To date, a number of gas-liquid flow reactors and devices for laboratory-scale preparation have been reported, and a brief review of the field has recently been conducted. It is worth noting that those technologies based on the use of permeable membranes facilitate the transfer of gases into the liquid stream, allowing the use of highly toxic and harmful gases in a controlled, safe and reliable manner. Recently, a gas-liquid flow reactor using a semi-permeable Teflon AF-2400 membrane (a copolymer of tetrafluoroethylene and perfluorodimethyldioxane) has been developed. The reactor has been shown to successfully perform a range of organic conversions under flow conditions using several commonly used reaction gases, such as Glaser HAY oxidative coupling homogeneous and heterogeneous hydrogenation of alkynes (O2) during ozone decomposition of olefins (O3), carboxylation of Grignard reagent with carbon dioxide (CO2), two-dimensional alkoxycarbonylation using carbon monoxide (Co), Mizoroki Heck reaction using ethylene (C2H2), syngas (Co H2, 1:1 mixture) for hydroformylation of styrene, and the preparation of thiourea and paal knorr pyrrole formations using ammonia (NH3). The reactor itself has also evolved through some gradual developments. Eventually, two methods were found to measure and quantify the amount of gas absorbed into the stream. For the absorption of hydrogen, we have developed a camera-enabled pixel counting device. It consists of a stream (colored with a red dye) that passes through a breathable membrane and is exposed to hydrogen at different high pressures. Once the hydrogen-rich flowing solution is returned to atmospheric pressure, degassing is inevitably observed, whereby a colorless gas plug and a red liquid plug flow through a suitable digital camera. The computer uses this information and a simple ** to calculate the "red" pixels present in the image, thus quantifying the amount of colorless hydrogen released relative to the background experiment. For ammonia, we prefer to perform a real-time** colorimetric titration of the gas-rich stream using an acidic stream and a bromocresol green indicator at known concentrations. The initial orange acidic solution (mixed with a gas-rich stream at the junction mixer) has a gradual increase in flow velocity until neutralization occurs by reaction with ammonia, at which point the flow parameters can be used to calculate the concentration of ammonia in the stream. Interestingly, as a result of this work, we can make some observations about the response of the system. First, by separating the gas introduction zone from the reaction zone, we can independently control the temperature of each process zone, which allows for higher ammonia absorption through initial cooling, and then by heating the reactor, increasing the reactivity to shorten the reaction time. 2h It was soon recognized that many variations in the basic reactor configuration were possible, so that they could study various parameters: gas introduction, solvent pressure, solvent dielectric constant, gas pressure, throughput potential, and robustness of the setup. A good understanding of these factors can greatly influence the optimization of reaction protocols using these devices. The use of ammonia under intermittent conditions is not without its associated problems, such as the discharge of excess gases into the waste and the need for pressurized reaction vessels, while the use of pre-prepared ammonia solutions is often wasteful and limited to a small range of solvents. In this work, we explored the opportunities created by these tube-in-tube reactor systems for the use of ammonia reactions, in particular how to achieve the multigram, two-step telescopic synthesis of the anti-inflammatory agent farnetinazole after optimization. In earlier studies of a range of gases, the tube-in-tube reactor was configured in such a way that the liquid flow was through the internal AF-2400 pipe, while the reaction gas was introduced through the external pipe (B, Figure 1). In our work with ammonia as a reactive gas, we explored the opposite configuration (a, Figure 1). We believe that this may improve the temperature control of the gas input process, as the outer liquid layer will provide better heat transfer and possibly higher gas concentrations, resulting in higher conversion and yield of subsequent reactions. Leadbeater and colleagues used the same configuration for alkoxycarbonylation reactions using an alternative tube sleeve device, so the ability to heat the reactor was considered critical to the process, although the nature of the membrane material was not disclosed. In the current study, we compare two possible configurations of the ammonia reaction. In addition, we evaluated other parameters, such as reactor temperature, flow rate, and solvent polarity, to better understand the role and importance of each variable in ammonia absorption using these tube-in-tube gas-liquid flow reactors. 
Figure 1The key parameters of the tube-in-tube reactor are two possible configurations. Typically, the reactor consists of two concentric tubes connected by a sealed stainless steel T-piece (Swagelok). PTFE Outer Tube (O.D. 3.)2 mm, inner diameter 16 mm) in a Teflon AF-2400 tube (1 mm O.D., 0.D. I.D.)8 mm, 2 m long) (a, Figure 2). The reactor coil is then wound around an aluminum mandrel (B and C, Figure 2) and placed on a cooling heating block (D, Figure 2), such as a Polar Bear Plus (PBP). Organizing the equipment in this way combines the advantages of a tube-in-tube reactor with the temperature control provided by the heater cooler unit at various temperatures (typically -89 to 150 °C). As far as safety is concerned, the unit is usually contained in a fume hood and piped to an appropriate cylinder that can be located remotely (or better yet, an integrated "sparklet" gas tank can be used instead). When the cylinder is open toward the device, use a moistened pH test strip to check the tight connection of the Swagelok fitting and carefully tighten any problematic connectors as needed (overtightening may damage the AF-2400 tubing). To prevent irreversible damage to the pipes, the maximum cut-off pressure of the flow is fixed at 5 bar above the back pressure regulator (BPR) using UniQSIS Binary Pump Module (BPM) software. This ensures that any potential blockage downstream of the tube casing manifold does not result in high liquid pressure and potential collapse of the internal AF-2400 pipe. 
Figure 2(a) Tube-in-tube gas-liquid reactor based on a semi-permeable membrane (internal AF-2400 tubing is filled with a red dye solution for demonstration purposes). (b) Aluminium mandrel. (c) Reactor wound around an aluminium mandrel. (d) The tube-in-tube reactor is placed on the cooling heating block of the Polar Bear Plus. As mentioned above, taking advantage of the alkalinity of ammonia, the concentration of ammonia in different solvent streams is measured using ** colorimetric titration. **Schematic diagram of the titration device is shown in Protocols 1. Here, the solvent is continuously pumped into the tube-in-tube reactor, and the reactor itself is placed on a heated cooling block. The output stream merges with the second stream at the Y-joint, which is composed of 012 M HCl in water containing bromocresol green as a pH indicator (pH below 3.)Orange at 8 with a pH above 5Blue at 4 and green at moderate pH). The organic and water streams are then mixed using a magnetic mixer**, which consists of four PTFE-coated magnetic stir bars in a 3 mm Omnifit column placed on a magnetic stirrer hot plate. The resulting solution is then passed through a section of tubing coiled around a white cylinder to facilitate the observation of color changes and to accurately determine the titration endpoint. A back pressure regulator must be installed at the end of the system to keep the gas in solution and ensure uniform conditions. Solvent and titrant solutions are pumped using a binary pump moduleUse the UniQSIS PC interface to control and monitor the flow rate of both streams as well as the temperature of the Polar Bear Plus. 
Scenario 1Flow titration settings and visual results** (PBP = Polar Bear Plus;.)BPR = Back Pressure Regulator) Initially, a reactor configuration (Figure 1, A) in which pressurized gas was introduced through an inner tube was evaluated for three solvents at different temperatures. After leaving the tube-in-tube reactor, the aerated solvent stream is streamed at 0A fixed flow rate of 5 ml min (298 s residence time) was mixed with an increasing stream of titrant solution until the titration endpoint was reached. Each measurement is repeated twice. With each change in parameters, the corresponding amount of time for the entire dwell is given in order to "flush" the reactor with new conditions before making measurements. The results in Figure 3 represent the ammonia concentrations (in mol L) of methanol, 1,2-dimethoxyethane (DME), and toluene over the temperature range of -20 to 80. 
Figure 3Effect of temperature on ammonia absorption (configuration a, Figure 1). It is clear from these experiments that the temperature of the reactor has a considerable influence on the permeation and absorption of gases. Obviously, a temperature of 0°C provides the highest concentration of ammonia of all three solvents. The overall trend of higher ammonia concentrations in the stream due to lower temperatures may be related to the greater solubility of the gas at low temperatures. However, the morphological changes of membranes cannot be ignored. In fact, Polyakov and Yaampolskii report that the permeability of Teflon AF-2400 decreases with increasing temperature, while the permeability of Teflon AF-1600 increases with increasing temperature. The general trend of ammonia concentration increasing with the increase of solvent dielectric constant indicates that polar intermolecular forces are important factors for osmotic dissolution. Obviously, this is a complex issue, and of course further work is needed to unravel the many influencing factors in order to truly understand the complex details of these systems. At temperatures below 0, we observed condensation of ammonia in the gas pipe. In fact, ammonia is in 3Liquefaction begins at a pressure of 5 bar, which occurs around 5. 11 This liquefaction process adds an additional complication to the calculation of the passage of molecules across the permeable membrane, namely the difference in the intermolecular bonding of ammonia in the gas and liquid phases: the enthalpy change of vaporization. The standard enthalpy of vaporization (ΔHVAP) for ammonia is 2335 kJ mol (for comparison, water = 40.)68 kJ mol, hydrogen = 046 kj/mol）。Therefore, when the tubular casing device cools beyond the optimal temperature (locally around 0), the ammonia exposed to the permeable membrane surface is proportionally present in liquid form, so the enthalpy of vaporization must be overcome to pass through the membrane barrier. In fact, it seems that this effect was observed in experiments, when passing through the point of liquefaction start, the concentration was significantly lower and the temperature lowered. However, this working hypothesis assumes that the molecule must be strictly in the gas phase in order to pass through the membrane. Figure 4 shows three plots of ammonia concentration at 0 °C versus residence time in a tube-in-tube reactor for methanol, DME, and toluene. These data show that for each solvent, the ammonia concentration is approximately proportional to the residence time, and in all cases solvent saturation is not reached. After a certain residence time, the saturation will provide a constant concentration, which can be demonstrated by the plateau in the diagram. The gradient of the line represents the rate at which the solvent absorbs ammonia. This absorption value can be used for any concentration of the desired residence time, provided that these times are before saturation begins (where the rate of absorption is no longer constant). 
Figure 4Plot of ammonia concentrations in MEOH, DME, and toluene versus 0 residence time (configuration A, Figure 1). Next, we reversed the reactor configuration to deliver ammonia to the outer tube and solvent to the inner tube (B, Figure 1). This can be achieved simply by repositioning the appropriate connector at the end of the pipe. The flow rate is then adjusted to provide the same residence time as earlier reactor configuration a (now 0.).2 ml min to provide a residence time of 300 seconds) and repeat the titration over the same solvent and temperature range. The results are shown in Figure 5. 
Figure 5Effect of temperature on ammonia absorption (configuration b, Figure 1). Under this arrangement, the data clearly shows that temperature still has a significant effect on ammonia absorption. Also, the optimal temperature seems to be concentrated around 0 again. Clearly, the gas outer jacket (which is actually static) provides significant heat transfer between the internal stream and the cold bath. This is perhaps not surprising given the relative size of the pipes. It is clear from the data that the configuration of the liquid in the inner tube provides a significantly higher concentration of ammonia at the same liquid residence time at any given temperature. This can be explained by the different liquid flow rates. Since we chose to maintain the same residence time in these initial configuration temperature comparisons, the liquid flow rate for configuration A (liquid in the outer tube) needs to be 2 faster than in the second configuration (liquid in the inner tube).5 times. This reflects the ratio of the volume of the liquid phase in each configuration. When the liquid phase concentration of ammonia is significantly lower than the saturation value, it can be expected that the net permeation rate across the membrane should not depend on the ammonia concentration in the liquid phase, but rather on the surface area between the phases. Using the inner and outer diameters of the AF-2400 tube, the relative values of the surface area of configurations A and B are 62., respectively83 cm and 5027 cm, therefore, we might expect a net ammonia permeability of approximately 1. per unit time for configuration A25 times. If configured A has a traffic rate of 25-fold (dilution factor is 2.)5 times), this equates to an ammonia concentration approximately 2 times higher in configuration B0x. If we assume that the liquid and gas phases extend into the amorphous structure of the membrane to the same extent (i.e., they meet in the middle), then we can ignore the thickness of the membrane, which has the same relative contact surface area in both configurations. The concentration will depend only on the flow rate ratio (25)。Using methanol data at 10 °C as an example, the measured ammonia concentration (372 and 130) ratio (287) is fairly close to this estimate. The relationship between ammonia concentration and residence time in the new configuration was also studied. As shown in Figure 6, there is a strong correlation between the concentration of methanol and the residence time, while for DME, a slight deviation from linearity can be seen at higher residence times. Interestingly, for toluene, at longer residence times, when the concentration is close to 2At 0 m, the ammonia concentration seems to begin to stabilize. However, for both methanol and DME, saturation was not reached;In the case of methanol, the ammonia concentration is measured at 010 ml min reaches 876 m (equivalent to a residence time of 10 minutes in a gas reactor). 
Figure 6Plot of ammonia concentrations in MEOH, DME, and toluene versus 0 residence time (configuration B, Figure 1). Although ammonia solutions in some solvents are commercially available, the choice of solvents and concentrations is quite limited. In addition, unless the container is completely sealed, the ammonia concentration can decrease significantly over time. It is important to note that ammonia can be loaded into an arbitrarily selected solvent (methanol, dimethyl ether, and toluene as shown here) using this conceptually simple and easy-to-assemble tube-in-tube reactor at a controlled and controllable concentration. This process provides an opportunity to explore new solvent options, including the use of solvent blends in a variety of reactions using ammonia. Based on the results obtained from the study of the two reactor configurations, the data indicate that the ammonia concentration obtained at a specific interfacial surface area (at pre-saturated concentrations) is inversely proportional to the flow rate. Similarly, the concentration of ammonia obtained at a specific flow rate (again, at a presaturated concentration) appears to be proportional to the interfacial surface area, and for a fixed residence time, the concentration is proportional to the ratio of the interfacial surface to the volume of the liquid. Using the data shown in Figure 1, the ratio is 25 for configuration a64 cm (1), which is 49 for configuration b77 cm^(−1)。In the latter configuration, a tube-in-tube reactor can be likened to a three-channel microreactor developed by Kim and colleagues, in which the reaction channels are exposed to gas from both sides, increasing the effective interface area between the liquid and gas streams. In this work, the authors demonstrated that a double-sided channel microreactor has better performance than a single-sided channel microreactor, as shown by the photosensitive oxidation reaction. In addition, a detailed mathematical analysis of these considerations has recently been published. The ability of our system to provide a stable and consistent concentration of ammonia solution over time was also evaluated. Over the course of 4 days, the reactor unit turned the power on and off every day, with the same parameter settings as in the previous day's experiment (DME, 0.).5 ml min flow rate and 25). Concentrations are checked daily by titration. The results obtained are shown in Figure 7. 
Figure 7Reproducibility and stability evaluation of ammonia concentrations in DME at 25 (flow rate 0.).5 ml/min；Configuration b). This clearly indicates that the ammonia concentration in the stream is relatively stable, showing consistency and reproducibility between days and from day to day (0.).97 ± 0.05 mol/l）。With evidence that there is a robust and reliable method to provide ammonia solutions of known concentrations on demand, as a procedure with a long lifespan, we next turn to the synthesis application of the device. Reactive gases are often used in large quantities (to drive the reaction to completion) because of their inherent function to easily remove from the reaction mixture, providing easier post-processing and a cleaner process. However, this default mindset is inherently wasteful, and is a particular problem when using more toxic gases that cannot be safely released into the atmosphere. Here, the excess gas has to be recirculated or somehow removed by scrubbing the exhaust stream, which is clearly not ideal from an economic point of view. Therefore, we sought to develop a synthetically useful reaction capable of using stoichiometric levels of ammonia. We first add ammonia to isothiocyanate to prepare thiourea in a continuous flow. Thiourea is an important component in the preparation of valuable materials, which has also been intensively studied in the field of molecular identification, and has recently been used as an effective catalyst in a series of organocatalytic reactions. The setup of the flow device for the initial optimization experiment is shown in Protocol2, where the use of sample loops is a prelude to a larger-scale continuous run. Stream methanol or DME at 05 ml min pumped into the tube-in-tube reactor (configuration B, Figure 1), stable at 0 (under these conditions, the concentration of MeOH is measured at 1.)44 m at a concentration of 137 m）。Add a one-milliliter sample loop upstream of the tube-in-tube reactor** into which a solution of alkyl isothiocyanate 1 is introduced. The gaseous liquid stream carrying the isothiocyanate is then fed into another reactor coil (PFA, inner diameter 2.) placed on the second Polar Bear Plus reactor heating block44 mm, 10 ml), the reactor coil is heated to 80 and then discharged from the system through a back pressure regulator. 
Scenario 2Flow settings for reaction optimization Initially, the concentration of the isothiocyanate solution injected into the sample loop is changed to find the maximum value that can be fully transformed during the residence time of the reactor. We note here that when performed under fractional flow conditions, (1) the segmented plug will undergo inevitable dispersion before the second (heated) coil, therefore, the stoichiometry of ammonia with the substrate will be different, (2) the presence of solutes in the solvent stream may alter ammonia absorption, and (3) ammonia may be consumed by reaction with isothiocyanates. Since the concentrations produced under these conditions are well below the saturation level, any effect of the reactive solute changing the equilibrium position (resulting in greater net absorption of ammonia) should be minimal, especially when the reaction requires the higher temperature of the downstream heating stage to proceed at a significant rate. Satisfactorily, DME concentrations are as high as 12 m and methanol concentrations up to 1A 6 m solution of isothiocyanate can be treated in such a way that it is completely converted to thiourea 2. The maximum concentration of injected reagent is greater than the measured concentration of ammonia (no reagent) and complete transformation is still observed, which supports the consideration (1). Due to the observation of small amounts of methyl thiocarbamate by-products in the case of methanol (<20%), so continue to use DME for further research. These preliminary results show that we can match the stoichiometry of reagents and reaction gases and achieve good product conversion. On this basis, we evaluated the reactor's ability to achieve a complete continuous flow of thiourea. Except for bypassing the loading loop and introducing the reagents through the stock solution, the setup is identical. In this case, we need to change the conditions to ensure complete conversion: isothiocyanate starts with 1Introduced in the form of a 0 m solution, the temperature of the second reactor was raised from 80 to 100 and the flow rate was from 05 ml min to 06 ml min (Based on the absorption rate, the changed residence time should provide 106 m ammonia), see Scheme 3. Under these conditions, the reaction was carried out for more than 5 hours, and after a simple concentration of solvent on a rotary evaporator, 343 grams of product required (equivalent to 6.)5 g hour of productivity), high yield, good purity (>95% by H nmr spectroscopy). This is in stark contrast to typical batch reaction procedures, as only 0. is required under continuous flow conditions06 Complete conversion can be achieved with an equivalent amount of excess ammonia, whereas in batch reactions, excess ammonia, possibly due to difficulty in controlling stoichiometry, is common in these reactions (typically 7.).0 equal parts of methanol). 
Scenario 3The process for continuous thiourea preparation is set up among other advantages associated with flow chemistry techniques, and the possibility of stretching the reaction to integrate the individual steps into a single process is attractive. We and others have validated this approach in the synthesis of drug molecules and natural products. This strategy has also been successfully applied to the preparation of useful building blocks using the integrated flow of branched aldehydes directly from aryl iodide using three gases. With these concepts in mind, we investigated the preparation of 2-aminothiazole in a stretchable manner, as the 2-aminothiazole moiety has been shown to be a key pharmacophore in many pharmaceutical active agents. As a representative of this family, farnetinazole is a known anti-inflammatory drug that has entered a phase II clinical trial for rheumatoid arthritis. Although Hantzch synthesis is a typical selection strategy, a number of methods for thiazole derivatives have been reported to date, including the flow method. However, despite their potential use for delivering a library of compounds, these flow methods rely on the use of microchannel reactors under electroosmotic flow control or the use of fluorinated solvents under a segmented flow procedure to deliver the product. Therefore, we designed a reactor structure for the preparation of fanetizole as shown in Scheme 4. The flow of thioure2-containing gas (05 ml min) on a Y-piece with a gas stream of 3-bromoacetophenone 3 (05 ml min) and the new gas stream passes through another reactor coil (PFA, I.D. 2.) located on the same heating block44 mm，5 ml）。After experimenting with various solvent combinations, we were pleased to find that using a mixture of dimethyl ether water DMF, each in a 3:1:1 ratio, avoided both phase separation of the 3-bromoketone stock solution and precipitation of the thiazole bromide product in the back pressure regulator. Therefore, the newly formed alkyl thiourea was further condensed with 3-bromoacetophenone in the second reactor coil to obtain the corresponding 2-aminothiazole farnetazole 4. The material precipitates out of the reaction solution after cooling (Inset, Scheme 4). In order to collect the flowing stream, we chose to abandon the first 30 minutes of the run and then collect it for 7 hours. This provides product 515g (yield 68%), which is thiazolium bromide salt 5 with excellent purity. The 1H NMR spectrum of farnetizole·HBR was 13A characteristic peak is shown at 81 ppm, corresponding to the hydrogen of the thiazoleium ring. The remaining material left in the solution is also characterized as the free base form of the drug substance. After liquid-liquid extraction, it is separated from the mother liquor and further obtained182 g of farnestrozole free base (31% yield). Overall, the total yield of both forms of farneshezole was 99% over a 7-hour run. Farnestrozole free base can be obtained from thiazole gallium salt by washing hydrogen bromide from a chloroform solution with distilled water.
Scenario 4The telescopic process setup including thiourea preparation and farnestrozole Hantzsch synthesis is shown in Figure 8. This arrangement can also be easily applied to the small- and medium-scale preparations of other biologically important thioureas and 2-aminothiazoles. **Titration methods can also be integrated into flow systems, enabling on-demand titration to check ammonia concentrations, which is particularly useful during longer reaction runs.
Figure 8** Complete apparatus for farnetrozole synthesis. The above systematic studies have led to a better understanding of these pipe-in-tube gas-liquid flow devices, including many operating parameters. The combination with the heater cooler block has shown that the temperature does affect the absorption of gases into the stream, and in the case of ammonia, for the three test solvents, in 3At a pressure of 5 bar, the temperature reaches a maximum of about 0. Below such temperatures, we observe what we consider to be the enthalpy effect of vaporization, which results in what is seen at temperatures below the beginning of liquefaction (35 bar at 5 bar). All of the solvents tested showed a clear overall trend over the temperature range, proving that the setup was reliable and robust. In addition, it is clear from the data that the configuration of the liquid in the inner tube and the gas in the outer tube is the most efficient for achieving the highest concentration (using the pipe size used). This result is consistent with the difference in the surface area-to-volume ratio between the two possible configurations, and indicates that the two configurations provide comparable heat exchange and temperature control capabilities. After determining the optimal operating parameters, we demonstrated the longevity and reliability of the method over a four-day period through multiple power-on/shutdown cycles. This capability was then used as a preparation platform to prepare 35 grams of thiourea product in an unattended manner and with negligible excess of gaseous reagents. Finally, a nested procedure was designed to deliver the drug substance at a substantially quantitative yield and a useful scale (70 g) without any chromatographic purification. The process developed in this work is attractive in terms of safety, as the volume of gas in the reactor is small in any given situation (245 ml）。As a result, the risks associated with the use of conventional high-voltage batch equipment are avoided. Finally, all the materials used to build the tube-in-tube reactor were commercially available, and the same reactor was used throughout the job, allowing more than 300 data measurements and many reaction cycles to be reliably performed. General steps. Reagents and solvents are commercial grade and are used as per the standards provided. Reactions were monitored using Merck silica gel 60 F-254 thin layer plates by thin layer chromatography by UV fluorescence (= 254 nm) or acid potassium permanganate staining. H NMR and 13C NMR data were recorded by Bruker **ance (400 MHz for H NMR and 100 MHz for C NMR) spectrometer, using the residual solvent peak as the internal control (CDCL3 = 7.).26 ppm(h) and 770 ppm（c；dmso-d6 = 2.50 ppm(h) and 394 ppm（c））。The data are reported as follows: chemical shift ppm (δ multiplex (s = singleplex, d = double, t = triple, q = quadruple, quint = quint, m = multiplex, br = wide signal or a combination thereof), coupling constant (hz), integral. Infrared spectra were obtained on a PerkinElmer Spectrum One Fourier transform infrared spectrometer using the Universal ATR sampling accessory, and the absorption was measured in reciprocal centimeters. The melting point was recorded on an Stanford Research Systems MPa100 (Optimelt) melting point instrument and was not calibrated. High-resolution mass spectrometry (HRMS) was performed using a Waters Micromass LCT Premier Spectrometer using a positive ESI time-of-flight analysis with an error of less than 5 ppm from the theoretical calculations. Mesoscale continuous flow preparation of thiourea. Using Uniqsis BPM, start with 0At a rate of 6 ml min, 2-phenylethyl isothiocyanate 1 solution (1.)0 m, 1,2-dimethoxyethane) pumped into a tube-in-tube reactor (L = 2 m, af2400) and replaced with 35 bar of ammonia (gas in the outer tube) pressurized. The gas reactor is placed on a Polar Bear Plus cooling block at 0 °C and the outgoing reaction stream is fed into a 10 mL PFA reaction coil (inner diameter = 2.).44 mm), which is placed on the heating block of the second Polar Bear Plus at a temperature of 100 . The output stream passes through a back pressure regulator (p = 6 bar) and after a steady state (30 min), collected in a round-bottom flask5After 3 hours of the sample, volatiles were removed under reduced pressure to obtain a white solid thiourea 2 (34.) with a fixed yield yield3 g;NMR purity greater than 95%). Mesoscale telescopic flow synthesis of farnetizole (4). Using Uniqsis BPM, start with 0At a rate of 5 ml min, 2-phenylethyl isothiocyanate 1 solution (10 m, 1,2-dimethoxyethane) pumped into a tube-in-tube reactor (L = 2 m, af2400) and replaced with 35 bar of ammonia (gas in the outer tube) pressurized. The gas reactor is placed on a Polar Bear Plus cooling block at 0 and the outgoing reacting stream is fed into a 10 mL PFA reaction coil (inner diameter = 2.).44 mm) and place on another 100 Polar Bear Plus heating block. The output stream is mixed at the Y-junction with a solution of 3-bromoacetophenone 3 (10 m,1,2-dimethoxyethane water n,n-dimethylformamide 3:1:1) with a pump speed of 05 mL min, the resulting stream is fed into a 5 mL PFA reaction coil (inner diameter = 2.).44 mm) and place on the heating block of the second Polar Bear Plus at 100 °C. The output stream passes through a back pressure regulator (p = 6 bar) and samples are collected in a beaker for 7 h after steady state (30 min). After cooling to room temperature, white crystals were precipitated from the solution and then filtered in a Bü Chner funnel to give fanetizole-HBR (5) in 68% (51.).5 g;NMR purity >95%), white crystals. After extraction with methylene chloride (3 150 ml), the combined organic layer was washed with water (5 200 ml) to remove residual DMF. The organic layer was then dried on MgSO4, filtered, and concentrated to dryness to give nitrofurantidazole free radicals (4) with a yield of 31% (18.).2 g;NMR purity >95%), white solid. In addition, after recrystallization with cold methanol and washing the chloroform solution with water, fanetizole free base (26% yield) can also be isolated from the mother liquor. Finally, fanetizole free base can be quantitatively obtained from thiazole salts by washing (5 times) with water (5 times) the solution of fanetizole-hbr in chloroform. E-mail: li@fulinsujiaocom