Dynamics of NO and N2O decomposition over Cu-ZSM-5 under transient reducing and oxidizing conditions

Journal of Catalysis 196, 271–286 (2000)

doi:10.1006/jcat.2000.3046, available online at http://ww w .idealibrar y .com on

Dynamics of NO and N₂O Decomposition over Cu–ZSM-5 under

Transient Reducing and Oxidizing Conditions

Mahesh V. Konduru¹and Steven S. C. Chuang²

Department of Chemical Engineering, The University of Akron, Akron, Ohio 44325-3906

Received March 17, 2000; revised August 28, 2000; accepted August 29, 2000

been extended to N₂O decomposition due to the fact that

nitrous oxide (N₂O) is not only a by-product of NO decom-

position and reduction but also an effective greenhouse gas

and ozone-destructing reagent (4–16). Cu–ZSM-5 has been

known to exhibit the highest activity for NO decomposition

and also high N₂O decomposition activity (17–30). Despite

its effectiveness for the NO and N₂O decomposition reac-

tion, Cu–ZSM-5 is not resistant to oxidants, water, or sulfur

(27, 28). In the present study, we have focused on investi-

gating the mechanisms of NO and N₂O decomposition on

Cu–ZSM-5 so that the knowledge of active adsorbates and

sites can be utilized to design and synthesize novel catalysts

that would exhibit not only high activity but also deactiva-

tion resistance.

N₂O and NO decomposition pathways on Cu–ZSM-5 have been

investigated by monitoring the adsorbate dynamics and changes in

reactant and product concentrations using infrared spectroscopy

(IR) and mass spectrometry (MS) under transient reducing and

oxidizing conditions. Transient reducing and oxidizing conditions

were produced by the H₂, CO, and O₂pulses into either N₂O or NO

streams. IR and MS studies under the transient conditions revealed

that adsorbed O produced from N₂O exhibited different reactivity

and dynamics from the adsorbed O produced during NO decompo-

sition. The differences in reactivity and dynamics of the adsorbed O

were evidenced by the fact that (i) adsorbed O from N₂O interacted

with CO/H₂to produce two humps in the H₂O and O₂concentration

proﬁles; adsorbed O from NO reacted with CO/H₂that led to only

one hump in the H₂O and O₂concentration proﬁle, and (ii) addi-

tion of the O₂pulse led to reaction of O₂with adsorbed O from NO,

resulting in oxidation of Cu⁺in Cu⁺(NO) to Cu²⁺in Cu²⁺(NO₃);

addition of O₂did not lead to any reaction with adsorbed O from

N₂O decomposition. N₂O decomposition is proposed to proceed via

Cu⁺–ON₂, Cu²⁺O , and Cu²⁺O –ON₂with Cu⁺–ON₂serving as a

Extensive infrared (IR) studies have reported

that Cu⁺(NO) (1814 cm ¹), Cu²⁺(NO) (1910 cm ¹),

Cu²⁺O (NO) (1890–1905 cm ¹), NO⁺(2123 cm ¹), and

bridging and chelating Cu²⁺(NO₃) (1623, 1610, and

1565 cm ¹) are the major adsorbates observed during NO

decomposition at 77–673 K (21–32). Our recent transient,

site-poisoning, and site-promotion studies have revealed

that (i) NO⁺is a spectator, (ii) N₂and O₂formation

proceeds through separate rate-limiting steps, and (iii)

inhibition of adsorbed O migration is the major deactiva-

tion pathway for O₂formation (32). These studies further

suggested that Cu⁺is the active site and that Cu⁺(NO)

precursor for N₂formation and Cu²⁺

O as a precursor for O₂for-

mation. NO decomposition proceeds via Cu⁺(NO), Cu²⁺O , and

Cu²⁺(NO₃) with Cu⁺(NO) serving as a precursor for NO dissocia-

tion. Cu⁺in Cu⁺(NO) is different from that of Cu⁺in Cu⁺–ON₂.

The former may be associated with Al(OH)₄of the zeolite, the latter

c

with Si(OH)₄.

2000 Academic Press

Key Words: N₂O decomposition; NO decomposition; Cu–ZSM-5;

in situ IR; temperature-programmed reaction; step switch; pulse

switch; adsorbate dynamics; adsorbate reactivity; transient reduc- and Cu²⁺(NO₃) are the active adsorbates that participate

ing conditions; transient oxidizing conditions; reaction mechanism;

adsorbed O; active sites; site environment.

in the catalytic cycle of NO decomposition.

In contrast to extensive studies of NO decomposition,

very few in situ IR studies on N₂O decomposition over

Cu–ZSM-5 have been reported. The N₂O decomposition

mechanism and its relation to NO decomposition have not

been established unambiguously. The current study investi-

gates the mechanisms of both NO and N₂O decomposition

over Cu–ZSM-5 by probing the adsorbate reactivity and

dynamics as well as the active sites with transient pulsing

of H₂, O₂, and CO into the reaction systems. Since both

NO and N₂O decomposition reactions are redox reactions

that may involve both Cu⁺and Cu²⁺sites on Cu–ZSM-

5, probing the reactions with a reductant (i.e., H₂or CO)

could lead to the reduction of Cu²⁺to Cu⁺; probing with

1. INTRODUCTION

Catalytic nitric oxide (NO) decomposition has been a

subject of extensive studies due to its promise for the ef-

fective control of NO_xemission (1–3). The attention has

¹Present address: Oak Ridge National Laboratory, Oak Ridge,

Tennes`see.

²To whom correspondence should be addressed. E-mail: schuang@

uakron.edu.

271

0021-9517/00 $35.00

c

Copyright

2000 by Academic Press

All rights of reproduction in any form reserved.

272

KONDURU AND CHUANG

an oxidant (O₂) could result in the oxidation of Cu⁺to 2.3. Temperature-Programmed Reaction Studies

Cu²⁺, creating a perturbation in the adsorbate concentra-

Temperature-programmed reaction (TPR) studies in-

tion and in the product formation processes. In addition,

volved the exposure of Cu–ZSM-5-523 to a 5% NO in He

the reductants and oxidant are also expected to react with

stream (NO = 2.5 cm³/min and He = 47.5 cm³/min) at 373 K

some of the adsorbates. Changes in adsorbate as well as re-

followed by heating the reactor from 373 to 743 K at a heat-

actant/product concentrations were monitored by IR and

ing rate of 10 K/min. The Cu–ZSM-5-523 catalyst, active

mass spectrometry (MS), respectively. This study revealed

for pulse NO and N₂O decomposition, was also exposed

the nature of catalytic cycles for N₂O and NO decompo-

to a 5% N₂O in He stream (N₂O = 2.5 cm³/min and He =

sition, distinguished the adsorbates and sites that are ac-

47.5 cm³/min) at 333 K and was subjected to a temperature

tive for the N₂O and NO decomposition reactions, and ex-

increase (TPR) from 333 to 773 K at 10 K/min. The FTIR

amined the role of oxidizing and reducing reagents in the

and MS monitored the changes in the adsorbate and

reactions.

gaseous efﬂuent concentration during the TPR studies,

respectively.

2. EXPERIMENTAL

2.1. Catalyst Preparation

Prior to the TPR studies, IR spectra of each of the catalyst

samples were collected in ﬂowing He at each speciﬁc tem-

perature. These spectra were denoted as the background

spectra. The background spectrum at the speciﬁc tempera-

ture was subtracted from the IR spectrum collected at the

corresponding temperature during the TPR studies. The re-

sulting IR spectra, shown in this study, hence do not have

any contribution from the characteristic vibrations of the

catalyst sample or from background spectrum changing

with temperature. The background spectra and the spectra

obtained during the TPR studies were taken by co-adding

32 scans, which were collected in 4 s. The spectrum at each

temperature during the TPR was collected within a 1 K

variation in the temperature. Hence, the co-added spectrum

collected during the TPR represents the average change in

adsorbate intensity during a 1 K increase in temperature.

Cu–ZSM-5-523, with a 523% copper exchange, was pre-

pared by the repeated ion exchange of Cu–ZSM-5-83 (83%

copper exchange), which was prepared by Johnson Matthey

and supplied by the Catalyst bank of Sandia National Lab-

oratories. The percentage copper exchange is deﬁned as the

molar ratio of the amount of Cu to that of Al multiplied by 2

[% Cu exchange = 2 (mol of Cu/mol of Al)] (27). Cu–ZSM-

5-83 was added to a 0.004 M copper acetate solution at a pH

of 7 and stirred for 24 h. The solution was ﬁltered and the

solid deposited on the ﬁlter paper was washed with distilled

water and dried in air at 373 K for 24 h. Inductively coupled

plasma (ICP) emission spectroscopy (Galbraith Laborato-

ries, Knoxville, TN) determined Si/Al to be 25 and Cu/Al

to be 2.615 in Cu–ZSM-5-523.

2.4. Transient Reduction and Oxidation Pulse Studies

2.2. Reactor System

One cubic centimeter (1 cm³) of H₂, O₂, and CO was

The catalyst, pressed in the form of a self-supporting thin separately pulsed with a 6-port gas chromatograph (GC)

pellet of 15–25 mg, was placed in the in situ IR reactor injection valve into a steady-state 5% NO in He ﬂow over

cell directly in the path of the IR beam (33). Three addi- Cu–ZSM-5-523 at 673 and 723 K. Introduction of the H₂

tional catalyst pellets were crushed into ﬂakes and placed and CO (well-known reductants) pulses allowed Cu–ZSM-

in the reactor outlet that was in the close vicinity of the self- 5-523 to be exposed to a reducing environment for a short

supporting pellet. The total amount of catalyst placed in the period of time while the O₂pulse exposed the catalyst to

IR cell was 105–110mg. A Magna (Nicolet Instruments) 550 an oxidizing environment for a short period of time under

Fourier transform IR spectrometer (FTIR) recorded the IR NO reaction conditions. The H₂, O₂, and CO pulses were

spectra and a Pfeiffer Prisma QMG 200 (Pfeiffer Vacuum also injected into a 5% N₂O in He ﬂow over Cu–ZSM-5-

Technology) MS recorded the changes in the concentration 523 at 673 and 723 K. The H₂, O₂, and CO pulse into NO

of the reactants and products during the study. No spec- and N₂O studies involved the introduction of three sep-

1

tra below 1400 cm are presented in this study because arate 1 cm³of H₂/O₂/CO pulses into the 5% NO in He

below this value a low signal-to-noise ratio was observed; and 5% N₂O in He streams at 673 K over Cu–ZSM-5-

additionally, the use of CaF₂rods and windows cut off IR 523. Each of the three pulses resulted in similar changes

transmission below 1000 cm ¹. The spectral change in the in the adsorbate intensity and MS proﬁles of gaseous re-

900 to 1000-cm ¹range associated with extralattic oxygen actants and products, indicating good reproducibility. The

can only be observed with a very thin Cu–ZSM-5 disk di- IR and MS monitored the changes in the adsorbate and

luted with KBr. The possible complication of KBr and the reactor efﬂuent concentrations during the H₂, O₂, and CO

need to use costly H₂O-resistant and high-temperature IR pulse study under NO and N₂O reaction conditions, respec-

transmission windows precluded investigation of dynamic tively. Along with the IR spectra and MS proﬁles, the vari-

behavior of the extralattic oxygen.

ation in the peak intensity (peak height) of the adsorbate

NO AND N₂O DECOMPOSITION ON Cu–ZSM-5

273

species was also tabulated and displayed with respect to

time.

2.5. NO/N₂O Co-Adsorption and Reaction Studies

NO/N₂O co-adsorption and reaction studies were carried

out by pulsing 1 cm³of N₂O into 5% NO over autoreduced

Cu–ZSM-5-523 at 673 and 723 K and pulsing 1 cm³of NO

into 5% N₂O over Cu–ZSM-5-523 at 323, 673, and 723 K.

Introducing a small amount of N₂O into NO ﬂow and a

small quantity of NO into N₂O ﬂow is expected to result in

the perturbation of the sites and the adsorbates present on

the sites. Monitoring the adsorbate dynamics and concen-

tration proﬁles of reactants and products during these pulse

studies would shed light on the role of N₂O in the NO de-

composition pathway and also aid in determining whether

NO and N₂O decomposition on Cu–ZSM-5 occur on the

same sites.

2.6. Fragmentation of NO, N₂, CO, CO₂, N₂O, and NO₂

The MS proﬁles of the gaseous species were distinguished

from each other by the m/e ratios: m/e = 30 for NO, m/e =

28 for N₂, m/e = 32 for O₂, m/e = 44 for N₂O, m/e = 46 for

NO₂, m/e = 12 for CO, m/e = 22 for CO₂, m/e = 18 for H₂O,

and m/e = 17 for NH₃. The contributions of the N₂O frag-

ment to NO (m/e = 30) and N₂(m/e = 28) proﬁles, of NO₂

to NO (m/e = 30), and ofH₂O (m/e = 18) to NH₃(m/e = 17)

were accounted for in the present study. Calibration factors

for each gaseous species, except H₂O and NH₃, were ob-

tained by pulsing known amounts of the gases into the MS.

The calculated calibration factors were utilized for convert-

ing the MS intensities of the reactants and products to their

respective molar rates.

FIG. 1. IR spectra and MS proﬁles collected during the temperature-

programmed reaction (TPR) over Cu–ZSM-5-523 in a 5% NO in He

stream from 373 to 753 K.

3. RESULTS

adsorbates were detected, indicating that the rate of des-

orption of adsorbates and the rate of product formation

from adsorbates are signiﬁcantly higher than the adsorbate

3.1. NO and N₂O TPR

Figure 1 shows the IR spectra and the change in the con- formation rate.

centrations of the reactor efﬂuent obtained during the 5%

The IR spectra and changes in the concentration of the

NO in He TPR from 373 to 773 K (10 K/min) over Cu–ZSM- reactor efﬂuent obtained during the 5% N₂O in He TPR

5-523. Flowing 5% NO at 373 K produced adsorbed N₂at over Cu–ZSM-5-523 from 333 to 773 K are shown in Fig. 2.

2158 cm ¹, Cu²⁺(NO) at 1900 cm ¹, bridging Cu²⁺(NO₃) at Exposure of Cu–ZSM-5-523 to 5% N₂O at 333 K led to the

1

1623and 1612cm ¹, and chelatingCu²⁺(NO₃) at 1567 cm

appearance of prominent IR bands at 2287 and 2240 cm

1

as shown in Fig. 1a. Increase in temperature from 373 to and a weak band at 1623 cm along with the gaseous

743 K led to changes in both the adsorbate and reactor ef- N₂O bands. The assignment of the adsorbate N₂O bands

ﬂuent concentrations: NO decomposition to N₂O and O₂in the 2200- to 2300-cm ¹region has not been established

commenced at 473 K (Fig. 1b) where Cu²⁺(NO) showed a unambiguously on Cu–ZSM-5 while the 1623-cm band

1

substantial drop in its IR intensity, Cu²⁺(NO₃) exhibited was previously assigned to the bridging Cu²⁺(NO₃) species

a slight decrease in its intensity, and Cu⁺(NO) emerged. based on the NO adsorption studies on Cu–ZSM-5 (31, 32).

The Cu⁺(NO) intensity paralleled the rate of N₂O forma- The present study attempts to assign bands in the 2200 to

tion with the depletion of Cu⁺(NO) coinciding with that of 2300-cm ¹region based on the obervations reported during

N₂O, suggesting that NO adsorbed on Cu⁺may serve as a N₂O adsorption studies on transition metal zeolites, metal

precursor for N₂O formation. Selective NO decomposition oxides, and metals. The high electron density at the ter-

to N₂and O₂occurred at 753 K where no IR observable minal N and O atoms allows N₂O to bond to the zeolites,

274

KONDURU AND CHUANG

1

the 2240-cm band to the O-bound N₂O on Cu⁺(Cu⁺–

ON₂). The corresponding N–O vibrations in the lower fre-

quency region of 1235–1310 cm ¹could not be observed in

the present study as they were overlapped by the presence

of gaseous N₂O bands in the same region and were low in

intensity when compared to the N–N vibrations.

A temperature increase from 333 to 423 K depleted

Cu⁺–ON₂at 2240 cm ¹and decreased Cu⁺–N₂O species at

2287 cm ¹and led to an increase in the bridging Cu²⁺(NO₃)

species at 1623 cm ¹while there was no change in the con-

centration of the gaseous reactants and products. An addi-

tional temperature increase led to the eventual depletion of

Cu⁺–N₂O at 573 K, the formation of adsorbed N₂at 473 K

and its depletion at 623 K, start of N₂O decomposition and

N₂formation at 573 K, and start of O₂formation at 643 K.

High rates of N₂O decomposition were achieved at 773 K

where no IR observable adsorbateswere detected. The N₂O

IR-TPR results have, for the ﬁrst time, showed that N₂O

adsorbs on Cu–ZSM-5 via both N and O as Cu⁺–N₂O and

Cu⁺–ON₂as well as a bridging Cu²⁺(NO₃) species at 333 K

and have also conﬁrmed that N₂O starts to decompose to

N₂and O₂at 573 K and reaches a conversion of 100% at

773 K over Cu–ZSM-5.

The TPR results show that Cu–ZSM-5-523 exhibits a

higher decomposition activity for N₂O than NO; N₂O de-

composition during TPR produces greater amounts of N₂

and O₂compared to NO decomposition in the same temper-

ature range; O₂desorption (523 K) starts at a lower temper-

ature during NO decomposition than that of N₂O (643 K).

The different O₂desorption temperatures suggest that dif-

ferent forms of adsorbed O were produced from NO and

N₂O adsorption and surface reactions.

FIG. 2. IR spectra and MS proﬁles collected during the temperature-

programmed reaction (TPR) over Cu–ZSM-5-523 in a 5% N₂O in He

stream from 333 to 773 K.

3.2. H₂Pulse Studies

metal oxides, and metals via either the N or O end. The

Figures 3 and 4 show the IR spectra, MS proﬁles of

1

band at 2224 cm corresponds to the N–N vibration; the gaseous reactants and products, and adsorbate peak inten-

1285-cm ¹band signiﬁes the N–O vibration for N₂O in the sity proﬁles collected during the ﬁrst 1 cm³of the H₂pulse

gas phase (34). N₂O adsorbed on ZnO and ZrO₂via the O into a 5% NO in He stream over Cu–ZSM-5-523 at 673 K. It

1

end, giving bands at 2237–2242 and 1233–1255 cm (34– should be noted that the IR spectra taken during the pulse

39), on the -chromia surface via N, giving bands at 2305 were collected every 3.6 s and the IR intensity proﬁles of

and 1339 cm ¹, and via O, leading to bands at 2238 and the adsorbates during the entire duration of the pulse are

1

1237 cm (40), and on NaA, CoA, NiA, and CuA via N, plotted in Fig. 4. The spectra shown in Fig. 3 do not rep-

producing bands at 2258–2272 and 1282–1304 cm ¹, and via resent all the collected spectra but illustrate the selected

O, giving bands at 2223–2250 and 1240–1282 cm ¹(41). It few that show the prominent changes during the H₂pulse

has also been reported that N₂O bonds to the Cu(I) cations switch. Exposure of Cu–ZSM-5-523 to 5% NO at 673 K,

1

on CuA via N, producingbandsat 2230–2250and 1289cm

,

prior to the H₂pulse, led to the formation of bridging and

and via O, leading to bands at 2223 and 1282 cm (41). chelating Cu²⁺(NO₃) at 1623 and 1567 cm ¹, Cu⁺(NO) at

1

Hence, as a general rule, bands observed in the 2224-, to 1814 cm ¹, Cu²⁺(NO) at 1907 cm ¹, and adsorbed NO⁺at

2260 and 1200- to 1285-cm (lower than 1285 cm ¹) re- 2123 cm as shown in Fig. 3 as well as produced a 8.7%

gions have been assigned to N₂O adsorbed via O while the NO conversion as illustrated in Table 1. Introduction of 1

bands in the 2224- to 2300 and 1285- to 1350-cm ¹regions cm³of H₂(a) decreased Cu⁺(NO), Cu²⁺(NO), bridging and

1

have been assigned to N₂O adsorbed via N. Based on these chelating Cu²⁺(NO₃), and adsorbed NO⁺, (b) increased

1

studies, the band observed in the present study at 2287 cm

NO conversion and N₂, O₂, and N₂O formation, and (c)

is assigned to the N-bound N₂O on Cu⁺(Cu⁺–N₂O) and led to the formation of H₂O (gas) at 3505 cm (42) and

1

NO AND N₂O DECOMPOSITION ON Cu–ZSM-5

275

FIG. 3. IR spectra taken during the ﬁrst 1 cm³of H₂pulse into a 5% NO stream over Cu–ZSM-5-523 at 673 K.

adsorbed NH₃species at 3360. The adsorbed NH₃species their respective original intensities as the pulsed H₂exited

at 3360 can be assigned to NH₃bound to Cu, based on the the reactor cell.

IR spectra of NH₃species bound to various Cu, Co, Pt,

The IR spectra in Fig. 3 and IR peak intensity proﬁles in

Mn, and Ni complexes, which give bands in the 3327- to Fig. 4d indicate a sharp increase and decrease of H₂O (gas)

1

3345-cm range (42). Figure 4 shows that the initial de- while the MS proﬁles (Fig. 4b) showed the gaseous H₂O

crease of all the NO adsorbate species and gaseous NO lagging behind the rest of the gaseous species signiﬁcantly.

was followed by an increase in Cu⁺(NO) and N₂, N₂O, and The lag in the MS proﬁle could be a result of the delay of

O₂formation during the continued decrease of Cu²⁺(NO), H₂O (gas) in entering the MS analysis chamber from the

bridging Cu²⁺(NO₃), and NO⁺. The increase in Cu⁺(NO), IR cell due to adsorption/desorption of H₂O on the walls

to an intensity higher than that prior to the H₂pulse, ac- of the reactor system lines entering the MS.

companied with a decrease in the intensity of the Cu²⁺

The role of H₂in NO decomposition can be summarized

species (Cu²⁺(NO) and Cu²⁺(NO₃)) reﬂects an increase in as causing the reduction of Cu²⁺(NO₃) and Cu²⁺(NO) to

the number of Cu⁺sites that may be promoting NO conver- Cu⁺, NH₃, and H₂O, promotion of O₂desorption and for-

sion and giving high rates of N₂, O₂, and N₂O formation. The mation, and an increase in the number of Cu sites, leading

adsorbate and reactant/product concentrations returned to to increases in the NO decomposition rate. The H₂pulse

results also reveal that the Cu²⁺for Cu²⁺(NO) is different

from that of Cu²⁺(NO₃) as they are affected to a different

extent by H₂addition. Cu⁺–NO parallels N₂O formation

TABLE 1

similar to that observed during NO TPR.

NO and N₂O Conversion Prior to 1 cm³of H₂, CO, and O₂Pulses

into 5% NO and 5% N₂O in He Flow over Autoreduced Cu–ZSM-

5-523 (W/F = 0.12 g s cm ³) at 673 K

The ﬁrst 1 cm³of H₂pulse into a 5% N₂O stream over

Cu–ZSM-5-523 led to changes in the IR intensities of the

adsorbate species as well as MS proﬁles of the gaseous re-

actants and products as shown in Figs. 5 and 6. Prior to

the H₂pulse, the 5% N₂O ﬂow over Cu–ZSM-5-523 pro-

duced gaseous N₂O at 2235 cm ¹and bridging and chelating

Cu²⁺(NO₃) at 1622 and 1560 cm ¹as shown in Fig. 5 as well

as led to a 9.6% N₂O conversion (Table 1). Pulsing 1 cm³of

H₂led to the complete depletion of bridging and chelating

% Conversion

H₂

CO

O₂

NO

N₂O

8.7

9.6

12.8

8.9

8.7

7.9

276

KONDURU AND CHUANG

observation of OH is a manifestation of the O ion, which

has been found to involve homolytic splitting of hydrogen

(45) and partial oxidation of methane (46). As the pulsed H₂

exited the reactor, the adsorbed OH species completely dis-

appeared while the bridging and chelating Cu²⁺(NO₃) and

all the gaseous reactants/products returned to their original

concentrations.

The most obvious role of H₂in both NO and N₂O decom-

position is the reduction of Cu²⁺to Cu⁺and promotion of

N₂formation. The distinction in the effect of H₂on the

product formation is the difference in the reactivity of ad-

sorbed oxygen produced from these two reactions. Oxygen

from NO tends to undergo desorption to produce gaseous

O₂while oxygen from N₂O tends to react with adsorbed H₂

to produce H₂O.

3.3. O₂Pulse Studies

Figures 7 and 8 show the IR spectra, change in concentra-

tion of reactants and products, and variation in adsorbate

peak intensity collected during the ﬁrst 1 cm³of O₂pulse

into a 5% NO in He stream over Cu–ZSM-5-523 at 673 K.

The adsorbate spectra prior to the O₂pulse were essentially

identical to those prior to the H₂pulse as shown in Fig. 3.

The O₂pulse (a) decreased Cu⁺(NO) and N₂O formation,

(b) increased bridging and chelating Cu²⁺(NO₃), NO

conversion, and NO₂formation, and (c) did not affect

Cu²⁺(NO), N₂formation, and NO⁺asshown in Figs. 7and 8.

These observations can be explained by oxidation of adsor-

bates and surface from Cu⁺to Cu²⁺due to addition of O₂.

A decrease in Cu⁺(NO) with N₂O further supports the pos-

tulation that Cu⁺(NO) could serve as a precursor for N₂O

formation. Furthermore, the mirror image intensity proﬁles

of Cu⁺(NO) and bridging and chelating Cu²⁺(NO₃) reﬂect

the oxidation of Cu⁺sites to Cu²⁺, leading to the conversion

of part of Cu⁺(NO) to bridging and chelating Cu²⁺(NO₃)

during O₂addition. The decrease in NO concentration and

increase of NO₂is a result of the reaction of NO with O₂to

produce NO₂as well as a consequence of the co-adsorption

of NO and O₂to produce the Cu²⁺(NO₃) species.

FIG. 4. MS intensity proﬁles of reactor efﬂuent and peak intensity

proﬁles of adsorbate species collected during the ﬁrst 1 cm³of H₂pulse

into a 5% NO stream over Cu–ZSM-5-523 at 673 K.

Cu²⁺(NO₃), the formation of adsorbed OH species at 3629

and 3583cm ¹(43, 44), an increase in NO conversion and N₂

formation, and a decrease in O₂formation. The adsorbed

OH species (3629 and 3583 cm ¹) reached their maxima

at two separate instances during the H₂pulse as shown in

Fig. 6e. Gaseous H₂O (Fig. 6b) also exhibited two humps

in the concentration proﬁles similar to the adsorbed OH

species. Careful examination of lead/lag relationships be-

tween Cu^{2 +}(NO₃), NO, O₂, and OH proﬁles suggests that

the ﬁrst OH peak in Fig. 6e may be due to the reaction of

Cu²⁺(NO₃) with adsorbed hydrogen and the second due

to the reaction of O produced from N₂O adsorption and

O present on ZSM-5 with adsorbed H₂. The presence of

H₂inhibits the recombination of the adsorbed O species to

In contrast to the pronounced effect of O₂on NO decom-

position, the O₂pulse into 5% N₂O at 673 K did not cause

any variation in adsorbate and product concentration. The

lack of changes in the adsorbate and product concentration

can be attributed to the lack of interaction between the ad-

sorbed O from O₂and the adsorbed O species produced

from N₂O decomposition.

3.4. CO Pulse Studies

Figures 9 and 10 show the IR spectra, MS proﬁles of re-

form O₂as seen by the decrease in O₂formation in Fig. 6a. actants and products, and the variation of adsorbate peak

The adsorbed O species instead react with H to adsorbed intensities with time collected during the ﬁrst 1 cm³of CO

OH species. The formation of adsorbed OH also prevents pulse into a 5% NO in He stream over Cu–ZSM-5-523 at

the re-oxidation of Cu⁺to Cu²⁺sites, inhibiting the forma- 673 K. The CO pulse into 5% NO produced changes in both

tion of the bridging and chelating Cu²⁺(NO₃) species. The the adsorbate and reactor efﬂuent concentrations. CO entry

NO AND N₂O DECOMPOSITION ON Cu–ZSM-5

277

FIG. 5. IR spectra taken during the ﬁrst 1 cm³of H₂pulse into a 5% N₂O stream over Cu–ZSM-5-523 at 673 K.

depleted bridging and chelating Cu²⁺(NO₃), increased NO the Cu²⁺sites to Cu⁺, which promotes the N₂O decompo-

conversion and N₂, O₂, and N₂O formation, and led to the sition to N₂and O₂. The O₂, produced during the CO pulse

1

formation of CO₂at 2345 cm ¹, Cu⁺(CO) at 2157 cm

Cu⁺(CO)₃at 2192 cm ¹, and adsorbed CO at 2108 cm

,

(Fig. 12a), exhibited two humps in the concentration pro-

ﬁles, suggesting that the CO pulse promoted the desorption

1

(20, 28, 32). Cu⁺(NO) initially decreased upon CO addition of two different types of adsorbed O species.

but increased to an intensity higher than that prior to CO

The CO pulse onto Cu–ZSM-5-523 at 673 K promotes N₂

pulse during the continued decrease of bridging and chelat- and CO₂formation during the NO and N₂O decomposition

ing Cu²⁺(NO₃). The formation of Cu⁺(CO) and Cu⁺(CO)₃reactions. CO addition also promoted O₂formation with

and depletion of Cu²⁺(NO₃) indicates the reduction of the desorption of two kinds of adsorbed O species during

Cu²⁺to Cu⁺; depletion of Cu²⁺(NO₃) and Cu⁺(NO) upon N₂O decomposition and one type of O₂during NO decom-

CO entry reﬂects the reaction of adsorbed (NO₃) species position. The bridging and chelating Cu²⁺(NO₃), produced

with CO; an increase in NO conversion and O₂formation during NO and N₂O decomposition reactions, decreased

in the presence of Cu⁺(CO) conﬁrms that Cu⁺plays a vital upon pulsing of CO, indicating the reduction of the Cu²⁺

role in NO decomposition. The major roles of CO are to sites to Cu⁺, which adsorbed CO similarly as Cu⁺(CO) and

reduce Cu²⁺to Cu⁺, increasing NO decomposition, to re- CO_adsunder both NO and N₂O decomposition conditions

act with adsorbed nitrato species on Cu²⁺, and to adsorb as as evidenced in Figs. 9 and 11.

Cu⁺(CO) and Cu⁺(CO)₃.

Figures 11 and 12 show the IR spectra, MS proﬁles of re-

3.5. NO/N₂O Adsorption

actants and products, and the variation of adsorbate peak

intensities with time collected during the ﬁrst 1 cm³of CO

Figure 13 shows the IR spectra, MS proﬁles of the re-

pulse into a 5% N₂O in He stream over Cu–ZSM-5-523 actants and products, and the variation of adsorbates peak

at 673 K. Pulsing of 1 cm³of CO decreased bridging and intensities with time collected during a 1 cm³of N₂O pulse

chelating Cu²⁺(NO₃), increased N₂O conversion and N₂into a 5% NO stream over Cu–ZSM-5-523 at 673 K. An

and O₂formation, and produced CO₂(gas) at 2345 cm

,

N₂O pulse into 5% NO led to a decrease in Cu⁺(NO) and

1

Cu⁺(CO) at 2157 cm ¹, and adsorbed CO at 2140 cm ¹. The NO⁺, increase in bridging and chelating Cu²⁺(NO₃) and in

decrease in bridging and chelating nitrato species combined N₂and O₂formation, and little change in Cu²⁺(NO) and

with the formation of Cu⁺(CO) and promotion of N₂and NO as shown in Figs. 13a–13d. N₂O adsorbs onto the Cu⁺

O₂formation suggest that the introduction of CO reduced sites, leading to a decrease in Cu⁺(NO). The decrease in

278

KONDURU AND CHUANG

FIG. 7. IR spectra taken during the ﬁrst 1 cm³of O₂pulse into a 5%

NO stream over Cu–ZSM-5-523 at 673 K.

FIG. 6. MS intensity proﬁles of reactor efﬂuent and peak intensity

proﬁles of adsorbate species collected during the ﬁrst 1 cm³of H₂pulse

into a 5% N₂O stream over Cu–ZSM-5-523 at 673 K.

Cu⁺(NO) is accompanied by an increase in N₂and O₂for-

mation but with no signiﬁcant change in NO concentration,

suggesting that the additional N₂and O₂formed is from the

N₂O decomposition reaction rather than due to the promo-

tion of NO decomposition. Furthermore, IR spectra and

MS proﬁles of the reactor efﬂuents obtained (not shown)

from a separate study involving 1 cm³of N₂O pulse into He

at 673 K produced identical adsorbates with the same in-

tensity and the same N₂O, N₂, and O₂MS proﬁles as shown

in Fig. 13. In summary, the addition of N₂O into a 5% NO

in He stream at 673 K does not affect either the NO con-

version to N₂and O₂or the pulsing activity of N₂O to N₂

and O₂.

One cubic centimeter of NO pulse into a 5% N₂O in He

stream at 673 K (not shown) produced no changes in N₂O

(gas) in the IR spectrum, increased bridging and chelat-

ing Cu²⁺(NO₃), led to the formation of NO⁺, Cu⁺(NO),

and Cu²⁺(NO), and caused no changes in the reactant and

FIG. 8. MS intensity proﬁles of reactor efﬂuent and peak intensity

proﬁles of adsorbate species collected during the ﬁrst 1 cm³of O₂pulse

product concentrations. To observe the changes in N₂O into a 5% NO stream over Cu–ZSM-5-523 at 673 K.

FIG. 11. IR spectra taken during the ﬁrst 1 cm³of CO pulse into a 5%

N₂O stream over Cu–ZSM-5-523 at 673 K.

FIG. 9. IR spectra taken during the ﬁrst 1 cm³of CO pulse into a 5%

NO stream over Cu–ZSM-5-523 at 673 K.

FIG. 10. MS intensity proﬁles of reactor efﬂuent and peak intensity

proﬁles of adsorbate species collected during the ﬁrst 1 cm³of CO pulse

into a 5% NO stream over Cu–ZSM-5-523.

FIG. 12. MS intensity proﬁles of reactor efﬂuent and peak intensity

proﬁles of adsorbate species collected during the ﬁrst 1 cm³of CO pulse

into a 5% N₂O stream over Cu–ZSM-5-523.

280

KONDURU AND CHUANG

FIG. 13. (a) IR spectra, (b) MS intensity proﬁles of reactor efﬂuent, and (c) and (d) peak intensity proﬁles of adsorbate species collected during

the ﬁrst 1 cm³of N₂O pulse into a 5% NO stream over Cu–ZSM-5-523 at 673 K.

adsorbates (Cu⁺–N₂O and Cu⁺–ON₂) during the NO

products, and the variation of adsorbate peak intensities

with time collected during a 1 cm³of NO pulse into a

pulse into 5% N₂O, a similar study was conducted at 323 K

5% N₂O stream over Cu–ZSM-5-523 at 323 K. Intro-

duction of 1 cm³of NO led to a decrease in Cu⁺–N₂O,

where the N₂O adsorbates were observed clearly. Figure 14

shows the IR spectra, MS proﬁles of the reactants and

FIG. 14. (a) IR spectra, (b) MS intensity proﬁles of reactor efﬂuent, and (c) and (d) peak intensity proﬁles of adsorbate species collected during

the ﬁrst 1 cm³of NO pulse into a 5% N₂O stream over Cu–ZSM-5-523 at 323 K.

NO AND N₂O DECOMPOSITION ON Cu–ZSM-5

281

no change in Cu⁺–ON₂, N₂O_ads(2219 cm ¹), and show that Cu⁺–N₂O and Cu⁺–NO exhibit near-mirror-

bridged Cu²⁺(NO₃), and the formation of Cu⁺(NO), image responses to the NO pulse, suggesting that Cu⁺–

Cu⁺(NO)₂, and Cu²⁺(NO) as shown in Figs. 14a–14d. Dis- N₂O and Cu⁺–NO share the same Cu⁺sites and that Cu⁺–

tinct responses of the Cu⁺–N₂O and Cu⁺–ON₂species N₂O + NO ↔ Cu⁺–NO + N₂O is a reversible process. The

(Fig. 14b) to NO addition suggest that they are adsorbed on process also appeared to occur at 673 K as evidenced by

different Cu⁺sites. Near-mirror-image responses of Cu⁺– the decrease in Cu⁺–NO during the N₂O pulse into the NO

N₂O and Cu⁺(NO) (Fig. 14b) indicate that N₂O adsorbs as ﬂow as shown in Fig. 13.

Cu⁺–N₂O on the same site as the one on which NO adsorbs

Figures 14a and 14b also show that Cu⁺–ON₂, unlike

as Cu⁺(NO). Cu⁺–N₂O, which is depleted upon NO addi- Cu⁺–N₂O, was not affected by the introduction of the NO

tion (Figs. 14a and 14b), is expected to do the same during pulse at 323 K. Furthermore, the NO pulse into 5% N₂O at

the NO pulse into 5% N₂O at 673 K. The NO pulse into a reaction temperature of 673 K was found to have no ef-

5% N₂O, however, did not inhibit or promote N₂O decom- fect on the N₂O decomposition rate and product selectivity,

position, conﬁrming that Cu⁺–N₂O does not play a role in indicating that NO does not adsorb onto the active N₂O de-

the N₂O decomposition pathway; i.e., Cu⁺–N₂O is a spec- composition sites but may adsorb and replace adsorbates

tator adsorbate species. The spectator Cu⁺–N₂O may be present on sites not participating in the N₂O decomposi-

adsorbing on the same site as that of Cu⁺(NO) that is the tion reaction. The replacement of Cu⁺–N₂O by Cu⁺–NO

active adsorbate during NO decomposition on Cu–ZSM- at 323 K combined with the absence of the NO effect on

5, suggesting that NO and N₂O decomposition occurs via N₂O decomposition at 673 K suggests that Cu⁺–N₂O is a

separate active sites and adsorbates.

spectator adsorbate species during the N₂O decomposition

reaction. This inference would possibly, but not necessarily,

point to Cu⁺–ON₂as a possible active adsorbate for N₂O

decomposition.

Decomposition of N₂O to N₂and O₂is envisioned to

occur as indicated in the following pathways:

4. DISCUSSION

4.1. N₂O Decomposition

N₂O decomposition has been proposed to proceed via

the following steps (11):

Cu⁺–ON₂→ Cu²⁺

O

+ N₂

[3]

[4]

N₂O + S ↔ N₂O–S

N₂O–S → N₂+ O–S

2O–S ↔ O₂+ 2S.

Cu²⁺

O

→ Cu⁺+ O₂.

1

2

Reaction [3], the proposed pathway for N₂formation, is in

line with the observation that depletion of Cu⁺–ON₂was

followed by the formation of adsorbed N₂during N₂O TPR

(Fig. 2a). The reduction of Cu²⁺to Cu⁺led to the formation

where S is the active catalytic site. Although the proposed

steps provide a reaction pathway for N₂O decomposition,

they do not present information on the structure of active

adsorbates and the nature of active sites and offer little

theoretical basis in assisting the design of improved N₂O

decomposition catalysts. The key to developing a working

reaction mechanism, which would aid in the design of novel

highly active catalysts, is the knowledge of active adsorbates

and the nature of sites on which the active adsorbates re-

side. One objective of this study is to elucidate the type of

adsorbed N₂O involved in N₂O decomposition from the re-

sults of in situ IR studies during a CO, O₂, and H₂pulse into

N₂O.

of O₂from Cu²⁺

O as illustrated in reaction [4]. Quantum

chemical and statistical mechanical calculations have esti-

mated the Gibbs free energy (1G) of reaction [4] to be

24.2 kcal/mol at 773 K (47). An alternate pathway for

N₂formation has also been envisioned to occur as shown

in reaction [5], which is expected to compete with reaction

[3]. Although density functional theory calculations (48) in-

dicate that the energy barriers to reactions [3] and [5] are

comparable, observation of Cu⁺–ON₂(Fig. 14) and absence

of Cu²⁺O –ON₂in the present study point to reaction [3]

as the possible route for N₂formation.

IR results in Fig. 2 show that N₂O adsorbs on the Cu–

1

ZSM-5 surface sites via N and O as Cu⁺–N₂O (2287 cm

)

Cu²⁺O –ON₂→ Cu²⁺OO + N₂.

[5]

and Cu⁺–ON₂(2240 cm ¹), respectively, as shown in reac-

tions [1] and [2].

Reactions [3] and [4] may also be linked as consecutive

pathways in the N₂O decomposition mechanism as shown

in Scheme 1. Scheme 1 also includes the pathway for

Cu²⁺(NO₃) formation (reaction [6]):

Cu⁺+ N₂O ↔ Cu⁺–N₂O

Cu⁺+ N₂O ↔ Cu⁺–ON₂.

[1]

[2]

The distinction in catalytic properties between Cu⁺–N₂O

and Cu⁺–ON₂can be differentiated by their responses to

the NO pulse at 323 K (Fig. 14). Figures 14b and 14c

Cu²⁺O₂+ Cu⁺–ON₂→ Cu²⁺(NO₃) + Cu⁺–N [6]

Cu⁺–N + Cu⁺–N → Cu⁺+ N₂

[7]

282

KONDURU AND CHUANG

SCHEME 1

Both steps [6] and [7] remain in a speculative stage. Step and [9]:

[6] is postulated according to recent results of step switch

results of He to 5% N₂O at 323 K (49) that show that the for-

mation of Cu⁺–ON₂preceded that of Cu²⁺(NO₃), suggest-

ing that Cu⁺–ON₂may act as a precursor for Cu²⁺(NO₃)

Cu⁺+ NO ↔ Cu⁺(NO)

[8]

[9]

Cu⁺(NO) + Cu⁺–N → 2Cu⁺+ N₂O.

formation. The slow development of Cu²⁺(NO₃) can also

be observed during the H₂and CO pulse into N₂O ﬂow

(Figs. 6 and 12) in which Cu²⁺(NO₃) was ﬁrst removed by

the H₂and CO pulse and then gradually returned to the

initial intensity.

The correlation between Cu⁺(NO) and N₂O formation pro-

ﬁles, however, was not observed on poisoned catalysts (32).

This could be due to the lack of NO dissociation to adsorbed

N on these catalysts, which would prevent N₂O formation

but would not affect the surface concentration of Cu⁺(NO).

The N₂O produced from reaction [9] can follow the path-

wayshown in Scheme 1to produce N₂and O₂. Cu⁺(NO) can

also serve as a precursor for the formation of Cu²⁺(NO₃)

as shown in reaction [10]:

The most interesting results of the H₂/CO pulse studies

are the formation of two H₂O/OH concentration peaks dur-

ing the H₂pulse (Fig. 6) and the formation of two O₂peaks

during the CO pulse (Fig. 12). These observations suggest

the presence of at least two types of oxygen on the catalyst

surface under N₂O decomposition conditions. These oxy-

gen species can interact with adsorbed hydrogen, leading

to OH and H₂O formation during the H₂pulse, while a sig-

niﬁcant fraction of these oxygen species cannot be accessed

by CO and simply desorb as O₂upon reduction of Cu²⁺to

Cu⁺during the CO pulse.

Cu⁺(NO) + O₂↔ Cu²⁺(NO₃).

[10]

The formation of Cu²⁺(NO₃) from Cu⁺(NO) is evidenced

by the decrease in Cu⁺(NO) and increase in Cu²⁺(NO₃)

during the O₂pulse into NO (Figs. 7 and 8).

4.2.2. NO dissociation. Cu⁺has been considered as a

site for the dissociation of NO. The results from CO and H₂

pulse studies (Figs. 3, 4, 9, and 10) show that the formation

of N₂, N₂O, and O₂was enhanced in the reducing environ-

ment where Cu⁺is the prevailing copper site as evidenced

bythe high intensitiesofCu⁺(CO) and Cu⁺(NO). Cu⁺(NO)

may be involved in NO dissociation as shown in reaction

[11]:

Oxygen in Cu²⁺(NO₃) may contribute to CO₂forma-

tion since a sharp decrease in Cu²⁺(NO₃) intensity corre-

sponded to a rapid increase in CO₂and Cu⁺(CO) intensi-

ties (Figs. 11 and 12). Furthermore, CO reacted with Cu²⁺

,

leading to the reduction of Cu²⁺to Cu⁺. Since the zeolite

framework oxygen must stay intact and cannot be removed

byCO, the oxygen desorbed duringthe CO pulse must come

from extra framework lattice oxygen (ELO). The oxygen

species may associate with Cu in several different forms,

which will be further discussed in Section 4.3.

Cu⁺(NO) + Cu⁺→ Cu⁺–N + Cu²⁺O .

[11]

It must be noted that there is no direct evidence for the

occurrence of reaction [11]. Alternate pathways for NO

dissociation have been proposed to proceed as indicated in

reactions [12]–[14]:

4.2. NO Decomposition

The adsorbates observed during NO TPR are Cu⁺(NO),

Cu²⁺(NO), bridging and chelating Cu²⁺(NO₃), and NO⁺.

The role of these adsorbates in NO decomposition may

be elucidated from their relationships to N₂, N₂O, and O₂

formation during TPR and H₂/O₂/CO pulse studies.

Cu⁺(NO)₂→ Cu²⁺

O

+ N₂O

[12]

[13]

[14]

Cu⁺–ON + NO → Cu²⁺

2ZCu + 2NO → ZCuOCuZ + N₂O.

4.2.1. N₂O formation. The Cu⁺(NO) proﬁle correlates

well with the gaseous N₂O proﬁle during NO TPR (Figs. High Gibbs free energy of reaction (1G) at 773 K (47) and

1a and 1b), the H₂pulse into NO (Figs. 3, 4a, and 4b), and the lack of correlation between Cu⁺(NO)₂and N₂O forma-

the O₂pulse into NO (Figs. 7, 8a, and 8b). These results tion (31) suggested that reaction [12] is unlikely to be the

suggest that Cu⁺(NO) may be a precursor for N₂O for- pathway for the NO bond dissociation. Although the O-

mation, which may be written as shown in reactions [8] bonded Cu⁺–ON is not observed in the present study, the

NO AND N₂O DECOMPOSITION ON Cu–ZSM-5

283

SCHEME 2

highly reactive nature of such species (51) may not allow its have shown that another alternative for O₂formation is

detection and the reaction of Cu⁺–ON with NO (reaction through the decomposition of bridged Cu²⁺(NO₃) to Cu⁺

[13]) may be a plausible route for NO dissociation. The and NO and O₂(reaction [18]). The bridged Cu²⁺(NO₃)

NO dissociation pathway may also involve the oxocation could be formed via two routes: (i) reaction of NO with

species (50), ZCuOCuZ (Z represents the Al(OH)₄group Cu²⁺O₂, which is formed by the combination of a mobile O

of the zeolite; O donates extralattice oxygen, ELO.), as il- species with Cu²⁺

O produced from NO dissociation, and

lustrated in reaction [14]. NO dissociation via the (NO)₂(ii) interaction of NO with a CuO₂Cu species since such

precursor and N₂O intermediate is the most widely postu- a species would offer the possibility of two copper sites

lated pathway in NO decomposition (21–23, 26). However, acting as a bridge for the nitrate ligand. The dissociation

the signiﬁcant differences in the N₂, O₂, OH, and H₂O pro- of NO and formation of N₂, N₂O, and O₂during the NO

ﬁles during the H₂pulse in Figs. 3–6 revealed that N₂O is not decomposition reaction are summarized in Scheme 2 and

a major precursor for the decomposition of NO to N₂and Table 2.

O₂. The NO dissociation pathwaywithout N₂O involvement

remains to be investigated. In summary, Cu⁺is needed for

the dissociation of NO and the unique high activity of Cu–

4.3. Cu⁺Sites for Cu⁺–ON₂, Cu⁺–N₂O, and Cu⁺(NO)

ZSM-5 for NO decomposition is due to its ability to sustain

Cu in the Cu⁺state. The identity of adsorbates participating

in NO dissociation cannot be determined by the in situ IR

technique.

IR results from the present study revealed the dynamic

behavior of the NO and N₂O adsorbates but provide lit-

tle information on the nature of Cu⁺sites and its chemi-

cal/physical environment. The nature of the Cu⁺site may

be elucidated from the recent results of density functional

calculations for NO and N₂O adsorption (53) on two types

of Cu⁺sites: (i) Cu⁺(Al(OH)₄) (Z_AlCu⁺), i.e., the Cu⁺site

associated with Al, and (ii) Cu⁺(Si(OH)₄) (Z_SiCu⁺), i.e.,

the Cu⁺site associated with Si. Binding energy calculations

from this study indicate that the adsorption of NO onto

Z_AlCu⁺is more favorable than NO onto Z_SiCu⁺. Hence,

NO is envisioned to adsorb onto Z_AlCu⁺as shown in reac-

tion [19]:

4.2.3. O₂formation. The formation of O₂during NO

decomposition has been proposed to occur as shown in re-

actions [15]–[18] (31, 32, 47, 52).

Cu²⁺

O

→ Cu⁺+ O₂

[15]

[16]

[17]

1

CuO₂Cu ↔ 2Cu⁺+²O₂

1

2

CuO₂Cu ↔ CuOCu + O₂

Cu²⁺(NO₃) ↔ Cu⁺+ NO + O₂.

[18]

Z_AlCu⁺+ NO ↔ Z_AlCu⁺–NO.

[19]

Quantum mechanical and statistical calculations (47) have

conﬁrmed that reaction [14], with a Gibbs free energy of

reaction (1G) of 24.1 kcal/mol at 773 K, is thermody- Comparison of the calculated binding energies (53) for ad-

namically favorable. Density functional studies (52) indi- sorbed N₂O on various Cu⁺sites also revealed that N₂O

cated that two Cu²⁺

O

species can combine to form the adsorbs as Cu⁺–ON₂on Z_SiCu⁺more favorably than on

dioxocation, ZCuO₂CuZ, which would further decompose Z_AlCu⁺. NO pulse into 5% N₂O results (Fig. 14) from the

to regenerate Cu⁺and produce O₂(reactions [15] and present study show that Cu⁺–N₂O and Cu⁺(NO) adsorb on

[16]). Results from our previous NO TPD studies (31, 32) the same Cu⁺site. These results combined with the density

284

KONDURU AND CHUANG

TABLE 2

Proposed NO and N₂O Decomposition Mechanisms over Cu–ZSM-5

NO decomposition

N₂O decomposition

Adsorption

Cu⁺+ NO ↔ Cu⁺(NO)

Cu²⁺+ NO ↔ Cu²⁺(NO)

S–O + NO ↔ S–O–NO⁺

Cu⁺+ N₂O ↔ Cu⁺–N₂O

Cu⁺+ N₂O ↔ Cu⁺–ON₂

NO dissociation

Cu⁺(NO) + Cu⁺→ Cu⁺–N + Cu²⁺

O

Cu⁺–ON + NO → Cu²⁺

O

+ N₂O

N₂O and N₂formation

N₂formation

Cu⁺(NO) + Cu⁺–N ↔ 2Cu⁺+ N₂O

Cu⁺–ON₂↔ Cu²⁺

O

+ N₂

Cu⁺–ON + NO → Cu²⁺

O

+ N₂O

Cu⁺+ N₂O ↔ Cu⁺–ON₂

Cu⁺–ON₂→ Cu²⁺

O

+ N₂

Cu⁺–N + Cu⁺–N → 2Cu⁺+ N₂

O₂formation

Cu²⁺

O

↔ Cu⁺

+

¹O₂

Cu²⁺

O

↔ Cu⁺

+

¹O₂

2

CuO₂Cu → 2Cu⁺+ O₂

Cu²⁺O₂↔ Cu⁺+ O₂

Cu²⁺(NO₃) ↔ Cu⁺+ NO + O₂

Cu²⁺(NO₃) formation

Cu⁺(NO) + O₂↔ Cu²⁺(NO₃

)

Cu²⁺

O

+ Cu⁺–ON₂↔ Cu²⁺O –ON₂

Cu²⁺

O

+ O → Cu²⁺O₂

Cu²⁺O –ON₂↔ Cu²⁺O₂+ N₂

Cu²⁺O₂+ NO → Cu²⁺(NO₃

)

Cu²⁺O₂+ Cu⁺–ON₂↔ Cu²⁺(NO₃) + Cu⁺–N

Note. S: cationic site on ZSM-5; S–O–NO⁺is indicated as NO⁺in the present study.

ZCuOCuZ or ZCuO₂CuZ can replace all single Cu⁺sites in both mechanisms.

functional calculations suggest that Cu⁺for Cu⁺(NO) and associated O²could be the adsorbed O species produced

Cu⁺for Cu⁺–N₂O may be associated with Z_Al; the Cu⁺for from NO decomposition and O₂addition since the O₂pulse

Cu⁺–ON₂may be associated with Z_Si, supporting the fact into NO signiﬁcantly inhibited NO decomposition. The for-

that NO and N₂O decomposition may occur on Cu⁺sites mation of additional O², produced from the adsorption of

present in different environments.

the pulsed O₂, along with the already present O²from

NO decomposition should lead to the inhibition of O₂des-

orption and NO decomposition. Density functional studies

(48, 50) show that the O²species may be associated with

Cu(II)–O²–Cu(II), which can further be interconverted to

4.4. Nature of Oxygen in NO and N₂O Decomposition

The formation of O₂has been envisioned to occur as

shown in Scheme 1 during N₂O decomposition and as il-

lustrated in Scheme 2 during NO decomposition over Cu–

ZSM-5. The distinct difference in reactivity of adsorbed

O produced from N₂O and NO decomposition is summa-

rized in Table 3. N₂O decomposition is expected to produce

N₂and O upon N₂O dissociation. ESR studies (11) have

proved unambiguouslythat oxygen adsorbsasa reactive O

species during the N₂O decomposition reaction. O from

N₂O is expected to combine with another O to desorb as

gaseous O₂and reduce Cu²⁺to Cu⁺. Furthermore, the O

has been shown to migrate and transform to a more strongly

bonded O²(reaction [20]) in oxide catalysts under high

temperatures, depending on the type and concentration of

transition-metal ion associated with the catalyst (11):

TABLE 3

Differences in Reactivity of Adsorbed O from N₂O

and NO Decomposition

O from N₂O

O from NO

Experiment

decomposition

CO pulse (Figs. 10

and 12)

Two hump O₂proﬁle One hump O₂proﬁle

H

₂pulse (Figs. 3–6) Two hump H₂O/

OH proﬁle

One hump H₂O proﬁle;

formation of N₂O;

desorption of O₂

O₂pulse (Figs. 7

and 8)

No effect

Oxidization of Cu⁺in

Cu⁺(NO) to Cu²⁺in

Cu²⁺(NO₃); inhibition

of NO decomposition

O

+ e ↔ O²

.

[20]

NO and N₂O pulses O₂did not lag behind O₂proﬁle lags behind N₂,

into He (Refs. 30

and 49)

N₂and NO

N₂O, and NO

It is expected that O²is more difﬁcult to desorb than O

due its strong association with metal sites. This strongly

NO AND N₂O DECOMPOSITION ON Cu–ZSM-5

ACKNOWLEDGMENTS

285

Cu(II)–O²₂–Cu(II). Lack of spectroscopic information on

the oxocation and adsorbed oxygen species did not allow

veriﬁcation of the types of the oxygen species, which were

postulated based on density functional calculations. How-

ever, the postulation of different forms of oxygen, O and

Although the research described in this article has been funded wholly

by the United States Environmental Agency under assistance agreement

R823529-01-0 to the University of Akron, it has been subjected to the

Agency’s peer and administrative review and therefore may not neces-

O², from N₂O decomposition is indeed consistent with the sarily reﬂect the views of the Agency and no ofﬁcial endorsement should

be inferred. The authors would also like to acknowledge X. Kang at the

University of Akron for his help with some of the experiments.

observation of two O₂and H₂O peaks during the CO and

H₂O pulse studies (Figs. 6 and 12).

The activity of the Cu–ZSM-5 catalyst for NO and N₂O

decomposition is closely related to its ability to desorb oxy-

gen from the catalyst surface, allowing it to maintain the

REFERENCES

majority of its Cu sites in the Cu⁺state. The importance of

the Cu⁺state is illustrated in Figs. 9 and 11, which show that

Cu⁺(CO) is the dominant species during a CO pulse into

NO and N₂O. In contrast to the dominance of the Cu⁺state

in Cu–ZSM-5 under NO and N₂O decomposition condi-

tions, Cu²⁺is the major species on the CuO/Al₂O₃catalyst

(accompanied by a trace amount of Cu⁰and Cu⁺) which ex-

hibits little activity for both reactions (49). The unique ca-

pability of Cu–ZSM-5 can be attributed to the nature of the

interaction between Cu⁺and the zeolite framework; such

interactions cannot be probed by our infrared techniques.

1. Hightower, J. W., and Van Leirsburg, D. A., “The Catalytic Chemistry

of Nitrogen Oxides” (R. L. Klimisch and J. G. Larson, Eds.). Plenum,

New York, 1975.

2. Iwamoto, M., in “Future Opportunities in Catalytic and Separation

Technology” (M. Misono, Y. Moro-oka, and S. Kimura, Eds.), Elsevier,

Amsterdam, 1990.

3. Shelef, M., Chem. Rev. 95, 209 (1995).

4. Winter, E. R. S., J. Catal. 15, 144 (1969).

5. Fu, C. M., Korchak, V. N., and Hall, W. K., J. Catal. 68, 166 (1981).

6. Leglise, J., Petunchi, J. O., and Hall, W. K., J. Catal. 86, 392 (1984).

7. Swamy, C. S., and Christopher, J., Catal. Rev.-Sci. Eng. 34, 409

(1992).

8. Li, Y., and Armor, J. N., Appl. Catal. B 3, 55 (1993).

9. Kannan, S., and Swamy, C. S., Appl. Catal. B 3, 109 (1994).

10. Chang, Y.-F., McCarty, J. G., Wachsman, E. D., and Wong, V. L., Appl.

Catal. B 4, 283 (1994).

11. Kapteijn, F., Rodriguez-Mirasol, J., and Moulijn, J. A., Appl. Catal. B

9, 25 (1996).

12. Yamashita, T., and Vannice, A., J. Catal. 161, 254 (1996).

13. Pang, X. Y., Zeng, H. C., Wu, J. C., and Li, K., Appl. Catal. B 9, 149

(1996).

5. CONCLUSIONS

Addition of CO, H₂, and O₂into the steady-state ﬂow

of NO and N₂O over Cu–ZSM-5 not only produces tran-

sient reducing and oxidizing conditions but also leads to

changes in the reactivity and dynamics of adsorbates. NO 14. Turek, T., Appl. Catal. B 9, 201 (1996).

15. Shimokawabe, M., Hirano, K., and Takezawa, N., Catal. Today 45, 117

and N₂O decomposition produces differently adsorbed O

species: adsorbed oxygen produced from N₂O decomposi-

tion (i) reacts with the pulsed H₂to produce two humps in

the H₂O proﬁle, (ii) interacts with the pulsed CO to desorb

(1998).

16. Dandekar, A., and Vannice, M. A., Appl. Catal. B 22, 179 (1999).

17. Iwamoto, M., Yoko, S., Sakai, K., and Kagawa, S., J. Chem. Soc. Fara-

day Trans. 1, 77, 1629 (1981).

O₂with two humps, and (iii) did not interact with adsorbed 18. Iwamoto, M., Furukawa, H., and Kagawa, S., in “New Developments

in Zeolite Science and Technology” (Y. Murukama, A. Ichijima, and

J. W. Ward, Eds.), p. 943. Elsevier, Amsterdam, 1986.

19. Li, Y., and Hall, W. K., J. Phys. Chem. 94, 6145 (1990).

20. Aylor, A. W., Larsen, S. C., Reimer, J. A., and Bell, A. T., J. Catal. 157,

O produced from the O₂pulse; adsorbed O produced from

NO decomposition interacts with the pulsed CO and H₂

to (i) desorb O₂and H₂O with a single peak in concentra-

tion and (ii) interact with adsorbed O from the O₂pulse to

inhibit NO decomposition.

592 (1995).

21. Iwamoto, M., Yahiro, H., Mizuno, N., Zhang, W.-X., Mine, Y.,

Furukawa, H., and Kagawa, S., J. Phys. Chem. 96, 9360 (1992).

22. Giamello, E., Murphy, D., Magnacca, G., Morterra, C., Shioya, Y.,

Nomura, T., and Anpo, M., J. Catal. 136, 510 (1992).

23. Valyon, J., and Hall, W. K., in “Proceedings, 10th International

Congress on Catalysis, Budapest, 1992 (L. Guczi, F. Solymosi, and

P. Tetenyi, Eds.), p. 1341. Akade´miai Kiado´, Budapest, 1993.

24. Spoto, G., Zecchina, A., Bordiga, S., Ricchiardi, G., Marta, G.,

Leofanti, G., and Petrini, G., Appl. Catal B 3, 151 (1994).

25. Hoost, T. B., Laframboise, K. A., and Otto, K., Catal. Lett. 33, 105

(1995).

26. Cheung, T., Bhargava, S. K., Hobday, M., and Foger, K., J. Catal. 158,

301 (1996).

27. Iwamoto, M., Yahiro, H., Tanda, K., Mizuno, N., Mine, Y., and Kagawa,

S., J. Phys. Chem. 95, 3727 (1991).

28. Yan, J. Y., Lei, G.-D., Sachtler, W. M. H., and Kung, H. H., J. Catal.

161, 43 (1996).

29. Adelman, B. J., Beutel, T., Lei, G.-D., and Sachtler, W. M. H., J. Catal.

158, 327 (1996).

N₂O decomposition is proposed to proceed via Cu⁺–

ON₂, Cu²⁺O , and Cu²⁺O –ON₂with Cu⁺–ON₂serving as

a precursor for N₂formation and Cu²⁺

O as a precursor for

O₂formation. NO decomposition proceeds via Cu⁺(NO),

Cu²⁺O , and Cu²⁺(NO₃) with Cu⁺(NO) serving as a pre-

cursor for NO dissociation. Cu⁺in Cu⁺(NO) is different

from that of Cu⁺in Cu⁺–ON₂. The former may be associ-

ated with Al(OH)₄of the zeolite, the latter with Si(OH)₄.

The lack of O₂inhibition effect on N₂O decomposition

and the differences in the reactivity of adsorbed O from

N₂O and NO decomposition suggest that future studies

for the development of novel NO decomposition catalysts

should aim for the conversion of the adsorbed O²pro-

duced from NO decomposition to the type of adsorbed O

(O ) resembling those produced from N₂O decomposition.

286

KONDURU AND CHUANG

30. Hadjiivanov, K., Saussey, J., Freysz, J. L., and Lavalley, J. C., Catal. 42. Nakamoto, K., “Infrared and Raman Spectra of Inorganic and Coor-

Lett. 52, 103 (1998).

31. Konduru, M. V., and Chuang, S. S. C., J. Phys. Chem. B. 103, 5802

(1999).

32. Konduru, M. V., and Chuang, S. S. C., J. Catal. 187, 436 (1999).

33. Chuang, S. S. C., Brundage, M. A., Balakos, M., and Srinivas, G., Appl.

Spectrosc. 49(8), 1151 (1995).

34. Hussain, G., Rahman, M. M., and Sheppard, N., Spectrochim. Acta A

47, 1525 (1991).

dination Compounds,” 3rd ed. Wiley, New York, 1978.

43. Sa´rka´ny, J., Appl. Catal. A 188, 369 (1999).

44. Sa´rka´ny, J., J. Mol. Struct. 410–411, 137 (1997).

45. Sterrer, M., Diwald, O., and Knozinger, E., J. Phys. Chem. 104, 3601

(2000).

46. Ito, T., Wang, J., Lin, C., and Lunsford, J. H., J. Am. Chem. Soc. 107,

5062 (1985).

47. Trout, B. L., Chakraborty, A. K., and Bell, A. T., J. Phys. Chem. 100,

17582 (1996).

35. Miller, T. M., and Grassian, V. H., Colloids Surf. 105, 113 (1995).

36. Miller, T. M., and Grassian, V. H., J. Am. Chem. Soc. 117, 10969

(1995).

48. Schneider, W. F., Hass, K. C., Ramprasad, R., and Adams, J. B., J. Phys.

Chem. B 102, 3692 (1998).

37. Kustov, L. M., Smekalina, E. V., Uvarova, E. B., and Kazansky,

V. B., in “Zeolites: A Reﬁned Tool for Designing Catalytic Sites”

(L. Bonneviot and S. Kaliaguine, Eds.), Elsevier Science, 1995.

38. Miyata, H., Konishi, S., Ohno, T., and Hatayama, F., J. Chem. Soc.

Faraday Trans. 91, 1557 (1995).

39. Lin, J., Chen, H. Y., Chen, L., Tan, K. L., and Zeng, H. C., Appl. Surf.

Sci. 103, 307 (1996).

40. Borello, E., Cerruti, L., Ghiotti, G., and Guglielminotti, E., Inorg.

Chim. Acta 6, 45 (1972).

49. Kang, X., Ph.D. Preliminary Study, The University of Akron, Akron,

OH, 2000.

50. Goodman, B. R., Schneider, W. F., Hass, K. C., and Adams, J. B., Catal.

Lett. 56, 183 (1998).

51. Schneider, W. F., Hass, K. C., Ramprasad, R., and Adams, J. B., J. Phys.

Chem. B 101, 4353 (1997).

52. Goodman, B. R., Hass, K. C., Schneider, W. F., and Adams, J. B.,

J. Phys. Chem. B 103, 10452 (1999).

53. Schneider, W. F., Hass, K. C., Ramprasad, R., and Adams, J. B., J. Phys.

Chem. B 102, 3692 (1998).

41. Forster, H., and Remmert, M., J. Mol. Struct. 174, 357 (1988).

Article Doi

Dynamics of NO and N2O decomposition over Cu-ZSM-5 under transient reducing and oxidizing conditions

DOI: 10.1006/jcat.2000.3046

Source and publish data:

Authors:

Article abstract of DOI:10.1006/jcat.2000.3046

Full text of DOI:10.1006/jcat.2000.3046

Products guided by the article

Relevant to this article

Hot Product