REACTOR AND METHOD FOR MEASUREMENT OF SPATIALLY RESOLVED PROFILES IN PERMEABLE CATALYST BODIES

Abstract

A reactor for measurement of spatially resolved profiles in a single catalyst body includes, in the single catalyst body, a channel which accommodates a sampling capillary. The sampling capillary includes a sampling orifice for taking samples of a fluid phase passing the single catalyst body. By shifting the sampling capillary relative to the single catalyst body spatially resolved profiles of a parameter can be obtained.

Claims

1. A reactor for measurement of spatially resolved profiles of a parameter in a single catalyst body, comprising: a reactor chamber comprising at least one reactor wall defining the reactor chamber; at least one sampling capillary passing the reactor wall through an entry port and protruding into the reactor chamber, said at least one sampling capillary comprising at least one sampling orifice, said sampling orifice being arranged in the reactor chamber a holding device for holding a single catalyst body in the reactor chamber, wherein the sampling capillary is arranged such that the sampling capillary passes through the single catalyst body after the single catalyst body has been fixed to the holding device.

2. A reactor according to claim 1, wherein the sampling capillary is movable in at least one direction relative to the single catalyst body and a actuator is provided for movement of the sampling capillary or the single catalyst body.

3. A reactor according to claim 1, wherein a single catalyst body is provided on the holding device for holding a single catalyst body, a channel is provided in the single catalyst body and the sampling capillary is guided through the channel.

4. A reactor according to claim 1, wherein the sampling capillary is traversing the reactor chamber and is guided in a port provided in the reactor wall arranged opposite of the entry port of the sampling capillary in the reactor wall.

5. A reactor according to claim 1, wherein at least one sampling orifice is provided in a sidewall of the sampling capillary.

6. A reactor according to claim 1, wherein a window transparent to electromagnetic radiation is provided in a reactor wall.

7. A reactor according to claim 1, wherein a temperature-sensitive sensor is arranged in the reactor chamber, wherein the temperature-sensitive sensor has the form of a pyrometer fiber or of a thermocouple comprising a tip for sensing a temperature.

8. A reactor according to claim 7, wherein the pyrometer fiber or the thermocouple is arranged inside the sampling capillary and the tip of the pyrometer fiber or of the thermocouple is arranged at the sampling orifice of the sampling capillary.

9. A reactor according to claim 1, wherein the reactor comprises a sensor for collecting spectroscopic information.

10. A reactor according to claim 9, wherein the sensor for collecting spectroscopic information comprises a fiber transparent for electromagnetic radiation.

11. A method for detecting a spatially resolved profile of a parameter in a single catalyst body, wherein, i) A single catalyst body is provided, ii) A channel is provided in the single catalyst body; iii) at least one sampling capillary is inserted into the channel provided in the single catalyst body; iv) A fluid phase is flown through the single catalyst body; v) The sampling capillary is moved relative to the single catalyst body to a first position, a sample of the fluid phase is taken through the sampling orifice of the sampling capillary at the first position and the sample is analyzed for a first value of the parameter; vi) The sampling capillary is moved relative to the single catalyst body to at least a further position, a further sample of the fluid phase is taken through the sampling orifice of the sampling capillary at the further position and the further sample is analyzed for a second value of the parameter; vii) Step vi) is repeated to obtain a spatially resolved profile of the parameter.

12. A method according to claim 11, wherein a reactor according to any one of claims 1 to 10 is used and the single catalyst body is fixed to the holding device for holding a single catalyst body provided in the reactor

13. A method according to claim 11, wherein the sampling capillary is moved relative to the single catalyst body in a forward direction to detect a first spatially resolved profile and the sampling capillary is moved relative to the single catalyst body in a backward direction to obtain a second spatially resolved profile.

14. A method according to claim 11, wherein at least one further parameter is determined.

15. A method according to claim 14, wherein the at least one further parameter is determined at the position of the sampling orifice of the sampling capillary.

16. A method according to claim 11, wherein the single catalyst body is a heterogeneous catalyst.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0248] The Invention will be explained in more detail with reference to the accompanying figures, wherein the figures display:

[0249] FIG. 1: an illustration of a reactor chamber of a reactor according to the invention.

[0250] FIG. 2: a diagram showing mole fractions of CO.sub.2 in % for the intact particle (a), the drilled particle with the capillary going through the whole channel (b) and the drilled particle with the capillary tip (c);

[0251] FIG. 3: simulated CO.sub.2 mole fractions inside the catalytic cylinder. For the pristine particle (short dash), the drilled particle with the capillary inside without sampling (short dot), sampling with the side orifice of a 10 μm ID capillary (solid) and with a 40 μm ID capillary (dash), as well as the tip of a 10 μm ID capillary (short dash dot);

[0252] FIG. 4: Spatially resolved mole fraction of CO.sub.2 for T.sub.in=173 (□), 182 (•) and 191° C. (Δ) with 3.2% O.sub.2 and 5.4% CO, measured with the tip of the capillary (a), (b) shows a repetition of the experiment;

[0253] FIG. 5: Particle temperature for different O.sub.2 mole fractions with T.sub.in=173° C. and x.sub.CO=5.4% by increasing (black) and decreasing the latter (red) (right hand side). Spatial profiles of CO.sub.2, CO and O.sub.2 for the same inlet conditions (left hand side), the upper one in the diffusion limited state, the lower one in the reaction limited state;

[0254] FIG. 6: Particle temperature profiles (top) and the corresponding CO.sub.2 signal at m/z=44 amu (bottom); for Re=5.7 without reaction (left), 2.2% O.sub.2 and 1% CO (middle) and 3.4% O.sub.2 and 1% CO (right);

[0255] FIG. 7: Spatially resolved mole fraction of CO.sub.2 (□) and O.sub.2 (•) with 3.2% O.sub.2, 5.4% CO and T.sub.in=170° C., measured with the side orifice (red) and the tip of a 5 μm ID capillary (blue)

[0256] FIG. 8: Temperature of the particle T.sub.P for increasing (lower branch) and decreasing (upper branch) the inlet temperature T.sub.in

[0257] FIG. 9: a single catalyst body with a sampling capillary being inserted in a channel;

[0258] FIG. 10: a sectional view of a single catalyst body with a sampling capillary traversing the single catalyst body;

[0259] FIG. 11: a detail of FIG. 10;

[0260] FIG. 12: a sectional view of a single catalyst body with a sampling capillary having a sampling orifice at a distal end;

[0261] FIG. 13: a detail of FIG. 12;

[0262] FIG. 14: a sectional view of a single catalyst body with a sampling capillary traversing the single catalyst body, wherein a fiber is introduced into the sampling capillary;

[0263] FIG. 15: a sectional view of a single catalyst body with a sampling capillary having a sampling orifice at a distal end, wherein a fiber is introduced into the sampling capillary;

[0264] FIG. 16: a sectional view of a single catalyst body with a sampling capillary having a sampling orifice at a distal end, wherein a fiber is introduced from an opposite of a channel provided in the single catalyst body.

DETAILED DESCRIPTION

Materials and Methods

[0265] Catalyst

[0266] The catalyst used in this study is a platinum-coated, porous alumina cylinder. The alumina support (length=5 mm, diameter=5 mm) was kindly provided by Sasol Germany GmbH (Hamburg, Germany). The particle has the following properties: density is 1168 kg m.sup.−3, porosity is 0.55, BET surface area is 203 m.sup.2 g.sup.−1, average pore size is 57 Å, and thermal conductivity 1 W m.sup.−1 K.sup.−1. Surface area and pore size were determined with a Quantachrome autosorb IQ 2 (Anton Paar GmbH, Graz, Austria); thermal conductivity with a C-Therm TCiTM thermal conductivity analyzer (C-Therm Technologies Ltd., Fredericton, Canada). A channel with a diameter of 300 μm was drilled through the middle of the circular surface of the particle. Afterwards, the 3 wt. % platinum/alumina catalyst was prepared by incipient wetness impregnation with an aqueous solution of H.sub.2PtCl.sub.6*6H.sub.2O (˜40% Pt, Carl Roth GmbH, Karlsruhe, Germany) as precursor, and dried in a desiccator. The catalyst was reduced in a tubular furnace with 5% H.sub.2 in N.sub.2 (V.sub.total=450 ml min.sup.−1) at a temperature of 500° C. for 5 h.

[0267] Set-Up

[0268] The reactor is designed to measure concentration profiles inside a single catalyst particle in a defined flow and temperature field. FIG. 1 shows a sketch of the reactor. The reaction chamber has a width and height of 2 cm, as well as a length of 6 cm. During operation, inlet temperature is kept constant by a PID controller (Eurotherm, Worthing, UK). The temperature in the reactor is measured with a type K thermocouple (TMH, Maintal, Germany). Additionally, by drilling a hole with a diameter of 300 μm and a depth of 2 mm into the particle, it is possible to place a thermocouple (250 μm diameter) directly inside the former—and thus measure the temperature within the particle.

[0269] The reactor is equipped with a glass window on top, to allow the use of optical methods (e.g. Raman spectroscopy). Therefore, the setup is not adiabatic, since the presence of the window causes heat losses. Reactant gases (CO and O.sub.2, as well as Ar as inert) are fed via calibrated mass flow controllers (Brooks instruments, Hatfield, USA). The gases are mixed, and subsequently pass a flow straightener before entering the reactor (FIG. 1, inlet).

[0270] The catalyst is mounted on a magnesia rod (FIG. 1, holder) with an alumina-based adhesive (Ceramabond 569, Kager Industrietechnik, Ditzenbach, Germany) to withstand high temperatures. It is possible to orient the particle at different angles with respect to the flow. Sampling is done with a fused silica capillary (OD 130 μm, ID 10 μm, Polymicro Technologies, Phoenix, USA) which goes through the drilled hole in the catalyst, horizontally to the flow. The capillary exits through the reactor wall to a port, which is connected to a positioning system with μm resolution to move the capillary through the catalyst particle. From the port, the sampled gas is transferred to a quadrupole mass spectrometer (QMS). Sampling inside the reactor is either possible with the tip of the capillary or a side orifice in the middle of the capillary. For the second technique, the other side of the capillary exits through the opposite reactor wall (shown in FIG. 1) and the tip of the capillary is sealed. The orifice has a diameter of 15 μm and is drilled with a Focused Ion Beam Microscope (FEI Strata 205, TSS Microscopy, Beaverton, USA). In this way, it is possible to measure the gas phase concentration in the reactor on both sides of the particle. In contrast, when using the tip of the capillary, only one side is measured, as it is not always possible to reinsert the capillary inside the particle without stopping the reaction and unmounting parts of the set-up.

[0271] Data Analysis

[0272] Data were acquired with a QMS with secondary electron multiplier (HAL 510, Hiden Analytical, Warrington, UK). Due to the low sampling volume, pressure in the QMS vacuum chamber did not exceed 1.5*10.sup.−7 mbar. To achieve a higher sensitivity, the samples were fed directly into the ionization head and the mid axis potential of the probe was increased, with the latter resulting in a loss of mass resolution. Therefore, m/z intervals around the strongest peak of every species (CO 28, CO.sub.2 44, O.sub.2 32, Ar 40) are determined and the intensity is integrated in this interval for evaluation. All species are calibrated by using the argon peak area as internal standard.

[0273] To account for the very sensitive response to pressure changes in the QMS, species were calibrated before an experiment. The peaks at m/z=28 and m/z=32 have a background signal originating from N.sub.2 (˜10% of the Ar peak) and O.sub.2 (˜10% of the Ar peak), which is not constant. No helium peaks were detected when testing all connections for leakages. The leaking rates might be minimal, but since aperture consists of several connectors, they add up in the background as a result, in combination with a very low sampling rate.

[0274] Model Equations

[0275] A model of the reactor is set up with Comsol Multiphysics® including the drilled particle and the sampling capillary. Standard CFD equations for conservation of mass, momentum and energy are used and can be found in the COMSOL Multiphysics® 5.4 reference manual. Flow in the reactor is laminar (Re.sub.p≈5.7). The particle is set as a solid phase, therefore convection inside is neglected. Furthermore, the gap between capillary and particle surface inside the particle is part of the fluid phase and it is assumed that the adhesive underneath the particle is not permeable. The reaction source terms in the solid phase use a Langmuir-Hinshelwood kinetic from Shisha and Kowalczyk [11]. Gas phase reactions can be neglected at the conditions in this study [12]. Mixture-averaged mass-based diffusion coefficients D.sub.im (equation 1) are calculated out of the binary diffusion coefficients D.sub.ij of all species for the fluid phase. D.sub.ij is calculated according to the Chapman-Enskog theory. In the solid phase, an effective diffusion coefficient D.sub.i,eff accounting for Knudsen diffusion, tortuosity τ, and porosity ∈ is calculated (equation 2) [13].

[00001]$\begin{array}{cc}\mathrm{fluid}{D}_{\mathrm{im}}=\frac{1-w_i}{M_n{.Math.}_{\underset{j\ne i}{i=1}}^n\frac{w_j}{M_j{D}_{\mathrm{ij}}}}& \left(1\right)\\ \mathrm{solid}{D}_{i,\mathrm{eff}}=\frac{1}{\left(D_{\mathrm{im}}^{-1}+D_{i,K}^{-1}\right)}\frac{ϵ}{\tau }& \left(2\right)\end{array}$

[0276] Density is calculated according to ideal gas law and thermodynamic data are calculated with the Shomate equation and parameters are taken from the NIST database [14]. The capillary is implemented as a non-permeable wall with no slip boundary condition on the surface. Only on a small circular area in the size of the ID of the capillary, an outlet volume flow is defined, to include the sampling.

[0277] This volume flow through is calculated with the Hagen-Poiseuille equation for gases (equation 3) [15]; where p.sub.1 represents the pressure inside the reactor, p.sub.2 that of the vacuum system, d the inner diameter of the capillary, and I the length of the latter. The resulting flow rate is 0.5 μl min.sup.−1. Since the flow regimen in the sampling system changes from viscous to molecular flow, the equation overestimates the sampling volume, to account for a worst-case scenario. Sa et al [20] measured a flow rate in a 75 μm ID capillary that was less than 10% of the calculated one with equation 3.

[00002]$\begin{array}{cc}{\stackrel{.}{V}}_{\mathrm{vis}}=\frac{\pi }{256}\frac{1}{\eta }\frac{d^4}{l}\frac{p_1^2-p_2^2}{p_1}& \left(3\right)\end{array}$

[0278] For CO oxidation, very steep temperature and concentration gradients at the surface can occur, and CO can be consumed within less than 100 μm [3]. Therefore, 10 prism layers of 10 μm are applied inside the particle at all surfaces. For a better resolution of the boundary layers in the gas phase, 15 prism layers with a layer thickness ratio of 1.2, starting at 15 μm, are implemented. Elements in the gap between capillary and particle are between 5 and 150 μm, resulting in 98000 elements; the particle consists of 340130 elements. Overall, 1.2 million elements are created.

Results and Discussion

[0279] Validation of the Set Up

[0280] The proposed set up aims to study the reactions taking place locally in a single catalyst particle; in doing so, accessibility of measurements inevitably disrupts original mass and heat transfer, through two ways:

[0281] The gap between the capillary and the particle, allowing the former to move;

[0282] The Gas Sampling Volume.

[0283] Simulations were performed to investigate the influence of these two disruptions, specifically on the CO oxidation. As for the gap between the capillary and the particle, FIG. 2 shows simulations of the pristine catalytic particle (left), and of the capillary with orifice going through the complete channel (right). The influence of the drilled channel over the CO.sub.2 concentration gradient is directly observable in the figure; this is induced by an increased diffusion of educts through the gap. The average diffusional flux per unit area is about ten times higher in the channel than in the particle. The average convective fluxes per unit area are one hundredth of the diffusive ones, and thus can be neglected. In other cases, this behavior is expected to change: with larger pores, and therefore, negligible Knudsen diffusion, the difference between diffusional fluxes in the particle and in the gap will decrease. Conversely, with higher Reynolds numbers, the boundary layer around the particle will decrease, and the convective fluxes in the gap might increase accordingly.

[0284] FIG. 3 shows the CO.sub.2 fraction through the particle along the top of the capillary in four different scenarios: a) no channel, b) capillary going through the whole channel without sampling, c) sampling with the tip of a 10 μm ID capillary, d) sampling with the side orifice with a 10 μm ID capillary, and e) sampling with the side orifice and a capillary of 40 μm ID. As already discussed before, the presence of the drilled channel leads to a lower CO.sub.2 fraction in the center (6.9% instead of 7.0%) of the particle. Moreover, the concentration drops closer to the center than when taking into consideration a pristine particle, due to diffusional fluxes. By enabling the sampling, the convective fluxes in the gap are increased, but in scenario d) the diffusive ones are still dominating. Therefore, the profile is not affected significantly; in the graph both profiles are nearly identical. In scenario e), the convective fluxes from the sampling become dominant and educts are sucked into the channel, resulting in a CO.sub.2 drop from 7.0% to 5.9%, due to dominating convective fluxes. During scenario c), the fluxes are higher on the open side of the channel and more educts enter from this side, resulting in lower CO.sub.2 concentrations and a shift of the maximum concentration to the side where the capillary enters the particle.

[0285] To reduce the invasiveness of the method, the fluxes inside the gap between capillary and the particle need to be reduced. For this a smaller diameter of the hole is necessary, however, the dimension of the capillary, the particle and the material of the catalyst are limiting the aspect ratio (ratio of hole depth to diameter) to about 25:1, with some methods resulting in unwanted conical holes. In case of a reduced flux inside the gap, more volume would be sucked out of the particle; hence, a reduced sampling flow might be needed. The kinetics of CO oxidation strongly depend on the preparation method of the catalyst [16]; furthermore, they are determined for large quantities of catalyst, where local, particle-level phenomena (e.g. Pt distribution) are averaged. Local changes of the catalyst will have a major influence on the concentration profiles of a single particle. Therefore, the simulations were performed before the experiments with a literature-based kinetic, to show the feasibility of the measurement method.

[0286] Profile Measurements—Tip of Capillary

[0287] In a first series of experiments, the profiles in the particle were measured for T.sub.in=173° C., 182° C., and 191° C. with 3.2% of oxygen at a Re.sub.p of 5.7. FIG. 4 shows the spatially resolved mole fraction of CO.sub.2 and a rerun of the experiment; the particle area is highlighted in grey. At an inlet temperature of 173° C., the CO.sub.2 concentration in the center (r=0 μm) reaches 7.0% and then drops symmetrically at both sides, until it reaches zero outside the particle (r<−2500 μm, r>2500 μm). The temperature inside the reactor increased by 2 K (measured 2 mm above the particle). At 182° C., the maximal CO.sub.2 fraction is only 0.2% higher and it decreases less sharply towards the inside the particle, with a temperature increase of 7 K. By increasing the temperature further, the catalyst ignites and changes its state from low conversion reaction-limited to high conversion diffusion-limited, in accordance with a significant temperature rise. At an inlet temperature of 191° C., the temperature inside the particle increased by 46 K. CO.sub.2 reaches a maximum of 6.6%, which plateaus, and then decreases towards the outside of the particle (r=2000 μm). In this case only the outer shell (˜500 μm) of the catalyst is used, which results in a low efficiency of the catalyst. Furthermore, a boundary layer built up around the particle; subsequently, 1.5 mm into the gas phase the CO.sub.2 fraction has not reached zero, hence the reaction is film-diffusion limited.

[0288] A low- and a high-active regime that changes within a few degrees is typical for CO oxidation. The low activity regime is caused by adsorbed CO on the platinum surface—which is poisoning the surface [17], [18]. In the high-active regime, the surface is nearly free of CO and is partially oxidized. This state is induced by low temperatures and high CO concentrations [18].

[0289] In order to demonstrate the reproducibility of our measurement and the validity of our set up, the measurement series has been repeated (FIG. 4b), after flushing the reactor only with argon. For all three experiments of the second measurement, CO.sub.2 fractions in the center are higher compared to the first run: nevertheless, the pathway of both curves is similar. The particle ignites at an inlet temperature of 195° C. instead of 191° C., but both experiments reach the same gas phase temperature of 246° C. inside the reactor. Hegedus et al. [19] could reproduce the ignition temperature within □□□° C., and observed a shift in ignition to higher temperatures for catalysts that have been used more than once.

[0290] By decreasing the inlet temperature again, the catalyst particle stays in the ignited state for temperatures even lower than the ignition temperature, until it extinguishes and falls on the lower branch. Heating up the particle again, it will stay on the lower branch until its ignition temperature is reached again, resulting in a hysteresis, which was observed also by previous authors [17], [19]. An example for the hysteresis due to temperature variations can be found in the supporting information [0301].

[0291] The same behavior can be observed by increasing the O.sub.2 concentration. FIG. 5 shows on the right-hand side the resulting hysteresis of the temperature measured inside the particle (T.sub.P), by increasing (black line) and decreasing (red line) the O.sub.2 concentration. On the left-hand side, the mole fractions for CO, O.sub.2 and CO.sub.2 are shown for the same conditions, but the upper one is in the ignited state and the lower one in the extinguished state. In the high-active state, a boundary layer of about 3.5 mm built up, resulting in a surface concentration of about 6% CO.sub.2, whereas in the low-active state the boundary layer is only one mm thick and the CO.sub.2 concentration on the surface is only 1%. The gradient over the film surrounding the particle is one of the reasons for the multiplicity, which is indicated by the criterium from Luss [20], since the conditions of the reaction are out of the region for uniqueness

[0292] (Table 1). The film diffusion limitations are not the sole responsible for the observed behavior. The CO poisoning of the Pt surface causes a negative reaction order. Therefore, for certain conditions, the reaction rate increases with decreasing CO concentration. This might result in effectiveness factors greater than one (they are defined as the ratio of reaction rate on the surface to the reaction rate inside the particle), and consequently in multiple steady states. The criterion from Luss [21]

[0293] (Table 1) for uniqueness indicates that for CO concentrations inside the particle that are less than half of the concentration on the surface, this will be the case. A look at FIG. 5 confirms this assumption for both shown profiles. Furthermore, effectiveness factors greater than one are possible in non-isothermal particles for very exothermic reactions, where the temperature inside the particle is increasing. However, in our case the Prater number is lower than 0.005 and therefore, temperature difference between catalyst surface and catalyst core is not significant. The Prater number multiplied by the surface temperature results in the maximal temperature difference between catalyst surface and catalyst core. Indeed, the criterion from Luss and Amundson [22] confirms that multiple steady states can be neglected because of this phenomenon.

TABLE-US-00001 TABLE 1 Uniqueness criteria for the three different phenomena causing multiplicity Phenomenon Criterion for uniqueness This study fulfilled Reference Non-isothermal βγ << 1 βγ = 0.025 ✓ [22] particle Kinetic [00003]$n\ge \left(n-1\right)\frac{c}{c^S}$ [00004]$n=-1\mathrm{and}\frac{c}{c^S}<0.5$ x [21], [23] Boundary layer [00005]$\mathrm{\beta \gamma }<4\left(\frac{{\mathrm{Bi}}_k}{{\mathrm{Bi}}_m}+\beta \right)$ [00006]$4\left(\frac{{\mathrm{Bi}}_k}{{\mathrm{Bi}}_m}+\beta \right)=0.012$ x [20], [23]

[0294] As already expected from the reaction order, by keeping the O.sub.2 concentration constant, the particle ignites with decreasing CO concentration. For all experiments, the ratio of CO:O.sub.2 for ignition was between 1.4 and 1.25. By decreasing this ratio further, the particle temperature starts to oscillate. As can be seen in the upper part of FIG. 6, by feeding only CO.sub.2 and argon into the reactor, the temperature fluctuates only by 0.1° C.—caused by the temperature control. In reaction regimes without oscillations, the temperature control has the same stability. With a CO:O.sub.2 ratio of 0.47, irregular temperature oscillations are measured, and by decreasing the ratio, further oscillations become regular with an amplitude of 2 K and a frequency of 2.5 min.sup.−1. Temperatures after changing the CO concentration and waiting for the steady state have been cut out of the graph for a better overview. The lower graph shows the corresponding measured intensity of the CO.sub.2 signal (m/z=44) corrected for background noise, measured 500 μm inside the particle. As can be seen by looking at the measurement without reaction, our measuring technique causes irregular oscillations in the signal. However, the signal behavior changes in regimes with oscillating temperature, showing a similar frequency as the latter. The absolute difference in the molar fraction of CO.sub.2 between the maximal and minimal intensity is about 0.1 mol-%. Even though it is expected that the slow diffusional processes around the particle (Knudsen diffusion, film diffusion) are reducing the amplitude, the main cause will be the sampling system. The equilibrium in the vacuum chamber for each sampling point is obtained after two minutes, and therefore maximal and minimal concentrations will not be measured.

[0295] CO oscillations have been observed on clean Pt structures under UHV conditions [24], but also on monoliths [25], single catalyst particles [17], [26] an Pt gauzes [17] under atmospheric pressure. Oscillations arise when the catalyst switches continuously between the high-active and the low-active state: that is not always possible, as can be seen in FIG. 5. With the ignition of the particle, the desorption of CO becomes so fast that it cannot fall back in the low-active state, where a high CO coverage is needed. On the other hand, by extinguishing the particle, the CO coverage will be so high that it cannot fall back in the state of fast desorption. Ertl et al. [24] determined a temperature of more than ˜530 K in case of the reaction getting stuck in the high-active regime, and a temperature of lower than ˜450 K for the opposite case, which is in agreement with our data. Therefore, oscillations can only occur when both the heat of reaction released in the high-active state increases the particle temperature by a few degrees; and at low CO concentrations (see FIG. 6).

[0296] Profile Measurements—Alternatives

[0297] To decrease the diffusive fluxes inside the drilled hole, and to avoid asymmetric boundaries around the particle by sampling with the tip, sampling with a side orifice on the capillary (3.2% O.sub.2, 5.4% CO, T.sub.in=170° C.) is tested. The mole fractions of CO.sub.2 and O.sub.2 can be found in red in FIG. 7. The blue profiles are measured under the same conditions, with the tip of a 5 μm capillary to reduce the invasiveness of the method. The sampling volume flow will be reduced to 1/16 compared to the 10 μm ID capillary, according to equation 3.

[0298] The CO.sub.2 profile measured with the side orifice is symmetric, whereas the profile measured with the tip looks slightly shifted: it is increasing slower from −2500 μm to 0 than from 0 to 2500 μm. This confirms the simulations from chapter [0279], where the CO.sub.2 is slightly shifted due to a higher diffusion of educts. Additionally, the measurement with the 5 μm capillary was performed a few days after the side orifice measurement, resulting in a reduced activity of the catalyst and explaining the lower CO.sub.2 ratios.

[0299] The O.sub.2 profiles of both measurements are fluctuating, particularly the one performed with the smaller capillary; this is caused by small changes in the O.sub.2 background signal. The measurement with the five μm capillary is even more sensitive to these, since the O.sub.2 background signal increased from 10% to 30% of the argon signal.

[0300] Furthermore, the O.sub.2 and CO calibration lines measured with the 5 μm-capillary result in a R.sup.2 of less than 0.99, which presents an issue in precise evaluation of these mole fractions. However, the CO.sub.2 fractions show that a reliable measurement is possible, but molecules on m/z=28 and m/z=32 should be avoided. Furthermore, the sampling time of every point has to be doubled, to reach a constant QMS signal.

[0301] Tests with a two μm capillary showed that a detection of CO.sub.2 is possible (R.sup.2=0.97), again increasing the sampling time, but the procedure will need to be optimized to get reliable results.

Concluding Remarks

[0302] An innovative method to elucidate the processes undergoing within a single catalyst particle has been introduced in this paper, allowing the measurement of spatially-resolved concentration profiles. Simulations showed that, due to diffusion through the drilled channel, product concentration profiles are slightly lower than in a pristine particle. Nonetheless, the sampling with a 10 μm ID capillary had a negligible effect. Importantly, experimental results could be reproduced, and show the potential of the method in determining mass transfer and reaction rate limitations, multiple steady states, as well as oscillations inside single catalyst pellets. Therefore, this work proved to be valuable in order to fully understand processes inside of a catalyst particle; in the future, the method developed within this paper will be employed to optimize catalyst properties based on the process' conditions.

Supporting Information

[0303] Ignition-Extinction

[0304] FIG. 8 is showing the hysteresis of the particle temperature for 3.2% O.sub.2 and 5.4% CO, in dependence of the inlet temperature. In this experiment, only the temperature of the particle and the temperature of the wall were measured; the latter was used for temperature control. The measurement was conducted at an early point of this study, where the optical access caused major heat losses. This explains why the particle temperature on the low temperature branch is lower than the inlet temperature.

[0305] Parameters

[0306] All parameters are calculated for the inlet conditions with x.sub.CO=5.4, x.sub.O2=3.2, and 170° C., according to chapter [0274] and the references listed in the table.

TABLE-US-00002 TABLE 2 Parameters used for calculations in Table 1 T [° C.] 170 V.sub.in [ml min.sup.−1] 450 x.sub.CO [%] 5.4 x.sub.O2 [%] 3.2 d.sub.p [m] 4.95E−03 D.sub.m [m.sup.2/s] 3.74E−05 D.sub.eff [m.sup.2/s] 1.42E−06 EA [J mol.sup.−1] 6.30E+04 [19] ΔH [J mol.sup.−1] 2.87E+05 [14] λ.sub.eff.sup.− [W m.sup.−1 K.sup.−1] 1.00 ρ.sub.f [kg m.sup.−3] 1.07 η [Pa s] 3.00E−05 c.sub.p [J kg.sup.−1 K.sup.−1] 552 Re u ρ.sub.f d.sub.v η.sup.−1 [—] 5.74 λ.sub.f [W m.sup.−1 K.sup.−1] 2.46E−02 Pr η c.sub.p λ.sub.f.sup.−1 [—] 0.673 Sc η ρ.sub.f.sup.−1 D.sub.m.sup.−1 [—] 0.750 Nu [00007]$\left(0.4\sqrt{\mathrm{Re}}+0.06\sqrt[\Xi ]{{\mathrm{Re}}^2}\right){\mathrm{Pr}}^{Q4}$ [—] 2.54 [27] Sh [00008]$0.664\sqrt{\mathrm{Re}}\sqrt[\Xi ]{\mathrm{Sc}}$ [—] 1.45 [28] h Nu λ.sub.f d.sub.v.sup.−1 [W/(m.sup.2 K)] 8.02 k Sh D.sub.m d.sub.v.sup.−1 [m/s] 6.95E−03 Bi.sub.h d.sub.v/2 h λ.sub.eff.sup.−1 [—] 0.023 [23] Bi.sub.m d.sub.v/2 k D.sub.eff.sup.−1 [—] 13.9 [23] Y −E.sub.A R.sup.−1 T.sup.−1 [—] 17.1 β −ΔH.sub.R D.sub.eff c.sub.CO.sup.S λ.sub.eff.sup.−1 T.sup.−1 [—] 1.47E−03 Symbols used Bi.sub.h d.sub.v h λ.sub.eff.sup.−1 2.sup.−1 [—] Biot number for heat transport Bi.sub.m d.sub.v k D.sub.eff.sup.−1 2.sup.−1[—] Biot number for mass transport c [m s.sup.−1] Thermal velocity c.sub.p [J kg.sup.−1 K.sup.−1] Heat capacity d [m] Capillary diameter D.sub.i,eff [m.sup.2 s.sup.−1] Effective diffusion coefficient D.sub.ij [m.sup.2 s.sup.−1] Binary diffusion coefficient D.sub.i,K [m.sup.2 s.sup.−1] Knudsen diffusion coefficient Mixture averaged, mass based, diffusion D.sub.im [m.sup.2 s.sup.−1] coefficients d.sub.v [m] equivalent spherical diameter h [W m.sup.2 K.sup.−1] Heat transfer coefficient l [—] QMS signal intensity k [m s.sup.−1] Mass transfer coefficient l [m] Length of capillary n [—] Reaction order n [mol] Amount of substance Nu d.sub.v h λ.sub.f.sup.−1 [—] Nusselt number p [Pa] Pressure Pr η c.sub.p λ.sub.f.sup.−1 Prandtl number T [K] Temperature T.sub.P [K] Particle temperature r [m] Radial position Sc η ρ.sub.f.sup.−1 D.sub.m.sup.−1[—] Sh d.sub.v k D.sub.m.sup.−1[—] Sherwood number t [s] Time {dot over (V)} [m.sup.3 s.sup.−1] Volume flow w [—] Mass fraction β −ΔH.sub.R D.sub.eff c.sub.CO.sup.S λ.sub.eff.sup.−1 T.sup.−1 [—] Prater number Y −E.sub.A R.sup.−1 T.sup.1 [—] Arrhenius number ε [—] Particle porosity η [kg m.sup.−1 s.sup.−1] Dynamic viscosity λ.sub.s [W m.sup.−1 K.sup.−1] Thermal conductivity ρ.sub.f [kg m.sup.−3] Fluid density T [—] Tortuosity Abbreviations ID Inner diameter IR Infrared OD Outer diameter QMS Quadrupole mass spectrometer

[0307] FIG. 9 schematically displays a single catalyst body 1. Single catalyst body 1 has a filled cylinder shape. Through single catalyst body 1 is passing sampling capillary 2. Sampling capillary 2 is hollow such that a fluid stream can be guided through the interior of sampling capillary 2. Sampling capillary 2 is received in a channel traversing single catalyst body 1. Sampling capillary 2 can be shifted back and forth relative to the single catalyst body as indicated by double arrow 3. A fluid stream 4 is entering single catalyst body 1 and is passing the porous body 1 such that a fluid stream is generated in the volume of single catalyst body 1.

[0308] FIG. 10 displays a sectional view of a single catalyst body 1 as shown in FIG. 9 with a sampling capillary 2 traversing single catalyst body 1. A small gap is formed between the outer surface 5 of sampling capillary 2 and side wall of channel 6 traversing single catalyst body 1. The gap is kept as small as possible to reduce fluid streams leaking into channel 6 through the distal ends of channel 6 but large enough to allow shifting of sampling capillary 2 relative to the single catalyst body.

[0309] FIG. 11 displays a detail marked by circle “A” in FIG. 10. Sampling capillary 2 comprises a sampling orifice 7 provided in the side wall of sampling capillary 2. Fluid phase passing the porous volume of catalyst particle 1 enters channel 6 provided in catalyst particle 1 through the side wall of channel 6. A small sample of the fluid phase can be collected through sampling orifice 7 to be guided through the hollow interior space of sampling capillary 2. The fluid sample is guided through sampling capillary 2 to an analytical instrument (not shown) to be analyzed qualitatively and/or quantitatively.

[0310] FIG. 12 shows a sectional view of a single catalyst body with a sampling capillary having a sampling orifice at a distal end. In such embodiment sampling capillary 2 is introduced into a channel 6, e.g. a blind hole, provided in single catalyst body 1. Sampling capillary 2 is accommodated in the side wall of channel 6 passing single catalyst particle 1.

[0311] FIG. 13 shows a detail marked in FIG. 12 by circle “B”. Fluid phase passing the porous phase of single catalyst body 1 enters channel 6 provided in single catalyst body 1 through the sidewalls of channel 6. The fluid phase, marked by arrows, then enters sampling capillary 2 through sampling orifice 8 arranged at a distal end of sampling capillary 2.

[0312] FIG. 14 shows a sectional view of a single catalyst body 1 with a sampling capillary 2 traversing the single catalyst body 1, wherein a fiber 9 is introduced into the sampling capillary 2. The tip of fiber 9 is arranged close to sampling orifice 7 provided in a side wall of sampling capillary 2.

[0313] Fiber 9 can be a thermocouple or a resistance thermometer for determination of a temperature inside the reactor chamber or a pyrometer fiber to collect and guide thermal radiation to a detector outside of the reactor for temperature measurement. Fluid phase entering the interior hollow space of sampling capillary 2 gets in close contact with fiber 9 and, therefore, temperature of the fluid phase as present at a place of sampling orifice 7 can be determined. In such embodiment, fiber 9 takes the form of a thermocouple or a resistance thermometer. When a pyrometer fiber is used as fiber 9 radiation emitted by the porous phase of single catalyst body 1 interacts with fiber 9 and the temperature of the solid porous phase of single catalyst body 1 can be determined.

[0314] According to another embodiment fiber 9 can be a fiber made of a material transmissive or transparent for electromagnetic radiation and that guides radiation from a radiation source (not shown) that is e.g. arranged outside the reactor chamber to a place inside the reactor chamber. The radiation enters the transparent fiber 9 on one end and exits the fiber 9 at the opposite end. The radiation is directed onto the surface of single catalyst body provided inside the reactor chamber to interact with the sample. In particular, the radiation exiting the fiber is directed onto the wall of channel 6 provided in the single catalyst body 1 for receiving the sampling capillary 2. After interaction with the material of the single catalyst body 1, radiation reflected, scattered or emitted by the single catalyst body 1 is collected and guided to a processing unit (not shown) where the collected radiation is processed to obtain e.g. a spectrum. For guiding the collected radiation reflected, scattered or emitted by the sample to the processing unit, a transparent or transmissive fiber can be used.

[0315] FIG. 15 shows a sectional view of a single catalyst body 1 with a sampling capillary 2 having a sampling orifice 8 at a distal end, wherein a fiber 9 is introduced into the sampling capillary 2. The tip of fiber 9 is arranged at sampling orifice 8 of sampling capillary 2. Fiber 9 can be used in the same manner as described for the embodiments displayed in FIG. 14.

[0316] FIG. 16 shows a sectional view of a single catalyst body 1 with a sampling capillary 2 having a sampling orifice 8 at a distal end, wherein a fiber 2 is introduced from an opposite end of a channel 6 provided in the single catalyst body 1. As described for the embodiments shown in FIGS. 14 and 15, fiber 9 can be used to determine a temperature of a fluid phase or of a solid phase of the single catalyst body 1 or can be used to obtain spectroscopic data by use of a transparent fiber and a suitable source of electromagnetic radiation.

REFERENCES

[0317] [1] W. Zhang, K. E. Thompson, A. H. Reed, and L. Beenken, “Relationship between packing structure and porosity in fixed beds of equilateral cylindrical particles,” Chem. Eng. Sci., vol. 61, no. 24, pp. 8060-8074, December 2006. [0318] [2] A. G. Dixon, “CFD study of effect of inclination angle on transport and reaction in hollow cylinder catalysts,” Chem. Eng. Res. Des., vol. 92, no. 7, pp. 1279-1295, July 2014. [0319] [3] G. D. Wehinger, F. Klippel, and M. Kraume, “Modeling pore processes for particle-resolved CFD simulations of catalytic fixed-bed reactors,” Comput. Chem. Eng., vol. 101, pp. 11-22, June 2017. [0320] [4] B. Partopour and A. G. Dixon, “An integrated workflow for resolved-particle packed bed models with complex particle shapes,” Powder Technol., vol. 322, pp. 258-272, December 2017. [0321] [5] R. Horn, N. J. Degenstein, K. A. Williams, and L. D. Schmidt, “Spatial and temporal profiles in millisecond partial oxidation processes,” Catal. Lett., vol. 110, no. 3-4, pp. 169-178, September 2006. [0322] [6] O. Korup et al., “Catalytic partial oxidation of methane on platinum investigated by spatial reactor profiles, spatially resolved spectroscopy, and microkinetic modeling,” J. Catal., vol. 297, pp. 1-16, January 2013. [0323] [7] F. Küster et al., “In-situ investigation of single particle gasification in a defined gas flow applying TGA with optical measurements,” Fuel, vol. 194, pp. 544-556, April 2017. [0324] [8] L. L. Hegedus and E. E. Petersen, “An Improved Single-Pellet Reactor to Study the Interaction of Kinetics with Mass Transfer Effects in Heterogeneous Catalysis,” Ind. Eng. Chem. Fundam., vol. 11, no. 4, pp. 579-584, November 1972. [0325] [9] C. Chmelik et al., “One-Shot Measurement of Effectiveness Factors of Chemical Conversion in Porous Catalysts,” ChemCatChem, vol. 10, no. 24, pp. 5602-5609, 2018. [0326] [10] R. H. Venderbosch, W. Prins, and W. P. M. van Swaaij, “Platinum catalyzed oxidation of carbon monoxide as a model reaction in mass transfer measurements,” Chem. Eng. Sci., vol. 53, no. 19, pp. 3355-3366, October 1998. [0327] [11] R. C. Shishu and L. S. Kowalczyk, “The Oxidation of Carbon Monoxide on Supported Platinum,” Platin. Met. Rev., vol. 18, no. 2, pp. 58-64, April 1974. [0328] [12] M. Hettel, C. Antinori, and O. Deutschmann, “CFD Evaluation of In Situ Probe Techniques for Catalytic Honeycomb Monoliths,” Emiss. Control Sci. Technol., vol. 2, no. 4, pp. 188-203, October 2016. [0329] [13] J. Solsvik and H. A. Jakobsen, “Modeling of multicomponent mass diffusion in porous spherical pellets: Application to steam methane reforming and methanol synthesis,” Chem. Eng. Sci., vol. 66, no. 9, pp. 1986-2000, May 2011. [0330] [14] P. Linstrom, “NIST Chemistry WebBook, NIST Standard Reference Database 69.” National Institute of Standards and Technology, 1997. [0331] [15] K. Jousten, Ed., Wutz Handbuch Vakuumtechnik, 11th ed. Vieweg+Teubner Verlag, 2012. [0332] [16] R. H. Venderbosch, W. Prins, and W. P. M. van Swaaij, “Platinum catalyzed oxidation of carbon monoxide as a model reaction in mass transfer measurements,” Chem. Eng. Sci., vol. 53, no. 19, pp. 3355-3366, October 1998. [0333] [17] H. Beusch, P. Fieguth, and E. Wicke, “Thermisch and kinetisch verursachte lnstabilitäten im Reaktionsverhalten einzelner Katalysatorkorner,” Chem. Ing. Tech., vol. 44, no. 7, pp. 445-451, April 1972. [0334] [18] J. Singh, M. Tromp, O. V. Safonova, P. Glatzel, and J. A. van Bokhoven, “In situ XAS with high-energy resolution: The changing structure of platinum during the oxidation of carbon monoxide,” Catal. Today, vol. 145, no. 3, pp. 300-306, July 2009. [0335] [19] L. L. Hegedus, S. H. Oh, and K. Baron, “Multiple steady states in an isothermal, integral reactor: The catalytic oxidation of carbon monoxide over platinum-alumina,” AIChE J., vol. 23, no. 5, pp. 632-642, September 1977. [0336] [20] D. Luss, “Luss, D, Proc. 4th Int. Symp. Chem. React. Eng., Heidelberg, 1976,” 1976. [0337] [21] D. Luss, “Sufficient conditions for uniqueness of the steady state solutions in distributed parameter systems,” Chem. Eng. Sci., vol. 23, no. 10, pp. 1249-1255, October 1968. [0338] [22] D. Luss and N. R. Amundson, “Uniqueness of the steady state solutions for chemical reactor occurring in a catalyst particle or in a tubular reaction with axial diffusion,” Chem. Eng. Sci., vol. 22, no. 3, pp. 253-266, March 1967. [0339] [23] G. F. Froment, Chemical reactor analysis and design, 3. ed. Hoboken: Wiley, VTM-305, 2011. [0340] [24] G. Ertl, P. R. Norton, and J. Rustig, “Kinetic Oscillations in the Platinum-Catalyzed Oxidation of Co,” Phys. Rev. Lett., vol. 49, no. 2, pp. 177-180, July 1982. [0341] [25] J. Sá et al., “SpaciMS: spatial and temporal operando resolution of reactions within catalytic monoliths,” Analyst, vol. 135, no. 9, pp. 2260-2272, 2010. [0342] [26] J. (J H. I. of P. C. and E. Kapicka and M. (Prague I. of C. T. Marek, “Oscillations on individual catalytic pellets in a packed bed: CO oxidation on Pt/Al sub 2 O sub 3,” J. Catal. USA, vol. 119:2, October 1989. [0343] [27] W. Wagner, Wärmeübertragung, 7., überarb. und erw. Aufl. Vogel Buchverlag, 2011. [0344] [28] D. S. Christen, Praxiswissen der chemischen Verfahrenstechnik: Handbuch für Chemiker und Verfahrensingenieure, 2., bearbeitete und ergänzte Auflage. Heidelberg: Springer, 2010.