K-Doped CoâMnâAl Mixed Oxide Catalyst for N2O Abatement from

Article pubs.acs.org/IECR

K‑Doped Co−Mn−Al Mixed Oxide Catalyst for N2O Abatement from Nitric Acid Plant Waste Gases: Pilot Plant Studies K. Pacultová,*,† K. Karásková,‡ F. Kovanda,§ K. Jirátová,∥ J. Šrámek,⊥ P. Kustrowski,# A. Kotarba,# Ž . Chromčaḱ ová,† K. Kočí,†,‡ and L. Obalová†,‡ Institute of Environmental Technology, and ‡Faculty of Metallurgy and Materials Engineering, VŠB-Technical University of Ostrava, 17. listopadu 15, 708 33 Ostrava, Czech Republic § University of Chemistry and Technology, Prague, Technická 5, 166 28 Prague, Czech Republic ∥ Institute of Chemical Process Fundamentals of the ASCR, v. v. i., Rozvojová 135, 165 02 Prague 6, Czech Republic ⊥ Chemoprojekt Chemicals, s.r.o., Třebohostická 14, 100 31 Praha 10, Czech Republic # Faculty of Chemistry, Jagiellonian University, Ingardena 3, 30-060 Krakow, Poland †

ABSTRACT: The K-doped Co−Mn−Al mixed oxide deN2O catalyst was prepared by calcination of Co−Mn−Al layered double hydroxide, subsequent impregnation with KNO3, and shaping into tablets of 5 mm × 5 mm. Tablets were tested for N2O decomposition in pilot scale reactor connected at the bypassed tail gas from the nitric production plant downstream of the SCR NOx/NH3 catalyst with the aim to perform kinetic measurements. Special attention was paid to study the changes in the catalytic performance, which was monitored using both the long-term catalytic test and the comparison of physical−chemical properties of the fresh and used catalyst. The changes in the surface composition, caused by the time-on-stream operation for 112 days in the pilot reactor, provided stable catalytic performance; average value of N2O conversion of 90 ± 6% at 450 °C was kept (GHSV = 11 000 m3 mbed−3 h−1). The obtained kinetic data were applied in modeling of a fullscale reactor for the N2O emissions abatement and can be used for estimation of the amount of the catalyst necessary for obtaining required N2O conversion in the target plant.

1. INTRODUCTION Nitrous oxide is formed as an undesirable byproduct during catalytic oxidation of ammonia in nitric acid production plants. Taking into account the global production of the nitric acid and the high global warming potential of N2O, much emphasis is currently placed on the development of methods for N2O emissions abatement. Depending on the location in the process, the approaches of lowering of N2O from nitric acid plants can be classified into three groups: (i) prevention of N2O formation in the ammonia burner, (ii) N2O removal from NOx gases between the ammonia burner and the absorption column, and (iii) N2O removal from the tail gas downstream (low temperature deN2O, tertiary abatement). The choice of suitable deN2O technology depends on the current arrangement of the nitric acid plants. Since they can be radically different, each approach is necessary to be developed individually. While the first method covers changes in the burner geometry and conditions of the process, the N2O removal downstream burner requires either thermal or catalytic decomposition of the already formed N2O into N2 and O2 (eq 1). As the tertiary technology is an end-of-pipe one, no interference with the nitric acid production process is the advantage of this option. Only a small number of catalysts for low temperature deN2O application have already been commercialized, and also only © XXXX American Chemical Society

few studies devoted to the catalysts tested in pilot plants were reported.1−4 The resistance against water and oxygen inhibition, high performance in the presence of NOx (NO + NO2), and long-term stability in wet acidic environment are crucial requirements for this catalyst. Nowadays, the most known available tertiary abatement technology is the EnviNOXN2O manufactured by Süd Chemie and C-NAT catalyst prepared by CRI, both also encompassed in CDM projects.5 However, the offer of commercial catalysts suitable for low temperature N2O decomposition available at the market is still very limited.3 Our recent research activities have been focused on development of a catalyst for the low temperature N2O decomposition, covering the catalysts development and preparation, their laboratory and pilot plant-scale testing including the long-term activity tests and assessment of usage of such a catalyst in the full scale plant based on the mathematical model of pilot plant reactor. Received: March 29, 2016 Revised: June 8, 2016 Accepted: June 13, 2016

A

DOI: 10.1021/acs.iecr.6b01206 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research N2O → N2 +

1 O2 2

mm) shown in Figure 1a. The catalyst was produced by ASTIN Catalysts and Chemicals, Ltd., Czech Republic.

(1)

From a wide group of catalytic systems proposed for this process in literature, the cobalt-containing mixed oxides prepared from layered double hydroxide (LDH) precursors have been found to be very efficient in N2O decomposition.6−11 Layered double hydroxides (also called hydrotalcite-like compounds or hydrotalcites) are layered materials consisting of positively charged hydroxide layers separated by interlayers composed of anions and water molecules. The chemical composition of LDHs can be expressed by general formula [MII1−xMIIIx(OH)2]x+[An−x/n·yH2O]x− where MII and MIII are divalent and trivalent metal cations, An− is an n-valent anion, and x usually has values between 0.20 and 0.33. After heating at moderate temperatures, LDHs give finely dispersed mixed oxides of MII and MIII metals with high surface area and good thermal stability. We have tested different combinations of metal cations in hydroxide layers at constant MII/MIII molar ratio of 2 (MII = Co, Ni, Cu, Mg; MIII = Al, Mn, Fe; An− = CO32−). Optimization of chemical composition resulted in obtaining the LDH-related Co−Mn−Al mixed oxide with Co:Mn:Al molar ratio of 4:1:1, which showed high activity and stability in wet gas containing oxygen.12 The catalytic activity of cobalt-containing mixed oxides can be improved by their modification with potassium promoter.13−16 Our recent results showed that the optimum content of potassium in Co−Mn−Al mixed oxide is 0.9−1.6 wt % K.15 The direct correlation of the catalytic activity with the catalyst electron work function revealed that the beneficial effect of K is mainly of electronic origin.17 Models of catalytic reactor for N2O abatement from waste gas from HNO3 production plant using intrinsic kinetic data over grains of Co−Mn−Al mixed oxide18 and/or kinetic data obtained over laboratory prepared Co−Mn−Al mixed oxide tablets19 were reported earlier. In the present work, the Co−Mn−Al mixed oxide modified by potassium promoter was prepared in large scale for the first time and tested in real conditions. Results of N2O catalytic decomposition in the pilot plant reactor installed at the bypassed tail gas from the nitric acid production plant (BorsodChem-MCHZ, Czech Republic) are shown covering activity and selectivity. Due to the known thermal instability of alkali promoters, special attention was paid to study the changes in the catalytic performance, which was monitored using both the long-term catalytic test and the comparison of physical−chemical properties of the fresh and used catalyst. The study of the kinetics of the de N2O decomposition directly in the pilot plant was also done. The obtained kinetic data are used for modeling of full scale reactor for the N2O emissions abatement.

Figure 1. Photos of K-doped Co−Mn−Al mixed oxide catalyst (a), pilot plant catalytic reactor (b), and pilot plant catalytic setup (c).

2.2. Catalyst Characterization. The contents of Co, Mn, Al, and K in the catalyst were determined by atomic absorption spectroscopy (AAS) after sample dissolution in hot hydrochloric acid (35 wt %) and appropriate dilution with distilled water. The concentrations of the metal cations were determined using a SpectrAA880 instrument (Varian). Powder X-ray diffraction (XRD) patterns were recorded using a Seifert XRD 3000P instrument with Co Kα radiation (λ = 0.179 nm, graphite monochromator, goniometer with the Bragg−Brentano geometry) in 2θ range from 10° to 80° with a step size of 0.05°. Qualitative analysis was performed with the HighScore package (PANalytical, version 2.0a, The Netherlands). Surface area and mesoporous structure of prepared catalysts were determined by physisorption of nitrogen using Micromeritics ASAP 2020 instrument after catalysts drying at 105 °C for 24 h under vacuum of 1 Pa. The adsorption−desorption isotherms of nitrogen at −196 °C were treated by the standard Brunauer−Emmett−Teller (BET) procedure to calculate the specific surface area SBET. The surface area of mesopores Smeso and the volume of micropores Vmicro were determined by t-plot method. Pore-size distribution (pore radius 1−100 nm) was calculated from the desorption branch of the adsorption− desorption isotherm by the advanced Barrett−Joyner−Halenda (BJH) method. The Lecloux−Pirard standard isotherm20,21 was employed for the t-plot as well as the pore-size distribution evaluation. An AutoPore 9266 was used to determine catalyst total pore volume, bulk density, and pore size distribution and a helium pycnometer AccuPycII 1340 to measure true (helium) catalyst density. Temperature-programmed reduction (TPR-H2) measurements of the calcined samples (0.025 g) were performed with a H2/N2 mixture (10 mol % H2), flow rate of 50 mL min−1, and linear temperature increase of 20 °C min−1 up to 500 °C. A change in H2 concentration was detected with a mass spectrometer Omnistar 300 (Pfeiffer Vakuum). Reduction of the grained CuO (0.16−0.315 mm) was performed to calculate absolute values of the hydrogen consumed during reduction. The TPR experiments were evaluated using OriginPro 8.0 software with an accuracy of ±5%. The X-ray photoelectron spectra (XPS) of the catalysts were measured on a Prevac photoelectron spectrometer equipped with a hemispherical VG SCIENTA R3000 analyzer. The studied samples were loaded through a load lock (where pressure better than 3 × 10−7 mbar was achieved) into an ultrahigh vacuum analytical chamber with a base pressure of 5 × 10−9 mbar. XPS measurements were taken with a monochromatized aluminum source AlKα (E = 1486.6 eV) and a low energy electron flood gun (FS40A-PS) to

2. EXPERIMENTAL METHODS 2.1. Catalyst Preparation. The Co−Mn−Al LDH precursor with Co:Mn:Al molar ratio of 4:1:1 was prepared by coprecipitation of corresponding nitrates in an alkaline Na 2 CO 3 /NaOH solution at 25 °C and pH 10. The concentrations used were 40 and 106 mol/L for NaOH and Na2CO3, respectively. The resulting suspension was filtered off, washed with water, dried at 105 °C, and calcined for 4 h at 500 °C in air. The resulting mixed oxides were milled, impregnated with KNO3, recalcined, and formed into tablets (5 mm × 5 B


Article

Industrial & Engineering Chemistry Research Table 1. AAS and XPS Analysis of the Fresh and Used K Doped Co−Mn−Al Mixed Oxide Catalysts XPS (mol %), AAS (wt %) catalyst fresh useda a

AAS XPS AAS XPS

molar ratio

Co

Mn

Al

K

O

C

Co:Mn:Al

K:Co

Co:Mn

Co:Al

Mn:Al

45.0 18.14 49.9 17.45

9.28 3.34 8.51 4.83

5.16 8.68 5.20 10.78

1.25 1.71 1.27 2.26

nd 54.13 nd 60.20

nd 13.99 nd 4.47

4:0.88:1.0 4:0.7:1.9 4:0.73:0.91 4:1.1:2.5

0.042 0.094 0.038 0.130

4.52 5.43 5.47 3.61

3.99 2.09 4.39 1.62

0.88 0.38 0.80 0.45

For 112 days.

2.4. Reactor and Kinetics Models. A pseudo-homogeneous one-dimensional model of an ideal plug flow reactor in isothermal regime at steady state was used for the modeling of pilot plant fixed bed reactor (eqs 2−5).22 The first-order rate law (eq 4) was supposed for the kinetics of N2O decomposition in the excess of oxygen and constant concentration of water vapor and NOx.23 Evaluated kinetic parameters determined from experiments over the catalyst tablets in the pilot plant reactor describe the reaction rate including internal diffusion hindering effect.

compensate charge on the surface of nonconductive samples. Peaks were recorded with constant pass energy of 100 eV for the survey and high resolution spectra. The binding energies were referenced to the C 1s core level (284.6 eV) from hydrocarbon contaminations. The composition and chemical surrounding of sample surface were investigated on the basis of the areas and binding energies of K 2p, Mn 2p, Co 2p, O 1s, Al 2p, and C 1s photoelectron peaks. Fitting of the high resolution spectra was provided through the CasaXPS software. The stability of potassium was investigated by the Species resolved thermal alkali desorption method (SR-TAD). The experiments were carried out in a vacuum apparatus with a background pressure of 10−7 kPa. The samples (calcined at 500 °C) in the form of wafers (10 mm diameter, 100 mg weight) were heated up from room temperature to 600 °C in a stepwise mode at the rate of 5 °C min−1. The desorption flux of potassium atoms was determined by means of a surface ionization detector, whereas the flux of ions was determined with an ion collector. During the measurements, the samples were biased with a positive potential (+10 and +100 V for atoms and ions, respectively) to quench the thermal emission of electrons and additionally, in the case of ions, to accelerate them toward the collector. In this way, the possibility of reneutralization of ions by thermal electrons outside the surface is effectively eliminated. In all measurements, the resulting positive current was directly measured with a Keithley 6512 digital electrometer. 2.3. Measurements of N2O Catalytic Decomposition in the Pilot Plant Reactor. Pilot plant catalytic measurements of N2O decomposition were performed in a fixed bed stainless steel reactor (0.31 m internal diameter) (Figure 1b) in the temperature range from 350 to 450 °C and inlet pressure of 0.6 MPa. The reactor was connected at the bypassed tail gas from the nitric acid production plant downstream the SCR NOx/ NH3 catalyst (Figure 1 c). The catalyst tablets (69.1 kg weight, 62.5 cm bed high, 1334 kg m−3 bed density) was placed on a stainless steel grate, sieve, and bed of ceramic spheres (diameter of 8 mm) with hight of 5 cm. On the layer of catalyst, again ceramic spheres (hight of 1 cm), sieve, and last layer of ceramic spheres (hight of 6.5 cm) was placed. The feed to the reactor varied between 300 and 600 kg h−1 and contained typically 400−700 ppm of N2O together with oxygen, water vapor, and low concentration of NO, NO2, and NH3 (0−70 ppm of NOx, 0−30 ppm of NH3). The variable composition of gas mixture at the reactor inlet was due to the fact that the inlet gas was the real waste gas from nitric acid plant downstream the SCR NOx/ NH3 unit. The infrared and chemiluminescence online analyzers were used for analysis of the gas at the catalyst bed inlet and outlet: Sick (N2O), Horiba (NO, NO2), ABB modul Uras 26 (N2O), and ABB Limas11 (NO, NO2, and NH3). The reactor was equipped with online monitoring of concentrations of all measured gas components, temperature in the catalyst bed, and pressure drop.

material balance of component A (A = N2O): p0 M Aρbed dXA =r 0 dz RT cA0ṁ

−

impulse balance: cD =

(2)

dp c ρv 2 = D , dz dekv

⎤ (1 − ε) ⎡ 368(1 − ε) + 1.24⎥ ⎢ 3 ⎣ ⎦ Re ε r = kcA ,

kinetic equation:

stoichiometry:

k = k 0 e−Ea /(RT )

⎛ p⎞ cA = cA0(1 − XA )⎜⎜ ⎟⎟ ⎝ p0 ⎠

(3) (4)

(5)

This simple model was sufficient for the description of the pilot plant reactor because (i) the change of volumetric flow along the reactor can be neglected due to the low N2O concentration, (ii) temperature gradients in the catalyst bed are not expected on account of a low heat release during the reaction, and (iii) axial diffusion was negligible as it was proven by the Bodenstein number (Bo = 955 > 100). Moreover, the plug flow conditions and homogeneous distribution of the gas residence time were validated because the criterion Lbed/dp > 100 was met. Polymath software was used for solving the system of ordinary differential eqs (eqs 2−5). For the evaluation of kinetic parameters (k0, Ea) from the measurements in the pilot plant reactor, only mass balance was considered while the impulse balance could be neglected due to the pressure drop lower than 10 kPa.24 Integrated form of mass balance was used for k calculation (eqs 6 and 7), and then k0 and Ea were determined from Arrhenius plot ln k* versus 1/T. ln

w 1 = k* ṁ 1 − XA

k* = k C

(6)

p0 M RT 0

(7) DOI: 10.1021/acs.iecr.6b01206 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

3. RESULTS AND DISCUSSION 3.1. Characterization of Fresh and Used Catalyst. Chemical composition of potassium-promoted Co−Mn−Al mixed oxide catalyst determined from chemical analysis is summarized in Table 1. Molar ratio Co:Mn:Al in the fresh catalyst slightly differed from the demanded values: Mn and Al contents were lower than intended, very probably due to loss of metals during the washing of coprecipitated precursor. Such effect was not observed during the following impregnation step of the calcined precursor with potassium nitrate solution. Therefore, the potassium content in the fresh catalyst as well as the K/Co molar ratio (0.042) determined by AAS corresponded well to the value adjusted in alkali nitrate solution during impregnation (0.04). The catalyst taken from the pilot reactor showed a slightly higher concentration of metals (excepting Mn). This effect could be caused by additional removing of possible rests of anions present in the calcination product. Moreover, the fresh catalyst contained small amount of graphite (about 3 wt %), which was added to the milled calcination product to avoid problems with tools of a pelletization machine; according to the XRD and XPS results, graphite was entirely removed during the catalytic measurements. The molar ratio K/Co found in the used catalyst (0.038) is somewhat lower than that in the fresh catalyst. The finding can also indicate that losses of potassium from the catalyst surface during its exposition to the relatively high reaction temperatures are negligibly small. Powder XRD patterns of the fresh and used K doped Co− Mn−Al mixed oxide catalysts are compared in Figure 2. The

mixed oxide with spinel structure (denoted as S in Figure 2) was found in both samples. The sharp diffraction line at 31° 2θ corresponding to graphite (denoted as C in Figure 2) was clearly apparent in the powder XRD pattern of the fresh catalyst. This line was not observed in the used catalysts, confirming combustion of graphite during the catalytic test. Practically no changes in positions and intensities of spinel diffraction lines were observed when powder XRD patterns of the fresh and used catalysts were compared. The spinel lattice parameters of 8.092 and 8.081 Å were evaluated for the fresh and used K doped Co−Mn−Al mixed oxide samples, respectively. The slight decrease in the spinel lattice parameter could be connected with a subtle structural ordering of the nonstoichiometric Co−Mn−Al mixed oxide, obtained after thermal decomposition of the LDH precursor, during the longterm exposition of the catalyst to reaction temperature in the catalytic reactor. The crystallite sizes evaluated from powder XRD data using the Scherrer equation showed no significant change as well (the values of 57 and 61 Å were found for the fresh and used catalysts, respectively). Thus, from the point of view of the phase composition, the catalyst can be considered under tested conditions as stable. Characteristic values of porous structure are summarized in Table 2. Specific surface area of the fresh K doped Co−Mn−Al mixed oxide tablets was 93 m2 g−1, total pore volume determined by mercury porosimetry 0.196 g cm−3, average pore radius 4.6 nm, ρHg = 2.30 g cm−3, and ρHe = 4.62 cm−3. After 112 days (2688 h) of work, surface area decreased only negligibly from 93 to 87 m2 g−1. Only slight changes were also observed in pores characterization. Pore size distribution vs pore radius can be seen in Figures 3 and 4. In the fresh catalyst,

Figure 3. Mesopore size distribution of the fresh and used K-doped Co−Mn−Al mixed oxide catalyst determined from desorption branch of N2 physical adsorption isotherm.

no pores with radius higher than 100 nm were found, while in the used catalyst such pores were identified. The findings are in line with the revealed increase in mercury density (from 2.30 to 2.36 g cm−3) confirming mild extinguishing of small pores in the catalyst. The increase in helium density from 4.62 to 4.98 g

Figure 2. XRD patterns of the fresh and used K-doped Co−Mn−Al mixed oxide catalyst: C, graphite; S, spinel phase.

Table 2. Textural Parameters and Reducibility of the Fresh and Used K Doped Co−Mn−Al Mixed Oxide Catalysts

a

catalyst

SBET (m2 g−1)

Vmeso (cm3 g−1)

Vtot (cm3 g−1)

ρHg (g cm−3)

ρHe (g cm−3)

porosity

Ra (nm)

TPRb (mmol H2/g)

fresh usedc

93 87

0.213d 0.205d

0.196 0.219

2.30 2.36

4.62 4.98

0.48 0.54

4.6 4.7

5. 16 5.73

Average pore radius R = 2Vmeso/SBET b25−500 °C. cFor 112 days. dSingle point adsorption total pore volume. D


Article


Co0 and to reduction of MnIV to MnIII oxides.12,15,25,26 The presence of potassium in the mixed oxides provoked growing of an easily reducible peak (Tmax of about 307 °C). This reduction peak was particularly distinct in the samples containing higher amount of potassium,13,15,26 and it is tempting to ascribe it to the reduction of CoIII. A shoulder at 200 °C could be evolved by reduction of compounds originated during calcination from KNO3, e.g., KNO2. Formation of KO2 or K2O2 cannot also be excluded. The reduction pattern of the spent catalyst is almost the same as that of the fresh catalyst. Only very slight decrease in the first shoulder intensity and a slight shift of the second shoulder Tmax to higher temperatures can be noticed together with narrower form of the main peak which could indicate formation of the more uniform particles after 112 days. The findings also confirm the quantitative data shown in Table 2. The amount of reducible compounds in the range 25−500 °C decreased after using the catalyst in the pilot reactor only by 6%, which means that no fundamental changes were recognized in the reducibility of the fresh and used catalyst. X-ray photoelectron spectroscopy (XPS) allows gaining data about surface composition of samples and chemical state of elements in a near-surface region. Both fresh, industrially manufactured K doped Co−Mn−Al mixed oxide catalyst and catalyst used in the pilot plant reactor presented in this work did not show any significant changes in the shape and width of the K 2p, Mn 2p, Al 2p, and Co 2p photoelectron spectra in comparison with the spectra of laboratory prepared sample published in our recent work27 (not shown). The binding energies of the photoemission lines of the fresh and used catalysts (Table 3) confirmed in both samples the oxidation states of surface manganese (Mn4+, Mn3+) and cobalt (Co3+, Co2+). Quantitative data about surface element composition determined from the XPS measurements indicated difference between surface and bulk concentrations of the element constituents in fresh catalyst (Table 1). The surface of catalyst particles was substantially enriched in Al; surface concentrations of both Co and Mn are roughly by 60% lower than their bulk concentrations. The substantial increase in the surface concentration of potassium compared to the bulk was found. The Co/Mn ratio changed pointing to specific surface enrichment in manganese content at the expense of cobalt. The findings are likely connected with the mobility of the metal ions, roughly corresponding to melting points of metals. When we analyzed data in both tables presenting the XPS results (Tables 1 and 3), we can conclude that the surface of K promoted Co−Mn−Al mixed oxide changed during the catalytic pilot plant tests. Reduction of Co3+ to Co2+ and manganese cations is observed. The reduction of Co is mainly manifested by changes in peak areas related to both kinds of Co ions (from 0.98 to 1.46), whereas the reduction of Mn is manifested by decrease in Mn 2p3/2−O 1s distance. Moreover, shifts of Mn 2p and Co 2p BE to the lower values supported this fact. The changes in the surface concentrations caused by the catalyst at 112 days of use in the pilot reactor were also observed. The surface K concentration increased more than

Figure 4. Pore size distribution of the fresh and used K-doped Co− Mn−Al mixed oxide catalyst determined from mercury porosimetry.

cm−3 was connected with removal of rest of anions from the calcination product and burning off the graphite added to the pelletized catalysts. The small discrepancies in the total and mesopore volumes of the fresh catalyst (Vtot < Vmeso) can be caused by the inhomogenity of the tablets, from the point of view of the tablets density, used for physisorption measurements. In any case, the mentioned changes in porous structure between fresh and used catalysts were within experimental error (10%). Temperature-programmed reduction patterns of both catalysts, i.e., before and after using the catalyst in the pilot plant reactor, are shown in Figure 5. The reduction pattern of

Figure 5. TPR-H2 patterns of the fresh and used K-doped Co−Mn−Al mixed oxide catalyst.

the fresh catalyst showed a main broad peak with Tmax of 422 °C. The peak was accompanied by two shoulders appearing at 200 and 307 °C. Reduction of the catalyst was intentionally finished at 500 °C, in order to protect the mass spectrometer detector from its contamination by alkali metal. The main reduction peak can be ascribed to reduction of CoIII → CoII →

Table 3. Binding Energies of the Selected Photoemission Lines of the Fresh and Used K Doped Co−Mn−Al Mixed Oxide Catalysts catalyst fresh used

K 2p3/2 293.5 292.9

296.3 295.7

Co 2p3/2 780.2 779.5

781.6 781.1

Co 2p3/2−1/2

Co2+/Co3+

15.3 15.3

0.98 1.46 E

Mn 2p3/2 641.6 641.0

643.2 642.5

O 1s 530.2 530.1

532.2 532.0

Al 2p

Mn 2p3/2−O 1s

73.9 73.3

111.8 111.3


Article

Industrial & Engineering Chemistry Research twice its initial value. Only slightly higher surface concentration of Mn at the expense of cobalt was observed, while enrichment of the surface by Al was not practically changed during the catalyst work. The increase in K surface concentration during pilot plant experiment could be the reason for observed BE shifts of all components to lower values. The stability of alkali promoter was studied by SR-TAD of fresh catalyst (Figure 6). From the temperature changes of the

Figure 7. Dependence of N2O conversion on space time (points, experimental data; dashed line, model) on K-doped Co−Mn−Al mixed oxide catalyst (512 ± 135 ppm of N2O, 26 ± 16 ppm of NOx, 9 ± 3 ppm of NH3, p = 0.6 MPa).

performed in recent years demonstrated that N2O decomposition is inhibited by nitrogen oxides (NO and NO2), oxygen (up to 3 mol % O2, at higher content almost no effect was observed23), and water vapor,15 so each variation of the inhibiting compounds present in the feed gas directly influenced N2O decomposition efficiency and makes the evaluation of such data more complicated. Variation in NOx concentration was really observed during online NOx analysis at the catalyst bed inlet; oxygen and water vapor contents were not measured. However, it is generally known that they are present in such a tail gas. The influence of NH3 in the feed gas was studied in the individual experiment performed in the laboratory conditions over K doped Co−Mn−Al mixed oxide tablets. The experiment proceeded in the model gas mixture containing 1000 ppm of NH3 + 4 mol % O2 + 3 mol % H2O at VHSV = 50 mL min−1 mLcat−1 (not shown). It was found that ammonia oxidized into N2O, NO, NO2, and probably N2 (not measured) during reaction, which means that also ammonia can contribute to the fluctuations of NOx concentration present in the feed or increase of N2O concentration at the outlet. However, in the range of the extent of the inlet fluctuations of NOx concentrations, there was seen no measurable origin of NOx during N2O decomposition over the studied catalyst and catalytic decomposition can be considered as selective process. The N2O catalytic decomposition was described by the firstorder kinetic equation at given process conditions and values of Ea = 97776 ± 4554 J mol−1 and k0 = 3.17 × 108 kg kgcat−1 h−1 were determined (Figure 8) by simple linear regression analysis with least-squares approach with determination coefficient of 0.85. The presented temperature dependence of N 2O conversion on the space time, shown in Figure 7, could be used for estimation of the amount of the catalyst necessary for obtaining required N2O conversion in the individual plant. Long-term stability of K doped Co−Mn−Al mixed oxide catalyst during the pilot plant testing is shown in Figure 9. Very slow increase in N2O conversion can be distinguished, and average value of 90 ± 6% at 450 °C was determined. This

Figure 6. Potassium desorption flux as a function of temperature (SRTAD) from fresh calcined K-doped Co−Mn−Al mixed oxide catalyst.

atomic and ionic fluxes of potassium it is clearly seen that catalyst is stable up to 500 °C with respect to alkali loss, and in spite of the mobility of potassium discernible from the XPS results, there is no detectable desorption of alkali from the catalyst surface in the temperature region of the operation conditions of the deN2O unit. However, at higher temperatures, desorption of alkali from the samples during catalytic reaction is always possible and the overheating of the catalyst should be avoid. The domination of the atomic flux in the desorption process indicates that the work function of the catalyst is relatively low, which is beneficial for the N2O decomposition process, as shown for similar catalytic system elsewehere.16,17 3.2. N2O Catalytic Decomposition in Pilot Plant Reactor. Results of kinetic experiment, when the inlet total gas flow varied between 300 and 600 kg h−1, are shown in Figure 7. We obtained a big set of data from the pilot plant measurements covering all the variety of temperatures but also all the variety of the N2O concentrations as well as other compounds with different inhibiting effects like NOx and NH3. At first, data with higher NOx and/or NH3 concentrations (NOx > 70 ppm, NH3 > 30 ppm) were omitted, since changes in these concentrations could influence obtained N2O conversions as will be explained further. Afterward, we have tried to avoid outlaying data from the point of view of the measured temperature; however this elimination of selected data did not lower the scattering of the obtained N2O conversions. For that reason we decide to use all of them as is in Figure 7, in spite of broad range of temperatures accuracy, since the scattering is probably caused by the fluctuations of other compounds present in the feed gas and cover the effect of the temperature fluctuations. Our laboratory experiments F


Article


Figure 8. Arrhenius plot: N2O decomposition on K-doped Co−Mn− Al mixed oxide catalyst in the pilot plant test (400−700 ppm of N2O, 26 ± 16 ppm of NOx, 9 ± 3 ppm of NH3, p = 0.6 MPa).

Figure 10. Temperature dependence of N2O conversion over Kdoped Co−Mn−Al mixed oxide catalysts in the laboratory test (GHSV = 3000 m3 mbed−3 h−1, 1000 ppm of N2O, 5 mol % O2, 2 mol % H2O, 200 ppm of NO).

gas with total gas flow of 30 000 m3/h (Table 4). The obtained results seem to be very good; the catalyst bed height of 1.3 m is Table 4. Inlet and Calculated Parameters for N2O Abatement in Waste Gas from HNO3 Production inlet parameters pressure temperature volume flow

600 000 Pa 450 °C 30 000 m3/h (NTP)

N2O concentration kinetic constant

0.07 mol %

a

Figure 9. Time on stream pilot plant test (GHSV = 11 000 m3 mbed−3 h−1, p = 0.6 MPa).

b

increase could be connected with changes of surface composition observed by XPS or small catalyst attrition. The activity of deN2O catalyst taken out from the reactor after 112 days was also controlled in the laboratory test. The results are shown in Figure 10, and these are in accordance with pilot plant results and confirm successful scale up and stability of catalyst performance. 3.3. Modeling of Full Scale Reactor. Evaluated kinetic parameters determined from experiments over the catalyst tablets in the pilot plant reactor describe the reaction rate including internal diffusion hindering effect. At the assumption that these data are not influenced by macrokinetic phenomena of the pilot plant reactor, since it was supposed that influence of external diffusion on the reaction rate was covered by higher internal diffusion resistance,28 they were directly used for reactor scale-up. By simultaneous solving of eqs 2−5, K doped Co−Mn−Al mixed oxide catalyst amount of 3200 kg was determined as necessary for 90% N2O conversion (450 °C, 0.6 MPa) in waste

27.4 kg kgcat−1 h−1 (first-order)

calculated parametersa,b catalyst weight catalyst bed high catalyst bed volume porosity of catalyst bed pressure drop

3200 kg 1.3 m 2.3 m3 0.7 10 kPa

Calculated for reactor diameter of 1.5 m and 90% N2O conversion. Catalyst tablets of 5 mm × 5 mm.

necessary for the reactor with given diameter of 1.5 m leading to the catalyst bed volume of 2.3 m3. The amount of the N2O decomposed per year is 328 100 kg from 364 240 kg of N2O released from such a plant in the case without purification technology. Implementation of this technology to the given nitric acid plant would lead to the green gas emissions saving of 98 t year−1 of CO2 equivalent. However, it is very difficult to compare the obtained results with other commercial technologies, since there are only limited amounts of results of pilot or full scale plant tests presented in the scientific literature.1,3−5 To the best of our knowledge, the only results that can be used for comparison are those from the group of Inger,1 where 70% N2O conversion was achieved in pilot plant operations over multicomponent cobalt spinel catalyst at 400 °C, GHSV ≈ 6000 m3 mbed−3 h−1 and p > 5 bar, since the mostly published process for tertiary N2O abatement is the EnviNOX-N2O, whose difference is in the need of reducing agent unlike our study. G


Article


4. CONCLUSIONS The preparation of a multicomponent K-doped Co−Mn−Al mixed oxide catalyst for low-temperature N2O decomposition was successfully reproduced in pilot plant scale. The pilot plant catalytic measurement of N2O decomposition was performed in a fixed bed reactor connected at the bypassed tail gas from the nitric production plant downstream the SCR NOx/NH3 catalyst. High performance in the N2O removal was reached. The changes in the surface composition, caused by the time-onstream operation of the catalyst for 112 days in the pilot reactor did not negatively affect the catalytic performance. The obtained experimental data were evaluated by applying the first-order kinetic equation and using it for calculation of fullscale reactor parameters. The presented data can be used to assess the necessary amount of a real catalyst for N2O emission lowering in different plants operating at specific conditions. The obtained results confirmed the successful scale-up of the deN2O mixed oxide catalyst suitable for low temperature application, thus opening the possibility of the implementation of this technology to real nitric acid plants.

■

Lbed Lp M n ṁ p, p0, p0 Peax = [(3 × 107)/Re21 + 1.35/Re1/8]−1 r Re = vdpρ/η T, T0 v XA

AUTHOR INFORMATION

Corresponding Author

ε

*Phone: +420 596997327. E-mail: e-mail:katerina.pacultova@ vsb.cz.

ρ

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

ρbed ρHg

Notes

The authors declare no competing financial interest.

ρHe

ACKNOWLEDGMENTS Financial support of the Technology Agency of the Czech Republic (Project TA 01020336), Czech Science Foundation (Project P106/14-13750S), and EU structural funding Operational Programme Research and Development for Innovation Project CZ.1.05/2.1.00/19.0388 is gratefully acknowledged.

η

■ ■

Bo = PeaxLbed/dp cA cA0 cD dp Ea k k* = k[p0M/(RT0)]

k0

REFERENCES

(1) Inger, M.; Wilk, M.; Saramok, M.; Grzybek, G.; Grodzka, A.; Stelmachowski, P.; Makowski, W.; Kotarba, A.; Sojka, Z. Cobalt Spinel Catalyst for N2O Abatement in the Pilot Plant OperationLongTerm Activity and Stability in Tail Gases. Ind. Eng. Chem. Res. 2014, 53, 10335−10342. (2) Inger, M.; Kowalik, P.; Saramok, M.; Wilk, M.; Stelmachowski, P.; Maniak, G.; Granger, P.; Kotarba, A.; Sojka, Z. Laboratory and pilot scale synthesis, characterization and reactivity of multicomponent cobalt spinel catalyst for low temperature removal of N2O from nitric acid plant tail gases. Catal. Today 2011, 176, 365−368. (3) Hoang, M.; Hill, A. J. Industrial Emission Treatment Technologies. In Municipal and Industrial Waste Disposal; InTech, 2012; DOI: 10.5772/29773; http://www.intechopen.com/. (4) Groves, M. C. E.; Sasonow, A. Uhde EnviNOx® technology for NOX and N2O abatement: a contribution to reducing emissions from nitric acid plants. J. Integr. Environ. Sci. 2010, 7, 211−222. (5) Lee, S.-J.; Ryu, I.-S.; Kim, B.-M.; Moon, S.-H. A review of the current application of N2O emission reduction in CDM projects. Int. J. Greenhouse Gas Control 2011, 5, 167−176. (6) Kannan, S.; Swamy, C. S. Catalytic decomposition of nitrous oxide over calcined cobalt aluminum hydrotalcites. Catal. Today 1999, 53, 725−737. (7) Pérez-Ramírez, J.; Kapteijn, F.; Moulijn, J. High activity and stability of the Rh-free Co-based ex-hydrotalcite containing Pd in the catalytic decomposition of N2O. Catal. Lett. 1999, 60, 133−138. (8) Alini, S.; Basile, F.; Bologna, A.; Montanari, T.; Vaccari, A. Preparation of stable catalysts for N2O decomposition under industrial

ABBREVIATIONS USED

A

kc

■

height of catalyst bed (m) catalyst particle height (m) molar weight of waste gas (kg mol−1) reaction order weight gas flow (kg h−1) pressure, normal pressure, pressure at the reactor inlet (Pa) Peclet number reaction rate (mol h−1 kg−1) Reynolds number thermodynamic temperature, normal temperature (K) superficial velocity (m s−1) Conversion of component A porosity of the catalyst bed density of gas (kg m−3) density of catalyst bed (kg m−3) apparent sample density (kg m−3) True sample density (kg m−3) dynamic viscosity of waste gas (Pa s)

reactor cross section area (m2) Bodenstein number concentration of component A (mol m−3) concentration of component A at the reactor inlet (mol m−3) resistance coefficient for Re/(1 − ε) < 500 catalyst particle diameter (m) activation energy (J mol−1) kinetic constant (m3 h−1 kg−1) kinetic constant (kg kgcat−1 h−1) mass transfer coefficient (m s−1) pre-exponential factor (kg kgcat−1 h−1) H


Article

Industrial & Engineering Chemistry Research conditions. In Studies in Surface Science and Catalysis; Gaigneaux, E., Poncelet, Eds.; Elsevier, 2000; Vol. 143, pp 131−139. (9) Chang, K. S.; Song, H.; Park, Y.-S.; Woo, J.-W. Analysis of N2O decomposition over fixed bed mixed metal oxide catalysts made from hydrotalcite-type precursors. Appl. Catal., A 2004, 273, 223−231. (10) Pérez-Ramírez, J.; Overeijnder, J.; Kapteijn, F.; Moulijn, J. A. Structural promotion and stabilizing effect of Mg in the catalytic decomposition of nitrous oxide over calcined hydrotalcite-like compounds. Appl. Catal., B 1999, 23, 59−72. (11) Pérez-Ramírez, J.; Mul, G.; Xu, X.; Kapteijn, F.; Moulijn, J. A. Co-based ex-HTlc for the decomposition of N2O: Tailoring catalysts for active and stable operation. In Studies in Surface Science and Catalysis; Corma, A., Luis, J., Eds.; Elsevier, 2000; Vol. 130, pp 1445− 1450. (12) Obalová, L.; Pacultová, K.; Balabánová, J.; Jirátová, K.; Bastl, Z.; Valásǩ ová, M.; Lacný, Z.; Kovanda, F. Effect of Mn/Al ratio in Co− Mn−Al mixed oxide catalysts prepared from hydrotalcite-like precursors on catalytic decomposition of N2O. Catal. Today 2007, 119, 233−238. (13) Cheng, H.; Huang, Y.; Wang, A.; Li, L.; Wang, X.; Zhang, T. N2O decomposition over K-promoted Co-Al catalysts prepared from hydrotalcite-like precursors. Appl. Catal., B 2009, 89, 391−397. (14) Li, Q.; Meng, M.; Tsubaki, N.; Li, X.; Li, Z.; Xie, Y.; Hu, T.; Zhang, J. Performance of K-promoted hydrotalcite-derived CoMgAlO catalysts used for soot combustion, NOx storage and simultaneous soot−NOx removal. Appl. Catal., B 2009, 91, 406−415. (15) Obalová, L.; Karásková, K.; Jirátová, K.; Kovanda, F. Effect of potassium in calcined Co−Mn−Al layered double hydroxide on the catalytic decomposition of N2O. Appl. Catal., B 2009, 90, 132−140. (16) Zasada, F.; Stelmachowski, P.; Maniak, G.; Paul, J.-F.; Kotarba, A.; Sojka, Z. Potassium Promotion of Cobalt Spinel Catalyst for N2O DecompositionAccounted by Work Function Measurements and DFT Modelling. Catal. Lett. 2009, 127, 126−131. (17) Obalová, L.; Maniak, G.; Karásková, K.; Kovanda, F.; Kotarba, A. Electronic nature of potassium promotion effect in Co−Mn−Al mixed oxide on the catalytic decomposition of N2O. Catal. Commun. 2011, 12, 1055−1058. (18) Obalová, L.; Jirátová, K.; Karásková, K.; Kovanda, F. Simulation of N2O Abatement in Waste Gases by Its Decomposition over a KPromoted Co-Mn-Al Mixed Oxide Catalyst. Chin. J. Catal. 2011, 32, 816−820. (19) Obalová, L.; Jirátová, K.; Karásková, K.; Chromčaḱ ová, Ž . N2O catalytic decompositionFrom laboratory experiment to industry reactor. Catal. Today 2012, 191, 116−120. (20) Lecloux, A.; Pirard, J. P. The importance of standard isotherms in the analysis of adsorption isotherms for determining the porous texture of solids. J. Colloid Interface Sci. 1979, 70, 265−281. (21) Sing, K. S. W.; Williams, R. T. Physisorption Hysteresis Loops and the Characterization of Nanoporous Materials. Adsorpt. Sci. Technol. 2004, 22, 773−782. (22) Fogler, H. S. Elements of Chemical Reaction Engineering, 3rd ed.; Prentice Hall PTR: Upper Saddle River, NJ, 1999. (23) Obalová, L.; Fíla, V. Kinetic analysis of N2O decomposition over calcined hydrotalcites. Appl. Catal., B 2007, 70, 353−359. (24) Pachulski, A.; Schödel, R.; Claus, P. Kinetics and reactor modeling of a Pd-Ag/Al2O3 catalyst during selective hydrogenation of ethyne. Appl. Catal., A 2012, 445−446, 107−120. (25) Obalová, L.; Jirátová, K.; Kovanda, F.; Pacultová, K.; Lacný, Z.; Mikulová, Z. Catalytic decomposition of nitrous oxide over catalysts prepared from Co/Mg-Mn/Al hydrotalcite-like compounds. Appl. Catal., B 2005, 60, 289−297. (26) Jirátová, K.; Mikulová, J.; Klempa, J.; Grygar, T.; Bastl, Z.; Kovanda, F. Modification of Co−Mn−Al mixed oxide with potassium and its effect on deep oxidation of VOC. Appl. Catal., A 2009, 361, 106−116. (27) Obalová, L.; Karásková, K.; Wach, A.; Kustrowski, P.; Mamulová-Kutláková, K.; Michalik, S.; Jirátová, K. Alkali metals as promoters in Co−Mn−Al mixed oxide for N2O decomposition. Appl. Catal., A 2013, 462−463, 227−235.

(28) Klyushina, A.; Pacultová, K.; Obalová, L. Advantage of the single pellet string reactor for testing real-size industrial pellets of potassiumdoped CoMnAl catalyst for the decomposition of N2O. React. Kinet., Mech. Catal. 2015, 115, 651−662.

I


K-Doped CoâMnâAl Mixed Oxide Catalyst for N2O Abatement from

Recommend Documents

K-Doped CoâMnâAl Mixed Oxide Catalyst for N2O Abatement from