Detailed Chemical Kinetic Modeling of Methylamine in Supercritical

We constructed a detailed chemical kinetics model (DCKM) for the oxidation and pyrolysis of methylamine in supercritical water (400-500°C). The model...
1 downloads 0 Views 143KB Size
Ind. Eng. Chem. Res. 2005, 44, 9785-9793

9785

KINETICS, CATALYSIS, AND REACTION ENGINEERING Detailed Chemical Kinetic Modeling of Methylamine in Supercritical Water Kenneth M. Benjamin† and Phillip E. Savage* Department of Chemical Engineering, University of Michigan, Ann Arbor, Michigan 48109-2136

We constructed a detailed chemical kinetics model (DCKM) for the oxidation and pyrolysis of methylamine in supercritical water (400-500°C). The model contains 72 species and 603 elementary free-radical and molecular reactions. This DCKM for methylamine oxidation is the first to include peroxy radical chemistry, which becomes important at these lower temperatures. The major products from oxidation in supercritical water are predicted to be formamide, ammonia, CO2, and formic acid. The major products predicted for methylamine hydrothermolysis are ammonia, HCN, methane, water, CO2, and H2. The activation energies are predicted to be 47 and 51 kcal/mol for oxidation and thermolysis, respectively, in supercritical water. The main route for methylamine removal during oxidation is OH attack. The main route to ammonia is hydrolysis of formamide. Predictions from the DCKM, which is built upon homogeneous chemistry, are not consistent with published experimental results for either methylamine pyrolysis or oxidation in supercritical water. This inconsistency is due primarily to heterogeneous catalytic reactions occurring in the experimental system, but also to uncertainties in the homogeneous mechanism and its kinetics parameters. Reaction pathway and sensitivity analyses indicate that several of the most important elementary steps are ones with estimated kinetic parameters, due to a lack of information in the combustion literature. These modeling results highlight the need for an improved understanding of the mechanism and kinetics for organonitrogen chemistry during gas-phase oxidation in this low-temperature region. Additionally, they point to the need for experimental data for methylamine reactivity in supercritical water in the absence of heterogeneous reactions. Introduction Supercritical water oxidation (SCWO) is a process that oxidizes organic compounds in an aqueous environment above the critical point of water (Tc ) 374 °C, Pc ) 218 atm). Organic compounds and permanent gases (such as oxygen) are completely miscible in supercritical water, so the oxidation occurs in a single-phase, aqueous environment. The high temperature gives fast reaction kinetics and a high selectivity to complete oxidation products.1 For incineration of nitrogen-containing organic compounds, byproducts can include NO and NO2. At the lower SCWO temperatures (compared to incineration), however, thermodynamics favors nitrogen conversion into N2 and N2O, rather than NO and NO2.2 In addition, whatever NOx is formed will remain in the supercritical water, rather than being emitted to the atmosphere. In this manner, SCWO is self-scrubbing and is considered by the EPA to be a totally enclosed treatment method.3 These features make SCWO an attractive alternative to technologies such as wet-air oxidation or incineration, which suffer from masstransfer limitations and hazardous byproduct formation from incomplete combustion, respectively. * To whom correspondence should be addressed. Tel.: (734) 764-3386. Fax: (734) 763-0459. E-mail: [email protected]. † Present address: Department of Chemical and Biological Engineering, The State University of New York at Buffalo, Buffalo, NY 14260-4200.

The creation of mechanism-based detailed chemical kinetics models (DCKMs) for the SCWO of individual compounds has been an active area of research (see, e.g., refs 4-8). Such models are appealing because they not only provide insight into the governing chemistry, but also provide the firmest chemical foundation for making predictions. The working assumption has been that oxidation chemistry in supercritical water is comparable to gas-phase combustion chemistry in the same temperature region. Thus, one can use mechanisms and kinetics of elementary steps in the combustion literature as a foundation for a mechanism-based model for SCWO. Of course, one must account for the higher pressure and water density in SCWO, as described in detail elsewhere.9 To date, the focus in this field has been primarily on simple molecules (such as hydrogen, carbon monoxide, methane, and methanol) and aromatic compounds (such as phenol and benzene), which contain only C, H, and O atoms.4-20 A number of models have successfully predicted various features of SCWO behavior for these compounds. This previous work and success provide a solid background and foundation for studying other oxidation chemistries, such as that of nitrogen-containing compounds, in supercritical water. Nitrogen-containing compounds are present in energetic materials, wastes from chemical processes, and biological wastes, all of which have been treated by SCWO. This article

10.1021/ie050926l CCC: $30.25 © 2005 American Chemical Society Published on Web 11/02/2005

9786

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005

presents the first detailed chemical kinetics model (DCKM) for the oxidation and pyrolysis of a nitrogencontaining organic compound, methylamine, in supercritical water. Background The combustion literature provides a few mechanistic models for methylamine oxidation and pyrolysis,21-26 but typically for temperatures much higher than those of interest in SCWO. Nevertheless, these studies are a useful starting point for developing a DCKM for SCWO. They provide many of the reaction steps, kinetics data, and thermochemical data that might be needed in an SCWO model of methylamine. Basevich21 presented a detailed chemical kinetics model for H-C-O-N chemistry and applied it to methylamine combustion at 1160 K in the gas phase. The mechanism included only nine elementary reactions that involved methylamine and its associated radicals (CH3NH, CH2NH2, etc.). The author acknowledged that such a simple mechanism cannot capture the complete, complex chemistry and should be regarded as a “first approximation”. Gardiner and co-workers22-24 performed shock tube and modeling studies of both methylamine pyrolysis and oxidation in the gas phase. Their reaction mechanism comprised 141 elementary steps, and it was based on the validated Miller and Bowman27 mechanism for C-H-N-O chemistry. Kinetics parameters for about 10 of the reactions were adjusted to provide better agreement between model calculations and experimental results. Although this work represented the first full model for methylamine oxidation, the temperature range studied (1260-1600 K) is much higher than that of SCWO (about 700-800 K). It is likely that the reaction mechanism and pathways will be different at the lower temperatures and that additional steps will become important for SCWO conditions. Williams and Fleming25 investigated methylamine as a dopant in low-pressure methane flames in the temperature range 600-2200 K. They used the mechanism of Hwang et al.,23 along with some additional reactions. Modifying the kinetics of several reactions in the Hwang mechanism improved the “predicted” amounts of NH2, an important radical in nitrogen oxidation chemistry. The most extensive effort to date to model methylamine oxidation is the work of Kantak et al.26 They present a reaction mechanism comprising 350 elementary reactions and 65 species. The five submechanisms in the overall mechanism are (1) the C1 mechanism containing 125 reactions, (2) 53 H-C-O-N reactions from Miller and Bowman,27 (3) the 113-reaction HCN oxidation mechanism from Glarborg and Miller,28 (4) 45 reactions involving CH3NH2 proposed by the authors, and (5) 14 reactions taken from other literature sources. Of the 14 reactions in this last submechanism, some came from the mechanism of Hwang et al.23 Only five of the 10 reactions with kinetics adjusted by Hwang et al. to fit their data are left adjusted by Kantak et al., however. The model was used to predict previously published results for methylamine oxidation in a flow reactor at 1160 K at subatmospheric pressure. The model did an excellent job of predicting methylamine disappearance and a fair job of predicting the evolution of major nitrogen- and carbon-containing products. This background section shows that there have been only a small number of modeling and experimental

studies on methylamine oxidation in the gas phase and none in the temperature range of interest for SCWO. As a result, the kinetics and mechanism are not entirely certain. To date, every model proposed for methylamine oxidation adjusted the kinetics of some steps to resolve differences with experimental results. Also, few sensitivity analyses have been done to identify important reactions and species. Additionally, as mentioned above, the temperature range wherein DCKMs for methylamine oxidation have been validated is much higher than that of SCWO. A different set of reaction pathways might become important at the lower temperatures of interest in SCW. Additional differences in the pathways between gas-phase oxidation and SCWO can be anticipated because of the higher pressure and abundance of water in SCWO. These factors combine to make constructing a DCKM for methylamine SCWO applications a challenging endeavor. Model Development This section describes the approach and sources used to construct the detailed chemical kinetics model for methylamine chemistry in supercritical water. The mechanism is based on the SCWO and combustion literature. Brock29 constructed and validated a DCKM with 22 species and 150 elementary, reversible, freeradical reactions to describe the SCWO of C1 compounds. We used all of Brock’s steps and their associated kinetics except for CH3OH + OH ) CH2OH + H2O because an experimental rate constant30 is now available for this reaction in SCW. To describe the nitrogen chemistry, we took elementary steps and/or kinetics from the combustion literature. The components used in developing the model for nitrogen chemistry in SCWO are (1) the nitrogen submodel of Kantak et al.;26 (2) additional methylamine reactions;23,25,31 and (3) additional reactions for ammonia,32-34 hydrogen cyanide,28,35 and general fuel nitrogen chemistry.36-43 For some reactions, the kinetics recommended by Dean and Bozzelli37 were adopted rather than the kinetics given in the references above. We estimated44 high-pressure rate constants when these were not available in the literature for unimolecular reactions and certain bimolecular reactions that have pressuredependent kinetics. In addition to the steps noted above, this model for methylamine SCWO also included low-temperature methylamine oxidation reactions based on peroxy radical chemistry.45 The present model is the first DCKM to include these for methylamine oxidation, and we believe they are essential at the lower temperatures employed in SCWO. Peroxy radicals have been shown to be important for both methane7,11 and benzene17,18 SCWO. The kinetics for these steps in the methylamine model were estimated by analogy with ethane oxidation46 because no better data are available in the literature. Including peroxy radical chemistry in the SCWO DCKM admits the possibility of forming formamide and, by its hydrolysis, formic acid. Therefore, the DCKM needed to include reactions of these compounds. Our model includes formamide thermal decomposition47 and hydrolysis. Kinetics for hydrolysis were taken to be the same as the experimental kinetics reported for acetamide hydrolysis in SCW.48 We also used experimental rate constants for formic acid decomposition49 and oxidation (by analogy with acetic acid) in SCW.50-52

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005 9787 Table 1. Polynomial Coefficients for the Heat Capacity Equation (Eq 2) coefficient

HOCN

HNCO

HCCO

HCNO

HCNN

a1 a2 a3 a4 a5 a6 a7

3.786 6.887 × 10-3 -3.215 × 10-6 5.172 × 10-10 1.194 × 10-14 -2.827 × 103 5.633

3.631 7.303 × 10-3 -2.281 × 10-6 -6.613 × 10-10 3.622 × 10-13 -1.559 × 104 6.195

2.252 1.766 × 10-2 -2.373 × 10-5 1.728 × 10-8 -5.066 × 10-12 2.006 × 104 1.249 × 101

2.647 1.275 × 10-2 -1.048 × 10-5 4.414 × 10-9 -7.575 × 10-13 1.930 × 104 1.073 × 101

2.524 1.596 × 10-2 -1.882 × 10-5 1.213 × 10-8 -3.236 × 10-12 5.426 × 10+4 1.168 × 10+1

Note that none of the previous models for methylamine oxidation included formamide formation and hydrolysis, and none included formic acid formation and decomposition. Because they lack these species and steps that can be important under SCWO conditions, they cannot be expected to give accurate predictions of the kinetics or product distribution during SCWO. To adapt gas-phase oxidation kinetics to the highpressure, high-water-density environment of SCW,6 we incorporated falloff behavior and broadening for pressure-dependent (unimolecular and chemically activated) reactions and used collision efficiencies for water because it was the dominant collision partner in reactions. The Supporting Information available on the Internet provides a table that lists each step in the present DCKM, its associated kinetics, and the source of the kinetics parameters. The model contains 72 species and 603 elementary, homogeneous reactions. With regard to thermochemistry, we calculated equilibrium ratios of partial pressures (to determine reverse reaction rates) according to the equation

(

KP ) exp

)

∆Srxn ∆Hrxn R RT

(1)

The values of ∆Srxn and ∆Hrxn are calculated from the standard entropies and heats of formation, along with the heat capacities of each reactant and product in a given chemical reaction. The constant-pressure heat capacity for each species is fit to the polynomial

Cpi0 R

N

)



aniT(n-1)

(2)

n)1

For most of the molecules and radicals in the mechanism, heat capacity data existed, and the coefficients had been determined already.37,53,54 For some of the species (HOCN, HNCO, HCCO, HCNO, and HCNN), however, there were no experimental heat capacity data. We used THERM55 and the NIST Structure and Properties56 computer program to estimate heat capacities and standard-state enthalpies and entropies via a group-additivity approach. Table 1 lists the coefficients in eq 2 for these species. Model Predictions The DCKM was used to predict the outcome of methylamine oxidation and pyrolysis in supercritical water and to elucidate the underlying chemistry. We used sensitivity analysis to identify important reactions in the mechanism. Normalized sensitivity coefficients were calculated according to the expression

Sij )

∂ ln xi ∂ ln kj

(3)

Figure 1. Product concentrations from methylamine SCWO at 410 °C as predicted by our DCKM. Conditions: 410 °C, 249 atm, [CH3NH2]0)3.31 mmol/L, [O2]0)17.9 mmol/L, and [H2O]0)8270 mmol/L. (a) Major products. (b) Minor products.

where Sij is the normalized sensitivity coefficient for species i and reaction j, xi is the mole fraction of species i, and kj is the rate constant of reaction j. The value of Sij indicates the relative change in the mole fraction of species i that would result from some relative change in the rate constant for reaction j. The larger the absolute value of the sensitivity coefficient, the more sensitive the calculated concentration of species i is to the kinetics of step j at the particular conditions being examined. In addition to a sensitivity analysis, we also performed a net reaction rate analysis. The forward, reverse, and net rates for each elementary reaction were calculated and compared. This procedure identifies the reactions that proceed with the highest rates and therefore carry most of the reaction traffic. A reaction rate analysis also reveals the net direction (forward or reverse) in which each elementary reaction proceeds. From these calculations, one is able to elucidate the molecular connectivity of the reaction pathways, as well as the instantaneous rate of each step.

9788

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005

Figure 2. Major reaction paths during methylamine SCWO. Conditions: 410 °C, 249 atm, t ) 20.1 s, [CH3NH2]0 ) 3.31 mmol/L, [O2]0 ) 17.9 mmol/L, [H2O]0 ) 8270 mmol/L, and conversion ) 0.33. Thickness of arrows indicates relative rate of each step. Numbers next to the arrows are the relative net reaction rates for each step.

All of the DCKM calculations were carried out using the CHEMKIN II software package53,57 for gas-phase chemical kinetics. In this section, we present model predictions and, in some cases, compare them to experimental data.45,58,59 Methylamine Oxidation in Supercritical Water. Figure 1a shows the concentrations of major products predicted by the DCKM for SCWO at 410 °C. The model predicts that formamide is the exclusive primary product at short times. The analytical methods used in our experimental investigation of methylamine SCWO58 would not have allowed us to identify formamide, if it were a product of methylamine SCWO. The methods used did allow essentially complete closure of the nitrogen atom balance, however, so we do not believe that formamide, if formed, built up to measurable yields in those experiments. At longer times, the predicted formamide concentration reaches a maximum and then decreases as the concentrations of NH3 and CO2 increase. Figure 1b shows the concentrations of the minor products as predicted by the DCKM. Formic acid appears at short times and then reacts away. The model predicts CO to be present only in negligibly low yields. Nitrogen-containing products present in low yields include HONO, NO2, and N2O. A net reaction rate analysis was performed for methylamine SCWO at 410 °C and 20.2 s, conditions representative of those used in an experimental study.58 The conversion calculated at these conditions was 33%. The results were used to construct Figure 2, which shows the fastest reaction pathways for methylamine disappearance. The dominant route for CH3NH2 disappearance is hydrogen abstraction by OH.

for removal of hydrocarbons and oxygenates during SCWO. The kinetics favor abstraction at the carbon atom rather than the nitrogen atom (kR152/kR155 ) 3.5 at 410 °C).37 From the CH3NH intermediate, products such as H2CNH, HCN, HONO, NO, NO2, and N2O are obtained. Only one of these products (N2O), however, was observed during methylamine SCWO.58 In high-temperature combustion, the methylamine radicals formed above would eliminate a hydrogen atom by β-scission to form methylenimine, H2CNH. Lowtemperature oxidation chemistry, however, favors O2 addition

CH3NH2 + OH f CH2NH2 + H2O

(R152)

CH3NH2 + OH f CH3NH + H2O

(R155)

We found no additional support in the literature for this type of step for methylamine or, by analogy, for ethane oxidation. Once formamide (NH2CHO) is formed, the model provides two possible routes to NH3. The first is thermal decomposition.

R152 refers to reaction number 152 in the DCKM in the Supporting Information. We apply this naming convention to all of the elementary reactions discussed in the text. H abstraction by OH is also the main route

CH2NH2 + O2 f NH2CH2O2

(R550)

over β-scission. At 410 °C and an oxygen concentration of 17.9 mmol/L, kR550[O2]/kβ-scission ) 2.6 × 106. Hence, the CH2NH2 radicals formed in R152 are converted to peroxy radicals, NH2CH2O2, in R550. The model permits three main fates for the peroxy radical. These pathways include bimolecular reactions that form NH2CH2O via the loss of an O atom, hydrogen abstraction by the peroxy radical to form the peroxide (NH2CH2OOH), and unimolecular loss of OH to produce formamide.

NH2CH2O2 (+ M) f NH2CHO + OH (+ M) (R558) It is this last fate that has the fastest kinetics under SCWO conditions, so formamide formation is favored. We note here that an additional path for this radical has, at least on one occasion, been proposed.45

NH2CH2O2 f NH3 + CO + OH

NH2CHO (+ M) f NH3 + CO (+ M) (R559)

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005 9789 Table 2. Reactions with the Largest Sensitivity Coefficients for Methylamine during SCWO at 410 °C, 249.3 atm, [CH3NH2]0 ) 3.31 mmol/L, [O2]0 ) 17.9 mmol/L, [H2O]0 ) 8270 mmol/L, and t ) 30 s normalized sensitivity coefficient

reaction CH3NH2 + OH ) CH2NH2 + H2O H2O2 (+M) f OH + OH (+M) OH + H2O2 ) HO2 + H2O CH3NH2 + O2 ) CH3NH + HO2 CH3NH2 + OH ) CH3NH + H2O NH2CHO + H2O f NH3 + HCOOH

(R152) (R148) (R150) (R162) (R155) (R560)

-0.551 -0.330 0.072 0.025 0.019 0.011

The kinetics47 for this step are too slow, however, to form ammonia in high selectivities during low-temperature oxidation of methylamine (gas-phase or in SCW). The second route is a water-assisted reaction that leads to the formation of NH3 and formic acid.

NH2CHO + H2O f NH3 + HCOOH (R560) The kinetics for this reaction were taken to be the same as the experimental kinetics reported for acetamide hydrolysis in SCW.48 This rate is much faster than that for thermal decomposition, and this pathway is predicted to be largely responsible for NH3 formation in SCW. This reaction might also be important for gasphase combustion, provided that the water concentration is high enough and the intrinsic kinetics fast enough. The peroxy radical chemistry outlined above provides a route to NH3, possibly in the high selectivities observed experimentally in SCWO58 and low-temperature gas-phase oxidation.45 Previous DCKMs for methylamine oxidation, which omit this chemistry, would not predict a high selectivity to ammonia. Rather, HCN and NOx are the main nitrogen-containing products. We next performed a sensitivity analysis to determine the steps that had the greatest influence on the calculated CH3NH2 concentration. Reactions with the largest sensitivity coefficients appear in Table 2. The reactions identified in Table 2 are either part of the aforementioned reaction pathway for CH3NH2 consumption or steps that modify the OH radical concentration, which affects the CH3NH2 consumption rate. In particular, the reactions

H2O2 (+ M) f OH + OH (+ M)

(R148)

and

OH + H2O2 f HO2 + H2O

(R150)

have been found to be important in other SCWO modeling efforts.4,7-9,29 We used methylamine disappearance data predicted at 390, 410, and 450 °C to calculate a pseudo-first-order rate constant at each temperature. Fitting these rate constants to the Arrhenius equation leads to values of log A ) 30.6 ( 1.0 s-1 and E ) 47 ( 1 kcal/mol. Methylamine Pyrolysis in Supercritical Water. Figure 3 shows the concentrations of the major products predicted from methylamine pyrolysis in SCW at 410 °C. At 2 h of reaction time, the model predicts a conversion of 2.2%. The model predicts that the most abundant nitrogen-containing products will be HCN and NH3. The major carbon-containing products, in addition to HCN, include CH4 and CO2. The analytical methods

Figure 3. Concentration of major products from methylamine thermolysis in SCW as predicted by DCKM: 410 °C, 250 atm, [CH3NH2]0 ) 33.3 mmol/L, and [H2O]0 ) 8270 mmol/L.

used in our experimental investigation of methylamine hydrothermolysis59 allowed identification and quantification of only NH3. A net reaction rate analysis was performed for methylamine pyrolysis in supercritical water. Results and the specific reaction conditions used in the calculations appear as Figure 4. There are three main pathways for methylamine disappearance. The fastest involves hydrogen abstraction from methylamine (by OH, CH3, and H) to form CH3NH.

CH3NH2 + OH f CH3NH + H2O

(R155)

The CH3NH intermediate then eliminates a hydrogen atom by β-scission to form H2CNH

CH3NH f H2CNH + H

(R356)

which exists in near-equilibrium with HCN and H2. At this point, HCN is hydrated to form formamide, NH2CHO. NH2CHO, in turn, undergoes hydrolysis to form NH3 and formic acid, HCOOH. Finally, HCOOH decomposes to produce CO2. The second fastest path for methylamine disappearance involves the reaction of methylamine with a hydrogen atom to form NH3 and a CH3 radical.

CH3NH2 + H f CH3 + NH3

(R163)

The final path for methylamine disappearance is similar to the first in that it involves hydrogen abstraction by CH3 and H. In this path, though, the abstraction is from the carbon atom to form CH2NH2.

CH3NH2 + CH3 f CH2NH2 + CH4

(R159)

This intermediate has a smaller rate constant for β-scission than does CH3NH, so it builds up to higher concentrations than CH3NH. In fact, reaction of CH2NH2 with water to re-form methylamine is the fastest removal step for this intermediate. This result (along with the HCN hydration and NH2CHO hydrolysis mentioned above) shows that the water is actively participating in the chemistry as a reactant. This incorporation of H atoms in water into final products has been noted previously for thermolysis in supercritical water.60 On the basis of these dominant reaction paths, the products predicted to form from methylamine

9790

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005

Figure 4. Major reaction paths during methylamine pyrolysis in SCW: 410 °C, 250 atm, [CH 3NH2]0 ) 33.3 mmol/L, [H2O]0 ) 8270 mmol/L, 120 min, and conversion ) 0.022. The thicknesses of the arrows indicate the relative rates of the steps. Numbers next to the arrows are the relative net reaction rates for the corresponding steps. Table 3. Reactions with the Largest Sensitivity Coefficients for Methylamine during Pyrolysis in SCW at 410 °C, 249.3 atm, [CH3NH2]0 ) 33.3 mmol/L, [H2O]0 ) 8270 mmol/L, t ) 7200 s, and Conversion (DCKM) ) 0.022 normalized sensitivity coefficient

reaction CH3NH2 + H ) CH3 + NH3 CH3NH2 (+M) ) CH3 + NH2 (+M) CH3NH2 + OH ) CH3NH + H2O CH3NH (+M) f H2CNH + H (+M)

(R163) (-R151) (R155) (R356)

-0.0120 -0.0109 -0.0043 -0.0032

thermolysis in supercritical water are NH3, HCN, CH4, CO2, H2O, and H2. A sensitivity analysis for methylamine pyrolysis in SCW identified the reactions in Table 3 as the ones that most strongly affected the calculated concentration of CH3NH2. These important reactions either appear in the reaction pathway as steps for CH3NH2 disappearance or modify the H radical concentration, which influences the CH3NH2 removal rate. These normalized sensitivity coefficients are very small, however, indicating that it is difficult to alter greatly the calculated methylamine concentrations given the allowable uncertainties of the associated kinetic parameters. We used methylamine disappearance data predicted at 386-500°C to calculate a rate constant at each temperature. We used a 0.66-order rate equation, as was determined experimentally.59 Fitting these rate constants to the Arrhenius equation leads to the values log A ) 23.8 ( 2.8 M0.34 s-1 and E ) 51 ( 4 kcal/mol. Comparison with Experimental Data When doing detailed modeling studies, one typically seeks to validate the model by comparing DCKM predictions and experimental data. One major difficulty in making these comparisons for the present system is the absence of experimental data that reflect the action of homogeneous reactions alone. The methylamine

SCWO experiments58 were conducted in a flow reactor made from Hastelloy C-276 (57% Ni). The SCW hydrothermolysis experiments59 were conducted in 316 stainless steel (12% Ni) batch reactors. Nickel and nickel oxide both catalyze cleavage of the C-N bond in methylamine,61,62 and we had suspected that heterogeneous catalysis contributed to the experimental results obtained. Recent experiments in our laboratory have confirmed catalysis by nickel during methylamine hydrothermolysis in supercritical water. Experiments were conducted at 410°C in quartz tubes. Nickel powder (about 1 mg) was added to some reactors and omitted from others. The methylamine conversion at 120 min was 100% in the reactors with added Ni but only 16% in the reactors with no added metal. In both experiments, the ammonia selectivity was essentially 100%. We will report more fully on this topic in due course. At present, though, this finding leads us to conclude that the chemistry observed in previous experiments on methylamine reactivity in supercritical water, which were conducted in reactors with nickel-containing walls, occurred by a combined heterogeneous/homogeneous mechanism or perhaps even an exclusively heterogeneous mechanism. Given the intrusion of surface reactions into these previous experimental studies, one should not expect the DCKM to accurately predict these previous results. Therefore, we will show no comparisons between DCKM predictions and experimental results for methylamine reactions in SCW. The interested reader is referred elsewhere63 for these comparisons. Older data exist for methylamine oxidation in the gas phase45 at low temperatures. These experiments were done in glassware, so catalysis by metals is absent. The authors showed, however, that the reactor surface-tovolume ratio affected the measured rate. Hence, heterogeneous reactions contributed to the observed behavior in this system as well. Nevertheless, this system

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005 9791

Figure 5. Experimental results45 and DCKM predictions for the variation of product yields from methylamine oxidation in the gas phase at 350 °C.

is the best one available for comparing model and experimental results for methylamine oxidation. These experiments were also done near ambient pressure, so thermodynamic nonidealities omitted from the model should not be a complicating factor here. Figure 5 shows DCKM predictions and experimental data45 for methylamine oxidation at 350 °C. The model matches the experimental methylamine yields during the first 30 s. At longer times, though, the model does not capture what appears to be a sigmoidal curve, and it predicts slower CH3NH2 conversion than seen experimentally. The model predicts almost no NH3 formation during the first 30 s, but the experiments revealed ammonia to be the major nitrogen-containing product at these short times. The experimental ammonia selectivity is over 90% at low conversions. Rather than appearing as ammonia, the reacted nitrogen in the model appears as formamide at short times. At longer times, the formamide decomposes to form NH3, and hence, the NH3 yield predicted by the model increases with time. The DCKM results show that formamide disappearance (by both thermal decomposition and hydrolysis) is too slow to account for the experimentally observed formation of NH3 in high selectivities. It is interesting to note that Cullis and Willsher45 presented a mechanism for methylamine oxidation that was able to quantitatively describe the increase in reactor pressure with time during their constant-volume batch reactor experiments. The rate constants in their mechanism were treated as empirical parameters, though, so their model was correlative rather than predictive. That mechanism showed the peroxy radical decomposing in a single step to form ammonia

NH2CH2OO f NH3 + CO + OH We found no support for this type of step elsewhere in the more recent chemical literature, so we excluded it from our DCKM. Rather, the fastest path for NH2CH2OO in our DCKM is formamide formation.

NH2CH2O2 (+ M) f NH2CHO + OH (+ M) (R558) If the step Cullis and Willsher proposed actually occurs and is faster than the one in our model (R558), then our model should predict a lower methylamine reactivity (because the predicted OH concentration is too low) and a lower ammonia yield (because the model predicts NH2CHO rather than NH3), which are precisely the dis-

crepancies observed. It is possible that better agreement between the model and experimental results for lowtemperature gas-phase oxidation of methylamine could be achieved by including this step. Further investigation into this step is warranted. It might be a key to understanding the high yields of ammonia found during methylamine oxidation, both in the gas phase and in SCW. While acknowledging the inability of the present DCKM to predict the experimental results reported by Cullis and Willsher, we also point out that other DCKMs for methylamine oxidation in the gas phase are also unable to predict these results. This failure points to the need for a better understanding of methylamine oxidation in the gas phase at these lower temperatures. The mechanism in this region is uncertain, and the kinetics for several important steps are estimated. Our DCKM is the first to include peroxy radical chemistry, which is critical in this temperature range. Unfortunately, the kinetics for many of these steps and even the precise mechanism are not known with enough certainty to allow much confidence in the model results. We did not anticipate that such would be the state of affairs when we initiated this work. Outlook and Conclusions (1) This work presents the first DCKM for methylamine oxidation at low temperatures and methylamine reactivity in SCW. The model predicts that formamide is the exclusive primary product from oxidation reactions. It reacts further to produce formic acid and ammonia. Formic acid decomposes to give CO2. OH attack on methylamine and the formation and subsequent reactions of peroxy radicals are central features of the oxidation reaction network. (2) The DCKM developed in this article is the most complete mechanism to date for methylamine oxidation at low temperatures. Even so, this DCKM is not able to describe fully methylamine oxidation in the gas phase. This mismatch between model and experimental results indicates that the mechanism and kinetics of low-temperature gas-phase oxidation are not yet fully understood. More progress in this area, especially with respect to nitrogen-containing peroxy radical chemistry, is required before application to SCWO conditions is likely to be successful. (3) Catalysis by nickel, and perhaps by other metals, is important for methylamine reactivity in SCW. Because of this influence, the existing literature data for methylamine reactions in SCW, which were obtained in nickel-containing reactors, cannot be used to validate a DCKM. There is a need for experimental data for this system that are free of the influence of metal-catalyzed reactions. (4) The modeling results presented herein show that formation and decomposition of formamide, though the fastest path to ammonia in the DCKM, is not likely the fastest path to ammonia in low-temperature gas-phase oxidation experiments. The decomposition of formamide is too slow to give the high initial ammonia selectivities seen experimentally. To our knowledge, the literature contains no plausible mechanisms for NH3 formation in high selectivity during low-temperature methylamine oxidation. The direct decomposition of the peroxy radical that Cullis and Wilsher45 used to fit their experimental data might offer an explanation for high ammonia yields if such a step is chemically reasonable. This gap in the

9792

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005

combustion literature had apparently gone unrecognized, perhaps because little work has been done on the low-temperature oxidation of organo-nitrogen compounds in the past 50 years. (5) The present DCKM for methylamine is large and complex. A smaller, simpler model would be easier to work with and hence more likely to be used. Efforts to reduce the number of steps and number of species in the model would be worthwhile. Targets for simplification include the steps currently in the DCKM that are important at high temperatures but not likely to be important at the lower temperatures of interest in SCWO. Acknowledgment We thank Dr. John Barker for helpful insights regarding the kinetics and mechanisms of low-temperature oxidation chemistry and nitrogen chemistry. Matt Jorgenson, Christine Whitlock, and Craig Comisar provided the experimental results for methylamine reactions in SCW in quartz reactors with and without added nickel. We gratefully acknowledge partial financial support of this work by Grant CTS-9903373 from the NSF. Supporting Information Available: The detailed chemical kinetics model used in the simulations is available free of charge via the Internet at http:// pubs.acs.org. Literature Cited (1) Savage, P. E.; Gopalan, S.; Mizan, T. I.; Martino, C. J.; Brock, E. E. Reactions at Supercritical Conditions: Applications and Fundamentals. AIChE J. 1995, 41, 1723. (2) Killilea, W.; Swallow, K.; Hong, G. The Fate of Nitrogen in Supercritical Water Oxidation. J. Supercrit. Fluids 1992, 5, 72. (3) Thomason, T.; Hong, G.; Swallow, K.; Killilea, W. The MODAR Supercritical Water Oxidation Process. In Innovative Hazardous Waste Treatment Technology Series: Thermal Processes; Freeman, H., Ed.; Technomic Publishing: Lancaster, PA, 1990; p 31. (4) Holgate, H. R.; Tester, J. W. Fundamental Kinetics and Mechanisms of Hydrogen Oxidation in Supercritical Water. Combust. Sci. Technol. 1993, 88, 369. (5) Holgate, H. R.; Tester, J. W. Oxidation of Hydrogen and Carbon Monoxide in Sub- and Supercritical Water: Reaction Kinetics, Pathways, and Water Density Effects. 2. Elementary Reaction Modeling. J. Phys. Chem. 1994, 98, 810. (6) Brock, E. E.; Savage, P. E. Detailed Chemical Kinetics Model for Supercritical Water Oxidation of C1 compounds and H2. AIChE J. 1995, 41, 1874. (7) Dagaut, P.; Demarcillac, B. D.; Tan, Y.; Cathonnet, M.; Boettner, J. C. Chemical Kinetic Modeling of the Supercritical Water Oxidation of Simple FuelssH2, CO and CH4. J. Chim. Phys. PCB 1995, 92, 1124. (8) Dagaut, P.; Cathonnet, M.; Boettner, J. C. Chemical Kinetic Modeling of the Supercritical Water Oxidation of Methanol. J. Supercrit. Fluids 1996, 9, 33. (9) Alkam, M. K.; Pai, V. M.; Butler, P. B.; Pitz, W. J. Methanol and Hydrogen Oxidation Kinetics in Water at Supercritical States. Combust. Flame 1996, 106, 110. (10) Webley, P.; Tester, J. W. Fundamental Kinetics of Methane Oxidation in Supercritical Water. Energy Fuels 1991, 5, 411. (11) Savage, P. E.; Yu, J.; Stylski, N.; Brock, E. E. Kinetics and Mechanism of Methane Oxidation in Supercritical Water. J. Supercrit. Fluids 1998, 12, 141. (12) Webley, P.; Tester, J. W. Fundamental Kinetics of Methanol Oxidation in Supercritical Water. In Supercritical Fluid Science and Technology Johnston K., Penninger, J., Eds.; ACS Symposium Series No. 406; American Chemical Society: Washington, D.C., 1989; p 259.

(13) Brock, E. E.; Oshima, Y.; Savage, P. E.; Barker, J. R. Kinetics and Mechanism of Methanol Oxidation in Supercritical Water. J. Phys. Chem. 1996, 100, 15834. (14) Gopalan, S.; Savage, P. E. Reaction Mechanism for Phenol Oxidation in Supercritical Water. J. Phys. Chem. 1994, 98, 12646. (15) Gopalan, S.; Savage, P. E. Phenol Oxidation in Supercritical WatersFrom Global Kinetics and Product Identities to an Elementary Reaction Model. ACS Symp. Ser. 1995, 608, 217. (16) Henrikson, J. T.; Chen, Z.; Savage, P. E. Inhibition and Acceleration of Phenol Oxidation by Supercritical Water. Ind. Eng. Chem. Res. 2003, 42, 6303. (17) DiNaro, J. L.; Howard, J. B.; Green, W. H.; Tester, J. W. Elementary Reaction Mechanism for Benzene Oxidation in Supercritical Water. J. Phys. Chem. A 2000, 104, 10576. (18) DiNaro, J. L.; Howard, J. B.; Green, W. H.; Tester, J. W. Analysis of an Elementary Reaction Mechanism for Benzene Oxidation in Supercritical Water. Proc. Combust. Inst. 2000, 28, 1529. (19) Brock, E. E.; Savage, P. E.; Barker, J. R. A Reduced Mechanism for Methanol Oxidation in Supercritical Water. Chem. Eng. Sci. 1998, 53, 857. (20) Savage, P. E.; Rovira, J.; Stylski, N.; Martino, C. J. Oxidation Kinetics for Methane/Methanol Mixtures in Supercritical Water. J. Supercrit. Fluids 2000, 17, 155. (21) Basevich, V. Ya. Chemical Kinetics in the Combustion Processes: A Detailed Kinetics Mechanism and its Implementation. Prog. Energy Combust. Sci. 1987, 13, 199. (22) Higashihara, T.; Gardiner, W. C.; Hwang, S. M. ShockTube and Modeling Study of Methylamine Thermal Decomposition. J. Phys. Chem. 1987, 91, 1900. (23) Hwang, S. M.; Higashihara, T.; Shin, K. S.; Gardiner, W. C. Shock Tube and Modeling Study of Monomethylamine Oxidation. J. Phys. Chem. 1990, 94, 2883. (24) Lifshitz, A.; Bidani, M.; Carroll, H. F.; Hwang, S. M.; Fu, P. Y.; Shin, K. S.; Gardiner, W. C. Ignition of Monomethylamine. Combust. Flame 1991, 86, 229. (25) Williams, B. A.; Fleming, J. W. Radical Species Profiles in Low-Pressure Methane Flames Containing Fuel Nitrogen Compounds. Combust. Flame 1997, 110, 1. (26) Kantak, M. V.; De Manrique, K. S.; Aglave, R. H.; Hesketh, R. P. Methylamine Oxidation in a Flow Reactor: Mechanism and Modeling. Combust. Flame 1997, 108, 235. (27) Miller, J. A.; Bowman, C. T. Mechanism and Modeling of Nitrogen Chemistry in Combustion. Prog. Energy Combust. Sci. 1989, 15, 287. (28) Glarborg, P.; Miller, J. A. Mechanism and Modeling of Hydrogen-Cyanide Oxidation in a Flow Reactor. Combust. Flame 1994, 99, 475. (29) Brock, E. E., Detailed Chemical Kinetic Modeling of the Supercritical Water Oxidation of Simple Hydrocarbons. Ph.D. Thesis, The University of Michigan, Ann Arbor, MI, 1997. (30) Feng, J. B.; Aki, S. N. V. K.; Chateauneuf, J. E.; Brennecke, J. F. Abstraction of Hydrogen from Methanol by Hydroxyl Radical in Subcritical and Supercritical Water. J. Phys. Chem. A 2003, 107, 11043. (31) Jodkowski, J. T.; Ratajczak, E.; Fagerstrom, K.; Lund, A.; Stothard, N. D.; Humpfer, R.; Grotheer, H. H. Kinetics of the Cross Reaction Between Amidogen and Methyl Radicals. Chem. Phys. Lett. 1995, 240, 63. (32) Lindstedt R. P.; Lockwood, F. C.; Selim, M. A. Detailed Kinetic Modeling of Chemistry and Temperature Effects on Ammonia Oxidation. Combust. Sci. Technol. 1994, 99, 253. (33) Lindstedt R. P.; Selim, M. A. Reduced Reaction Mechanisms for Ammonia Oxidation in Premixed Laminar Flames. Combust. Sci. Technol. 1994, 99, 277. (34) Lindstedt R. P.; Lockwood, F. C.; Selim, M. A. A Detailed Kinetic Study of Ammonia Oxidation. Combust. Sci. Technol. 1995, 108, 231. (35) Pedersen, L. S.; Glarborg, P.; Dam-Johansen, K. A Reduced Reaction Scheme for Volatile Nitrogen Conversion in Coal Combustion. Combust. Sci. Technol. 1998, 131, 193. (36) GRI Mechanism 3.0, anonymous FTP site GRI_MECH at CRVAX.SRI.COM, 1999, or see Internet site http://www.gri.org (accessed March 2003). (37) Dean, A. M.; Bozzelli, J. W. Combustion Chemistry of Nitrogen. In Gas-Phase Combustion Chemistry; Gardiner, W. C., Jr., Ed.; Springer-Verlag: New York, 2000; p 125.

Ind. Eng. Chem. Res., Vol. 44, No. 26, 2005 9793 (38) Tsang, W.; Herron, J. T. Chemical Kinetic Data Base for Propellant Combustion. I. Reactions Involving NO, NO2, HNO, HNO2, HCN and N2O. J. Phys. Chem. Ref. Data 1991, 20, 609. (39) Ko, T.; Fontijn, A. High-Temperature Photochemistry Kinetics Study of the Reaction H + NO2 ) OH + NO from 296 to 760 K. J. Phys. Chem. 1991, 95, 3984. (40) Cohen, N.; Westberg, K. R. Chemical Kinetic Data Sheets for High-Temperature Reactions. Part II. J. Phys. Chem. Ref. Data 1991, 20, 1211. (41) Clyne, M. A. A.; Thrush, B. A. Rates of the Reactions of Nitrogen Atoms with Oxygen and with Nitric Oxide. Nature 1961, 189, 56. (42) Tsang, W., Chemical Kinetic Data Base for Propellant Combustion. II. Reactions Involving CN, NCO, and HNCO. J. Phys. Chem. Ref. Data 1992, 21, 753. (43) Hanson, R. K.; Salimian, S. Survey of Rate Constants in the N/H/O System. In Combustion Chemistry; Gardiner, W. C., Jr., Ed.; Springer-Verlag: New York, 1984; p 361. (44) Senkan, S., Detailed Chemical Kinetic Modeling: Chemical Reaction Engineering of the Future. Adv. Chem. Eng. 1992, 18, 95. (45) Cullis, C. F.; Willsher, J. P. The Thermal Oxidation of Methylamine. Proc. R. Soc. London Ser. A 1951, 209, 218. (46) Barbieri, G.; Di Maio, F. P.; Lignola, P. G. Low and Intermediate Temperature Ethane Combustion Modeling. Combust. Sci. Technol. 1994, 98, 95. (47) Kakumoto, T.; Saito, K.; Imamura, A. Thermal Decomposition of Formamide: Shock Tube Experiments and ab Initio Calculations. J. Phys. Chem. 1985, 89, 2286. (48) Lee, D. S.; Gloyna, E. F. Hydrolysis and Oxidation of Acetamide in Supercritical Water. Environ. Sci. Technol. 1992, 26, 1587. (49) Yu., J.; Savage, P. E. Decomposition of Formic Acid Under Hydrothermal Conditions. Ind. Eng. Chem. Res. 1998, 37, 2. (50) Boock, L. T.; Klein, M. T. Lumping Strategy for Modeling the Oxidation of C1-C3 Alcohols and Acetic Acid in HighTemperature Water. Ind. Eng. Chem. Res. 1993, 32, 2464. (51) Larson, C. W.; Stewart, P. H.; Golden, D. M. Pressure and Temperature Dependence of Reactions Proceeding via a Bound Complex. An Approach for Combustion and Atmospheric Chemistry Modelers. Application to HO + CO ) (HOCO) ) H + CO2. Int. J. Chem. Kinet. 1988, 20, 27.

(52) Galano, A.; Alvarez-Idaboy, J. R.; Ruiz-Santoyo, M. E.; Vivier-Bunge, A. Rate Coefficient and Mechanism of the Gas-Phase OH Hydrogen Abstraction Reaction from Formic Acid: A Quantum Mechanical Approach. J. Phys. Chem. A 2002, 106, 9520. (53) Kee, R.; Rupley, F.; Miller, J. The CHEMKIN Thermodynamic Data Base; Report SAND87-8215; Sandia National Laboratories: Albuquerque, NM, 1991. (54) Anderson, W. R. Heats of Formation of HNO and Some Related Species. Combust. Flame 1999, 117, 394. (55) Ritter, E.; Bozzelli, J. THERM: Thermodynamic Property Estimation for Gas-Phase Radicals and Molecules. Int. J. Chem. Kinet. 1991, 23, 767. (56) NIST Structure and Properties, version 2.0; National Institute of Standards and Technology (NIST): Gaithersburg, MD, 1994. (57) Kee, R.; Rupley, F.; Miller, J. CHEMKIN II: A FORTRAN Chemical Kinetics Package for the Analysis of Gas-Phase Chemical Kinetics; Report SAND89-8009; Sandia National Laboratories: Albuquerque, NM, 1990. (58) Benjamin, K. M.; Savage, P. E. Supercritical Water Oxidation of Methylamine. Ind. Eng. Chem. Res. 2005, 44, 5318. (59) Benjamin, K. M.; Savage, P. E. Hydrothermal Reactions of Methylamine. J. Supercrit. Fluids 2004, 31, 301. (60) Ederer, H. J.; Kruse, A.; Mas, C.; Ebert, K. H. Modeling of the Pyrolysis of tert-Butylbenzene in Supercritical Water. J. Supercrit. Fluids 1999, 15, 191. (61) Schoofs, G. R.; Benziger, J. B. Reactions of Organonitrogen Molecules with Ni(100). J. Phys. Chem. 1988, 92, 741. (62) Chang, C.; Khong, C.; R. Saiki, R. Temperature-Programmed Reaction of Methylamine on the Ni(100) Surface. J. Vac. Sci. Technol. A 1993, 11, 2122. (63) Benjamin, K. M. Nitrogen Chemistry in Supercritical Water. Ph.D. Thesis. University of Michigan, Ann Arbor, MI, 2004. (64) Kramer, A.; Mittelstadt, S.; Vogel, H. Hydrolysis of Nitriles in Supercritical Water. Chem. Eng. Technol. 1999, 22, 494.

Received for review August 10, 2005 Revised manuscript received September 27, 2005 Accepted September 28, 2005 IE050926L