kPa	Conversions of dodecene mixture up to 77% and yields of primary product (ketones) up to 45%.	[61]
16	Stainless steel microreactors, D = 4 mm, L = 30 mm	Au/Al2O3, 125–500 μm	Hydrogenation of functionalized nitroaromatics	Liquid: substituted nitrobenzene in toluene, Gas: H2, QL = 0.5 mL/min, QG = 8–60 mL/min T: 60–110 °C and P: 10–20 bar	A high selectivity to the target product (>92%) and almost 100% conversion of substrate were obtained	[93]
17	6 parallel stainless steel microreactors, D = 3.5 mm	Pt-HHDMA/C, 200–400 μm	Hydrogenation of functionalized nitroaromatics	Liquid: functionalized nitroaromatics in THF, Gas: H2 QL = 0.3–3 mL/min, QG = 3–60 mL/min T: 30–90 °C, P: 1–60 bar,	Pt-HHDMA/C exhibited superior activity, chemo selectivity, and leaching resistance compared with Pt−Pb/CaCO3	[97]
18	Stainless steel capillary, D = 4.6 mm, L = 50 mm	Pd/C powder, 6–20 μm	Hydrogenation of 2-(4-nitrophenyl) butanoic acid	Liquid: NBA, Gas: H2, QL = 0.8–1.2 mL/min, QG = 11.3 NmL/min, T: 0–30 °C, P: 1–2.5 Mpa	Reaction kinetics were determined	[3]
19	Capillary, D = 2.2 mm, L = 200 mm	Pt-Pd/Al2O3 catalyst, 70–200 μm	Hydrogenation of biphenyl	Liquid: biphenyl in tetradecane, Gas: H2, QL = 0.5–5 g/h, QG = 5–50 mL/min T: 140 °C, P: 5 MPa	Intrisic kinetic values were obtained in microreactors, similar to those from the autoclave experiments	[88]
20	PTFE capillary, D = 0.8 mm, L = 30 cm	Au–Pd/TiO2, 53–63 μm	Oxidation of cinnamyl alcohol	Liquid: cinnamyl alcohol in toluene, Gas: O2 (in N2), T: 80–120 °C, P: 4 bar	Catalyst performance was tested and deactivation chracteristics were revealed	[91]
21	Silicon/glass chip, H × W × L = 0.3 mm × 0.6 mm × 190 mm	Au-Pd/TiO2, 53–63 μm	Aerobic oxidation of benzyl alcohol	Liquid: benzyl alcohol, Gas: O2 QL = 0.003 mL/min, QG = 0.6 mL/min, T: 120 °C	MIm-prepared catalyst demonstrated long-term stability for over 75 h and achieved around 80% conversion of benzyl alcohol with 60% selectivity to benzaldehyde	[20]

^a D and L are the microreactor diameter and (bed) length, respectively. H, W, and L are the height, width, and length of the rectangular microchannel, respectively. ^b Q_L is the flow rate of liquid phase, and Q_G is the flow rate of gas phase.

Table 4. An overview of selected gas–solid reaction systems reported in packed bed microreactors.

Entry	Reactor a	Catalyst	Reaction	Operational Conditions	Results	Reference
1	Silicon chip, H × W × L = 300 μm × 625 μm × 20 mm	Activated carbon, 53–73 μm	Phosgene synthesis	Gas: mixed gas with molar ratio at CO/Cl2 = 2:1, T: 220 °C, P: 132 KPa at inlet	Kinetic operation and data were obtained, phosgene productivity at 9.3 kg/year	[43]
2	Stainless steel slit-type microreactor, W = 0.508 mm	Co-Re/Al2O3, 45, 150 μm	FT synthesis	Gas: syngas with molar ratio at H2/CO = 1–2.5, T: 224 °C, P = 10–35 atm,	CO conversion of 90.3%, CH4 selectivity of 3.4%, productivity of 2.14 g C2+/(gcat·h) were obrained under P = 35 atm, H2/CO = 2, and GHSV = 22,886 h−1	[75]
3	Stainless steel capillary, D = 1.4 mm, L = 320 mm	Co-Pt/Al2O3	FT synthesis	Gas: syngas with molar ratio at H2/CO = 2, T: 493 K, P: 20 bar	Chain growth probability was 0.92 with C5+ productivity of 1.1–1.4 g/(gcat·h), slow catalyst deactivation occured at 1% per day	[24]
4	Stainless steel plate-type microreactor, H × W × L = 0.8 mm × 8 mm × 60 mm	Cu-ZnO/Al2O3, 50–120 μm	Methanol synthesis from syngas	Gas: syngas with a molar ratio: H2/CO/CO2/N2 = 65:25:5:5, T: 235–265 °C, P: 80 bar	Methanol selectivity exceeding 98.7%	[82]
5	Not specified	Pd/In2O3/SBA-15 catalyst	Methanol synthesis from CO2	Gas: mixed gas with a molar ratio: H2/CO2/N2 = 64:16:20, T: 260–360 °C, P: 5 MPa, GHSV = 15,000 cm3/(gcat·h)	Methanol selectivity of 83.9%, CO2 conversion of 12.6%, and STY of 1.1 × 10−2 mol/(gcat·h)	[85]
6	Stainless steel slit-type micoreactor, L = 6 mm	CuO–ZnO– Al2O3 and γ-Al2O3, 50–80 μm	DME synthesis	Gas: syngas with a molar ratio at H2/CO/CO2/N2/ CH4 = 56:28:5:5:6, T: 220–320 °C, P: 50–70 bar	An DME yield up to ca. 50% at ca. 90% CO conversion was achieved	[87]
7	Silicon-glass chip, H × W = 0.42 mm × 2 mm	Pd/Al2O3, 53–90 μm	Methanol complete oxidation	Gas: mixed gas (O2, CH4, N2, He) with molar ratio at O2/CH4 = 2–4, T: 250–350 °C, P: 1.3 bar at outlet	Rapid and reliable kinetic model screening within 48 h	[106]

^a D and L are the microreactor diameter and (bed) length, respectively. H, W, and L are the height, width, and length of the rectangular microchannel, respectively.

Table 4. An overview of selected gas–solid reaction systems reported in packed bed microreactors.

Entry	Reactor a	Catalyst	Reaction	Operational Conditions	Results	Reference
1	Silicon chip, H × W × L = 300 μm × 625 μm × 20 mm	Activated carbon, 53–73 μm	Phosgene synthesis	Gas: mixed gas with molar ratio at CO/Cl2 = 2:1, T: 220 °C, P: 132 KPa at inlet	Kinetic operation and data were obtained, phosgene productivity at 9.3 kg/year	[43]
2	Stainless steel slit-type microreactor, W = 0.508 mm	Co-Re/Al2O3, 45, 150 μm	FT synthesis	Gas: syngas with molar ratio at H2/CO = 1–2.5, T: 224 °C, P = 10–35 atm,	CO conversion of 90.3%, CH4 selectivity of 3.4%, productivity of 2.14 g C2+/(gcat·h) were obrained under P = 35 atm, H2/CO = 2, and GHSV = 22,886 h−1	[75]
3	Stainless steel capillary, D = 1.4 mm, L = 320 mm	Co-Pt/Al2O3	FT synthesis	Gas: syngas with molar ratio at H2/CO = 2, T: 493 K, P: 20 bar	Chain growth probability was 0.92 with C5+ productivity of 1.1–1.4 g/(gcat·h), slow catalyst deactivation occured at 1% per day	[24]
4	Stainless steel plate-type microreactor, H × W × L = 0.8 mm × 8 mm × 60 mm	Cu-ZnO/Al2O3, 50–120 μm	Methanol synthesis from syngas	Gas: syngas with a molar ratio: H2/CO/CO2/N2 = 65:25:5:5, T: 235–265 °C, P: 80 bar	Methanol selectivity exceeding 98.7%	[82]
5	Not specified	Pd/In2O3/SBA-15 catalyst	Methanol synthesis from CO2	Gas: mixed gas with a molar ratio: H2/CO2/N2 = 64:16:20, T: 260–360 °C, P: 5 MPa, GHSV = 15,000 cm3/(gcat·h)	Methanol selectivity of 83.9%, CO2 conversion of 12.6%, and STY of 1.1 × 10−2 mol/(gcat·h)	[85]
6	Stainless steel slit-type micoreactor, L = 6 mm	CuO–ZnO– Al2O3 and γ-Al2O3, 50–80 μm	DME synthesis	Gas: syngas with a molar ratio at H2/CO/CO2/N2/ CH4 = 56:28:5:5:6, T: 220–320 °C, P: 50–70 bar	An DME yield up to ca. 50% at ca. 90% CO conversion was achieved	[87]
7	Silicon-glass chip, H × W = 0.42 mm × 2 mm	Pd/Al2O3, 53–90 μm	Methanol complete oxidation	Gas: mixed gas (O2, CH4, N2, He) with molar ratio at O2/CH4 = 2–4, T: 250–350 °C, P: 1.3 bar at outlet	Rapid and reliable kinetic model screening within 48 h	[106]

^a D and L are the microreactor diameter and (bed) length, respectively. H, W, and L are the height, width, and length of the rectangular microchannel, respectively.

3. Challenges and Outlook

Packed bed microreactors as a miniaturized platform have shown great potential in executing sustainable chemical transformation and high-throughput catalyst testing efficiently due to their distinct process control/intensification merits compared with the conventional macroscale reactors. However, the hydrodynamics of reaction systems (particularly complex gas–liquid–solid or liquid–liquid–solid systems) and their underlying influence on the reaction behavior/optimization therein are not largely understood. The dominance of capillary force under prevailing conditions in packed bed microreactors renders hydrodynamic characteristics (largely) different from their macroscale counterpart. Meanwhile, the unfavorable flow phenomena (e.g., flow maldistribution, channeling, incomplete catalyst wetting) frequently encountered in conventional reactors also exist for microreactors [4,88]. Thorough fluid dynamic investigation and reaction characterization are still required for a fundamental understanding and precise control of microreactor performance towards their broader future application [107].

One of the primary concerns in using packed bed microreactors is the high pressure drop penalty, caused by the small dimensions of both microchannels and packing. This concern is more prevalent when highly viscous fluids or multiphase fluid systems are present. This necessitates the use of high-pressure pumps or flow controllers to maintain the desired flow rates and, in turn, increases energy consumption and operational costs, potentially offsetting the benefits in reaction efficiency enhancement in microreactors [29]. Thus, careful dimensioning of packed bed microreactors and a rational reaction operation are deemed necessary to ensure an optimized use of energy that matches the desired process performance, requiring dedicated experiments as well as computational modeling. In gas–liquid fluid systems, simplifying them to a single-phase system (e.g., using an upstream contactor before packed beds for gas dissolution [92]) can effectively lower pressure drop and flow complexity, though at the cost of a limited gas supply. Innovations like incorporating catalyst (coated) pillars [108] and single pellet string microreactors [109,110] are also anticipated to mitigate pressure drop increases while ensuring uniform flow dynamics. Moreover, the risk of microchannel clogging is another concern [111], particularly in reactions involving the formation of solid products (e.g., cokes or humins) that deposit on the catalyst [48] or the formation of (by-)products that can precipitate out of the solution. Mitigating this issue would necessitate frequent catalyst regeneration in situ or, more preferably, effective reaction/process control strategies to prevent solid formation.

It can be also challenging to scale up packed bed microreactors to meet industrial production needs, i.e., increasing the production from lab scale (in g/h or g/min) to industrial scale (in kg/h or even higher) [112]. When using the numbering-up strategy (i.e., operating numerous microchannels in parallel), fluid maldistribution, different catalyst loading, and the resulting non-uniform concentration would make it difficult to achieve consistent product quality in every single packed bed microreactor. Thus, a proper fluid distributor and catalyst filling approach should be considered during the scale-up process. If a great number (e.g., thousands) of microchannels need to be fabricated and precisely operated, both the capital and operational costs are anticipated to rise substantially [112,113]. The microreactor can alternatively be scaled up through the sizing-up approach, which involves increasing the microchannel dimension (length and/or diameter). Elongating the microreactor length enhances productivity by allowing for a longer residence time; however, a significant limitation could arise from the increased pressure drop. On the other hand, although a higher throughput is attained under a lower pressure drop when the microchannel diameter is increased, the inherent advantages of microreactors due to miniaturization may be compromised. Therefore, thorough consideration is needed for an appropriate reactor dimensioning to balance the targeted productivity with the requisite transport property within the scaled-up microreactor system [113,114]. Additionally, potential issues such as microreactor fouling and material cracks can arise after extended operation, posing challenges for reactor maintenance or replacement (especially when a large number of microreactor modules are operated in parallel [26,38,40]).

In terms of the costs of applying microreactor technology in practice, the microstructured devices could own no advantage in installation cost (capital cost) compared to conventional reactors [115,116]. The costs for fabricating and repairing microreactors, particularly those demanding high precision and durability, can be prohibitively high, raising concerns about their economic feasibility [84,117]. In contrast, the operation cost of microreactors (labor, waste treatment, transport, etc.) can be optimized or reduced due to the superior reaction performance caused by process intensification and the more compact and autonomous system (which allows remote operation) [115,116,118]. Additionally, the numbering-up strategy can decrease the overall costs, making microreactor technology a commercially attractive tool in the long run. Integrating microreactors into existing conventional plants can be substantially more cost-effective than developing a fully dedicated and customized microreactor plant from scratch [116]. The modularity and portability of microreactors greatly enhance their compatibility with existing industrial plants, allowing for relatively seamless integration into existing infrastructures and minimizing the need for major overhauls. To date, only a limited number of cases have been evaluated for the economic viability of microreactors. As a result, the full-scale deployment of microreactor technology in industrial production is still in an ongoing developmental phase. More research and practical trials are necessary to fully understand their economic implications and optimize the implementation potential of microreactor technology [119].

To adhere to the principles of green and sustainable chemistry, the further reaction scope with packed bed microreactors should focus more on the valorization of renewable resources (such as biomass and CO₂) and circular chemical synthesis. Integration with novel chemistry routes like photochemistry [120] and electrochemistry [121] is anticipated to offer a more eco-friendly approach. There exist other novel approaches for achieving process intensification, such as microwave reactors [122], acoustic reactors [123], and plasma reactors [124]. The combination of packed bed microreactors with these novel approaches holds great promise towards an even more efficient and sustainable means to further optimize reaction performance.

4. Conclusions

Microreactor technology with the integration of packed bed designs represents a significant advancement towards developing green and sustainable chemistry and processes. Various packed bed microreactor designs have been developed for catalytic reactions, which involve different fluid systems ranging typically from single-phase (gas or liquid) fluid systems to more complex multiphase (e.g., gas–liquid or liquid–liquid) fluid systems in the presence of packed catalysts or inert particles (Figure 2). For gas–liquid–solid three-phase systems (Table 3), reaction applications include, among others, the selective oxidation of alcohols, hydrogenation of nitroaromatics, and direct synthesis of H₂O₂. Improved reaction performance (e.g., high product yield, selectivity, or productivity) was widely reported in packed bed microreactors, often under kinetic operation as a result of superior mass transfer. The small voids and good heat transfer capacity in packed bed microreactors have been shown to enable a safer production environment even under explosive conditions.

The superior mixing efficiency of microreactors is also effective in liquid–liquid–solid reaction systems (Table 2). This was demonstrated for applications like biodiesel synthesis and the transformation of biomass derivatives such as C6 sugars to furanics, where reaction optimization and promising intensification potential were achieved. Gas–solid reaction systems (Table 4) are commonly used in the production of liquid fuels at high temperatures from gases like syngas or CO₂–H₂ mixture. Despite the highly exothermic nature of reactions such as Fischer–Tropsch and methanol syntheses, microreactors can maintain isothermal conditions in the packed bed.

Packed bed microreactors offer an exceptional platform for intrinsic kinetic studies, chemistry optimization, and catalyst evaluation/screening in widely employed hydrogenation and oxidation reactions. Particularly, the capability to operate multiple parallel beds under varying conditions enables the comprehensive and rapid catalyst or reaction parameter screening while minimizing material consumption and waste generation. Recently, the integration of packed bed microreactor systems with advanced automation technologies has further revolutionized experimental workflows, significantly enhancing research efficiency and convenience in catalytic process development.

Despite these potentials, challenges such as complex reaction hydrodynamics, high pressure drop, maintaining long-term catalyst stability, scale-up to meet industrial production, and optimizing cost-effectiveness may hinder the broader adoption of packed bed microreactors in the industry. To cope with these challenges, key areas of future researches are suggested to include unravelling multiphase flow mechanisms, optimized match between energy use and reaction efficiency (e.g., via fine tuning of packing size and bed/microreactor configuration and reaction operation), more precise parameter control over catalyst/reaction behavior (in single and parallel packed beds), and feasibility study in industrial practices. With the evolvement of microreactor technology, it is envisaged that these challenges will be well addressed through continued research and development and translated into vast opportunities for realizing the full potential of packed bed microreactors in advancing sustainable chemical process development in the near future.