Catalytic reduction of nitrogen oxides 1. The reduction of NO

Catalytic reduction of nitrogen oxides 1. The reduction of NO

335 Applied Catalysis, 18 (1985) 335-352 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands CATALYTIC REDUCTION OF NITROGEN 1...

2MB Sizes 0 Downloads 33 Views

335

Applied Catalysis, 18 (1985) 335-352 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

CATALYTIC REDUCTION OF NITROGEN 1. THE REDUCTION

C.U.

Odenbrand,

Ingemar

Department

Sten T. Lundin

of Chemical

Lund University,

Technology,

Institute

22March

and Lars A.H. Andersson

Chemical

of Science

P.O. Box 124, S-221 00 Lund,

(Received

OXIDES

OF NO

Center,

and Technology,

Sweden.

1985, accepted

26 June

1985)

ABSTRACT The selective catalytic reduction of NO with NH3 has been studied over V205/Si02-Ti02 catalyst. The influence of side reactions has been investigated in order to find appropriate conditions for the kinetic investigation. An exponential rate expression has been proposed and the reaction orders for NO, NH3 and 02 have been determined. The influence of pore diffusion was shown to be important and experimentally determined textural parameters have been used to determine the intrinsic rate parameters.

INTRODUCTION Lately

public

serious

effects

oxides.

These

therefore

opinion caused

effects

in Western

by pollution were first

for stationary vestigated

from these

sources

this subject

quite

150 commercial cerning monly

SCR plants

the factors

used catalysts

shown by the Swedish Using catalyst

governing

lytic reduction

for reducing

in Japan,

in operation. the reaction

between

known.

authorities

has prompted

on homogeneously

our aim was to increase of NO, with

standards have in-

reduction

(SCR)

[I] and in 1983 over

this, basic

NO and/or

knowledge

con-

NO2 and NH3 over cominterest

this study at our department.

co-precipitated the understanding

Si02-Ti02

0 1985 Elsevier Science Publishers B.V.

as a model

of the selective

concentrates

deal with NO2 and mixtures

well.

0166-9834/85/$03.30

of SOx and

researchers

This fact plus the increased

NH,. This first article

of NO and forthcoming.‘articles-will

and

emission

catalytic

method

Despite

and nitrogen

the amounts

stringent

Selective applied

aware of the

and in Japan,

and so Japanese

substantially.

.

by sulphur States

The most

is not well

V205 supported system,

on methods

the most widely

were

more and more

in the United

two countries.

are applied

of NO, using NH3 has become

has become

of the atmosphere

noted

the bulk of information

NO, has originated

Europe

cata-

on the reduction of NO and NO2 as

336

EXPERIMENTAL Catalyst

SECTION

preparation

The model to a method

catalyst,

20 wt% V205 supported

developed

353 K, and diluted The solution and further

by Shikada

with

50 cm3 cont.

113 cm3 concentrated

by

HCl was cooled

Riedel-de-Haen)

was added

was added

by heating

of chloride

ions. After drying

and calcinated

was performed

in oxalic

Catalyst

characterization

The catalyst using

porosity before

acid

support

Apparatus, Figure

intermittent

catalyst

proper

(Philips)

of catalyst

The total

apparatus

ported

after

materials

in a separate

inserted

t0

was then V205.

(R.G. 99% Riedel-

use was

density

investi-

the pore volume,

using

the

the same methods

was measured

as

in the adsorp-

of the materials

is placed

an isothermicity

was

pressure

line to the reactor.

get

in a coil

any preoxidation

flow type system.

(AISI 2343).

radially, along

and which

The reac-

inside

bottles

four

allows

which

axis.

The

is working

a gas chromatograph

oven,

1 K. via high precision

valves

and measured

at the low pressure

NH3 in N2 is mixed of these

with

the reactor

of the reactor

to within

by micrometer

way. The final mixing

steel

points

at the bottom

from storage

they have been preheated not

free

was crushed,

of 4.45 mm and is equipped

The flows of N2 and NO in N2 are mixed

tor in order

area,

is a conventional

at several

The reactor

ensures

the flow is regulated

in a similar

until

and mercury,respectively.

which

thermocouples

temperature

The gases are supplied

ported

surface

and after

were determined

isopropanol

on a SS screen

conditions.

and this arrangement

meters.

pH of 7. The

20 weight%

of NH4V03

(R.G.

(R.G.99.5%

and procedure

is supported

ure regulators;

1

was preci-

water

support

to reach

before

and all gas lines were made of stainless

under downflow

a final

distilled

solution

The true and apparent using

the reaction

0.5 mm Thermocoax

273 K, Tic14 250 g urea

reaching

drying

with an aqueous

tor is 26 cm long and has an inside diameter

measurements

before

times with

of the catalyst

helium.

analysis

The reactor

to pH

of Si and Ti and the support

hours

and the catalyst

by pycnometry

1 shows

to 288 K.

(R.G. 99.5% Riedel-de-Haen).

[31. The true density using

stirring.

for 24 h at 393 K the catalyst

N_2-adsorption techniques.

tion equipment

and cooled

HCl (R.G. 37% krck)

gentle

several

and the pore size distributions

also measured

according

1 hour at 863 K in a stream of air. Si02-Ti02

steps with

de-Haen)

gated

and washed

for

in several

The impregnation

slowly with

to 363 K for several

was decanted

impregnated

water

to 273 K and then 65 cm3 cold,

to the solution

precipitate

sieved

was prepared

(purum ROTH) was melted,

to 273 K.

99.5% Riedel-de-Haen)

pitated

1500 cm3 hot, 343 K, distilled

was acidified cooled

on Si02-Ti02,

130 g Na2Si03*5H20

[Zl.

panel

and trans-

with 02 and trans-

two gas streams

inside the oven,

press-

by rota-

takes

just ahead

of NO to NO2 and to avoid

place,

of the reac-

any

homogenous

337

rstuv

Figure

1

Experimental

(c) Low pressure (g) Filter.

setup;

regulator.

(a] Gas storage

(d) Micrometer

(h) Flow regulator.

(k) Z-Way magnetic

valve.

(n) NH3 scrubber.

reaction 200kPa

between

(1) High-low

the cata?yst the reactor mixtures

and after

in N2 j-3500

method,

the actual

(e) Rotameter.

The pressure valve

in the reactor at the exit.

by two high precision

the rotameters

and pressure.

lines are connected

and magnetic

valves

(398 K) teflon

experiments

showed

NO2 converter resulted serted

the instrument,

system

gas stream from

directed

when

to the NO, meter

is semiautomatic

is metered

in order

the sum of inlet flows to the ana>yzer

Preliminary over the

in the gas stream. NH3 scrubber

NH3 before

flow.

This

was

in-

analysis.

on the Analyzer

and the flow through

fil-

in a

NO, analyzer.

NO was reacting

to remove

and the sample

Dr to a waste

Through

by difference.

Therefore,an

via pressbuttons

tritharotameter

gas

of the components

is transported

NH3 was present

of NO2 concentration.

to

with corrections

955 Chemiluminiscence

NO and NO,, and NO2 is calculated

values

at the entrance

calculated either

inside

connected

from analyzed

to the reactor.

the gas mixture

Model

around drop over

recommended calcuJation

that as much as 15% of the incoming

in erratic

The sampling sample

line to a Beckman

measures

meters

The inlet concentrations

Two sampling

heated

is normally

are calculated

ters,

The NO, meter

gas.

(s) NH3 in N2.

The pressure

pressure

can thus be obtained. flow regulators

(r) Oz.

valve.

Valve.

(m) Calibration

(q) Vent.

the bed. Both NH3 and 1'40are taken

through

for temperature,,, viscosity

(f) Closing

(j) Back pressure

regulator.

ppm>. Using the manufacturer-s

flows

regulator.

(u) NO2 in N2- (v) N2.

by a micrometer

bed is measured before

pressure

(p) Manometer.

the components.

and is regulated

valve.

lb) High pressure

(i) 3-Waymagneticvalve.

(0) NOx-meter.

(t) NO in N2.(T).Thermocouple.

bottle.

panel.

The

the reactor

is

The sample flow is

?jne maintaining

d

CUtXtdf7t

flow

of

338

1 NTP h -' (NTP=273 K and 101 kPa) through the reactor. Reaction condi-

about 60 tions,

unless

and balance

cited

otherwise,

CALCULATIVE

PROCEDURES

Evaluation

of reaction

The conversion 0.425-0.710

mm. The rate of reaction

this evaluation

method

determined

was calculated

bed length

approximation.

by Meier

is 4 particle

below

approximation

30%,the

below

order

Diffusion

designing effects

sent system. portant

the catalytic

of different

The small value

result

mass

of NO of

Since the majority of backmixing

and

When the concentrabe used with errors

for NH3.

reactor,calculations

were

performed

in order

and heat transport

steps

suspected

in the pre-

of the reactor

from these calculations.

steps was applied lations

the rate

limitations

Before minimize

in

so the effect

effects

can be neglected.

of NH3 up to 40% could

of the low reaction

The error

[41 using

results.

tion of NH3 was limiting,conversions 1% because

the total

1% at conversions

diameters

with Meiers

at conversions

due to the differential

of

of NO and the catalyst

to be less than

is less than 5% in concordance

approximation with a diameter

from this conversion,

reactor

as described

below and was shown

of all rates were obtained error

was calculated

the differential

less than 30%. The catalyst backmixing

reactor

with 23 mg catalyst

the inlet concentration

the reactor,

This was done using

expression

600 ppm of NO, 700 ppm NH3, 2% 02,

rates and the differential

of NO was measured

flow rate through weight.

were as follows:

N2.

to the experimental

were based on the highest

diameter

(0.445 cm) was the most

Mears

[51 tests

for effects

results

presented

in this paper.

measured

reaction

to

im-

of transport The calcu-

rate and the abbreviations

are

those used by Meat-s. The measured for the main

rate was 6.78.10-6

reaction

-406 kJ mol-'.

(intraparticle)

AHlr2

__p

<

(17) below was calculated

The following

interphase

particle,

first.

steps. Mears

630 K. The heat of reaction

from tabulated

steps could

and interparticle

are dealt with

data

be important,

The gradients

criteria

161 to be

namely,

within

intra-

the particle

for an isothermal

particle

Qg

gradient,

for our system T-T,,

of the kinetics.

of the high intrinsic for absence

is;

(I)

and gives 0.002 < 0.108 The temperature

criteria

transport

at

E

XTS

evaluation

mol NO<'g,l,

and shows that

is only 0.01 K which

Low effectiveness

activity

observed.

of concentration

the particle causes

factors

are expected,

Kr2/C p s Deff=113

gradients

within

is isothermal.

no problems

and this violates the particle.

in the

though,

because

the

If the reac-

339

tion is assumed

to follow

first order

kinetics

in NO concentration

then the Thiele

modulus; OS = r

gives

m-l’

kaC

-V-Spi

0

eff an effectiveness

a value of 10.6 yielding

dients

of mass

are thus concluded

lysis of the kinetics The gradients

-AHRr

equals

Intraparticle

and must be included

gra-

in the ana-

to be small,

the particle

(interphase)

are treated

next.

that is;

.0162 = G.15Rglb E

.0025 <

temperature

0.5 K and would

calculations

the film around

are shown

hTp= b The calculated

to be important

of 0.26.

below.

across

The heat effects

factor

gradient;

not influence

the measured

rates more than 0.7%.

In these

a formula;

h = $as

0.36.Re

0.94

P taken from Gliddon fer coefficient, then means

iT.r -p'bWkc Mears almost

171 was used. The Re number

kc, between

gas and particle

equals

28.5 and gives a mass

of 52.6 ems -' (de Acetis

trans-

[83). This

that

0.15 = 0.15 equals = 0.15 n criteria

is that Rrp/Cb*kc

all of our experiments

(6) < 0.15/n

this criteria

for mass

will

gradients

below

hold and external

5 %.

gradients

In of

mass can also be neglected. In the scale of the reactor of axial

dispersion

and should

talc

That

The length

than the calculated

the freedom

of the catalyst

from effects

bed is 0.21 cm

value;

(7)

cf

= 0.15 cm for XNo = 0.23 and 0.20 cm for XNo = 0.30

is, no axial

30%. However, rion.

transport)

In 5 a

L

can be estimated.

be greater

L = dpy

(interparticle

higher

dispersion conversions

effects

should

be expected

in the differential

for conversions

reactor

violate

below

this crite-

340 at the reactor

The heat effects

hHR,RLbe

>>

= 0 119

Thus the heat resistance no radial

gradients

should

reactor

reaching

gradients

temperature

the center

Derivation

on the catalyst

of the kinetics

when

exist

Carberry

[9] states

the radial

gradient

at the reactor

inserted

aspect

across

bed

Since we

through

bed we do not expect

ratio

the catalyst

wall.

radially

that

the reactor

any problem

with

the

model

expression

used

= k'CNOa'CNH3

b.C

is shown

02

The rates were measured

and intrinsic

rNOapp

The

= n-rNOintr

The Bischoff

for the reduction in equation

must be included

of mass.

The method

for

of NO with NH3

9.

(9)

as a function

were obtained

rates.

steps which

gradients

c

tures from 470 to 580 K varying orders

heat transport

are the internal

below.

of the kinetic

The kinetic

factor.

could

by thermocouples

so will be described

reaction

is significant.

are expected

the only mass and/or

in the analysis

rNO

(8)

at the wall.

In conclusion doing

because;

= 0.0142

is 7.8 so the temperature

be low but a large gradient

measure wall

at the wall

of temperature

R,/do > 3-4. Our value

are important

0.4R,T,./E J *! [1+8(rp/Ro)Biwl



keTw

wall

of concentration

the concentration

from a linear

intrinsic

at 5 different

of one component

regression

rates were obtained

program

tempera-

at a time. The

for the appearent

from equation

10.

= n*kiCND ai.CNHbi.CDCi 3 2

approximation

(10)

[lo] was used for calculation

In this method the Thiele modulus

is formulated

of the effectiveness

as;

(11) where

‘NOs

I = 21

DeNO'rNO'dCNO

(12)

0

In the computations

performed

the integration,

the integral

NO in most

but could

cases

limiting

component

scribed

by equation

by a microcomputer

was evaluated

Newton-Cotes

for the limiting

also be NH3 in some experiments.

the concentration 13.

using

of NH3 inside

reactant.

method

for

This was

When NO was the

the catalyst

particle

was de-

341

so

3

'NH3

(13)

.(C NCI-~NO~)+~NH~~

DeNH3 The concentration sent

in excess

gible.

of 02 was set equal to its surface

in all experiments

The effectiveness

3 n = &coth The program

factor

and thus

value

its gradient

was calculated

since oxygen

was expected

from equation

was pre-

to be negli-

14. (14)

([email protected])-l/[email protected]) determines

the reaction

order

from the linear

regression

of expres-

sions of the type;

ln(rNO/n) The apparent

reaction

order was obtained

value of ki was computed determined

from another

was then computed intrinsic orders

RESULTS

by setting intrinsic

order

data.

and the same procedure

for NO was obtained.

and rate constants

were

again

mannor

factor

until a constant

procedures

In an iterative

for NH3 and 02

The effectiveness

applied

Similar

n = 1. From In k. the orders

were applied

the intrinsic

when

reac-

obtained.

AND DISCUSSION

X-Ray

powder

nal composition any other

of the catalyst diffraction

phases

could

with

be detected.

showed

the size 75-150 silica

[It].

studies.

Table

0.65 cm3g-'.

X-Ray microanalysis composition

of these macropores

char-

was built

up of porous

par-

found

as ours.

character using

were not large as will be

pore structure had a surface

very nicely

of the catalyst.

Si and 95% Ti using

support

with

data obtained by N2 adsorption 2 -1 area of 320 m g and a pore volume investigations

area of 338 m '2g -' for 50/50 mol%

lines by the amorphous

by the amorphous

nm. This size was in the same range as normally

support

the same conditions

No peaks from SiO2 or

caused

In the surface layer pores of the same size as the par-

This compared

who found a surface

SiO2-TiO2.

This was probably

from solution.

1 summarizes

The catalyst

peaks from V205 and Ti02 only. The nomi-

that the catalyst

were obsei-ved. The amount

shown below.

showed

V205 on 50/50 mol%

precipitated

SEM investigations

in amorphous

studies

was 20 weight%

acter of SiO2-TiO2

under

first

(10) using

set of rate-concentration

as above

reaction

Characterization

ticles

from equation

for NH3 and 02 were determined.

tion orders

ticles

(15)

= In k'+a-lnCNOS

They explained

SiO2-ii02

the absence

et al.

The surface This

semiquantitative

of the support

indicated

dried and calcined of X-ray

diffraction

measurements

of the

was found to contain

a surface

of

[12]

of the precipitate.

SEM can yield

this method.

by Tanabe

enrichment

5%

of Ti02.

342 TABLE

'

Catalyst

characteristics

Si02-Ti02

20% V205/Si02-Ti02

Property Particle

density

Solid density BET surface

(gcmm3)

area

Pore diameter Average

(gem-")

(m2g-')

at max

(nm)

pore diameter

Pore volume

(d<45nm)

Pore volume

(d<174nm)

Figure

2 shows

The amount

area,is

(cm3g-') (cm3g-')

the pore size distribution

of pores larger

ure. The majority tribution

(nm)

1.37

0.99

3.12

2.74

'21

32'

10.2

10.2

'0.4

8.0

0.310

0.645

0.3'3

0.646

for the used catalyst

and the support.

than 20 nm was very small and is not shown

of pores are below

at IO nm. The average

13 nm in the support

pore size, computed

8 nm and is not much different

in the fig-

with a peak in the dis-

from pore volume

and surface

from the peak value.

.os,

.04^m Gi .03‘E s $

.02-

b .Ol-

O0

10

S PORE

Figure

2

Pore size distribution

DIAMETER

20

15 (nm)

from N2 adsorption

for;(a)

The support

and

(b) The used catalyst.

The used catalyst of the support.

and the pore volume smallest

thickness

V205

since

most of the pore size distribution

in Figure

is believed

a surface

surface

2. Some new pores

to cover

V205 is visible

the V205 monolayer

with V205 decreased

the surface

to 0.31 cm 3 g -'. This was primarily

pores as seen

also formed.

yielded

retained

The impregnation

area

by filling

the

13 to 16 nm were

to more than a monolayer

in the X-ray diffraction based on the oxidized

area to 120 m 2 g -1

accomplished

in the range

the pore walls

characteristic

spectra.

surface

Calculation

state model

area of 136 m 2g -' for 20% V205. This was sliyhtly

above

[131 the

of

343

344

measured

value

and suggests

SEM photographs particle

colored

ness. This surface tion:

of used catalyst

at a magnification

are lightbrown

that some V205 is also present

and are covered

layer

but show clearly

The composition the elements Figure

3b shows the surface

structures.

areas

evenly

for the surface

The measured

values

diffusivity

pm particle

yields

the composi-

only be taken as an indication primarily

of V205.

particle.

and sometimes

fibre

show

the inner parts of the particle.

like structure

The more

On this surface

there

areas with quite compact is shown

in Figure

the same values

compact

3c

obtained

area has a composition

is with less Ti in it. area,

particle

from SEM photographs

density were

for NO and NH3 using the method

used

and average

pore radius

in the calculation

described

by Reid

of

[14].

and side reactions understanding

of NO to N2, the individual temperature

Determination

3.6 g 20% V205/Si02-Ti02, 2% 02.

morphology

of surface

to get a better

for a suitable

4

3a shows a cut

layer of 4-8 urn thick-

layer consists

throughout

layer of the cut particle.

as well as the information

reduction

should

of: 8% Si, 20% Ti and 72% V, about

of 6% Si, 7% Ti and 87% V, that

In order

values

of a used catalyst

The area with the porous

Stoichiometry

surface

and X-ray microanalysis

that the surface

of different

and has a composition

effective

3. Figure

of the 710-850

of the inner parts are 7% Sii 72% Ti and 21% V. Dot images

to be distributed

are different

in Figure

by a yellow

is very porous

9% Si, 23% Ti and 69% V. These

of composition

Figure

are shown

of 2500. The inner parts

as multilayers.

of the complex

steps were

range for kinetic

of the stoichiometry dp 6.85-1.4

studied

studies

reactions

to the

and conditions

were determined.

for the reduction

mm, SV 14000h-',

leading

separately

of NO!

426 K, NOin 600 ppm,

345

between

The reaction

The most

literature.

4NH3+6NO

used reactions

----) 5N2+6H20

4NH3+4N0+02

In this study

NO and NH3 have been described

commonly

e

4N2+6H20

NO and 2% O2 the concentration

NH3/N0.

The dotted

accuracy

(17)

of the relatively of catalyst

with a particle

pressures

large

of 0.85-1.40

yields

of catalyst

an

NH3/N0

ratio

in concordance

at high values

used

the result

is the limiting

mole

with

ratio

of

equation

by the de-

of NH3/NO

(14000 h-l). That

to convert

~JJI

of 600 ppm

of the ratio

of NH3 can be explained

space velocity

of a flow type reactor

yielding

The deviations

used is not large enough

ure 4, when the amount

diameter

of NO as a function

a I:1 mole

of the rotameter.

amount

27,18,20,22

of NH3 was varied

at low partial

effect

Simulation

Ref.

the conversion

line represents

(17). The deviation creased

4 shows

in the

(16)

of 14000 h-'. With an inlet concentration

velocity

0.05 to 1.15. Figure

ways

Ref. 15,16,19,21,23

3.6 g 20% V205/Si02-Ti02

was used at a space

in several

are:

are an

is, the

all of the NO introduced.

shown factor

by the solid line in Figat high conversions.

a % f: x m x

20.

10 ‘.

i

O!--M-E450

SD0

31 TEMPERATURE

Figure

5

Oxidation

of NH3.

(K)

(a) With 23 mg catalyst

SV 2 . l-IO6 h-' and 7% 0 2, 800

ppm

NH3.

dp 0.425-0.710

(b) Empty reactor

and 1.8% 02, 620 ppm NH3.

In the presence of oxygen NH3 can be oxidized to NO according which

is thermodynamically

5

NH3+-4 O2 -NO?

3

feasible

between

to reaction

(18)

above 445 K [24].

H20

With an inlet NH3 concentration temperatures

mm,

(18) of 800 ppm the oxidation

470 and 670 K at a space

velocity

was studied

of 2.1.106

at various

h-' and with

an

346 oxygen

concentration

maining

zero below

of 7%. Figure

670 K. In the same figure 620 ppm NH3 and remaining rapidly

the formation

above

investigations

could

without

any disturbances

reactor

rate of oxidation

570 K yielding These

be performed

31 ppm at

is shown.

only

At

1.6 % at 670 K

of NH3 to NO increases

an activation

energy

results

us to conclude

allow

in the temperature

from this particular

of NO re-

reaching

of NH3 to NO reached

at any temperature.

kinetic

concentration

temperatures

of NO in an empty

570 K. The catalytic

temperature

No NO2 was detected

the outlet

at higher

1.8% 02 the conversion

nil below

with

5 shows

570 K increasing

of 116 kJ mol-'. that the

range 470 to 570 K

side reaction.

f t

TEMPERATURE

Figure

6

SV 26000

Oxidation

of NO.(a)

(b) Empty reactor.

h-l,

The reaction

in a system

out a catalyst

in order

or as a result

of a surface

secondly

Figure with

only

to determine

was formed

1.9 g catalyst

NO2 formed creased

increased

amount

V205/Si02-Ti02. activation diffusion

either

with temperature

energy

of 28 kJ mol as will

system.

be corrected

reaching

was obtained.

be shown below.

and

within

outlet.

the

In anempty

both at the inlet independent

of re-

was zero. This meant

The amount

for in further of 26000

h-'

of NO2 wa,s

experiments.

, the amount of

8% of inlet NO at 650 'K. The in-

by catalytic

for NO2 formed -1

of NO2 was almost

a space velocity

of NO2 was thus obtained

limitations

of NO2 at the reactor

of NO over the reactor

and could

system

or catalytically

IO ppm NO2 was detected

in the sampling

with and withhomogeneously

of the sampling

homogeneously

This amount

and thereby

When correction

if any NO2 was formed

in some parts

and the conversion

1.5 % of NO introduced

Using

firstly

reaction

and at the outlet.

that the NO2 detected

mm,

of NO, O2 and N2 was studied

6 shows the concentration

temperature

dp 0.71-0.85

NOin 630 ppm, 2% D2.

2% O2 and 627 ppm NO about

to the reactor actor

1.91 g catalyst,

consisting

if any NO was reacting

reactor. reactor

With

(K)

oxidation

in the sampling

of NO to NO2 over system

was made,an

This low value was caused

by pore

341

/

550

Figure

7

Oxidation

(b) Conversion

Figure

of NO.

(a) Selectivity

the conversion

The conversion

The selectivity

for NO2 was constant

all the NO reacting

6.

to N02. These

for NO2 formation

with temperature

0.7 from 500 to 660 reactions

over

to 9% at 660 K. K. That

are therefore

is, not

proposed:

(191

Se1 NO2 =

0%

(20)

4N0 -----)2N20+02

Se1 NO2 =

0%

(21)

3N0 --)

Se1 NO2 =

33%

(22)

All of these (22) with observed. but will

N2+02

N20+N02.

are thermodynamically

55% reacting However,

reaction

be dealt with

NH3 in the temperature

NO to N2 andO

(22) also gives

before

NE0 which

over V205 catalysts

range 570-870

has been shown

in the range 870-1070

of reaction

give a total

(19) and

selectivity

of 70% as

is not analyzed

in this work,

work.

of NH3 makes

over NiO

A combination

to (19) would

in further

at 620 K, and the presence (21) has been observed

feasible.

according

N20 has been detected

tures

at-around

of N02.

Se1 NO2 = 100%

2N0 --+

with

as in Figure

was low,increasing

was oxidized

2N02

for the formation

of NO and the selectivity

V205/Si02-Ti02.

2N0+02 -

(K)

of NO. Same conditions

7 shows

700

650

600 TEMPERATURE

[26] at 570-870

to occur

in the reduction

K. The selectivities

comparison

difficult.

were

K over CuO

for the direct

7%

(20) and

decompostion

[271. Usually

decomposition

of NO

low, around

Reactions

K and the direct

at 640-760

K are needed

[15,25]

of

tempera-

on NO [B].

348 Reduction

of NO with NH,

Integral perature. perature

conversions.

Figure

range

500-630

Figure

decreased

8

at still

Reduction h-'.

higher

(a) With

of the conversions

below decreased

h-l. Recalculation

in Figure

velocity

yearly

cost. The catalyst

crease

the catalyst

in Figure

8. Kawasaki

reduction

used hare has much

8 some nitric

was very low below

ity was observed

mm,

h-'.

700 ppm, 2% 02.

using th-e kinetics

de-

of 10000 yielded Heavy

Industry

conversions

report

reductions

represented

higher

activity

plants

in

85% of the total and would

thus de-

cost considerably.

When the activity

conclude

0.71-0.85

ND, 247 ppm and 620 K for their commercial

1979 [291. At that time the value of the catalyst

the amount

was catiged by the

mm, SV 590000

pressure

630 K,

as shown above.

1.91 g catalyst,dp

0.71-0.85

h-l), the conver-

to 60, 80 and 89% at 470, 500 and 525 K and

with a space

of 80% at SV 10500 with.inlet

above.

The maximum

high temperatures

to 101 kPa total

conversion

than the ones shown

of tem-

over the tem-

at around

The same flow as in (a). NOin 600 ppm, NH3in

SV 27000

As shown

(590000

reach a maximum

temperatures.

of NH3 at these

of NO with NH3.

termined

higher

space velocity

(b) With 0.10 g catalyst,dp

(c) Empty reactor.

Recalculation

and greater

with temperature,

rate of oxidation

Sv 27000

complete

I

sion was shown to increase

increased

of NO as a function

was virtually

K.

Using 0.101 g of catalyst

and slowly

8 shows the conversion

h-l the conversion

At a SV of 27000

oxide was converted

570 K and only

of the catalyst

below 550 K and only

that homogeneous

was tested

2% NO was converted

of NO with NH3 to any considerable

reactor,

but

5% at 680 K and the same flow as

carrier

or wall catalyzed

in an empty

reactions extent,

at SV 127000

no activ-

at 580 K. Thus we can

do not contribute especially

to the

below 570 K.

349

The kinetics Figure

in the presence

of NO at several pearence

of v^? i the rate of reduction

9 shows

temperatures

at different

temperatures

of NO. At low concentrations is most

apparent

between

of NO as a function

with

the

d

linear

dependence

at

bend down towards

curves

at low temperatures

of the concentration

473 and 583 K. The curves

and at 583 K there

have similar high

origin.

ap-

concentrations

The linear

is almost

no linear

part part

at all. The reaction

order

for NO was determined,

is displayed

in Table

temperature.

The intrinsic

Figure

10 shows

Both curves

2. The apparent

the apparent

have a similar

of 3 to 5 if the pore fusion

limitations.

Figure

9

Figure

IO

There

are not very many

kinetic

slightly

to minimize

(c) 531,

with

in the same manner.

rate of reduction

of NO at 531 K. by a factor

the effects

of pore dif-

of NO with NH3 on the (d) 498 and (e) 473 K.

1.9% 02, SV 2.1.IO6

h-l.

limitations.

constants

Early

investigations

in the temperature

for NO.

(a) Intrinsic

rate.

sions.

The main

over the whole

criticism

of the kinetics

against

temperature

using

have used too narrow

a

the kinetics

have been determined

on much

[24,301

an apparent

report

range 473-673

integral

the procedure range.

of this reac-

Some authors

In these early investigations

were evaluated

orders

investigations

and in most cases

catalysts.

l:st order

thorough

in the literature.

for NO of 0.3-0.5

ported

(b) 547,

of pore diffusion

range of concentrations

order

increasing

rate.

tion on V205 catalysts

less active

be designed

mm, NH3in 860 ppm,

and the result

and the rate could be increased

of the rate of reduction

The influence

(b) Apparent

is 0.7-0.8

above,

also increasing

and the intrinsic

could

of NO at (a) 583,

dp 0.425-0.710

order

is 0.5-0.6

appearence

system

The dependence

concentration

order

as described

This

K. Later

Inomata

the reaction

conversion

data with

I311 re-

orders

and

high conver-

is the use of constant is especially

reaction

reaction

true for temperatures

above

570 K where

side reactions

der for NO and 0:th order ture. Stringaro

[25] introduced

into consideration

Figure

11

12

The dependence

11 shows

temperatures

the rate

NO, the curvature

creases

(b) 556,

rate

to a constant apparent

order

at 540 K. The ent one

increases

value.

temperature.

of NO, with

In a recent

order

report

Wong

temperature.

with

order

reaction

for the

for NH3 in-

order

obtains

to 0.27 at 583 K. This negative

of NH3 at low temperatures.

At higher

obtained

values

below

5%

the rate levels

are displayed

the same trend

in Table

this constant

but is higher

off

2. The value

than the appar-

for NO and NH3.

[331 determined

assuming

reaction

can be

rate expression.

a term accounting

on the rate. At concentrations

orders

of

But, as was the case for This behaviour

from 0.27 to 0.10 and reaches

to the orders

then the rate constant,

h-'.

The rate of reduction

The intrinsic

adsorption

follow

NH3 on the (e) 475 K.

a Langmuir-Hinshelwood

with concentration.

for 02 decreases

(e) 472 K.

than was the case for NO. At low

and increases

of oxygen

NH3 on the

(d) 499 and

show that the apparent

The reaction

intrinsic

in contrast

with

a rate expression

at low temperatures

12 shows the effect

the

slowly

of NH3 and thereby

1341 developed

taking

.

on NH3 concentration.

increases

is also a sign of the strong

oxygen

h-'

on NH3 concentration.

is independent

of the plots

0.1 to 0.27 with

values

Figure

(c) 530,

much more

of NH3. Our result

from

negative order

concentration

by adsorption

In fact Miamoto adsorption

of NO with

(c) 518, (d) 497 and

of the rate of reduction

the dependence

with

l:st orof tempera-

type rate expressions

mm, NOin 840 ppm, NH3in 890 ppm, SV 2.1~10~

NO increases

[321 assumed independent

side reactions.

(b) 544,

of O2 at (a) 577,

dp 0.425-0.710

explained

orders

mm, NOin 990 ppm, 1.8% 02, Sv 2.1.106

concentration

Figure

reaction

of the rate of reduction

of NH3 at (a) 581,

dp 0.425-0.710

Figure

Miyamoto

important.

Langmuir-Hinshelwood

also the possible

The dependence

concentration

become

for NH3 also with

constant

the reaction reaction

order

orders

for 02 at 470 K and

from 370 to 570 K. This

351 resulted

in an apparent

results

confirm

They

10 wt% V205 on Ti02. order

for O2 while

TABLE

2

Kinetic

order

for 02 of 0.25 and an intrinsic

our observations.

The catalyst

used

used O2 concentrations

we used concentrations

in their

of 0.15-1.7%

order

of 0.45. Their

investigation

was

to get the reaction

from 0.5 to 13%.

investigations.

Temperature

Apparent NO

258.2

Intrinsic

order

NH3 -

-

Intrinsic rate constant

NO

NH3

02

0.545

0.212

0.263

82

0.740

order

204.8

310.2

0.789

-

-

0.629

0.270

0.242

200.5

0.765

-

-

0.584

-0.152

0.420

225.1

0.738

-

-

0.538

0.054

0.386

273.2

0.782

-

-

0.617

0.243

0.261

245.7

0.273

-

0.578

0.205

0.302

420

308.1

0.239

-

0.614

0.286

0.242

5607

198.8

0.108

-

0.550

-0.135

0.431

271.7

0.266

-

0.592

0.206

0.264

224.2

0.169

-

0.565

-0.034

0.392

202.4

0.261

0.553

-0.137

0.413

303.5

0.105

0.612

0.268

0.242

225.7

0.203

0.566

0.060

0.386

282.7

0.095

0.600

0.253

0.228

256.9

0.122

0.584

0.209

0.255

The apparent versus

activation

i

energy

(Eaapp

l/T for 20% V205/Si02-Ti02

) determined

is 28 kJ mol-'

6597 2.433 24.74 1921

0.8939 571.5 9.322 0.6864 3930 40.13 1301 285.5

from plots of ln (rNOapp)

in our work and is close

to

29.7 kJ mol-' was 49.7

obtained by Wong 1331 for 10% V205/Ti02. On 1% V205/Si02-Ti02 -1 kJ mol and the intrinsic value (Eaintr) 60.3 kJ mol-' indicating

that there were catalyst

on

half the value orders

still

the other

some pore diffusion hand

of Eaintr

is severely

as observed.

for NO and NH3 also follow

limitations limited

The difference

the general

on this catalyst.

by diffusion

pattern

Eaapp

The 20%

and so Eaapp is about

between

apparent

for severe

and intrinsic

pore diffusion

limitations. The activation is 188 kJ mol-'

energy

cause of the variation parable

with

Hinshelwood

determined

and is much

higher

of reaction

the others.

from the intrinsic

orders

This variation

type rate expression

rate constants

than the ones obtained

would

with

temperature

instead

indicates

describe

data

before

in Table

[24,30,33].

our value

2

Be-

is not com-

that a Langmuir-

in a better

way.

Indeed

pre-

352 Timinary model.

calculations

show that better

The derivation

of an appropriate

for further tained

work

because

enough

correlations

could

rate expression

valuable

information

be obtained

using

of this type will

for design

purposes

this

be left was ob-

from our model.

ACKNOWLEDGEMENTS This work was supported We are also grateful Mrs Laila

Holm

by the National

to Mrs Birgitta

for typing

Swedish

Svensson

Board

for Technical

for technical

assistance

Development. and to

the manuscript.

REFERENCES J. Ando, EPA-600/7-83-027, May 1983. T. Shikada, K. Fujimoto, T. Kunugi and H. Tominaga, Ind. Eng. Chem. Prod. Res. Dev., 20 (1981) 91. C.U.I. Odenbrand and S.L.T. Andersson, J. Chem. Tech. Biotechnol., 32 (1982)365. H. Meier and G. Gut, Chem. Eng. Sci., 33 (1978) 123. D.E. Mears, Ind. Eng. Chem. Process Des. Develop., 10 (1971) 541. Selected Values of Properties of Chemical Compounds, Thermodynamic Research Center, Texas, 1984. B.J. Gliddon and R.R. Cranfield, Brit. Chem. Eng., 15 (1970) 481. r: J. de Acetis and G. Thodos, Ind. Eng. Chem.,52 (1960) 1003. J.J. Carberry, Chemical and Catalytic Reaction Engineering, McGraw-Hill, 9 New York, (1976) 533. K.B. Bischoff, A.1.Ch.E. Journal, 11 (1965)" 351. Kirk-Othmer Encyclopedia of Chemical Technology, Val.20, 3rd Ed., John Wiley & Sons, New York, (1982) 766. K. Tanabe, M. Ito and M. Sato, J.C.S. Chem. Comm., (1973) 676. J. Solid State Chem., 42 (1982) 263. 1: A. Andersson, 14 R.C. Reid, J.M. Prausnitz and T.K. Sherwood, The Properties of Gases and Liquids, McGraw-Hill, New York, (1977) 548. G.L. Baurle, S.C. Wu and K. Nobe, Ind. Eng. Chem. Prod. Res. Dev., 14 (1975)268. Progress,1 (1982) 84. 165 S.T. Darian, J.W. Eldridge and J.R. Kitrell, Environmental 17 S. Matsuda and A. Kato, Applied Catalysis, 8 (1983) 149. 18 S. Matsuda, M. Takeuchi, T. Hishinuma, F. Nakajima, T. Narita, Y. Watanabe and M. Imanari, J. Air Pollut. Control Assoc., 28 (1978) 350. 19 D.E. Powell and K. Nobe, Chem. Eng. Commun., IO (1981) 103. 20 P. Avila, J. Blanc0 and M. Moreno, Proc. 9th Iberoamerican Symp. on Catal., Lisboa (1984) 817. 21 L. Marzo, L. Fernandez and J.A. Gros, Proc. 9th Iberoamerican Symp. on Catal., Lisboa (1984) 826. 22 A. Kato, S. Matsuda, T. Kamo, F. Nakajima, H. Kuroda and T. Narita, J. Phys. Chem., 85 (1981) 4099. Progress, 1 (1982) 24. 23 P.M. Hirsch, Environmental 24 S.C. Wu, Dissertation Univ. California, Los Angeles, 75-21, 673 (1975). 25 J.P. Stringaro and G. Gut, Chimia, 35 (1981) 231. T.G. Alkhazov, G.Z. Gasan-Zade and T.K. Mukerdzhi, Kinet. Katal., 23 (1982) 89. 2'; E.R.S. Winter, J. Catal., 22 (1971) 158. 28 R.T. Yang and K.-T. Li, Ind. Eng. Chem. Prod. Res. Dev., 21 (1982) 405. Stearns-Roger,EPRI-CS-2713, Nov. 1982. 29 30 G.L. Baurle, S.C. Wu and K. Nobe, Ind. Eng. Chem. Prod. Res. Dev., 17(2) (1978) 117. 31 M. Inomata, A. Miyamoto and Y. Murakami, J. Catal., 62 (1980) 140. 32 A. Miyamoto, Y. Yamazaki, T. Hattori, M. Inomata and Y. Murakami, J. Catal., 74 (1982) 144. 33 W.C. Wong, Diss. Univ. California, Los Angeles, (1982).

: