B. Delman and G.F.

Froment (Editors), Catalyst Deactivation 1987 625

1987 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

POISONING AND PASSIVATION OF AMMONIA S-rnTHESIS CATALYST AT

VARIOUS TEMP~~ATURES

A.V. KRYLOVA, N.V. NEFEDOVA and T.M. PEEV

D.I.Mendeleyev Chemical Technological Institute, Moscow, USSR
Institute of Chemical Engineering, Burgas, Bulgaria

ABSTRACT
The mechanism of interaction of catalytic poisons (ox~gen,
carbon oxide and dioxide, water vapours and their mixtures)
with the ammonia synthesis catalyst has been studied in a wide
range of temperatures (550 + -195C) by using chemisorption,
isotopic methods, ~-resonance spectroscopy, non-stationary
catalysis, derivatography, and others. It has been found that
the poisoning of ammonia synthesis catalyst at intermediate
catalysis temperatures (~400C) is intensified, in particular,
due to the presence of non-hydrated forms of oxygen-containing
compounds on the surface. The results obtained were used for
developing the routes of reactivation of poisoned catalysts as
well as effective express methods for the elimination of catalyst
pyrophoricity (stabilization) at various temperatures (below the
catalysis temperature, intermediate and regular reaction tempe-
ratures)

INTRODUCTION
Oxygen and oxygen-containing compounds (H CO, CO are
20, 2)
well known poisons of the ammonia synthesis catalyst. It is
customary to assume that due to the presence of excessive oxygen
all oxygen-containing compounds in the process of synthesis
rapidly pass into H20, which is usually confirmed by the equiva-
lent poisoning effect of 100 ppm 02 and CO 2 or 200 ppm CO
and H20. However, not all experimental data on poisoning can be
interpreted in terms of this approach. A more detailed consider-
ation of the reasons for the well-known phenomenon of the
enhanced poisoning effect of oxygen-containing poisons at inter-
mediate catalysis temperatures (~400C) was of interest, es-
pecially in the context of the search for low temperature ammonia
synthesis catalysts. The phenomenon of mild surface oxidation at
the temperatures much lower than those of the reaction (20 ~ 95C)
have for a long time been used to eliminate pyrophoricity-passi-
vation [1] , which can be regarded as a particular case of re-
626

versible surface poisoning. Recently the possibility of using

the reversible poisoning of ammonia synthesis catalyst by oxy-
gen, CO 2 and mixture of oxidizers (C0 2, H20, 02) has been
proposed for the elimination of pyrophoricity (stabilization)
over a much wider ranSe of temperatures, for example, by air
treatment at -83 f -195C and above 400C, by CO 2 and mix-
tures of oxidizers (C0 2 + H2 0 , CO 2 + H2 0 + 02) at the tempera-
ture from -78C to -17C and from 100C to 500C (2-61
In the present paper the results obtained in the study of
the interaction of oxygen, oxygen-containing poisons (CO, CO 2,
H2 0 and their mixtures) with the catalyst at various catalysis
temperatures with respect to their poisoninb effect upon the
reaction and the use of the reversible catalyst poisoning ef-
fect for the elimination of pyrophoric properties have been
compared.

EXPERIW~~TAL LreTHODS
All the experiments were performed by using the standard
CA-IB industrial passivated catalyst. The grains of industrial
samples were ground in an air flow and a fraction 1-2 mm was
separated. The sample of the catalyst (0.2-4 g) was reduced in
a hydrogen flow at the usual space velocity of 5000 h- 1 at
a stepwise temperature increase and was kept for 1-3 hours,
with the final reduction temperature being equal to 500-530C.
Adsorption and desorption of oxygen-containing compounds at
various temperatures were studied in a pulse chromatography in-
stallation in helium and hydrogen flows using the method of
regeneration of sorption sites [71 For that purpose sorption
was carried out until the saturation with pulses reached the
range 0.1 - 1 cm3 at 20C, the velocity of the carrier-gas
being 20-100 cm3/min (as a rule, the partial pressure of the
adsorbate did not exceed 25 mm of m.c.). Then the catalyst
with the adsorbed gas was heated in the carrier gas flow for
certain time up to 50C, then cooled down and the second ad-
sorption was carried out at 20C. Using the same procedure
heatings were further carried out successively at much higher
temperatures (100, 150, 200, 300, 400, 500C), ~ach heating
was followed by cooling down to 20C and measuring the adsorp-
tion. The value of the gas adsorption after each heating made
it possible to determine the number of sorption sites regene-
rated at the temperature of heating. Comparison of regeneration
 627

of the sites in a~elium flow with that in a hydrogen flow alow-

ed one to identify and consider separately the processes of de-
sorption, thermal decomposition and hydration of the adsorbate.
The same installation was used for studying the effect of the
poisons upon the non-stationary catalysis. The reaction was
carried out by feeding the pulses of another reagent into nit-
rogen or hydrogen flows or the pulses of the nitrogen-hydrogen
mixture into a helium flow. The ammonia formed was absorbed by
water and titrated with sUlphuric acid. Regular oxygen pulses
were introduced into the carrier gas flow to study the poison-
ing. The catalysts were also treated with poisons in a flow-type
installation in a flow of purified and dried nitrogen contain-
ing admixtures of oxygen, carbon dioxide, mixtures of helium
and water vapours, mixtures of carbon dioxide and water vapours
or oxygen, mixtures of nitrogen with oxygen and water vapours
and mixtures of oxygen, carbon dioxide and water vapours of va-
rious composition. The rate of the interaction of the catalyst
wi th the oxidizer was characterized by a temperature increase
(pyrophoric heating) 4 T, C. Romomolecular and heteromolecular
oxygen exchanges were studied with the use of 0 18 isotope in
stationary circulation installation linked with a mass-spectro-
meter. The phase composition of the catalysts treated with dif-
ferent oxygen-containing compounds as well as their resistance
to disintegration of the surface layers were determined by using
-resonance spectroscopy and derivatography.

I. Interaction of the Catalyst with Poisons at Various

~emperatures. Reactivation of Poisoned Catalysts
Fig. 1 and 2 show the temperature dependences of the part
of chemisorption sites of 02r O and CO 2 regenerated in the flows
of R2 and Re. Analysis of the dependences has resulted in a
probable model of the poisoning process of the catalyst at dif-
ferent temperatures. In all the cases a higher extent of rege-
neration of sites in bydrogen(curves 1,3) is indicative of the
fact that the processes of poisons hydration playa decisive
role only at high reaction temperatures. In fact, chemisorbed
oxygen and carbon dioxide are completely hydrated on the cata-
lyst surface only at 500C, while under the experimental condi-
tions a complete hydration (a hundred per cent hydration) of CO
is not observed even at 500C. At lower temperatures (even at
LfOOOC) a considerable part of metallic centres is poisoned by
628

strongly bonded non-hydrated forms of poisons interaction (02'

CO 2, CO) with the catalyst. It should be noted that during the
a~Jonia synthesis, i.e., in the presence of hydrogen, the ca-
talyst will react with a mixture of poisons with water vapours
which result from hydration of the poison, for example, such
mixtures as 02 + H20, CO 2 + H20, etc. A simultaneous effect of
CO 2 + H20 + 02 and other mixtures is also possible. The fact
that the poisons cannot be removed from the surface in a helium
flow even at 500C show that they may strongly interact with
the catalyst resulting in the formation of chemical compounds.

r
80

60

40
2
20

roo 200 300 400 500 T,oC

Fig. 1. Rate of regeneration of chemisorption sites of 02 in
H2(1) and He (2)

The possible types of surface compounds of various poisons

with the catalyst have been studied in the experiments on treat-
ment of the reduced catalyst in a hydrogen-free flow of poisons
at various temperatures. Fig. 3 and Table 1 show the Mossbauer
spectra of the catalysts treated by the above-described proce-
dure as well as the samples of phase composition specified by
the data on the spectra. The data given show that the action of
poisons CO 2 and 02 at 350C identically results in the formation
of very close amounts of magnetite, while at 400C, unlike 02'
the interaction of CO 2 may result in the formation of both
magnetite and carbide. It should be noted that with the increase
of the temperature from 350 to 400C the degree of iron oxida-
tion in 02 decreases, while it sharply increases in CO 2 1~ese
data permit to conclude that similar compounds can be formed on
 629

t ,%
roo
80

60

40
20

roo 300 500 700 T ,c

Fig. 2. of CO and

e.,
I-
Z
::l
o

u~
" " "" ! ! Fe,5Q+
! ' ! , 8-Fe~C

Fig; 3. Mossbauer spectra of the CA-IB samples treated with

various poisons (the numbers of the curves correspond
to those given in Table 1)
the surface and poison the catalyst during the synthesis since
the rate of hydration of the poisons and these compounds is not
sufficiently high at intermediate temperatures. The catalysts
treated with a mixture of 02+H2 0 exhibited much higher de~ree of
oxidation (at 350"-17% l!'e 0 4,at 400o-33~;;), than those treated wi th
3
02 only (at 350"C-10% Fe 4,at 400"-83;). Also there can be ob-
30
served different effect of the increasing temperature: 02
decreases the oxidation of iron, a mixture of 02 + H20
630

'l'ABLE 1
Phase Composition of CA-IB Catalyst Samples ~reated with
Various Poisons

'l'emperature
of Poison -J
~-~e
n ~,
~e3
4 e -m
-~e3 -~e2
'I'empera ture
C Ene of the pro-
tective layer
treatment destruction

1%02+2.4% H2O 83 17
CO 90 10 108
2
350C CO 2+2.4%H2O+
+ 1% 02 78 18 4 91
CO 2+1.4% H2O 86 6 13
1% 02 89 11 90
----------------------------------------------------------------
1f.o 2+2 .4~~ H2O 67 33
400C CO 52 28 20 143
2
CO 2+2.4% H2O+
+1% 02 44 43 13
1~~ 02 82 8

sharply increases it. At 350 the interaction of a mixture of

CO 2 + H20 with the catalyst results in a decrease of the rate
of magnetite formation from 10 to 6 % with the simultaneous
formation of the carbide phase e
-Fe 2C in the catalyst. Ad-
dition of another poison (02) to the same mixture increases the
amount of magnetite formed, decreases the content and changes
the form of carbide. Hence, the effect of the poison mixtures
upon the catalyst is complicated and is not always of additive
character.
'l'hus , the nature and the amount of the compound s formed
depend not only on the interaction of poisons with the catalyst
but also upon the nature and composition of the poison mixture.
Analysis of the curves of thermal regeneration of sorption sites
and spectral data given with an account of the literature results
concerning the study of interaction of poisons with the catalyst
carried out by other methods permit to conclude that at inter-
mediate catalysis temperatures (~400C) the poisoning by dif-
ferent poisons should be of individual character for each parti-
 631

cular poison due to the blocking of various sites, different

changes in the electronic properties of the surface,different
temperature dependences and the nature of the surface compounds
which are formed upon the interaction of the catalyst with the
poisons and their mixtures with the vater vapours. The lower the
reaction temperature, the less the hydration rate of the poisons
and the greater is the contribution of the individual mechanism
of poisoning by each particular poison. The results obtained
allowed one to develop the routes for reactivation of the ca-
talysts poisoned by the poisons (studied at intermediate tempera-
tures) based on the increasing of the catalyst temperature for
hydrogenation of the forms of the strong interaction of poisons
with iron in the catalyst. The routes of reactiovation take in-
to account the temperature and composition of the poisoning
mixture which determine the probable composition of the surface
compounds of the catalyst.
At the intermediate synthesis temperatures the catalyst
exhibited an ability to considerably adsorb the reaction product,
ammonia, which, in our opinion, also can poison the process as
the temperature decreases [7] The implementation of the reac-
tion along the pulse route made it possible to increase the con-
versiun of the reagents, the effect becoming more pronounced as
the temperature decreased, and at 200C the effect increased
more than 102 times [8J Implementing the ammonia synthesis
through chromatographic route [8J no oxygen poisoning of the
catalyst was observed (see Table 2), which indicates a possibi-
lity of a decreased poisoning effect of oxygen upon the transi-
tion to certain non-stationary routes of the reaction implemen-

TABI& 2
~ffectof Oxygen upon the Rate of Reagent Conversion (oG) in
Chromatographic Regime at Vpulse/Vcat. 0.125

Tempe- NH .% in the absence NH .% in the presence

rature, 3 of oxygen 3 of oxygen
C --------------------onr--------------------------~----
H in a N in a ABC in a H in a N in a ABC in a
N~ flow H~ flow He flow N~ flow H~ flow He flow
----------------------------------------------------------------
450 3.7 3.4 1.7 3.8 3.3 1.9
300 3.2 3.0 1.5 3.6 3.1 1.5
200 2.7 2.2 1.2 2.6 2.4 1.4
The ratio of the pulse volume to the catalyst volume
3H2 + N2
632

tation. The result can be interpreted by the poison adsorption

at the most active catalyst sites with the subsequent displace-
ment of ammonia and removal of the poison during the intervals
between the pulses.

2. Peculiarities of Interaction of the Catalyst with

Oxygen during Passivation. The Use of Effects of
heversible Poisoning by Oxygen-containing Poisons
for Elimination of Pyrophoricity (S~abilization)
of the Catalyst

Comparison of interaction of oxygen with the catalyst at

various temperatures (of poisoning and passivation) was carried
out by means of the method of isotopic exchange (91 Homomole-
cular and heteromolecular oxygen exchange was studied by using
a reduced sample obtained after keeping it in oxygen at various
temperatures as well as a sample in the oxide form. Fig. 4 and
5 show examples of kinetic dependences thus obtained. Homomole-
cular exchange was observed on the samples treated in a nitro-
gen-oxygen mixture in the temperature ranbe between -78 T 80C.
Exchange kinetic curves (Fig.4) are described by the equation
of the first order, though these curves do not pass across the
beginning of the coordinates. Therefore the reaction rates were
calculated against the linear part of the curves. The activa-
tion energy of the exchange calculated with respect to the rate
constants was equal to 8.5 kJ/mol.
_ g(1-F).fil'mole/t Q FE

1.2 I

0.8

0.4
-5

t: min
B'it;. 4. Kinetics of the homomolecular exchange on the passivat-
ed catalyst: 1 - 4 - 293 K, 5 - 195 K
 633

~he exchange reaction proceeds at low temperatures due to

the presence of passivated catalysts of weakly-bonded forms of
non-oxide oxygen on the surface. In the course of a series of
subsequent experiments the catalyst activity in the exchange
decreased (curves 1-4, Fig.4), which indicates a gradual transi-
tion of weakly-bonded oxygen into strongly-bonded ones. The
heteroisotopic exchange was also observed on the passivated
catalysts at -78 ~ 20C. ~he number of molecules participating
in a heteroexchange calculated by a change in isotope content
was 0.1-0.5 10 19 molecules/m2Fe.
-l9(f-F). fO-"mo!e/m 2Fe
4250 3400
1.2

0.8

0.4

40 60
1; min
Fig. 5. Kinetics of the isotopic exchange on the oxidized
catalyst

The fact that weakly-bonded oxygen was adsorbed on the

metallic surface was confirmed by the absence of low-temperature
exchange reaction on the catalysts in an oxide form. Homomolecu-
lar exchange on the catalysts in an oxide form (magnetite form)
takes place at a noticeable rate only at 300C. The exchange
kinetics (Fig.5) was described by the equation of the first
order. The exchange was followed by absorption of oxygen and
according to the data on the chemical analysis by oXidizing a
portion of Fe(II) into Fe(III). The activation energy of the
reaction in the temperature range between 300 and 500C was
equal to ""'125 kJ/mol and was close to the temperature known
from the literature data for Fe203. Heteroexchange is observed
in the same temperature range. Exchange on the catalysts parti-
ally oxidized at 300-500 0C proceeded with the characteristics
634

obtained for the sample in an oxide form.

~herefore, the forms of interaction of the catalyst with
oxygen at the temperatures of synthesis and passivation (20C)
are different. At the synthesis temperatures oxide oxygen is
formed. It was found that passivated catalysts contain weakly-
bonoed oxygen which occupies 5-10% of the metallic sites of the
surface and slowly passes into strongly-bonded oxygen. The total
amount of oxygen absorbed upon passivation is equal, according
to our data, up to 10 molecular layers on the metallic surface;
the most part of oxygen is strongly bonded (oxide, strong chemi-
sorption), embraces near-surface layers and the whole passivat-
ing layer appears to be non-homogeneous.
The lengthy passivation process was due to a slow transition
of weakly-bonded oxygen into strongly-bonded one and insufficient
stability of the protective structure was explained by the fact
that the sites of weak adsorption were unprotected, these sites
can be covered by the molecules of competing adsorbates of water
or even hydrogen in the proce ss of pas si va tion l 2] and serve
as initiating sites of oxidation upon the contact of the catalyst
with air.
It follows from the results obtained that stable protective
structures in the conditions of passivation are not formed as a
result of slow transition of weakly-bonded oxygen into strongly-
bonded one on the sites and a high rate of oxygen diffusion into
the catalyst volume leading to a bulk structure of the passivat-
ing layer. Therefore, it was feasible to choose such conditions
of stabilization under which the formation of weakly-bonded oxy-
gen does not occur and rate of interaction of oxygen with the
surface considerably increases the rate of oxygen diffusion into
the catalyst volume. Such conditions appear to be high concen-
tration of oxygen in the gas phase which provides high concen-
tration of oxygen on the surface and low temperature which pro-
vides a decrease of the diffusion rate, or high temperature at
which, as shown previously [6] in Fig. 6 the oxidation degree of
the catalyst by oxygen decreases. The study of interaction of
the catalyst with air at low temperatures, in fact, allowed one
to identify the range of temperatures providing the formation of
strong protective structures which help eliminate pyrophoricity.
;l'hese structures are formed on the catalysts studied in the
range of temperatures from -195 to -83C. ~11e studies resulted
 635

in an express method of' loVi temperature stabilization of the

pyr-ophor rc catalyst which practically provides instantaneous
formation of the stable pr ote c td ve layer. Elimination of the
Vleak chemisorption of oxygen due to interaction with oxygen at
hibh temperature served as a basis for effective high-tempera-
ture ~L+OOOC) stabilization of pyr ophor Lc systems with air and
oxygen [2,31
t1Tox. ~Tai'l.

100 375
80 300

60 225
40 150
20 75

-100 o 100 200 300 400 500

Fig. 6. Dependence of the value of pyrophoric heating of the

CA-IB catalyst upon the temperature of the oxidation start in
the flows: 1 - CO2 , 2 - H20, 3 - CO 2 + H20 + 02' 4 - air.

The dependences obtained in the study of reversible inter-

action of other poisons with the catalyst at various tempera-
tures provided a basis for the development of stabilization of
the catalysts by carbon dioxide and mixtures of oxygen-contain-
ing poisons at intermediate and low temperatures.
Fig. 6 lists the temperature dependences of pyrophoric
interaction of the catalyst with oxygen (air), water vapours,
carbon dioxide and mixtures of poisons, these dependences being
determined with respect to increasing temperatures of the ca-
talyst upon the reaction of pyrophoric oxidation (AT ). Com-
ox
parison of the curves indicates a sharp difference between the
action of oxygen and other oxidizers in the region of regular
and intermediate (~ 400C) synthesis temperatures. At regular
temperatures very small amount of oxygen interacts with the
catalyst, while carbon diOXide, water vapours and oxidizers
mixtures, on the contrary, intensivelJ' oxidize the catalyst
636

iron. At the temperatures below 400C and especially below

300C, on the contrary, the interaction of all the poisons with
the catalyst, except oxygen, is slight. In accordance with
these data it was found that the LIossbauer spectra of the
catalyst treated with air or oxygen at 200-300 oC indicate the
maximum concentration of magnetite, while after the treatment
at 500C the samples contain hardly identified traces of oxides
t61 .
It has been mentioned that comparison of the properties of
the catalysts treated with an individual oxidizer of their mix-
tures shows the effects of non-additivity of the mixtures action.
After treatment with certain poisons under definite conditions
the pyrophoric ammonia synthesis catalysts did not exhibit any
pyrophoricity upon contact with air, i.e., they stabilized. At
certain temperatures of interaction the synergistic effect is
observed. It means that when H20 and CO2, for example, do not
stabilize the catalyst (300C), a mixture of these poisons
readily stabilizes the catalyst at any temperature above 100C.
The new effective express methods of stabilization of the ca-
talysts with carbon dioxide and the oxidizers mixtures at va-
rious temperatures ranging from -78 to 500C were based on the
series of the results obtained with respect to the interaction
of the catalyst with CO 2 and the oxidizers mixtures (51
The derivatograms given in Fig. 7 indicate that the cata-
lysts stabilized with CO2 at intermediate temperatures are more
resistant to oxidation in the air than those passivated by means
of the traditional method.
600 r - - - - - - = - - ; - ; - - - : - - - - " " 7 !
500
400

300

200

roo
Fig. 7. Derivatograms of oxidation of 1 passivated, 2 stabilized
CO2 (To z 400C)
 637

Thus, it follows that the peculiarities of interaction of

the ammonia synthesis catalyst with CO 2, H20 and their mixtures
which are different from the peculiarities found upon interac-
tion of the catalyst with 02 demonstrated the possibilities
of effective elimination of the catalyst pyrophoricity at inter-
mediate and low temperatures. Reversibility of the surface
poisoning upon stabilization is confirmed by the fact that after
activation the catalytic activity is preserved in the reaction
mixture at regular temperatures of the reaction \51
In general, the study performed have shown the complexity
of the processes of interaction of various oxygen-containing
poisons and their mixtures with the catalyst surface under the
reaction conditions at intermediate temperatures. Under these
conditions each of these poisons or their mixtures can interact
with the catalyst according to an individual mechanism. The in-
dividual character of interaction is confirmed by various re-
gimes of effective stabilization of the pyrophoric catalyst with
different poisons or their mixtures and it should be taken into
account when a certain route of regeneration of poisoned cata-
lysts is chosen.

1 Catalyst Handbook with Special Reference to Unit Processes

in Ammonia and Eyc1rogen Manufac bure , \"Iolfe Sci. Book,
London. 1970, 222 pp.
2 A.V.Krylova, G.A.Ustimenko and N.S.Torocheslmikov, in
G.Poncelet, P.Grange and P.A.Jacobs (Ed.), Proc. 3rd Int.
Symp. Preparation of Catalysts III, Louvain-la-Neuve, Belgium,
September 6-9, 1982, ~lsevier, Amsterdam, 1983, pp.441-450.
3 A.V.Krylova, G.A.Ustimenko, N.V.Nefedova, T.M.Peev and N.S.
Torocheclmikov, Appl.Catal., 20 (1986) 205-213.
4 A.V.Krylova, N.V.Nefedova, T.M.peev, T.M.Stoilova, Kinetika i
Kataliz, 27 (1986) 520.
5 A.V.Krylova, N.V.Nefedova and N.S.Torocheshnikov, B.Delman
(Ed.), Proc. 4th Int.Symp. Scientific Base5 for the Prepara-
tion of Heterogeneous Catalysts, Louvain-la-Neuve, Belgium,
September 1-4, A7, Elsevier, Amsterdam, 1986, Preprint of
paper A7, pp. 1-10.
6 A.V.Krylova, T.M.Peev, G.A.Ustimenko, Kinetika i Kataliz,
27 (1986)
7 A.V.Krylova, A.A.Veselkova, N.S.Torocheshnikov, Proc. All-
Union Conference on Mechanism of Catalytic Reactions, Moscow,
USSR, 1978, Nauka, pp 226-233.
8 A.V.Krylova and N.S.Torocheshnikov, in O.V.Krylov (Ed.),
Proc. 4th All-Union Conference on the Mechanism of Catalytic
Reactions, Moscow, USSR, April 19-22, 1986,part II,pp.188-192.
9 A.V.Krylova, L.A.Kasatkina, V.V.Morozov, Proc.All-Union se-
minar"Isotopic methods in studying the catalysis mechanism"
Novosibirsk, USSR, 1980, preprint of paper 33, pp. 1-4.