Reaction pathways and the role of solvent in the hydrogenation of chloronitrobenzenes

Article abstract of DOI:10.1135/cccc20031819

Liquid phase hydrogenation of chloronitrobenzenes to corresponding chloroanilines over Pd on carbon (Pd/C) under mild reaction conditions was studied. On the basis of ¹H, ¹³C NMR, GC-MS and HPLC analyses of reaction mixtures, the reaction pathways were evaluated. The reduction of substrates proceeds via the formation of chloronitrosobenzenes and N-(chlorophenyl)hydroxylamines and mainly results in the formation of the chloroanilines and aniline. Aniline is formed by hydrogenolysis of chlorine (dechlorination) in benzene ring. Other compounds (mono- and disubstituted azobenzenes and azoxybenzenes) were also detected by GS-MS and HPLC (<3 mole %). The used solvent influences the reaction mixture composition and catalyst activity.

Full text of DOI:10.1135/cccc20031819

Hydrogenation of Chloronitrobenzenes
1819
REACTION PATHWAYS AND THE ROLE OF SOLVENT IN THE
HYDROGENATION OF CHLORONITROBENZENES
a
b
a1
Vieroslav KRÁTKY , Milan KRÁLIK , Alexander KASZONYI
,
Magdaléna STOLCOVÁa2, Lubomír ZALIBERA , Miroslava MECÁROVÁ
and Milan HRONEC
c
a
a3,
*
a
Department of Organic Technology, Slovak University of Technology, Radlinského 9,
812 37 Bratislava, Slovak Republic; e-mail: kaszonyi@chtf.stuba.sk, stolcova@chtf.stuba.sk,
hronec@chtf.stuba.sk
VÚCHT a.s., Nobelova 34, 836 03 Bratislava, c/o Department of Organic Technology,
Slovak University of Technology, Radlinského 9, 812 37 Bratislava, Slovak Republic;
e-mail: kralik@chtf.stuba.sk
Department of Physical Chemistry, Slovak University of Technology, Radlinského 9,
812 37 Bratislava, Slovak Republic; e-mail: zalibera@cvt.stuba.sk
1
2
3
b
c
Received Jun e 24, 2003
Accepted August 8, 2003
Liquid ph ase h ydrogen ation of ch loron itroben zen es to correspon din g ch loroan ilin es over Pd
on carbon (Pd/C) un der m ild reaction con dition s was studied. On th e basis of 1H, 13C NMR,
GC-MS an d HPLC an alyses of reaction m ixtures, th e reaction path ways were evaluated. Th e
reduction of substrates proceeds via th e form ation of ch loron itrosoben zen es an d N-(ch loro-
ph en yl)h ydroxylam in es an d m ain ly results in th e form ation of th e ch loroan ilin es an d an i-
lin e. An ilin e is form ed by h ydrogen olysis of ch lorin e (dech lorin ation ) in ben zen e rin g.
Oth er com poun ds (m on o- an d disubstituted azoben zen es an d azoxyben zen es) were also de-
tected by GS-MS an d HPLC (<3 m ole %). Th e used solven t in fluen ces th e reaction m ixture
com position an d catalyst activity.
Keyw ord s: Reaction path ways; Ch loron itroben zen es; Catalyst activity; Selectivity; Deactiva-
tion ; Catalyst surface h ydroph obicity; Solven t effects; Heterogen eous catalysis; Reduction s;
Palladium .
Man y efforts h ave been recen tly devoted to a m ore detailed description of
th e reaction path ways in th e h ydrogen ation reaction s of ch loron itro-
ben zen es (CNB)1–3. Th eir true com plexity sh own by Sch em e 1 is n orm ally
n ot visible but m an ifests itself wh en developin g th e process. Man y varia-
tion s in product com position could be obtain ed un der differen t con dition s.
Ch oice of a proper catalytic an d solven t system un der suitable reaction
con dition s provides th e possibilities to en h an ce th e reaction path ways lead-
in g to th e desired products, i.e. ch loroan ilin es (CAN).
Collect. Czech. Chem. Commun. (Vol. 68) (2003)
doi:10.1135/cccc20031819

1820
Krátky et al.:
If CAN sh ould be th e m ain product of th e h ydrogen ation reaction s of
correspon din g CNB, platin um an d n ickel catalysts are th e m ost widely
used2–4. Metals with th e m ost favourable selectivity pattern s are ruth en ium
an d iridium 5,6. Sim ple palladium catalysts are reported as n ot very selec-
tive7. Th ey are often m odified with an oth er m etal or by th e addition of
som e m etal salts4,6. Selective poison in g of catalysts with com poun ds con -
tain in g sulfur, n itrogen or ph osph orus atom s h as been also successful1–3
.
Due to th e relation sh ip between th e catalyst h ydroph obicity or h ydro-
ph ilicity an d th e used solven t, a special en viron m en t m ay be form ed n ear
th e active catalyst sites favourable to th e in ten ded reaction . Gen erally,
aprotic solven ts (e.g. toluen e, xylen e, N-m eth ylpyrrolidon e) give better re-
sults with respect to low dech lorin ation . However, protic solven ts (e.g.
m eth an ol, eth an ol, propan -1-ol, propan -2-ol, dieth yl eth er; tetrah ydro-
furan , dioxan e) usually allow h igh er reaction rates1,3
.
H2
NO2
NO2
HO
NH2
X
X
X
1
1a
6
H2
O
N
NO
N
X
X
5
NH
NH2
2
H2
X
X
8
H2
N
NHOH
X
X
3
H2
H2
N
X
X
7
H2
H2
NH NH
NH2
X
X
9
4
NH2
H2
10
12
H2
X
11
SCHEME 1
Reaction path ways of h ydrogen ation of ch loron itroben zen es
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

Hydrogenation of Chloronitrobenzenes
1821
As a con tin uation of our previous study in th e field of h ydrogen ation of
arom atic n itroben zen es8–14 we report h ere on our in vestigation of th e reac-
tion path an d th e role of solven t in th e h ydrogen ation reaction s of CNB.
We focused our atten tion on th e in fluen ce of protic an d aprotic an h ydrous
solven ts un der m ild reaction con dition s usin g palladium supported on car-
bon as a catalyst.
EXPERIMENTAL
Catalyst
An am oun t of 1 wt.% palladium supported on ch arcoal (Pd/C) was prepared by m ixin g a so-
lution of 0.83 g of PdCl2 (Safin a Jesen ice, Czech Republic; an alytical grade, 40 wt.% solu-
tion ) in 6.4 cm 3 of water with a solution of 0.33 g of Na2CO3. Th is m ixture was refluxed at
80 °C un der m ild stirrin g for 2 h an d th en cooled to 25 °C. Th e obtain ed suspen sion was
added to 150 cm 3 of water with 20 g of ch arcoal (Colorex, Hn usta-Likier, Slovak Republic;
SBET = 1210 m 2/g) an d th e m ixture was stirred at 25 °C for 1 h . Th en , 10 wt.% of aqueous
solution of Na2CO3 was added gradually durin g 2 h to m ain tain pH 9–10. After stirrin g for
10 m in , th e m ixture was filtered off an d th ree tim es rin sed with 50 cm 3 of m eth an ol. Th e
m oist Pd/C was dried at 120 °C to con stan t weigh t. Activation of th e catalyst was carried
out prior to use in m eth an ol with h ydrogen at 50 °C an d 1 MPa for 1 h . For an alytical pur-
poses, som e reduced catalysts were dried after reduction to con stan t weigh t.
Preparation of Ph en ylh ydroxylam in es
N-(4-Ch loroph en yl)h ydroxylam in e (N-4-CHA), N-(3-ch loroph en yl)h ydroxylam in e (N-3-CHA)
an d N-(2-ch loroph en yl)h ydroxylam in e (N-2-CHA) were syn th esised by th e m odified Blatt
m eth od15. A solution of 2.5 g of am m on ium ch loride in 40 m l of distilled water was m ixed
with a solution of 5 g of 4-CNB, 3-CNB or 2-CNB in 40 m l of m eth an ol. Th e m ixture was
stirred vigorously with a m ech an ical stirrer an d 6.2 g of zin c powder (85 wt.% purity) was
added durin g 15–20 m in . As th e reduction proceeded, th e tem perature rose to 60–65 °C. Af-
ter addition of all zin c powder, stirrin g con tin ued for an oth er 15 m in . Th e en d of reaction
was in dicated by a sligh t decrease in th e reaction tem perature. Wh ile still h ot, th e solution
was filtered with suction in order to rem ove zin c oxide, wh ich was wash ed with 50 m l of
h ot (60 °C) distilled water. Th e filtrate was cooled to 0 °C. Th e crude product crystallised out
as lon g wh ite n eedles. Th en , th e crude product was filtered by suction an d dissolved in di-
eth yl eth er. Th e prepared solution was cooled down again an d frozen water was filtered off.
In order to rem ove traces of water, a n egligible am oun t of an h ydrous calcium ch loride was
added to th e filtrate. After filterin g off th e in organ ics, dieth yl eth er was evaporated an d
crude N-x-CHA con tain in g a sm all am oun t of an ilin e (AN) was recrystallised from a m ixture
of ben zen e an d h exan e (1:5 by volum e). After filtration of N-x-CHA, th e wh ite coloured
crystals were dissolved in dieth yl eth er again . Th en dieth yl eth er was evaporated from m ix-
tures in an oven in n itrogen atm osph ere at 40 °C an d N-x-CHA crystals were dried to con -
stan t weigh t.
Th e purity of N-x-CHA was ch ecked by m eltin g poin ts an d NMR spectra. Such N-x-CHA
were used as stan dards for HPLC an d NMR in vestigation of th e reaction m ixtures.
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

1822
Krátky et al.:
Catalyst Ch aracterisation
Th e reduced an d dried catalyst was ch aracterised by elem en tal an alysis, X-ray powder dif-
fraction , scan electron m icroprobe an alysis (SEMA) an d by adsorption –desorption of n itro-
gen 16. Th e results are sum m arised in Table I.
Catalytic Tests
Th e experim en ts were carried out in a stirred batch reactor equipped with a device for m on i-
torin g th e h ydrogen con sum ption . A typical reaction m ixture con sisted of about 500 m g of
th e catalyst an d 6 m l of a solution of x-CNB in m eth an ol (MeOH), dieth yl eth er (Et2O),
tetrah ydrofuran (THF) or 1,4-dioxan e (DOX). Th e h ydrogen ation was stopped after reach in g
th e stoich iom etric con sum ption of h ydrogen (nH /nx-CNB, 3 m ol of H2 per 1 m ol of x-CNB)
2
or it was left proceedin g furth er (wh en in vestigatin g th e h ydrogen olysis of ch lorin e).
An alysis
Th e reaction m ixture was an alysed by a gas ch rom atograph y. CP Sil-5 CB colum n (0.35 m m ×
50 m ) an d a FID detector were used. An isoth erm al regim e (200 °C) was used in an alyses.
Th e products were iden tified by GC-MS (GCMS-QP5000 Sh im adzu). Ph en ylh ydroxylam in es
were an alysed usin g th e HPLC m eth od on
a Sh im adzu in strum en t equipped with a
SPD-M10Vp ph otodiode array detector described in ref.17. N-4-CHA, N-3-CHA, N-2-CHA an d
som e in term ediate products were an alysed by NMR. 1H an d 13C NMR spectra were recorded
on a Varian VXR-300 NMR spectrom eter at am bien t tem perature in CDCl3 solution s in 5 m m
tubes at 299.9 an d 75 MHz for 1H an d 13C, respectively. A spectral width of 4 kHz was ch o-
sen for 1H with 32 K data poin ts usin g a 45° pulse an d repetition tim e of 3 s. 13C NMR spec-
tra were m easured with th e 16 kHz spectral width an d 32 K data poin ts usin g a 45° pulse
an d WALTZ 16 decouplin g sequen ce. Stan dard Varian pulse sequen ce an d param eters were
used for m easurem en t of 1D APT spectra. Th e ch em ical sh ifts of com poun ds in recorded
NMR spectra were com pared with th ose from a h an dbook of NMR spectra, from ref.18 an d
1H an d 13C ACD NMR prediction s. Dech lorin ation was also estim ated by a con ven tion al
acid-base titration . All th e reaction m ixtures were an alysed for th e presen ce of palladium ,
wh ich was determ in ed by atom ic absorption spectroscopy (AAS)9.
TABLE I
Ch aracterisation of Pd/C catalyst
Property
Param eter
Surface SBET,L, m 2 g–1
1180/1210a
Con ten t of palladium , wt.%
Average size of Pd crystallites, n m
Type of support surface
1.02
6
h ydroph obic
a
Surface of th e ch arcoal with reduced palladium /surface with out palladium .
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

Hydrogenation of Chloronitrobenzenes
1823
RESULTS AND DISCUSSION
In th e course of catalytic h ydrogen ation of x-CNB th ere are four types of re-
action path ways, wh ich could proceed on active catalytic sites1–3 (Sch em e 1).
Th e first on e, th e reduction of x-CNB in n eutral an d acid en viron m en t (1),
proceeds via th e form ation of x-CNOB (2) an d N-x-CHA (3) as th e in term e-
diates yieldin g x-CAN (4). In th e secon d reaction path way favoured by basic
en viron m en t, in term ediates m ay un dergo con den sation reaction s to form
dich lorin ated 4-am in odiph en ylam in e (8), dich lorin ated azoxy- (5) an d azo-
ben zen es (7). Azoxy an d azo derivatives can be subsequen tly con verted to
dich lorin ated h ydrazoben zen es (9) an d x-CAN. x-CAN itself m ay arise in
such con den sation reaction s as well. Th e th ird type of reaction s in cludes
th e h ydrogen olysis of substituen ts in th e ben zen e rin g. Due to th is un -
wan ted reaction , th e cleavage of substituen ts from in term ediates an d prod-
ucts can yield a variety of oth er products n ot in volved in Sch em e 1. Th e
form ation of n itroben zen e (1a), AN (10), ch loroben zen e (11) an d ben zen e
(12) from correspon din g x-CNB an d x-CAN is sh own in Sch em e 1 as exam -
J
IPa
Value
Error
5J
4J
3J
4J
5J
3J
2
2
3
3
3
5
5
6
2
5
6
6
0.40
2.50
8.60
2.80
0.40
8.60
–
–
–
–
–
–
8
NH2
4
5
6
3
2
a
In teraction of proton s.
1
Cl
7
Proton
APb
δ, ppm
Error
2
3
5
6
8
1
1
1
1
2
6.99
6.46
6.46
6.99
4.77
–
–
–
–
0.95
b
Am oun t of proton s.
FIG. 1
Codin g of carbon s in 4-ch loroan ilin e (4-CAN) an d data used for th e sim ulation of NMR spectra
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

1824
Krátky et al.:
ples of such poten tial h ydrogen olysis reaction s. It is em ph asised in th e lit-
erature th at un der relatively m ild reaction con dition s (about 65 °C an d
atm osph eric pressure or 30 °C an d 0.1–0.5 MPa of h ydrogen ), h ydrogen o-
lysis of ch lorin e is preferred to th at of n itrogen in NO2 or NH2 groups6,7. As
th e fourth type of reaction , th e Bam berger’s rearran gem en t of N-x-CHA in
acid aqueous en viron m en t givin g am in och loroph en ol derivatives (6) is as-
sum ed19
.
Due to th e fact th at NMR results differed for various sam ples, we sim u-
1
lated H NMR spectra for 4-ch loroan ilin e. Figure 1 con tain s its carbon cod-
in g an d th e data used for th is sim ulation . As sh own in Fig. 2, th e spectra
dram atically depen d on th e resolution . With th e NMR apparatus used, th e
1
resolution of 0.5 Hz is possible to obtain . H NMR an alysis (Table II) of th e
reaction m ixture of 4-CNB in Et2O (35% con version ) revealed on ly very lit-
[6]
[6]
[2]
[6]
[6]
[6]
[2]
[6]
[2]
7.1
7.0
6.9
δ, ppm
FIG. 2
Fragm en t of sim ulated 1H NMR spectra (300 MHz) with ch aracteristic m ultiplet of four-spin
AA′BB′ system for h ydrogen s 2,6 of 4-CAN (Fig. 1) calculated with 0.1 (top), 0.5 (m iddle) an d
2.5 (bottom ) Hz resolution usin g 1H NMR ACD predictor
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

Hydrogenation of Chloronitrobenzenes
1825
tle in ten se sign als of two doublets (Fig. 3; 7.6–7.8 ppm ) related to 4-CNOB.
At 65% con version of 4-CNB (Fig. 4) it is possible to see on ly on e doublet
(in th e sam e ran ge of ch em ical sh ifts) related to an im purity. Practically n o
x-CNOBs were detected by HPLC (Fig. 5) at 33–36% con version . Th e th ird
in term ediate in th e reduction of th e n itro group is N-x-CHA. Th is com -
TABLE II
1H NMR (300 MHz) data for th e reaction m ixtures of 4-CNB h ydrogen ation (25 °C, 0.5 MPa,
[Pd] = 7.8 m ol/m 3, 1 wt.% Pd/C) in Et2O. Th e sam ples were m easured in oxygen -free CDCl3
at room tem perature
δ, ppm
Com poun d
Com poun d
J a, Hz
or group
Ha
Hb
N–H
1
2
3
4
5
4-CNB
7.44b
*
7.13b
7.02b
–
8.10b
*
6.86b
6.57b
–
–
8.70
*
Cl
4-CNOB
N-4-CHA
4-CAN
*
Ha
Hb
Ha
Hb
–
–
8.10
8.3
NHxOy
NH2, NHOHc
4.59, 4.78
4.783
a
b
Separation between adjacen t peaks of two “doublets” in th e AA′BB′ system . Cen ter of gra-
c
vity of two “doublets” in th e AA′BB′ system . Am in o an d N-h ydroxyam in o proton s overlap,
h en ce on ly on e peak appears in th e recorded spectrum after th e 35 an d 65% con version of
4-CNB (see Figs 3 an d 4). * Com poun ds in n egligible am oun ts.
H(I)-3,5
H(I)-2,6
5
3
6
2
1
4
Cl
NHxOy
NH2, NHOH
H(IV)-2,6
H(IV)-3,5
H(III)-2,6
H(III)-3,5
8.2 8.0 7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2 5.0 4.8
δ, ppm
FIG. 3
Partial 1H NMR spectra (300 MHz) of the reaction m ixture from the 4-CNB hydrogenation (25 °C,
0.5 MPa, [Pd] = 7.8 m ol/m 3, 1 wt.% Pd/C) in Et2O recorded after 35% con version of 4-CNB.
Measured in oxygen -free CDCl3 at room tem perature
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

1826
Krátky et al.:
poun d was detected by NMR an d HPLC in all th e sam ples. Accordin g to th e
proposed reaction path way, a furth er reaction step is th e h ydrogen ation of
N-x-CHA to x-CAN.
Due to a very fast in term olecular exch an ge of proton s between N-4-CHA
an d 4-CAN, th e sin glets of NH2 an d NHOH groups in 1H NMR spectra over-
lap (Figs 3 an d 4). Th e fact th at th e discussed sign als com prised con tribu-
H(IV)-3,5
H(IV)-2,6
5
3
6
2
NH2, NHOH
1
4
H(I)-3,5
H(I)-2,6
Cl
NHxOy
H(III)-2,6
H(III)-3,5
8.4 8.2 8.0 7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2 5.0 4.8 4.6
δ, ppm
FIG. 4
Partial 1H NMR spectra (300 MHz) of the reaction m ixture from the 4-CNB hydrogenation (25 °C,
0.5 MPa, [Pd] = 7.8 m ol/m 3, 1 wt.% Pd/C) in Et2O recorded after 65% con version of 4-CNB.
Measured in oxygen -free CDCl3 at room tem perature
3-CNB
N-3-CHA
AN
3-CAN
Condensed
products
f
e
d
c
b
a
5
10
t, min
FIG. 5
HPLC pattern s of th e reaction m ixtures of 3-CNB h ydrogen ation (25 °C, 0.5 MPa, [Pd] =
7.8 m ol/m 3, 1 wt.% Pd/C) in various solven ts at differen t con version s: a stan dard of N-3-CHA
in MeOH, b total con version of 3-CNB in MeOH, c pure 3-CNB in MeOH, d 3-CNB in THF at
33% con version , e 3-CNB in Et2O at 35% con version , f 3-CNB in DOX at 36% con version
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

Hydrogenation of Chloronitrobenzenes
1827
tion s of both groups was also supported by com parison of th eir in tegrals
an d 13C NMR (Table III).
NMR an d HPLC an alysis (Figs 3–5) proved th at th e reaction proceeds via
th e form ation of n itroso an d N-ph en ylh ydroxylam in e in term ediates an d
yields m ain ly x-CAN an d AN. Neverth eless, a n egligible am oun t of con -
den sed products (m on o- an d dich lorin ated azo- an d azoxyben zen es) was
also detected by HPLC (Fig. 5) an d GC-MS.
Precursors of th ese con den sed products are also eviden t from 1H NMR
spectra (Figs 3 an d 4; 8.2 an d 7.4 ppm ). 4,4′-Dich loroazoxyben zen e is a
com poun d detected by Mon tan ari et al.18, to wh ich sh ould be ascribed th e
sign als. Th e total am oun t of con den sed products in our experim en ts n ever
exceeded 3 m ole %. Hen ce, it is possible to con clude th at m ild reaction
con dition s, n eutral en viron m en t an d 1 wt.% Pd/C as th e catalyst (Table I)
are favourable to th e form ation of x-CAN obeyin g th e first reaction path -
way m en tion ed above.
Hydrogen ation reaction s of x-CNB are in fluen ced by dech lorin ation
wh ich results in th e form ation of AN (Figs 5 an d 6). Th e h ydrogen olysis of
ch lorin e can proceed in all reaction steps, but predom in an tly after reach in g
a stoich iom etric con sum ption of h ydrogen (Fig. 6). Th is fin din g allowed to
sim plify Sch em e 1 an d con sider th e two reaction s: (i) h ydrogen ation of
x-CNB to x-CAN an d (ii) dech lorin ation of x-CNB to n itroben zen e (NB) an d
its h ydrogen ation to AN. Dech lorin ation is assum ed to be th e rate-
determ in in g step. Both reaction s were taken as m odel reaction s for th e
treatm en t of kin etic data an d th e power-law m ath em atical m odel was ap-
plied12. Origin ally, our kin etic m odel also con sidered th e dech lorin ation of
x-CAN as th e th ird reaction . However, th e zero value of th e rate con stan t
for th is reaction with all th e x-CNB isom ers allowed to n eglect th is reaction
an d to state th at h ydrogen olysis of ch lorin e proceeds parallel with th e re-
TABLE III
13C NMR (75 MHz) ch em ical sh ifts (δ, ppm ) for th e reaction m ixtures of 4-CNB h ydrogen a-
tion (25 °C, 0.5 MPa, [Pd] = 7.8 m ol/m 3, 1 wt.% Pd/C) in Et2O. Th e sam ples were m easured
in oxygen -free CDCl3 at room tem perature
Com poun d
C1
C2
C3
C4
C5
C6
1
3
4
123.3
126.7
123.0
129.5
128.8
129.0
124.9
115.6
116.5
146.4
148.9
144.7
124.9
115.6
116.5
129.5
128.8
129.0
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

1828
Krátky et al.:
duction of th e n itro group. A very stron g ch em isorption of all th e tested
substrates on th e Pd surface was assum ed from a very low reaction order. It
was determ in ed by m ath em atical treatm en t of kin etic data th at th e Pd/C
catalyst is practically n ot deactivated durin g th e h ydrogen ation (zero value
of deactivation coefficien t). Th is statem en t h as been also supported by th e
zero con ten t of palladium in th e reaction m ixtures12. In all th e experim en ts
presen ted h ere, m ore or less th e sam e order of decrease of turn over frequen -
cies (TOF) was foun d: 3-CNB > 4-CNB > 2-CNB (Table IV). Th is is in accor-
dan ce with th e in fluen ce of th e electron -with drawin g effect of ch loro
substituen ts on th e reactivity. If th e Cl atom is in meta position , th e posi-
tive electron effect of th e ch loro substituen t m ay decrease th e en ergy bar-
rier of th e n ucleoph ilic attack of th e h ydride an ion on th e n itro group1–3
an d th us to accelerate th e reaction rate.
On th e oth er h an d, ch loro substituen ts in ortho an d para position s h ave
an opposite effect, wh ich results in a decrease in th e reaction rate. Ch an ges
in th e reactivity of x-CNB partly correspon d to th e Ham m ett con stan ts of
ch loro substituen ts in 3-CNB an d 4-CNB (0.37 an d 0.24, respectively20).
Th e in fluen ce of 2-CNB can n ot be explain ed by th e Ham m ett con stan ts,
because Cl an d NO2 groups in ortho position in teract.
Sign ifican t differen ces in specific activity an d selectivity of th e catalyst
wh en usin g diverse solven ts (Table V) were observed in th e h ydrogen ation
of all x-CNB isom ers. Th e TOF values (Table IV) in creased in th e order:
DOX < THF << Et2O < MeOH. If th e reaction s were carried out in MeOH or
3
2
1
0
0
1
2
3
4
5
6
–3
10 t, s
FIG. 6
Norm alised H2 con sum ption , nH /nx-CNB (in m ol/m ol), versus tim e (in s) in th e h ydrogen ation
2
of 1.0 M 2-CNB (ٗ), 1.0 M 3-CNB (᭺) an d 0.5 M 4-CNB (᭝) to x-CAN in m eth an ol (25 °C,
0.5 MPa, [Pd] = 7.8 m ol/m 3, 1 wt.% Pd/C)
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

Hydrogenation of Chloronitrobenzenes
1829
TABLE IV
Hydrogen ation a of x-CNB catalyst (startin g con cen tration of substrate was 1 m ol/dm 3; 25 °C,
0.5 MPa, [Pd] = 7.8 m ol/m 3, 1 wt.% Pd/C) in differen t solven ts
TOFb
m ol H2/m olPd s
nH /nx-CNB
m2ol/m ol
tR , s
Sx-CAN, %
Solven t
MeOH
Substrate
1 /
2
.
2-CNB
3-CNB
4-CNBd
2-CNB
3-CNB
4-CNB
2-CNB
3-CNB
4-CNB
2-CNBd
3-CNB
4-CNBd
1848
1146
1056d
2118
1386
1644
10200
3000
6420
9630
6720d
6420d
61.3
85.2
86.8d
97.9
96.7
97.5
99.2
94.2
96.7
96.7d
90.1
96.7d
0.10
0.17
0.10d
0.09
0.15
0.12
0.02
0.07
0.03
0.01d
0.02
0.02d
3.4
3
3
Et2OH
THF
3
3.1
3
2.3
2.9
3
DOX
2.5
3
3
a
tR , h alf-tim e of reaction of th e 50% of stoich iom etric h ydrogen con sum ption ; Sx-CAN, se-
1 /
2
b
lectivity to x-CAN. TOF (turn over frequen cy) calculated with t for th e 50% stoich iom etric
con sum ption of h ydrogen (1.5 m ol of H2 per 1 m ol of x-CNB). nH /nx-CNB n orm alised con -
c
2
sum ption of h ydrogen (th e fin al ratio of h ydrogen con sum ption to th e startin g am oun t of
d
x-CNB). Startin g con cen tration of substrate was 0.5 m ol/l.
TABLE V
Ch aracteristicsa of solven ts used
Solven t
ρ20, kg m –3
η20, cP
ε
µ, Cm
xH 105, kg/kg
2
MeOH
Et2OH
THF
791.3
713.3
0.59
0.24
0.55
1.37
32.7
1.70
1.15
1.63
0
4.5
8.3
4.1
2.1
7.58
888.0
4.33
2.25
DOX
1033.6
a
ρ20 den sity, η20 dyn am ic viscosity, ε perm ittivity, µ dipole m om en t an d xH
solubility of
,0.5
2
h ydrogen in various solven ts (20 °C, 0.5 MPa)
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

1830
Krátky et al.:
Et2O, th e h igh est reaction rates an d TOFs were foun d. Very low reaction
rates an d TOFs were foun d in experim en ts in THF an d DOX as solven ts
(Table IV). Th e m ost selective m edium for th e form ation of x-CAN was
Et2O, THF an d DOX. Th e lowest selectivities to x-CAN were obtain ed in
MeOH. Th e in fluen ce of reaction solven t on product selectivity is expressed
by th e order: Et2O > THF > DOX > MeOH. Th e effects of th e used solven ts
can be partly explain ed by th e differen ces in ph ysical an d ch em ical proper-
ties (Table V). Very low TOF in THF an d DOX can be ascribed to th eir
h igh er den sity an d viscosity. Due to th ese ch aracteristics an d rath er poor
dissolution of h ydrogen (m ain ly in DOX), con cen tration of h ydrogen in
th e n eigh bourh ood of catalytic sites is low an d, con sequen tly, th e reaction
rate is low. Com parable an d rath er h igh er values of TOF were ach ieved in
Et2O an d MeOH. For such effects of Et2O an d MeOH, th e m ain reason s
could lie in a low den sity an d viscosity of th e solven ts, wh ich facilitate th e
liquid tran sport. High er solubility of h ydrogen (see Table V) an d th e favour-
able diffusion properties m ake th e h ydrogen ation in Et2O or MeOH faster
th an in case of THF an d DOX.
Et2O was foun d to be th e m ost selective solven t (Table V). Sufficien t an d
still very h igh selectivities to th e desired products (x-CAN) were obtain ed in
THF an d DOX. Th e selectivities to x-CAN are very low in MeOH. As we
m en tion ed above, th e reaction m ech an ism of con version of th e n itro group
an d h ydrogen olysis of ch lorin e m ay be parallel or con secutive. If th e reac-
tion s are parallel, th e substrate m olecules m ust be orien ted with both reac-
tive groups (NO2 an d Cl) to th e catalyst surface. In solven ts wh ere th e
diffusion barriers are m in im ised, th e in fluen ce of perm ittivity (ε) of sol-
ven ts an d th eir polarity (dipole m om en t, µ) m ay also play an im portan t
role. Th e perm ittivity of MeOH is th e h igh est of all th e in vestigated sol-
ven ts, wh ich suggests a very poor ch an ce to establish th e n atural polaris-
ation between m olecules of substrate an d th e active catalyst sites th rough
NO2 favourably. More probably, th e in duced polarisation of substrate m ole-
cules takes place an d in teraction via both th e reactive groups occurs. Th is
in teraction results in decreasin g x-CAN selectivity wh ile th e production of
AN is in creased. In con trast, ε of Et2O is relatively low an d n atural polaris-
ation orien ts th e NO2 group of th e substrate m olecules towards th e catalyst
surface. Th is orien tation m akes NO2 be con verted preferably an d th e selec-
tivity to x-CAN can be h igh . Et2O is practically im m iscible with water
(sm all µ). Th us, Et2O togeth er with th e h ydroph obic ch aracter of Pd/C cata-
lyst en h an ces rem oval of th e reaction water from th e catalyst surface an d
th e ch em isorbed m olecules can be protected to dech lorin ation by th e n on -
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

Hydrogenation of Chloronitrobenzenes
1831
polar en viron m en t of Et2O. In a polar m edium of MeOH, th e reaction water
is m iscible with th e solven t an d so rem ain s in th e n eigh bourh ood of Pd
particles on th e catalyst surface. Th is causes a stron ger dech lorin ation of
ch em isorbed m olecules. THF an d DOX sh owed satisfactory selectivity to
x-CAN but th e h ydrogen ation s of substrates in th ese solven ts proceed very
slowly due to above effects of lower solubility of h ydrogen an d diffusion
lim itation s. Th ese solven ts h ave very low perm ittivity an d th us th ey can
provide a h igh selectivity to th e desired product. Sin ce DOX h as a zero di-
pole m om en t an d is im m iscible with water, it en sures a better selectivity to
x-CAN in com parison with MeOH, sim ilarly to Et2O (Table IV).
CONCLUSIONS
Hydrogen ation of ch loron itroben zen es (x-CNB) un der m ild reaction con di-
tion s over Pd/C catalyst proceeds via th e form ation of n itroso- an d N-ph e-
n ylh ydroxylam in e in term ediates an d yields correspon din g ch loroan ilin es
(x-CAN). However, a sm all am oun t of an ilin e was form ed an d traces of con -
den sed products (m on o- an d dich lorin ated azoxy- an d azoben zen es) were
also detected by HPLC an d GC-MS. Th e treatm en t of kin etic data revealed
th e reaction m ech an ism to be m ore parallel th an con secutive. Practically
n o deactivation of th e Pd/C catalyst was foun d. Th e stron g in fluen ce of sol-
ven ts on th e specific catalyst activity (expressed by TOF) an d selectivity was
observed. Dieth yl eth er was th e best solven t for h ydrogen ation of all x-CNB
from th e poin t of view of th e h igh est TOF an d product selectivity.
This work was supported by funds of the Slovak VEGA project No. 1/9142/02.
REFERENCES
1. Arnold H., Dobert F., Gaube J. in: Handbook of Heterogeneous Catalysis (G. Ertl,
H. Knozinger and J. Weitkamp, Eds), Vol. 5, p. 2165. Wiley–VCH, New York 1997.
2. Blaser H. U., Siegrist U., Steiner H., Studer M. in: Fine Chemicals Through Heterogeneous
Catalysis (R. A. Sheldon and H. van Bekkum, Eds), p. 351. Wiley–VCH, New York 2001.
3. Notheisz K., Bartok M. in: Fine Chemicals Through Heterogeneous Catalysis (R. A. Sheldon
and H. van Bekkum, Eds), p. 121. Wiley–VCH, New York 2001.
4. Coq B., Tijani A., Figueras F.: J. Mol. Catal. 1992, 71, 317.
5. Liu M. H., Yu W. Y., Liu H. F.: J. Mol. Catal. A: Chem. 1999, 138, 295.
6. Auer E., Freud A., Gross M., Hartung R., Panster P. in: Catalysis of Organic Reactions (F. E.
Herkes, Ed.). Marcel Dekker, New York 1998.
7. Yu Z. K., Liao S. J., Xu Y., Yang B., Yu D. R.: J. Mol. Catal. A: Chem. 1997, 120, 247.
8. Fišera R., M. Králik, Annus J., Krátky V., Zecca M., Hronec M.: Collect. Czech. Chem.
Commun. 1997, 62, 1763.
Collect. Czech. Chem. Commun. (Vol. 68) (2003)

1832
Krátky et al.:
9. Králik M., Fišera R., Zecca M., D’Archivio A. A., Galantini L., Jeřábek K., Corain B.:
Collect. Czech. Chem. Commun. 1998, 63, 1074.
10. Krátky V., Králik M., Hronec M., Zecca M.: Petroleum and Coal 2000, 42, 28.
11. Krátky V., Králik M., Hronec M.: Proc. XXXII Symposium on Catalysis, Prague, 2001.
12. Krátky V., Králik M., Mecárová M., Stolcová M., Zalibera L., Hronec M.: Appl. Catal., A
2002, 235, 225.
13. Králik M., Krátky V., Hronec M., Zecca M., Corain B.: Stud. Surf. Sci. Catal. 2000, 130,
2321.
14. Corain B., Králik M.: J. Mol. Catal. A: Chem. 2001, 173, 99.
15. Blatt A. H.: Org. Synth. 1948, 1, 445.
16. Horniaková J., Králik M., Kaszonyi A., Mravec D.: Microporous Mesoporous Mater. 2001,
46, 287.
17. Stolcová M., Grosková D., Hronec M.: Chromatographia Suppl. 2001, 53, 456.
18. Montanari S., Paradisi C., Scorrano G.: J. Org. Chem. 1999, 64, 3422.
19. Gao Y., Wang F., Liao S., Yu D.: React. Kinet. Catal. Lett. 1998, 64, 351.
20. March J.: Advanced Organic Chemistry. John Wiley and Sons, Inc., New York 1992.
Collect. Czech. Chem. Commun. (Vol. 68) (2003)