Kinetic Modeling of C3H6 Inhibition on NO Oxidation over Pt Catalyst

Exhaust after treatment for lean burn and diesel engine is a complex catalytic system that consists of a number of catalytic units. Pt/Al2O3 is often used as a model Diesel Oxidation Catalyst (DOC) that plays an important role to facilitate oxidation of NO to NO2. In the present study, we proposed a detailed kinetic model of NO oxidation as well as low temperature C3H6 inhibition to simulate temperatureprogrammed reaction (TPR) data for NO oxidation over Pt/Al2O3. A steady-state microkinetic model based on Langmuir-Hinshelwood mechanism for NO oxidation was proposed. In addition, low temperature C3H6 inhibition was proposed as a result of site blocking as well as surface nitrite consumption. The model can explain the experimental data well over the studied temperature range. Copyright © 2016 BCREC GROUP. All rights reserved


Introduction
The control of emission from gasoline lean burn and diesel vehicles is a complex chemical reaction engineering process.With the implementation of stringent emission standards in several countries such as the EURO 6 standard in Europe, improvement of exhaust after treatment system is needed.The lean exhaust after treatment technology is primarily aimed at eliminating the presence of CO, hydrocarbon (HC), particulate matters (PM) and oxides of nitrogen (NOx) in an exhaust gas outlet.For this purpose, lean exhaust after treatment consists of several units namely diesel oxidation catalyst (DOC), diesel particulate filter (DPF) and selective catalytic reduction (SCR) unit.
DOC is placed as the first catalytic unit with oxidative functionality by facilitating oxidation of HC, CO and NO.It consists of precious PGM components such as Pt and Pd which are dispersed in a high surface area washcoat support such as Al2O3.Oxidation of NO to NO2 in the DOC plays an important role since NO2 is beneficial for DPF regeneration and some SCR catalysts.In lean exhaust aftertreatment, NO oxidation reaction over Pt catalyst has been widely studied and it shows some unique char-acteristics.For example it is restricted by thermodynamics at high temperature, it possess hysteresis phenomenon during heating and cooling, it is structure sensitive and the reaction is inhibited by HC and CO at low temperature [1].Recently, it has also been reported that addition of H2 may boost the NO oxidation reaction over Pt catalysts [2,3].For this reason, kinetic studies of NO oxidation over Pt/Al2O3 have gained a lot of interests from a number of research groups.Mechanistic studies through microkinetic modeling of NO oxidation over Pt/Al2O3 have been widely available in literature for instance in [4][5][6][7][8].
Experimental studies of HC inhibition on NO oxidation has been reported in literature by using a number of HC compounds such as C3H6 [9,10], C3H8 [11], Dodecane [11,12] and mxylene [12].Among them, C3H6 is often used as the model HC compounds for DOC studies.The mechanism of C3H6 inhibition by competitive adsorption on the active sites [10] as well as C3H6 reaction with NO2 to NO [9,12] have been proposed in literature.
With a full DOC feed mixture, a number of global kinetic models are present in the literature based on the common Voltz model [13] obtained from lab tests [14,15] to the full-scale engine rig test [16].Recently, Khosravi et al. [14,15] reported a global kinetic model for the NO oxidation reaction in the presence of C3H6 over Pt and bimetallic Pt:Pd catalysts which includes C3H6-SCR, NO oxidation, C3H6 oxidation and N2O formation.In addition, the model also accounts for inhibition terms for C3H6 blocking as well as C3H6 reaction with NO2 to NO.They exercised the C3H6 influence on NO oxidation by coupling the rate of NO oxidation with C3H6 oxidation and inserting C3H6 inhibition term in the denominator part of their global rate expression.To gain a deeper understanding on the reaction mechanism, some microkinetic models have also been reported in literature by including an extensive number of elementary steps to simulate a full mixture of DOC feed including the influence of HC e.g.Koop and Deutschmann [17] and Sharma and Mhadeshwar [18].
In the present study, a kinetic model to describe the mechanism of low temperature C3H6 inhibition on NO oxidation over Pt/Al2O3 was constructed.The model here was aimed to represent mechanistic features of C3H6 inhibition however with relatively few reaction steps.It is hoped that the model here may provide useful insights to describe a larger mechanism especially H2 effect on NO oxidation over Pt catalyst which is part of our on going study as previ-ously reported in [3].

Experimental Method
The detail of catalyst preparation and flow reactor experiments were essentially the same as described in [3].Further, the catalyst sample had also been used in earlier study presented in [10] to ensure stable activities under the present work.Briefly repeated here, the catalyst powder of 1 wt.%Pt/Al2O3 was prepared by wet impregnation method.The powder was wash coated onto a cordierite monolith with the dimensions of 2 cm width and 2 cm length.The total catalyst deposited on the monolith was ca 0.5 g.For modeling purposes, Pt dispersion of 2% was used according to the CO chemisorption test in [10].The activity measurements were conducted in a heated horizontal flow reactor.The inlet gases to the reactor were supplied from a number of mass flow controllers (Bronkhorst).Gas FTIR analyzer (MKS 2000) was used to continuously measure the outlet gases from the reactor to measure NO, NO2, C3H6 in ppm level as well as H2O in percentage level.Pretreatment steps were conducted in O2 and H2 flow at 450 °C, respectively.
Temperature-programmed reaction (TPR) modes were used to measure the activity of NO oxidation in the presence and absence of C3H6 using a heating rate of 5 °C/min.The standard inlet gases were 500 ppm NO + 8% O2 + 0 or 200 ppm C3H6 + 5% H2O in Ar balance with the total flow of 3000 mL/min (corresponds to Weight Hourly Space Velocity, WHSV, of 562 h -1 or equivalent to Gas Hourly Space Velocity, GHSV, of ca.29000 h -1 ).The TPR of NO oxidation in the absence of C3H6 in this paper is designated as the base case.

Modeling Method
The reactor is modeled as a fixed bed reactor and discretized into a number of tanks in axial direction (tanks-in-series) as displayed in Figure 1.Here, the washcoat is divided into 8 segments (tanks) based on catalyst mass where an expansion factor of 10% for each tank is used to obtain a finer discretization in the front part of monolith.To achieve reasonable computational time, we simulated 14 temperature points as follows: 120, 150, 175, 200, 210, 220, 230, 240, 250, 260, 275, 300, 325 and 350 °C.Furthermore, some assumptions were made in the present model: steady state solution was reached for each temperature; the mass transfer from gas bulk flow to the washcoat and diffusion inside washcoat were neglected; it is under ki-
In each tank k (see Figure 1), the model solved the mole balances for each species i as described in Equation ( 1).
(1) Surface components balance for each surface species w in each tank k was described by transient equation shown in Equation (2).When steady state condition is achieved, the derivative term in Equation ( 2) turns to zero.
(2) Eventually, the reaction rate for each reaction step j is described according to Equations ( 3) and ( 4).All simulations were conducted with a standard personal computer and MATLAB software to solve equations ( 1)-( 4) using ODE 15s.For parameter estimation, lsqnonlin function was utilized with the objective function to minimize the SSE from residuals of NO2 and C3H6.

Rate expression and kinetic parameters
Table 1 shows the reaction mechanism used in the present model to describe NO oxidation reaction in the absence and presence of C3H6.As a starting point, a microkinetic model of NO oxidation based on Langmuir-Hinshelwood mechanism was used to fit the base case of NO oxidation data to yield kinetic parameters for steps 1 to 8. It should be mentioned that we do not include any coverage dependence on our rate expression to correct the activation energies (Table 1) as often presented in literature (e.g.[5,19]).
In order to decrease the number of free parameters during the estimation procedure, some parameters were determined from literature sources.Estimation of preexponential factors and activation energies for adsorption of NO, O2 and NO2 (step 1,3 and 5) were extracted from literature values of Bhatia et al. [6].Similarly, the preexponential factor and activation energy for surface reaction between NO * and O * in step 7 was also fixed to the value of Bhatia et al. [6].The kinetic parameters for step 8 were estimated from thermodynamics of overall NO oxidation reaction.In addition, preexponential factors for desorption of NO (step 2), O2 (step 4) and NO2 (step 6) were fixed at 1013 s -1 .
The influence of C3H6 on NO oxidation reaction is shown in step 9 to 12. To keep the model to be relatively simple and focused on the C3H6 inhibition effect on NO oxidation, other reac-

Temperature-programmed reaction of NO oxidation (base case)
Comparison between simulation and experimental TPR data of NO oxidation base case is presented in Figure 2. The activity measurement here only present the data up to 350 °C as NO oxidation reaction is commonly re-stricted by thermodynamics beyond this temperature as shown by the equilibrium line in Figure 2. As shown from the top panel of Figure 2, an excellent agreement was achieved through simulation of 14 experimental data within the studied temperature interval.
From the lower panel of Figure 2, the variation of surface species coverages as a function of temperature is presented.As seen here, NO * is the predominant species at low temperature as in agreement with the model of Bhatia et al. [6].By increasing temperature, it is apparent that NO oxidation reaction progressed more rapidly which gave higher coverage of NO2 * .Eventually, further increase in temperature caused the increase in O * coverage which is conceivable as the activation energy for O * is the highest among all steps (reaction 4).In addition, high coverage of O * at high temperature as presented here confirmed the presence of Pt-Oxide formation which is known over Pt catalyst.Pt oxide formation itself is known to cause deactivation and hysteresis effect on NO oxidation reaction as reported in [6,[20][21][22].

Modeling C3H6 inhibition on NO oxidation
The inhibition effect from the presence of minute concentrations of C3H6 on NO oxidation has been widely reported in literature as confirmed by the results presented in Figure 3.
Strong inhibition of C3H6 at low temperature is generally observed up to a temperature point where full conversion of C3H6 is achieved (at ca.220 °C).
Comparison between simulation and experimental data is also shown in Figure 3.The result from steady state modeling from 14 experimental data gave an excellent agreement with experimental data.With the aid of the lower panel in Figure 3, mean surface coverages of various species is presented as a function of temperature.As expected at low temperature, the predominant surface species is C3H6 * that blocks the active sites.As the temperature increases, the mean coverage of C3H6 * decreases and O * appears as the dominant species on the catalyst.The increase of NO2 outlet concentration (upper panel Figure 3) appears to be in parallel with the increase of O * which suggested that O * plays crucial role in the NO oxidation reaction.Comparing the NO2 yield above 220 °C where all C3H6 has been fully converted, it can be seen that the NO2 yield in the presence of C3H6 (Figure 3) is lower than the NO2 yield from the base case (Figure 2).Hence, it is likely that C3H6 plays a role to decrease the NO2 yield.For this reason, the present model includes reaction 12 to illustrate the role of C3H6 * to lower the NO2 yield by reducing NO2 * (surface nitrite) and it turns out that kinetic parameters for this reaction were sensi-tive to influence the profile of C3H6 and NO2.
Further evaluation can also be made to analyze the axial profile of NO2 gas concentration inside the monolith washcoat as shown in Figure 4. Below the light-off temperature for NO oxidation, the coverage of C3H6 * suppresses the NO2 yield in the entire monolith.When the temperature increases, the reaction front for NO oxidation initiated from the rear part of monolith.As the temperature increases, production of NO2 propagates to-   wards the front part of monolith.As a result, a larger part of the washcoat is utilized for production of NO2 in parallel with the increase of temperature.

Conclusions
An inhibition effect due to addition of C3H6 has a substantial influence on NO oxidation reaction over Pt/Al2O3 catalyst.A detailed kinetic model for NO oxidation (base case) has been constructed based on Langmuir-Hinshelwood mechanism using TPR data.The NO oxidation model was then extended to include inhibition effects due to addition of C3H6 on NO oxidation.Two mechanisms of C3H6 inhibition were included in the model: site blocking as well as surface nitrite consumption.The simulations show that the proposed model based on 7 fitted kinetic parameters can capture the experimental data well over the studied temperature range for both the base case and NO oxidation in the presence of C3H6.

Figure 1 .
Figure 1.Reactor model used in the present study where the monolith sample was simulated with tanks in series reactor model (1D model).

Figure 3 .
Figure 3.Comparison between simulation and experimental results from TPR of NO oxidation in the presence of C3H6 over 1 wt.%Pt/Al2O3 catalyst: gas concentrations (top panel) and surface species coverages (bottom panel).The inlet feed conditions: 500 ppm NO, 200 ppm C3H6, 8% O2, 5% H2O with Ar balance.

Figure 4 .
Figure 4. Model prediction for axial distribution of NO2 in the monolith Pt/Al2O3 catalyst for various temperatures.The inlet feed condition is the same as in Figure 3.

Table 2 .
The list of kinetic parameters used in the present model