4544
Ind. Eng. Chem. Res. 2006, 45, 4544-4553
Study of the Hydrogenation of Selected Nitro Compounds by Simultaneous Measurements of Calorimetric, FT-IR, and Gas-Uptake Signals Fabio Visentin, Graeme Puxty, Oemer M. Kut, and Konrad Hungerbu1hler* Safety and EnVironmental Group, Institute for Chemical and Bioengineering, ETH Zurich, 8093 Zurich, Switzerland
The hydrogenation of nitrobenzene (NB) and ethyl-4-nitrobenzoate (NEE) over a Pd/carbon (1 wt % Pd) catalyst has been studied using a new small-scale, pressure-resistant reaction calorimeter (CRC.v4) fitted with an integrated infrared-attenuated total reflection (FT-IR-ATR) probe. This new calorimeter exploits the principles of power compensation and heat balance in combination with IR spectroscopy and on-line gas-uptake measurements. Thus, the reactions can be followed by three independent signals based on different properties of the chemical components. It was shown that, in the temperature and pressure range studied, all three simultaneous measurements of the NB hydrogenation can be described by a simple empirical kinetic model. The presence of an electron acceptor substituent like that in NEE leads to a consecutive hydrogenation with the accumulation of the corresponding hydroxylamine as an intermediate. The different information content of the simultaneously measured signals allow a quantitative description of the hydroxylamine accumulation. The calculated concentration profiles were confirmed by off-line high-pressure liquid chromatography (HPLC) analysis. The same empirical kinetic model was used after appropriate expansion for a consecutive reaction. Introduction Reaction calorimetry has become one of the most powerful tools for the purpose of kinetic and thermodynamic screening in the early stages of process development. For this reason, a new reaction calorimeter exploiting a combination of the powercompensation and heat-balance principles1-3 in conjunction with IR spectroscopy and gas-uptake rate measurements was used to investigate heterogeneous hydrogenation of two different nitro compounds. The purpose was to demonstrate that the kinetic information needed at the early stages of chemical-process design can be acquired rapidly and consistently by these three measurements and that they can provide complimentary information. The new reaction calorimeter (CRC.v4) has a volume of 2545 mL and is equipped with an integrated IR-attenuated total reflection (IR-ATR) probe1,3 coupled to an FT-IR spectrometer to improve the information content of a single semibatch measurement of a reaction. A metal jacket is used instead of a glass device. Energy is removed from the jacket via Peltier elements. The combination of the power-compensation and heatbalance principles means no calibration of the heat-transfer coefficient is necessary. This principle is very attractive for small-scale reaction calorimeters because of the simplicity of the technique, as compared to conventional heat-balance implementations Moreover, the new reaction calorimeter is pressure-proof up to 30 bar. This allows reactions to be carried out under pressure and, in combination with the changes in pressure of an intermediate pressure reservoir, the determination of the gasuptake rate. To show the benefits of the CRC.v4 for monitoring multiphase reactions, three phase reactions under pressure were chosen. Hydrogenations of nitrobenzene (NB) and ethyl-4nitrobenzoate (NEE) were performed in order to demonstrate the ability and advantages of the monitoring the reactions with * Corresponding author. E-mail:
[email protected]. Tel.: +41 44 632 60 98. Fax: +41 44 632 11 89.
three different measurements:4 (1) calorimetry (for determination of the heat production); (2) IR spectroscopy (for determination of concentration profiles); and (3) gas consumption (H2 uptake). Simultaneous measurements of multiple signals allows deeper insight into complex reaction systems and may reveal effects that are not detectible from a single signal alone.1 The fitting of kinetic models to these data then leads to more rapid and complete characterization of the system in question. Because the purpose of this study was to highlight the benefits of simultaneously measuring multiple signal types for heterogeneous reaction monitoring, an empirical kinetic model was developed from a generally accepted kinetic model for hydrogenation reactions to represent the data. Modifications made to the standard model for the purpose of improving the model fit were empirical in nature, and no mechanistic interpretation will be offered. Catalytic hydrogenation of aromatic nitro compounds is an industrially important process for the introduction of amino functionality into pharmaceutical and agrochemical intermediates and in the polyurethane chemistry. Aromatic nitro compounds are very easily hydrogenated, and hydrogenations have been carried out under a wide range of conditions. They are known to be potentially hazardous reactions, especially because hydroxylamine intermediates formed are often thermally unstable and can disproportionate with a significant increase of the temperature and subsequent explosions.5 Without a reliable kinetic model, the maximum concentration of hydroxylamine accumulation is difficult to predict. During aryl-nitro hydrogenation, formation of the bimolecular azo and azoxy compounds is also possible.6 These compounds can in turn be hydrogenated to the arylamine along with formation of hydrazo compounds. The extent of azo and azoxy formation depends on temperature and accumulation of arylhydroxylamine.7 Depending on a number of factors discussed later, arylhydroxylamine can build up to significant concentrations during the hydrogenation. Accumulation of nitroso compounds has not been observed,8 although their transient presence may appreciably influence the course of reaction. In practice,
10.1021/ie0509591 CCC: $33.50 © 2006 American Chemical Society Published on Web 05/17/2006
Ind. Eng. Chem. Res., Vol. 45, No. 13, 2006 4545
nitro compounds often contain other functional groups that are to be either maintained or reduced as well. Commonly, for the catalytic hydrogenation of nitro aromatic compounds, the reactions are carried out in three-phase systems employing a metal catalyst such as Pd, Pt, Ni, or Cu. They are usually supported by active carbon, alumina, or silica. Other less common metals and supports are also used, especially for chemoselective and enantioselective reactions. The choice of catalyst often depends on what other functional groups are present on the substrate and on the desired product. The reactions are usually fast, and the observed rate is often limited by the rate of hydrogen transfer to the catalyst. Aliphatic nitro compounds are reduced much more slowly than aromatic ones, and higher catalyst loadings or relatively lengthy reduction times may be required. NB is the simplest aromatic nitro molecule suitable for this type of reaction, and it is liquid at room temperature. It can be easily dosed into the reactor by a pump. The catalytic hydrogenation of NB was chosen because it is commonly employed as a standard reference reaction for testing and comparing the activity of hydrogenation catalysts for a range of applications.9-12 Several reaction schemes are suggested in the literature.8,13-15 The second reaction considered was the catalytic hydrogenation of NEE. With this molecule, it is possible to accumulate ethyl-4-hydroxylaminobenzoate as an intermediate. Empirical mathematical models have been developed for semibatch isothermal reaction calorimetry measurements, and the experimental and simulated results have been compared. Much reference data about the kinetics of aryl-nitro compounds are available in the literature, but the problem is that the catalyst used is often not sufficiently characterized. Since the data are specific for a catalyst preparation, it is not possible to make a precise comparison of kinetic data with the literature. However, it is possible to compare reaction enthalpies, as they are independent of the kinetics. Experimental Section Materials. In the present study, all the hydrogenation reactions were carried out over commercially available Pd/C 1%, a solid powder of 1% in weight of palladium supported on active carbon with a particle size in a range of 3-10 µm provided by Acros Organic, Switzerland. The catalyst was ovendried but otherwise used as provided. The hydrogen gas was supplied by Pan Gas Zurich (CH), with 99.995% purity, and was used directly from the hydrogen reservoirs. The substrates NB (purity g 99.5% GC) and NEE (purity g98% (GC)) were both provided by Fluka AG Switzerland. The solvent used for the hydrogenation of NB was ethanol absolute (ACS g 99.8% GC from Scharlau), and that for the hydrogenation of NEE was methanol (ACS g 99.8% GC from Merck). All reagents and solvents were used as provided. Experimental Setup. The experiments were performed using a new pressure-resistant, small-scale reaction calorimeter (CRC.v4) that combines calorimetry with an FT-IR spectrometer and ATR probe that allows on-line measurement of the infrared absorbances. The device has been described in detail elsewhere.1-3 The sample volume of the reactor is 25-45 mL, it uses a metal block as an intermediate thermostat, and it is able to withstand pressure up to 30 bar. The maximum temperature of the reactor is ∼200 °C, and the minimum temperature depends only on the cryostat used. Consequently, for the described configuration of the CRC.v4, the practical range of working
temperatures is between -20 and 200 °C. Batch and semibatch reactions can be performed. Isothermal conditions are maintained using the power compensation principle. Peltier elements are coupled to an on-line feedback control to compensate the change of the overall heat transfer through the metal block during the measurement,16 making time-consuming calibration unnecessary. The Hastelloy reactor vessel is easily exchangeable and available with and without the FT-IR-ATR probe. The reactor is also fitted with a magnetically driven impeller with a four-blade Teflon stirrer capable of operating up to 2200 rpm. The reaction was carried out at a constant H2 pressure. The consumption of H2 in the hydrogenation reaction, as a function of time, was measured by the change in the H2 pressure in an intermediate gas reservoir and from the volume of the intermediate reservoir, the H2 compressibility factor, and the temperature of the gas. The measured uptake rate was monitored online by the pressure drop in the intermediate cylinder using the data acquisition software package LabView 6.1.17 The moles of gas consumed in a reaction (∆n) as function of time was estimated according to eq 1 from the change of the gas pressure in the intermediate vessel and from the volume of the intermediate vessel, the gas compressibility factor, and the gas temperature.
∆n )
(
)
Vres PH2i PH2f RTres zi zf
where z ) 1 + RH2PH2
(1)
where Vres is the volume of the reservoir in cm3, Tres is the temperature of the reservoir in K, R is the gas constant 83.14 cm3‚bar/mol‚K, PH2i is the initial and PH2f is the final pressure of the reservoir in bar, z is the compressibility factor in bar, and RH2 is the compressibility constant for hydrogen. The FT-IR-ATR spectroscopy measurements18 were carried out using a ReactIR 400019 system and the ReactIR 3.03 software package, both from Mettler Toledo. Principle component analysis (PCA) of the resulting multivariate data was done with the software ConcIR 3.0, also from Mettler Toledo. The ATR window was mounted directly in the base of the reactor vessel and coupled to the spectrophotometer. The acquisition of a single spectrum was carried out every 25 s during the reaction period. The progress of the reaction was also followed off-line by a high-pressure liquid chromatography (HPLC) analysis (HewlettPackard Series 1100 MSD) of samples withdrawn from the reaction mixture. To avoid further reactions,the catalyst was separated immediately by filtration. Quantitative concentration determination was done by calculating the area of the chromatographic peaks with an electronic integrator (HewlettPackard Series 1100). Experimental Procedure Hydrogenation of Nitrobenzene. It was assumed that the hydrogenation of NB follows the reaction shown in Figure 1. In this scheme the hydrogenation of NB is a simple reaction without any significant intermediate accumulation during the formation of aniline (AN).
Figure 1. Simplified reaction scheme for the hydrogenation of NB.
4546
Ind. Eng. Chem. Res., Vol. 45, No. 13, 2006
Table 1. Recipe Used for the Hydrogenation of NB solvent catalyst substrate ratio catalyst/substrate stirrer speed temp in the reactor pressure in the reactor
ethanol 35 mL Pd/C 1% nitrobenzene 4.1 g ) 0.0333 mol (3.4 mL in 25.5 s) 2.34 g‚mol-1 1200 rpma 35, 40, 45, 50, 55, 60 °C 13 bar (H2, const.)a
a The stirrer speed was varied between 600 and 1800 rpm and the pressure was variedbetween 0 and 22 bar to determine the nonmass-transfer-limited region.
The reactor was flushed with N2 at atmospheric pressure to remove the air in the reactor. To start the reaction, H2 was injected in the reactor as rapidly as possible, increasing the pressure to the desired value of 13 bar. Finally, the stirrer speed was increased to 1200 rpm. The reaction was followed for ∼1 h, until no further reaction was observed. The experiment was carried out three times. It should be noted that, because of the different procedure used for NEE, the results are not directly comparable to those for NB. For the NB reactions, liquid NB was dosed into the reactor with the catalyst already in the presence of H2. For NEE, because it is a solid at room temperature, it was part of the initial reactor charge, and H2 was dosed. This means that, for the NEE reactions, following the addition of H2 there was a time lag until the catalyst was saturated with H2 and reached maximum activity. Results and Discussion
Figure 2. Consecutive reaction scheme for the hydrogenation of NEE. Table 2. Recipe for the Hydrogenation of NEE solvent substrate catalyst ratio catalyst/substrate stirrer speed temperature of the reactor pressure in the reactor
methanol 35 mL ethyl-4-nitrobenzoate 6.5 g (0.033 mol) Pd/C 1% 5.15 g‚mol-1 1200 rpm 50 °C 13 bar (H2, const.)
The standard measurements were made according to the following procedure. First, the reactor was filled with 35 mL of ethanol absolute and 0.078 g of Pd/C 1%, the stirrer was initially set at 400-500 rpm, and the desired reaction temperature was set. In a second step, the reactor was flushed with N2 at atmospheric pressure to remove the air in the reactor and it was pressurized with H2. The H2 pressure in the reactor was increased from atmospheric pressure to the desired value of 13 bar. Afterward, the stirrer speed was increased to 1200 rpm. In a last step, 4.1 g (0.033 mol, 1.2 g‚ml-1) of NB was added with a constant dosing rate of 8 mL/min. To determine the energy introduced by dosing, qdos, the heat capacity of the feed (1.76 kJ‚kg-1‚K-1)20 was used. The reaction was followed for ∼1 h, until no further reaction was observed. All experiments were carried out three times under identical conditions. Hydrogenation of Ethyl-4-nitrobenzoate. It was assumed that the hydrogenation of NEE is a consecutive reaction with accumulation of the hydroxylamine intermediate (HEE) during the formation of ethyl-4-aminobenzoate (AEE) (see Figure 2).2 The reactor was filled initially with 35 mL of methanol, 0.170 g of Pd/C (same catalyst as for NB), and 6.5 g (0.033 mol) of NEE. The stirrer was set at 400-500 rpm, and the desired reaction temperature was set at 50 °C.
Hydrogenation of Nitrobenzene. As mentioned previously, the target of the reduction is the amine, but under specific conditions, an accumulation of the hydroxylamine intermediate can occur. Under the experimental conditions defined in Table 1, the observed accumulation of the hydroxylamine was