Role of base addition in the liquid-phase hydrodechlorination of 2,4-dichlorophenol over Pd/Al2O3 and Pd/C

Article abstract of DOI:10.1016/j.jcat.2004.05.003

The aqueous-phase batch hydrodechlorination (HDC) of 2,4-dichlorophenol (2,4-DCP), over 1 wt/wt% Pd/Al₂O₃ and Pd/C has been investigated with/without the addition of NaOH, NH₄OH, LiOH, KOH, RbOH, and CsOH; bulk solution pH spanned the range 1.5-13. The reaction was operated in the kinetic-controlled regime with 2-chlorophenol (2-CP) as the only intermediate partially dechlorinated product which reacts further to yield phenol; cyclohexanone was formed over Pd/Al₂O₃, but not Pd/C, prior to the completion of dechlorination. An increase in fractional dechlorination with the addition of base was observed and can be attributed to a suppression of HDC inhibition due to the HCl that is generated. The initial HDC activity and selectivity delivered by both catalysts were pH dependent and differences in response to bulk solution pH variations are discussed in terms of the nature of the reactive species in solution and the amphoteric behavior of the Pd supports. The Pd/Al₂O₃ catalyst is characterized by a high surface charge density while Pd/C bore a low basic functionality density on a high surface area carrier. A maximum initial HDC rate over Pd/Al ₂O₃ was attained at pH 7-9 while a higher pH (≥9) results in more effective HDC for Pd/C, effects that are linked to chloroarene dissociation and surface charge effects. In the case of Pd/C, the initial HDC rate with the addition of alkali metal hydroxide (0.074 moldm^-3) increased in the order Li⁺≈Na⁺+≈ Rb⁺+, a sequence that matches the adsorption affinity on a negatively charged carbon surface; HDC over Pd/Al ₂O₃ was insensitive to the nature of the alkali metal cation. The action of NH₄OH, as a weak base, is also considered.

Full text of DOI:10.1016/j.jcat.2004.05.003

Journal of Catalysis 225 (2004) 510–522
www.elsevier.com/locate/jcat
Role of base addition in the liquid-phase hydrodechlorination
of 2,4-dichlorophenol over Pd/Al O and Pd/C
2
3
∗
Guang Yuan and Mark A. Keane
Department of Chemical and Materials Engineering, University of Kentucky, Lexington, KY 40506, USA
Received 3 March 2004; revised 27 April 2004; accepted 4 May 2004
Available online 1 June 2004
Abstract
The aqueous-phase batch hydrodechlorination (HDC) of 2,4-dichlorophenol (2,4-DCP), over 1 wt/wt% Pd/Al O and Pd/C has been
2
3
investigated with/without the addition of NaOH, NH OH, LiOH, KOH, RbOH, and CsOH; bulk solution pH spanned the range 1.5–13. The
4
reaction was operated in the kinetic-controlled regime with 2-chlorophenol (2-CP) as the only intermediate partially dechlorinated product
which reacts further to yield phenol; cyclohexanone was formed over Pd/Al O , but not Pd/C, prior to the completion of dechlorination. An
2
3
increase in fractional dechlorination with the addition of base was observed and can be attributed to a suppression of HDC inhibition due to
the HCl that is generated. The initial HDC activity and selectivity delivered by both catalysts were pH dependent and differences in response
to bulk solution pH variations are discussed in terms of the nature of the reactive species in solution and the amphoteric behavior of the Pd
supports. The Pd/Al O catalyst is characterized by a high surface charge density while Pd/C bore a low basic functionality density on a high
2
3
surface area carrier. A maximum initial HDC rate over Pd/Al O was attained at pH 7–9 while a higher pH (ꢀ 9) results in more effective
2
3
HDC for Pd/C, effects that are linked to chloroarene dissociation and surface charge effects. In the case of Pd/C, the initial HDC rate with
−
3
+
+
+
+
+
the addition of alkali metal hydroxide (0.074 moldm ) increased in the order Li ≈ Na < K ≈ Rb < Cs , a sequence that matches
the adsorption affinity on a negatively charged carbon surface; HDC over Pd/Al O was insensitive to the nature of the alkali metal cation.
2
3
The action of NH OH, as a weak base, is also considered.
4

2004 Elsevier Inc. All rights reserved.
Keywords: Liquid-phase hydrodechlorination; Pd/C; Pd/Al O ; 2,4-Dichlorophenol; 2-Chlorophenol; Base addition
2
3
1
. Introduction
Contamination of water and soil by chlorophenolic com-
tion [6], catalytic processing represents a far more progres-
sive alternative. Catalytic hydrodechlorination (HDC) as a
reductive approach is now viewed as a promising emerging
technology [7] that, when compared with traditional oxida-
tion methods, presents the following advantages [8–10]:
pounds now represents a significant environmental burden
due their widespread industrial use and inherent toxicity and
persistence in the environment [1]. 2,4-Dichlorophenol (2,4-
DCP), the focus of this study, is a high-volume feedstock
chemical used in the synthesis of pharmaceuticals and the
herbicide 2,4-D (2,4-dichlorophenoxyacetic acid) of which
(i) low-temperature nondestructive transformation with no
directly associated NO /SO emissions;
x
x
(ii) no associated dioxin and dibenzofuran formation; and
2
6,300 tons were produced in the US in 1995 [2]. The large-
(
iii) selective Cl removal to generate recyclable products.
scale production and usage of 2,4-DCP has led to growing
concern regarding operation safety [3] and the treatment
of related effluent and polluted ground/surface water [4,5].
With increasingly more restrictive landfill legislation and the
possibility of hazardous by-product release during incinera-
HCl is generated during HDC and can act to poison the
catalytically active metal [11,12]. In the liquid-phase oper-
ation the presence of an inorganic base serves as a proton
scavenger, maintaining the catalytic metal in a reduced state
and limiting Cl interaction(s) [9,13]. In the aqueous-phase
HDC of 4-chlorophenol (4-CP) over Ru/C, the addition of
NaOH was also found to “activate” the reactant and neu-
*
Corresponding author. Fax: +1 (859) 323-1929.
E-mail address: makeane@engr.uky.edu (M.A. Keane).
0021-9517/$ – see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2004.05.003

G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
511
tralize acidic sites on the carbon support [14]. In the HDC
of chlorobenzene in methanol over Pd/AlPO4–SiO2, it was
found that the beneficial effects of NaOH depended on Pd
dispersion and support composition [15]. It is fair to state
that the consequences of base addition on liquid-phase HDC
activity/selectivity are still not well established and the role
of metal support is unclear. We have undertaken a funda-
mental study of base (NaOH, LiOH, KOH, RbOH, CsOH,
and NH4OH) addition in governing 2,4-DCP HDC, focusing
on the impact of pH on reactant/catalyst interactions at the
solid/liquid/gas interface. The effect of the support (activated
carbon vs alumina) in modifying the intrinsic Pd activity is
also assessed.
separate H2 flow for at least 10 min and the “preactivat-
ed” catalyst was then introduced into the reactor (time t = 0
for reaction), where the “preactivation” served to circumvent
the occurrence of an extended induction period (4–10 min)
that was observed for reaction with Pd/Al2O3 in the ab-
sence of a H2 precontact. The pH of the reaction mixture
was monitored continuously using a Dow-Corning pencil
electrode coupled to a data logging and collection system
(Pico Technology Ltd.). The initial 2-CP and 2,4-DCP con-
−3
centrations were 0.0570 and 0.0285 moldm , respectively,
−3
with a catalyst concentration (Ccat) = 0.3 gcat dm . The
Cl/Pd ratio was kept at ca. 1000 (mol/mol) throughout this
study and the initial [base]/[organic-Cl] was in the range 0–
1
.7 for NaOH, 0.2–1.3 for LiOH, KOH, RbOH, and CsOH
and 0–5.2 for NH4OH. A noninvasive liquid sampling sys-
tem via in-line filters allowed a controlled syringe removal
of aliquots (0.5 cm ) of reactant/product(s). Prior to analy-
2
. Experimental
3
2
.1. Materials
sis, the basic solution samples were neutralized with dilute
CH3COOH (ca. 0.2 moldm ).
−3
2
,4-DCP, 2-CP, LiOH, NaOH, KOH, RbOH, and CsOH
2
.3. Product analysis and activity/selectivity evaluation
(
all 99+%) were purchased from Aldrich Chemical Co.
and NH4OH (36.5–38%, AR) was supplied by Mallinckrodt
Baker, Inc.; all the chemicals were used as received. Cat-
alysts with the nominal loadings 1% wt/wt Pd/C and 1%
wt/wt Pd/Al2O3 (supplied by Aldrich) were sieved (ATM
fine test sieves) into batches of particle diameter < 325 mesh
The composition of the reaction/product mixture was an-
alyzed by gas chromatography (Perkin-Elmer Auto System
XL), employing an FID and a DB-1 J&W Scientific capil-
lary column (i.d. = 0.2 mm, length = 50 m, film thickness =
0
.33 µm). The relative peak area percentage was converted
(
45 µm). Stock 2,4-DCP and 2-CP solutions were prepared
to mol% using regression equations based on detailed cali-
bration and the detection limit corresponded to a feedstock
conversion < 0.4 mol%: overall analytical reproducibility
was better than ± 5%. The concentration of organic species
with deionized water (electronic resistance ꢀ 15 Mꢀ). The
concentration of the aqueous base solutions was determined
by standard acid titration.
(
2,4-DCP, 2-CP, and phenol) in the bulk liquid phase was de-
2
.2. Catalytic procedure
termined from the total mass balance in the reaction mixture
where the organic species were taken to be nonvolatile and
the effect of uptake on the supports was negligible [17–19].
All the liquid-phase HDC reactions were carried out in a
modified commercial stirred-glass reactor (Ken Kimble Re-
actors Ltd.) equipped with a H2 supply at a constant (Brooks
−3
The HCl produced/H2 consumption (moldm ) during re-
action were calculated from the molar balance based on GC
analysis. The conversion of 2,4-DCP (X2,4-DCP) is defined
by
3
−1
mass-flow controller) volumetric flow rate (250 cm min ).
There was no measurable conversion in the absence of the
H2 supply; it has been shown previously [16] that hydrolysis
of CPs in NaOH only occurs to a significant degree at el-
evated temperatures and [NaOH]/[CP] = 6. Liquid coolant
([2,4-DCP] − [2,4-DCP])
0
X2,4-DCP =
,
(1)
[
2,4-DCP]0
(
ca. 278 K) was used to condense all volatiles: loss of the re-
where the HDC selectivity (as a percentage) with respect to
-CP (S2-CP) is given as the mol% 2-CP in terms of the total
moles of product(s) formed, i.e.,
actor liquid contents in the H2 flow was negligible (< 0.5%
v/v). A glass impeller provided effective agitation at a stir-
ring speed of 1100 rpm. This choice of stirring speed and
H2 flow rate served to minimize transport limitations as de-
scribed elsewhere [17] along with a full description of the
catalytic reactor. A recirculator (Julabo) was used to stabilize
the reaction temperature at 303 ± 0.5 K. At the beginning of
each experiment, the 2,4-DCP or 2-CP aqueous solution was
2
[
2-CP]
S2-CP% =
× 100.
(2)
) ) was de-
[2,4-DCP]0 − [2,4-DCP]
The initial 2,4-DCP consumption rate ((R2
,4-DCP 0
termined using a pseudo-first-order linear regression from
temporal concentration profiles.
3
−1
charged and agitated in a He flow (50 cm min ). A known
amount of (aqueous) base solution was added to bring the
2.4. Catalyst characterization
3
total liquid volume to 0.1 dm and the temperature was al-
lowed to stabilize. The Pd/C catalyst was charged without
precontact with H2. In the case of Pd/Al2O3, a ca. 2 cm
catalyst slurry in deionized water was preactivated with a
The Pd content was determined by inductively coupled
plasma-optical emission spectrometry (ICP-OES, Vista-
PRO, Varian Inc.) from the diluted extract of aqua regia.
3

512
G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
The BET surface area was measured using the commercial
CHEM-BET 3000 (Quantachrome Instrument). After outgas
at 523 K for 30 min, at least 2 cycles of nitrogen adsorption–
desorption in the flow mode were employed to determine to-
tal surface area using the standard single-point BET method;
the calculated BET surface area is quoted as the average.
The Pd particle size distributions of the fresh and used cata-
lysts were determined by transmission electron microscopy
Table 1
Characterization data for the unused catalysts
Pd/C
1037
Pd/Al O
2
3
2
−1
)
cat
BET surface area (m g
160
Pd content (% wt/wt)
Surface-area-weighted average Pd
1.01
14.5
1.17
2.2
¯
particle size (d , nm)
Pd
Specific Pd surface area
34.4
8.5
227
2
−1 a
)
Pd
(S
, m g
(
TEM, JEOL 2000). At least 700 individual Pd particles
Pd
pH at the point of zero charge (pHpzc)
7.8
were counted for each catalyst and the mean Pd particle sizes
are quoted as the surface-area-weighted average particle size
a
¯
−3
.
SPd = 6/(ρdPd), ρ = 12.02 gcm
ꢀ
ꢀ
3
2
¯
¯
(
dPd) according to dPd =
nid /
nid [20], where ni
i
i
i
i
ꢀ
is the number of particles of diameter di and
ni > 700.
the surface will exhibit a higher affinity for cationic species
in solution. The experimentally determined pHpzc suggests
a weak basicity for both catalysts, diagnostic of the ampho-
teric nature of the supports. The basicity of Pd/Al O can
i
The pH associated with the point of zero charge (pHpzc)
for both catalysts was determined using the potentiometric
mass titration technique (PMT) [21]. Three different cata-
2
3
3
−3
lyst masses were immersed in 50 cm 0.1 mol dm NaCl to
which a known amount of 0.1 mol dm NaOH was added
to adjust the initial solution pH to ca. 11. A 0.1 mol dm
be attributed to the Al–OH group and is well established in
the aqueous chemistry of inorganic oxides [26]. In the case
of Pd/C, the presence of different (O or N containing) func-
tional groups allied to π-electron density at the carbon basal
planes may contribute to the overall weakly basic surface
[25,27]. The precise source and nature of these contribu-
tions is outside the remit of this paper, where the emphasis
is placed firmly on the role of surface charge and bulk so-
lution pH in determining HDC by supported Pd. Although
an identification of the surface functionalities may help in
understanding the surface/solute interaction(s), it has been
demonstrated [28,29] that a knowledge of surface charge
variation with solution pH is sufficient in elucidating the role
of surface chemistry in the adsorption of substituted phenols
on carbon [28] and oxides [29] in aqueous media.
−3
−3
HCl solution was used as titrant, added to the slurry which
was sealed in PET vials for 24 h with continuous agita-
tion in a He atmosphere. The acidity of the aqueous catalyst
slurries at equilibrium was measured with a pH meter (Corn-
ing, Model 440) equipped with a polymer-body, liquid-filled
combination electrode. Calibration before each pH measure-
ment employed standard buffer solutions (pH 4.0 and 10.0).
3
. Results
3
.1. Catalyst characteristics
The Pd content, BET, and specific Pd surface areas are
3.2. Effects of base addition on the HDC performance of
given in Table 1. The Pd phase on the lower surface area
Al2O3 support is in the form of smaller metal particles (pre-
dominantly 0–6 nm), i.e., higher specific Pd surface area.
The latter can be attributed to stronger Pd/support inter-
action(s) due to electron transfer from the Al2O3 support
Pd/Al O3
2
3.2.1. NaOH addition: HDC of 2,4-DCP and 2-CP
In order to investigate the role of NaOH addition, 2,4-
DCP HDC was conducted where the initial NaOH concen-
−3
[
22–24]. In the case of Pd/C, the supported Pd particles ex-
tration ([NaOH] ) spanned the range 0.012–0.09 moldm
0
hibit a broad size range (< 2 to > 20 nm) with a spherical
morphology that is indicative of weak Pd/support interac-
tion [22]. Hydrogen chemisorption measurements, reported
elsewhere [18], also revealed a higher metal dispersion as-
sociated with Pd/Al2O3. In an earlier study [17], we pre-
sented HDC activity/selectivity responses that are suggestive
of differences in the intrinsic HDC behavior of supported
Pd catalysts with different Pd content/dispersion. An explicit
consideration of Pd particle size effects is beyond the scope
of this paper which is focused on the impact of solution pH
in determining HDC performance of Pd (at a comparable
loading) on two supports. To this end, pHpzc is a critical cata-
lyst property, defined as the pH value at which the accessible
surface of the wetted catalyst particle possesses neither a net
positive nor a negative charge [25]. At a pH < pHpzc the cat-
alyst surface will bear a positive charge which favors inter-
action with anionic species. Conversely, where pH > pHpzc
([NaOH] /[organic-Cl] = 0.2–1.6); the resultant catalytic
0
0
results are summarized in Table 2. Typical temporal liquid-
phase compositions and pH profiles associated with reaction
at the two [NaOH] extremes are shown in Fig. 1. It can
0
be seen (Fig. 1a) that 2,4-DCP HDC proceeded in a step-
wise fashion with 2-CP as the sole partially dechlorinated
product. Cyclohexanone (resulting from phenol hydrogena-
tion) appeared as product before dechlorination had reached
completion but with low selectivity (ꢁ 5%). The absence of
4-CP as product may be explained on the basis of steric hin-
drance where the ortho-Cl is more resistant to H2 cleavage
[17,18,30]. The initial HDC rate ((RHDC)0) was raised in
the presence of base but declined with increasing [NaOH]0
while the fractional dechlorination (XCl after 120 min) was
significantly higher with an increase in [NaOH]0. The XCl
parameter is a useful indicator of HDC “stability” where
the lower XCl associated with reaction in the absence of

G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
513
Table 2
Effects of NaOH addition on 2,4-DCP HDC over Pd/Al O
2
3
a
b
c
d
[
(
NaOH]
[NaOH] /[organic-Cl]
(R
)
(R
)
/(R
)
X
Cl
Y
%
pH
0
−
0
0
HDC 0
2,4-DCP 0
HDC 0
C=O
3
−1 −1
)
cat
(0–120 min)
moldm
)
(mmol min
Cl
g
0
0
0
0
0
0
0
0
0
2.9
6.0
5.9
4.9
3.3
3.5
3.4
3.8
0.88
0.79
0.74
0.73
0.87
0.90
0.88
0.87
0.68
0.75
0.85
0.95
0.98
0.96
0.92
0.89
4.1
4.6
5.2
4.5
4.0
3.6
3.4
3.9
4.8–1.5
7.9–1.7
8.7–1.8
12.1–2.0
12.5–2.4
12.6–9.1
12.9–11.1
12.9–12.4
.012
.025
.037
.049
.058
.074
.090
0.23
0.44
0.65
0.85
1.02
1.30
1.58
a
Initial rate of consumption of 2,4-DCP, calculated from zero-order linear regression of the temporal [2,4-DCP] profiles.
b
Initial HDC rate, defined as the initial rate of Cl removal and calculated from zero-order linear regression of the temporal HCl product concentration
profiles.
c
Fractional dechlorination after 120 min.
Percentage yield of cyclohexanone after 120 min.
d
release of HCl during HDC (Fig. 1b); the drop of pH from
to 4 at ca. 6 min can be linked to the complete consump-
7
tion of NaOH. A consequent HCl inhibition of the HDC of
both 2,4-DCP and 2-CP can be observed after ca. 6 min
(
Fig. 1a). The pH associated with the higher [NaOH]0 ex-
hibited a slight decrease from 12.9 to 12.4, a response linked
to excess NaOH addition. The pH variations during HDC
(
1
from low to high [NaOH]0) span the broad range 12.9–
.7. As the pKa values of 2,4-DCP, 2-CP and phenol are,
respectively, 7.89, 8.56, and 10.0 [31], the nature of the re-
actants/products in bulk solution switches from chloropeno-
late/phenolate anions to the chlorophenolic/phenolicform as
the pH is lowered from 12 to 5. Moreover, the surface charge
of Pd/Al2O3, with an associated pHpzc = 7.8, is also pH de-
pendent. It is to be expected that the chloroarene/catalyst
interactions and HDC rates are pH sensitive and affected
by the addition of NaOH. In the absence of NaOH, the
initial pH is 4.8 and drops to 1.5 after 120 min 2,4-DCP
HDC (Table 1). Where pH ꢁ 5, the chlorophenolic species
dominates (> 99.9 mol/mol%) and interaction with a posi-
tively charged alumina surface is unfavorable. Consequently,
(
RHDC)0 rate was lower than that obtained in the presence
−3
of NaOH. Over the [NaOH]0 range 0.012–0.025 mol dm
the initial pH (8–9) is such that the reactant in solution
is a combined chlorophenolate/chlorophenolic species (11–
9
3 mol/mol% 2,4-DCP dissociation) and the Pd/Al2O3 sur-
face is weakly charged, accessible to both species. The cor-
responding (RHDC)0 exhibited a significant increase to at-
tain a maximum (Table 1). The lower ratio of initial 2,4-
DCP consumption to HDC rate ((R2,4-DCP)0/(RHDC)0) over
Fig. 1. HDC of 2,4-DCP over Pd/Al O : (a) liquid-phase composition
2
3
in terms of mol% 2,4-DCP (F,E), 2-CP (2,1), phenol (Q,P), and cy-
clohexanone (",!) as a function of time; (b) pH of the bulk solution
−3
[NaOH] = 0.012–0.037 mol dm , i.e., enhanced complete
0
−3
as a function of time; (1) [NaOH] = 0.012 mol dm
(open symbols);
0
dechlorination, suggests that the retention of 2-CP on the
−3
(
2) [NaOH] = 0.09 mol dm (solid symbols).
0
weakly charged surface favors full HDC to phenol. Over the
−3
higher [NaOH]0 range (0.037–0.090 moldm ), where the
corresponding initial pH > 12, the catalyst bears a negative
charge and presents an unfavorable surface to chlorophe-
−
3
base or dilute base addition ([NaOH]0 < 0.037 moldm
)
is indicative of catalyst deactivation due to HCl inhibition,
as demonstrated elsewhere for the liquid-phase HDC of
chlorobenzene [15]. The solution pH associated with the
dilute [NaOH] system decreased from ca. 8 to 2 with the
−
nolate/phenolate anions. The higher [OH ] inhibits HDC
with a resultant drop in (RHDC)0 and XCl but HDC effi-
ciency is still greater than that achieved in the absence of

514
G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
Fig. 3. Initial HDC rate ((R
) ) as a function of [NaOH] in the
HDC 0 0
HDC of 2,4-DCP (2) and 2-CP (1) over Pd/Al O : [2,4-DCP]
=
2
3
0
−3
−3
.
0
.0285 mol dm ; [2-CP] = 0.0570 mol dm
0
lectivity/conversion trend line at higher X2,4-DCP (Fig. 2a).
The latter can be linked to a common drop of pH be-
low 10 with increasing conversions (Fig. 2b). In the case
where the pH (after 120 min) > 11 ([NaOH]0 = 0.074–
−3
0
.09 moldm ), the selectivity trends overlapped with a
higher S2-CP maintained at greater X2,4-DCP. The increase
in S2-CP at higher pH suggests that the secondary conver-
sion of 2-CP to phenol was inhibited to a greater degree
−
at higher [OH ] when compared with the 2,4-DCP to 2-
CP step. In order to probe differences in 2,4-DCP and 2-CP
HDC behavior with NaOH addition, (RHDC)0 for 2,4-DCP
Fig. 2. HDC of 2,4-DCP over Pd/Al O : (a) selectivity with respect to 2-CP
and 2-CP at different [NaOH] are compared in Fig. 3. It
can be seen that the overall HDC response for 2-CP con-
2
3
0
(
S
) as a function of fractional 2,4-DCP conversion (X ) and
2,4-DCP
2
-CP
(
b) pH of the bulk solution as a function of X
; [NaOH] = 0.000
2
,4-DCP
0
version is similar to 2,4-DCP, i.e., increase in (RHDC)0 to
a maximum with the addition of NaOH and a subsequent
decline at higher [NaOH]0. Both (RHDC)0 maxima can be
linked to a more effective interaction between the chlorophe-
nolate anions and the weakly charged catalyst surface where
the initial solution pH is 7–9. The consistently higher HDC
rates for 2-CP, particularly evident where [NaOH]0 = 0–
(
0
E), 0.012 (+), 0.025 (×), 0.037 (P), 0.049 (2,1), 0.058 (∗), 0.074 ("),
−3
−3
.090 (—) mol dm ; [2,4-DCP] = 0.0285 mol dm ; solid and open
0
symbols represent repeated runs.
base. The latter is diagnostic of an overall greater reactivity
of chlorophenolate species for C–Cl H2 scission, as noted
elsewhere [14]. The return to higher (R2,4-DCP)0/(RHDC)0
−3
−3
0.025 moldm , suggests that (at a common starting Cl/Pd
ratio) the additional Cl substituent has a deactivating ef-
fect as noted elsewhere [9,30]. The drop in (RHDC)0 at
values at [NaOH]0 > 0.049 moldm suggests that the elec-
trostatic repulsion exhibited by the negatively charged cata-
lyst surface with respect to the chlorophenolate anions limits
the extent of complete dechlorination to phenol. The yield of
cyclohexanone (YC=O%) after 120 min HDC exhibited little
variation with NaOH addition; i.e., the subsequent hydro-
genation step is insensitive to solution pH.
−3
higher [NaOH]0 (ꢀ 0.057 moldm ) in both cases can be
attributed to electrostatic repulsion between the negatively
charged surface and the chlorophenolate anions at higher
pH (> 11). Comparing 2,4-DCP with 2-CP, the presence
of the second electron-withdrawing Cl substituent serves to
more effectively accommodate the negative charge of the
2,4-dichlorophenolate anion through delocalization over the
aromatic ring with a resultant weaker electrostatic repulsion
between the anions and the negative charged catalyst surface
than is the case for 2-CP. This may contribute to a lesser 2,4-
DCP HDC inhibition at higher pH, a response that is sup-
ported by the greater relative drop in 2-CP (RHDC)0, where
The addition of NaOH also impacted on HDC selectivity
as illustrated in Fig. 2, where the results of repeated runs are
included to demonstrate experimental reproducibility. The
lowest 2-CP selectivity (S2-CP) values were recorded in the
absence of base while NaOH addition where [NaOH]0 =
−3
0
.0–0.025 moldm (initial pH < 9) had no significant ef-
fect on the selectivity/conversion trends. With an increase in
−3
[
NaOH]0 (0.025–0.049moldm ), S2-CP values were higher
−3
at lower X2,4-DCP but declined to coincide on a common se-
[NaOH]0 ꢀ 0.057 moldm (Fig. 3) and is consistent with

G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
515
Table 3
Effects of NH OH addition on 2,4-DCP HDC over Pd/Al O : nomenclature as in Table 2
4
2 3
[
NH OH]
[NH OH] /[organic-Cl]
(R
)
(R
) /(R
2,4-DCP 0
)
X
Cl
Y
%
pH
4
0
4
0
0
HDC 0
HDC 0
C=O
−3
−1 −1
)
cat
(0–120 min)
(moldm
)
(mmol min
Cl
g
0
0
0
0
0
2.9
5.9
3.9
4.1
0.88
0.78
0.79
0.80
0.68
0.72
1.0
4.1
5.1
6.9
6.9
4.8–1.5
7.7–1.5
9.7–7.7
9.9–8.8
.013
.058
.074
0.23
1.02
1.30
1.0
a higher S2-CP for 2,4-DCP HDC at higher [NaOH]0 (see
Fig. 2a).
3
.2.2. NH4OH addition: HDC of 2,4-DCP
It has been reported [13,32] that NaOH, as a strong
base, can leach supported Pd into solution and damage the
pore structure of the support when used in liquid-phase
−3
chlorobenzeneHDC at high concentration (> 0.5 moldm ).
We did not detect, in our studies, any significant loss of Pd or
alteration to the BET surface area over the 120 min 2,4-DCP
HDC in the presence of [NaOH]. The utilization of weak
bases such as NH4OH, CH3COONa or organic amines in
liquid-phase reactions can, nonetheless, serve as viable al-
ternatives [9]. The use of CH3COONa and organic amines
has decided economic and environmental remediation draw-
backs while the application of NH4OH has been found to
be superior to NaOH in the liquid HDC of CPs over Pd/C,
leading to significantly higher HDC activity [9]. The results
of NH4OH addition on 2,4-DCP HDC over Pd/Al2O3 are
included in Table 3. As in the case of NaOH, HDC ac-
Fig. 4. Effect of concentration and nature of the alkali metal hydrox-
ide on the selectivity with respect to 2-CP (S %) as a function of
2
-CP
) in the HDC of 2.4-DCP
,4-DCP
the fractional 2,4-DCP conversion (X
2
−3
−3
over Pd/Al O : [AMOH] = 0.000 mol dm
(+), 0.013 mol dm
2
3
0
(LiOH (!), NaOH (E), KOH (P), RbOH (—), and CsOH (1)) and
−3
0.074 mol dm (LiOH ("), NaOH (F), KOH (Q), RbOH (×), and CsOH
−3
−3
(2)); [2,4-DCP] = 0.0285 mol dm
.
0
tivity was highest at lower [NH4OH]0 (0.013 moldm
)
and the reaction was inhibited at higher [NH4OH]0 (0.058–
−
3
0
.074 moldm ). The (RHDC)0 values at higher [NH4OH]0
3
2
.2.3. LiOH, KOH, RbOH, and CsOH addition: HDC of
,4-DCP
We have demonstrated that the liquid-phase HDC activity
are slightly greater than those associated with NaOH at the
same concentration (Table 2). This suggests a lesser in-
−
hibition associated with lower [OH ], where the pKb of
associated with Pd/Al2O3 is dependent on bulk solution pH
with rate inhibition under strongly basic conditions. While
NH4OH has proved to be a more effective additive, the pos-
sible role of the countercation was probed by examining the
action of the four alkali metal (AM) hydroxides (i.e., LiOH,
KOH, RbOH, and CsOH), taking two extreme initial con-
NH4OH = 4.75 and only ca. 1.6 mol% of NH4OH disso-
−3
ciates in water at [NH4OH] = 0.074 moldm compared
with a total dissociation of NaOH. At higher [NH4OH]0,
complete dechlorination (XCl = 1) was achieved, which rep-
resents greater efficiency than that associated with NaOH
−3
(
Table 2). The selectivity trends at every [NH4OH]0 over-
centrations ([AMOH]0 = 0.013 and 0.074 moldm ); the
lapped with that recorded in the absence of base. Such
an insensitivity to NH4OH addition (where pH < 10) sug-
gests that the 2-chlorophenolate anion may experience a
lesser electrostatic repulsion with the weakly negatively
results are given in Table 4. At a given [AMOH] each re-
0
action proceeded over a similar pH range with no significant
difference in Y
ever, the S2-CP associated with the higher [CsOH] is notably
greater (at X2,4-DCP > 0.4) when compared with the other
AMOH bases (Fig. 4), indicative of a more severe inhibi-
tion of the 2-CP to phenol step. At the higher [AMOH] , the
catalyst bore a negative charge with a consequent preference
for interaction with external cationic species. The adsorption
C=O
, (R
2,4-DCP 0
) /(R
HDC 0 Cl
) ratio or X . How-
0
−
charged catalyst surface: lower [OH ] does not limit the
2
-CP to phenol step to the same extent. The utilization of
−3
excess NH4OH (0.074 moldm ) was accompanied by a
narrow pH span (9.9–8.8) (Table 3) and ultimately led to
complete dechlorination. A slightly higher YC=O and lower
R2,4-DCP)0/(RHDC)0 was obtained at higher [NH4OH]0
> 0.058 moldm ) when compared with the NaOH system
0
+
(
enthalpies and affinity sequences for AM on a negatively
charged Al O surface are known [33] to decrease in the
order Li > Na > K > Rb > Cs ; i.e., the least fa-
−3
(
(
2
3
+
+
+
+
+
see Tables 2 and 3) and follows from enhanced HDC as a re-
sult of weaker repulsion between chlorophenolate/phenolate
anion(s) and the catalyst surface.
vorable surface interaction under basic conditions is with
external Cs . However, in aqueous solution Cs , with the
+
+

516
G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
Table 4
Effects of varying the nature and concentration of the alkali metal hydroxide addition on 2,4-DCP HDC over Pd/Al O : nomenclature as in Table 2
2
3
[
(
AMOH]
AMOH
(R
)
(R
) /(R
2,4-DCP 0
)
X
Cl
Y
%
pH
0
3
HDC 0
HDC 0
C=O
−
−1 −1
)
cat
(0–120 min)
moldm
)
(mmol min
Cl
g
0
0
2.9
0.88
0.68
4.1
4.8–1.5
.013
LiOH
NaOH
KOH
RbOH
CsOH
6.8
6.0
6.0
6.2
5.5
0.76
0.79
0.79
0.77
0.80
0.77
0.75
0.72
0.68
0.68
5.2
4.6
5.4
5.5
5.1
8.0–1.8
7.9–1.7
7.9–1.7
7.6–1.9
7.9–1.7
0.074
LiOH
NaOH
KOH
RbOH
CsOH
3.4
3.4
4.3
3.9
3.8
0.88
0.88
0.84
0.87
0.88
0.89
0.92
0.93
0.90
0.90
6.1
3.4
5.0
4.3
3.5
12.9–11.3
12.9–11.1
12.9–10.6
13.0–10.7
13.0–10.8
smallest hydrated ionic radius and greatest chemical polar-
izability (softness) [33,34] of the AM cations, can exhibit
a stronger interaction with chlorophenolate anions at higher
[
AMOH]0. This may result in the formation of an ion pair
+
between chlorophenolate anions and Cs cation at the inter-
face that serves to inhibit further hydrogen scission of the
ortho-substituted Cl resulting in the higher observed S2-CP
values.
3
.3. Effects of base addition on the HDC performance of
Pd/C
3
.3.1. NaOH addition: HDC of 2,4-DCP and 2-CP
In liquid-phase chloroarene HDC, it has been demon-
strated that Pd/C exhibits higher specific HDC activity than
Pd supported on oxide supports, an effect that has been at-
tributed to the greater adsorption capability of carbon [35]
while the involvement of H2 spillover may also be a con-
tributing factor [18,19]. A commercial Pd/C catalyst of com-
parable Pd loading (see Table 1) to the Pd/Al2O3 cata-
lyst was chosen to investigate the possible role of the sup-
port in modifying base addition effects on HDC. Over the
−3
[
NaOH]0 range 0.012–0.095 moldm ([NaOH]0/[organic-
Cl]0 = 0.2–1.7), Pd/C HDC performance is summarized
in Table 5. The temporal liquid-phase compositions and
pH profiles associated with the two [NaOH]0 extremes are
shown in Fig. 5. It can be seen (Fig. 5a) that the HDC
of 2,4-DCP over Pd/C also proceeds sequentially with 2-
CP as the sole partially dechlorinated product. In contrast
to Pd/Al2O3, there was no evidence of any cyclohexanone
formation over Pd/C during the course of HDC. The addi-
tion of NaOH significantly elevated both (RHDC)0 and XCl
Fig. 5. HDC of 2,4-DCP over Pd/C: (a) liquid-phase composition in
terms of mol% 2,4-DCP (F,E), 2-CP (2,1), and phenol (Q,P) as
a function of time; (b) pH of the bulk solution as a function of
(
Table 5) but, unlike Pd/Al2O3 (see Table 2), both these pa-
rameters attained the highest values at the higher [NaOH]0.
In common with Pd/Al2O3, the solution pH associated with
the lower [NaOH]0 dropped from ca. 8 to 2 with, in this in-
stance, an abrupt change at ca. 28 min (Fig. 5b). The pH
associated with the higher [NaOH]0 exhibited a lesser de-
crease from 12.6 to 11.9. It can be seen from the entries
in Table 5 that (RHDC)0 was largely unaffected by addition
−3
time; (1) [NaOH] = 0.012 mol dm
(solid symbols); (2) [NaOH]
=
0
0
−3
0
.095 mol dm (open symbols).
−
3
of dilute [NaOH]0 (ꢁ 0.012 moldm ), exhibiting a signif-
−3
icant increase when [NaOH]0 was raised to 0.062 mol dm
but was essentially invariant at higher concentrations, a re-

G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
517
Table 5
Effects of NaOH addition on 2,4-DCP HDC over Pd/C: nomenclature as in Table 2
[
(
NaOH]
[NaOH] /[organic-Cl]
(R
)
(R
) /(R
2,4-DCP 0
)
X
Cl
pH
0
−
0
0
HDC 0
HDC 0
3
−1 −1
)
cat
(0–120 min)
moldm
)
(mmol min
Cl
g
0
0
0
0
0
0
0
0
0
0.9
1.0
1.7
2.6
3.3
3.9
3.9
4.2
0.94
0.87
0.73
0.80
0.83
0.88
0.90
0.89
0.64
0.65
0.67
0.82
0.95
0.95
0.94
0.97
5.0–1.5
7.8–1.8
9.0–2.0
12.0–2.3
12.3–2.5
12.4–7.5
12.5–10.3
12.6–11.9
.012
.025
.037
.049
.062
.074
.095
0.23
0.44
0.65
0.86
1.08
1.30
1.67
Fig. 7. Initial 2,4-DCP (2) and 2-CP (Q) HDC rates ((R
over Pd/C as a function of [NaOH] ; [2,4-DCP] = 0.0285 mol dm
) )
HDC 0
−3
;
0
0
−3
[
2-CP] = 0.057 mol dm
.
0
As observed with Pd/Al2O3, S2-CP was sensitive to
[
NaOH]0, as shown in Fig. 6a, where the repeated runs
were again highly reproducible. However, the selectivity pat-
tern deviates significantly from that recorded for Pd/Al2O3
−
3
(
see Fig. 2a). Where [NaOH]0 = 0.000–0.012 mol dm
,
pH during the reaction < 8 (Fig. 6b) and the S2-CP vs
X2,4-DCP profiles coincide but are significantly higher than
those associated with Pd/Al2O3 at the same [NaOH]0. With
an increase in [NaOH]0 (0.025–0.037 mol dm ) HDC
over Pd/C delivered lower S2-CP values when the (tempo-
ral) reaction pH falls within 8–10, which are even lower
−3
than those associated with Pd/Al2O3. A further increase in
Fig. 6. HDC of 2,4-DCP over Pd/C: (a) selectivity with respect to 2-CP
) as a function of fractional 2,4-DCP conversion (X ) and
−3
[
NaOH]0 (0.049–0.062 mol dm ) with an associated pH
(
S
2
-CP
b) pH of the aqueous solution as a function of X
2,4-DCP
mainly > 10 served to raise S2-CP to a level similar to that
(
; [NaOH] = 0.000
2
,4-DCP
0
−3
(
0
F,E), 0.012 (+), 0.025 (2,1), 0.037 (",!), 0.049 (Q,P), 0.062 (∗),
where [NaOH]0 ꢁ 0.012 moldm , approaching the selec-
−3
−3
;
.074 (—), and 0.095 (×) mol dm ; [2,4-DCP] = 0.0285 mol dm
tivity associated with Pd/Al O at the same [NaOH] (see
0
2
3
0
solid and open symbols represent repeated runs.
Fig. 2a). At higher [NaOH]0, the distinctly higher values
of S2-CP (X2,4-DCP > 0.6) suggest an inhibition of the 2-
CP to phenol step, a response similar to that observed for
Pd/Al2O3 and which may be linked to the stronger electro-
static repulsion between 2-chlorophenolate anions and the
negatively charged surface. Comparing the direct 2-CP and
2,4-DCP HDC over Pd/C as a function of [NaOH]0 (Fig. 7),
the 2-CP (RHDC)0 exhibited a more significant increase
sponse that was matched by XCl. The lower value for the
−
3
ratio (R2,4-DCP)0/(RHDC)0 at [NaOH]0 = 0.025 moldm
is significant, as also observed in the case of Pd/Al2O3 (Ta-
ble 2), and suggests that the reaction pathway is sensitive to
a change in the fraction of (chloro-)phenolate and (chloro-)-
phenolic species (initial pH ca. 9).

518
G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
RbOH < CsOH. This sequence coincides with the adsorp-
+
+
tion affinity sequence of alkali metal cations (Li < Na <
+
+
+
K < Rb < Cs ) for a negatively charged carbon surface
33,34], suggesting a HDC pH dependence for Pd/C that is
[
the result of alkali metal interactions at the interface and
which is not evident for reactions involving Pd/Al2O3; i.e.,
the alkali metal cations have a distinct cocatalytic role dur-
ing HDC over Pd/C. Regardless, in common with Pd/Al2O3,
the selectivity trends observed for Pd/C are largely insensi-
tive to the nature of the AM hydroxides at a given [AMOH]0
(
Fig. 8). Moreover, as in the case of Pd/Al2O3 (Fig. 4), reac-
tion in the presence of the higher [CsOH]0 delivered consis-
tently greater S2-CP at each X2,4-DCP, an effect that we again
+
ascribe to Cs interaction with chlorophenolate anions that
hinders the 2-CP to phenol HDC step.
Fig. 8. Selectivity with respect to 2-CP (S
) as a function of frac-
-CP
2
tional 2,4-DCP conversion (X
) in the HDC of 2.4-DCP over
−3
(LiOH (1),
2
,4-DCP
4
. Discussion
−
3
Pd/C; [AMOH] = 0.000 mol dm
(+), 0.025 mol dm
0
−3
NaOH (E), KOH (P), RbOH (×), and CsOH (!)) and 0.074 mol dm
(
(LiOH (2), NaOH (F), KOH (Q), RbOH (—), and CsOH ("));
4.1. Mass-transport considerations
−3
[2,4-DCP] = 0.0285 mol dm
.
0
The relative importance of physical/chemical control in
the overall HDC of 2,4-DCP and 2-CP (with NaOH ad-
dition) over supported Pd has been the subject of previ-
ous reports [17,18,30]. Aqueous HDC of chlorophenols,
as a three-phase system, was identified as a fast reac-
tion over Pd/C and Pd/Al2O3 where any transport con-
straints at gas/liquid and/or liquid/solid interfaces and/or
the solid phase (intraparticle) can limit the true HDC rate
[17]. Hydrogen transport at the external/internal liquid/solid
interface was found to be the predominant physical con-
straint. It was confirmed that the reaction operated under
kinetic control with stirring speeds > 1000 rpm, H2 flow
−
3
where [NaOH]0 ꢁ 0.049 mol dm that may contribute to
the lower S2-CP values at lower [NaOH]0 shown in Fig. 6a.
−3
The lower 2-CP HDC rates at [NaOH]0 > 0.057 mol dm
can be linked to the higher S2-CP values where pH ꢀ 10, as
is the case with Pd/Al2O3.
3
.3.2. NH4OH addition: HDC of 2,4-DCP
Unlike the HDC response observed for Pd/Al2O3 (Ta-
ble 3), the addition of NH4OH proved to be less effective
Table 6) in terms of (RHDC)0 and XCl over Pd/C. The high-
(
−3
est (RHDC)0 was attained at [NH4OH]0 = 0.2–0.3 mol dm
3
−1
−3
(
0
initial pH ca. 10), while the addition of dilute NH4OH (ca.
rate > 150 cm min , [2,4-DCP]0 = 0.0475 mol dm
,
−3
−3
.01 mol dm ) resulted in a significant increase (by a fac-
[2-CP]0 = 0.095 mol dm , [NaOH]0/[organic-Cl]0 = ca.
−3
tor of 2) in the HDC activity of Pd/Al2O3. This difference
in response is again suggestive of a distinct pH dependence
of the (RHDC)0 delivered by both catalysts. In every HDC
operation where pH < 10.5, the selectivity with respect to
-CP associated with NH4OH addition was less than that
achieved without base, reaching a minimum at [NH4OH]0 =
1 mol/mol, Ccat ꢁ 0.5 g dm , and catalyst particle size
< 45 µm. Under these conditions, the initial HDC rates
−
1
−1
(T = 303 K) were 3.6 and 5.6 mmolg min for 2,4-
cat
DCP and 2-CP conversion, respectively, over Pd/C [18,30].
In the present study, where lower [2,4-DCP]0 and [2-
CP]0 (0.0285 and 0.057 mol dm ) and constant [organic-
2
−3
−3
−3
0
.05 mol dm
.
Cl]0/Ccat (Ccat = 0.3 g dm ) were employed, the (RHDC)0
values recorded for Pd/C were in the range 0.9–6.8 and
−1
−1
3
2
.3.3. LiOH, KOH, RbOH, and CsOH addition: HDC of
,4-DCP
2.4–7.1 molgcat min for 2,4-DCP and 2-CP HDC, re-
spectively; i.e., the same order of magnitude as the pre-
vious results and mass-transfer contributions can be taken
to be negligible. In the case of Pd/Al2O3, the highest
HDC initial rate recorded (T = 303 K) under an estab-
lished kinetic regime ([NaOH]0/[organic-Cl]0 = 1), [2,4-
−
−3
Taking [OH ]0 = 0.025 and 0.074 mol dm as two ex-
treme cases, 2,4-DCP HDC over Pd/C was conducted with
the addition of LiOH, KOH, RbOH, and CsOH and the re-
sults are compared with those generated for NaOH addition
in Table 7. At the lower [AMOH]0, the Pd/C HDC behav-
ior exhibited no significant dependence on the nature of the
AM hydroxide, as observed in the case of Pd/Al2O3 (Ta-
ble 4). At the higher [AMOH]0, the initial HDC rates ex-
hibited a decided dependence and the following sequence
of increasing (RHDC)0 emerges: LiOH ≈ NaOH < KOH ≈
−
3
−3
DCP]0 = 0.0475 mol dm , and Ccat = 0.5 g dm ) was
−1
−1
4.1 mmolgcat min [30]. The (RHDC)0 values recorded in
this study fall within 2.9–6.0 mmolg min and can again
−1
−1
cat
be taken to be free from physical transport constraints. The
observed responses to base addition reported in this study
can then be attributed to chemical/catalytic phenomena.

G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
519
Table 6
Effects of NH OH addition on 2,4-DCP HDC over Pd/C: nomenclature as in Table 2
4
[
NH OH]
[NH OH] /[organic-Cl]
(R
)
(R
) /(R
2,4-DCP 0
)
X
Cl
pH
4
0
4
0
0
HDC 0
HDC 0
−3
−1 −1
)
cat
(0–120 min)
(moldm
)
(mmol min
Cl
g
0
0
0
0
0
0
0
0
0
0
0
0.9
1.0
1.1
1.7
2.1
2.3
2.5
2.7
3.1
2.7
0.94
0.87
0.73
0.77
0.81
0.80
0.79
0.79
0.78
0.79
0.64
0.76
0.80
0.87
0.88
0.92
0.94
0.94
0.96
0.94
5.0–1.5
7.7–1.7
8.7–1.9
9.3–7.4
9.8–9.0
10.0–9.1
10.2–9.5
10.3–9.6
10.4–9.7
10.5–9.9
.013
.025
.050
.100
.125
.175
.200
.224
.298
0.23
0.44
0.88
1.75
2.19
3.07
3.51
3.93
5.23
Table 7
Effects of varying the nature and concentration of the alkali metal hydroxide addition on 2,4-DCP HDC over Pd/C: nomenclature as in Table 2
[
(
AMOH]
AMOH
(R
)
(R
) /(R
2,4-DCP 0
)
X
Cl
pH
0
3
HDC 0
HDC 0
−
−1 −1
)
cat
(0–120 min)
moldm
)
(mmol min
g
Cl
0
0
0.9
0.94
0.64
5.0–1.5
.025
LiOH
NaOH
KOH
RbOH
CsOH
1.5
1.7
1.0
1.3
1.2
0.82
0.73
0.86
0.85
0.89
0.79
0.67
0.68
0.68
0.67
9.2–1.9
9.0–2.0
8.6–1.9
8.5–2.1
9.1–2.1
0.074
LiOH
NaOH
KOH
RbOH
CsOH
3.7
3.9
5.5
5.3
6.8
0.82
0.90
0.76
0.82
0.84
0.93
0.94
0.98
0.96
1.0
12.6–10.4
12.5–10.3
12.5–10.0
12.5–10.0
12.7–10.1
sponse to changes in bulk solution pH. While the pHpzc
of Pd/Al2O3 (7.8) and Pd/C (8.5) are close, conversion of
+
the PMT titration data (used to determine pHpzc) into H
−
and OH uptake as a function of equilibrium solution pH
is revealing, as can be judged from the entries in Fig. 9.
+
−
The difference in H and OH uptake (on a surface area
basis) is such that a clear distinction in the behavior of
Pd/Al2O3 and Pd/C emerges. This is manifest in the higher
+
surface charge density (corresponding to the higher H or
OH uptake) of Pd/Al2O3 at the extreme pH values (< 6
−
or > 10). This should result in either a stronger electrosta-
tic attraction (pH < 6) or repulsion (pH > 10) with respect
to chlorophenolate anions in solution. Values for the pHpzc
and density of active surface hydroxyl groups (AlOH) for
γ -Al2O3 have been reported [26,29,36,37] in the range 6.9–
+
−
Fig. 9. H (open symbols) and OH (solid symbols) uptake on Pd/Al O
2
3
−2
9
.5 and 1.7–14 µmol m , respectively, which are consistent
(
Q,P) and Pd/C (F,E) as a function of the equilibrium solution pH;
Ccat = 6 gdm ; 0.1 moldm aqueous NaCl served as blank.
−3
−3
with our results for Pd/Al2O3. As hydrated Al2O3 exhibits
a weak chemical affinity for hydrocarbons [33,34], electro-
static rather than dispersion effects should dominate the in-
teraction between chlorophenol(s)/phenol and Al2O3, as has
been demonstrated in an FTIR analysis of phenolic uptake
(and pH dependence) on Al2O3 [29,38]. The surface charge
density is lower for Pd/C which possesses a high surface
area (Table 1) with an uptake capacity that necessitates pH
extremes (< 4 or > 10, see Fig. 9) to achieve surface satura-
4
.2. Effects of base addition: Pd/Al2O3 vs Pd/C
In this study, 2,4-DCP and 2-CP HDC was conducted
over a broad pH range (1.5–13.0) with the addition of var-
ious bases. An understanding of the amphoteric nature of
the catalyst surface is essential to explain the differences
in HDC behavior exhibited by Pd/C and Pd/Al2O3 in re-

520
G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
tion. The reported pHpzc for activated carbon falls within the
rather broad range 3–10, the actual value depending on the
carbon source and nature of the pretreatment/activation [25,
In the case of Pd/C, a higher pH (ꢀ 9) results in more ef-
fective HDC (Table 5), indicative of the beneficial effects of
chloroarene dissociation leading to higher (RHDC)0. In con-
trast to Pd/Al2O3, the (RHDC)0 delivered by Pd/C increased
with increasing [NaOH]0 (> 0.037 moldm ), a direct re-
sult of the difference in the chemical nature of Al2O3 and
carbon surfaces. The latter bears a lower charge density (less
electrostatic repulsion) and a stronger nonelectrostatic at-
traction with chlorophenolate anions in solution.
2
8,39,40]. To illustrate the heterogeneity of activated carbon
−3
surfaces, the HCl adsorption capacities of two commercially
available carbon samples have been reported to span the
−2
range 0.016–1.73 µmol m [25]. Our PMT titration/BET
data for Pd/C are consistent with a low density of basic func-
tional groups on a high surface area carrier. Consequently,
the possible interaction(s) between the carbon surface and
the chlorophenols/phenol is (are) [28,41–43]:
The difference in the selectivity response with respect
to the partially dechlorinated 2-CP is evident where pH <
1
0. In the absence of base, a lower S2-CP was delivered by
(
i) electrostatic attraction/repulsion associated with partic-
ipating charged species;
Pd/Al2O3 at a given X2,4-DCP; compare the entries in Figs. 2a
and 6a. This can be linked to the stronger electrostatic attrac-
tion between the 2-chlorophenolateanions and the positively
charged Al2O3 (pH < 5, see Fig. 2b), where the delocaliza-
tion of the negative charge on the 2-chlorophenolate anion is
less effective than the disubstituted arene due to presence of
the second electron-withdrawing para-Cl. Comparing the 2-
CP HDC performance over Pd/Al2O3 with 2,4-DCP (Fig. 3),
it is evident that the decline of (RHDC)0 for 2-CP is more
sensitive to the addition of NaOH (with a consequent re-
duction in positive surface charge density). By compari-
son, Pd/C exhibits a greater preference for interaction with
chlorophenolic species at low pH. In terms of adsorption
affinity, the uptake capacity of activated carbon for 2,4-DCP
has been found to be ca. 1.5–6 times of that for 2-CP [45,46].
This can be linked to the greater solubility of 2-CP (28.5
(
ii) nonelectrostatic attractions, i.e., dispersion (π–π elec-
trons) effects between the aromatic ring (reactant in
solution) and carbon surface, hydrogen bonding, and hy-
drophobic interaction.
The greater chemical affinity exhibited by carbon, when
compared with Al2O3 for interaction with hydrocarbons al-
lied to the lower surface charge density must translate into
a greater importance of nonelectrostatic contributions. This
is supported by the literature [14,28,39,41–44] dealing with
the pH dependence of chlorophenol(s)/phenol adsorption on
activated carbon. It is accepted [28,41,42] that uptake is fa-
vored under acidic conditions (pH < 3) where the interaction
between the positively charged surface and the chlorophe-
nolic/phenolic species is facilitated via the dispersion effect.
The adsorption capacity declines at pH close to the pKa of
chlorophenol/phenol to drop (at pH > 12) to 22–80% of the
uptake in acid media [14,39,41–44], an effect that is taken
to result from electrostatic repulsion between the negatively
charged carbon surface and the (chloro-)phenolate anions
−3
compared with 4.5 g dm for 2,4-DCP) but is also the re-
sult of a weaker dispersion force between the 2-CP and the
carbon surface. Under near neutral conditions (pH 8–10),
the decrease in S2-CP (see Figs. 2a and 6a) suggests im-
proved 2-CP conversion due to a more effective contribution
of electrostatic forces; 2-CP dissociation at pH 8–10 is ca.
22–96 mol/mol%. A comparison of 2-CP HDC performance
over Pd/C with 2,4-DCP (Fig. 7) reveals a more significant
[
28,41,42].
Features common to HDC over both Pd/C and Pd/Al2O3
−
3
are the increase in XCl with the addition of base (Tables 2–
), higher S2-CP at higher [NaOH]0 (Figs. 2a and 8a), and
increase of 2-CP HDC rates at [NaOH]0 ꢁ 0.037 moldm
with a subsequent decline at [NaOH]0 > 0.057 mol dm 3,
demonstrating the role of electrostatic forces in determining
2-CP reactivity on Pd/C.
−
7
the significant elevation of S2-CP with the addition of more
concentrated CsOH (Figs. 4 and 8). The first effect can be
attributed to a suppression of HCl poisoning, and the second
a result of the more severe inhibition of the 2-CP to phenol
step due to stronger electrostatic repulsion between the 2-
chlorophenolate anions and the negatively charge surface at
higher pH. The third response can be explained on the basis
The response to NH4OH addition over both catalysts fol-
lows the same trends established for NaOH. The higher
[NH4OH]0 required to influence HDC performance can be
attributed to the weaker basicity of NH4OH which demon-
strates further that HDC performance is governed by solu-
tion pH. An increase in HDC activity at higher [AMOH]0 in
the order of CsOH > RbOH ≈ KOH > NaOH ≈ LiOH was
only apparent for reactions over Pd/C, where the AMOH
can be considered to behave as a cocatalyst. The activity
+
of a Cs /chlorophenolate ion pair formation that limits the
degree of dechlorination of the sterically hindered ortho-Cl
substituent. Taking 2,4-DCP HDC over Pd/Al2O3, a maxi-
mum (RHDC)0 was attained at pH 7–9 with NaOH addition
+
(
Table 2). At pH < 5 the depletion of chlorophenolate anions
sequence matches that of increasing AM affinity for the
+
in solution limits HDC while at pH > 10 the development of
a negative charge on the support hinders reactant/catalyst in-
teraction(s) leading to C–Cl scission. Bulk solution pH close
to the chloroarene pKa and Al2O3 pHpzc is optimum, sug-
gesting that chlorophenolate anions are more reactive than
chlorophenolic species, as was proposed by Felis et al. [15].
negatively charged carbon surface where AM interactions
at the interface enhance (RHDC)0. Such an effect is not with-
out precedent in that it has been established [47,48] that
the presence of charge-transfer cations influences the hy-
droprocessing activity of supported Pd through a modifica-
tion of the supported metal site electron density. In the case

G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
521
of Pd/Al2O3, any beneficial effects due to increasing polar-
Notation
izability of the AM in solution is counterbalanced by the
reverse sequence of the chemical affinity with the alumina
surface and the resultant (RHDC)0 is largely insensitive to
the nature of the AM cation.
[
AMOH]0
Initial concentration of alkali metal
hydroxide (mol dm 3);
−
Concentration of base (moldm 3);
−
[
Base]
Concentration of catalyst (gcat dm 3);
−
Ccat
[
2-CP]0
Initial concentration of 2-chlorophenol
(
mol dm 3);
−
5
. Conclusions
dPd
Surface-area-weighted average Pd particle
diameter (nm);
In the aqueous-phase batch HDC of 2,4-DCP over 1
[2,4-DCP]0
[NaOH]0
Initial concentration of 2,4-dichlorophenol
(mol dm 3);
−
wt/wt% Pd/Al2O3 and Pd/C in the absence and presence of
base, the results of this study support the following conclu-
sions:
Initial concentration of NaOH (mol dm 3);
−
[Organic-Cl]0 Initial concentration of Cl associated with
the aromatic feed (mol dm 3);
pH associated with the point of zero
charge for the catalyst;
−
(
1) HDC over Pd/Al2O3 delivers a maximum (RHDC)0 at
pH 7–9 (NaOH addition), suggesting that both a deple-
tion of chlorophenolate anions in solution (pH < 5) and
the development of a negative charge on the support
pHpzc
(R2,4-DCP)0
Initial rate of consumption of
−1
−1
2,4-dichlorophenol (mmolg min );
cat
(
pH > 10) hinders catalyst/reactant interaction(s) that
(R
)
Initial hydrodchlorination rate
HDC 0
−
−1
mmolCl g min );
cat
1
facilitate C–Cl scission. In the case of Pd/C, a higher
pH (ꢀ 9) results in more effective HDC due to the
stronger dispersion forces between the aromatic ring of
the chlorophenolate anions and the carbon surface.
2) An increase in the ultimate fractional dechlorination
with the addition of base is observed for both catalysts
and can be attributed to a lessening of any inhibition
due to the HCl product. Appreciably more concentrated
NH4OH, as a weak base, is required to significantly el-
evate HDC but a complete dechlorination of 2,4-DCP is
possible.
(
S2-CP
Hydrodechlorination selectivity with
respect to 2-chlorophenol, defined by
Eq. (2) (dimensionless);
Fractional dechlorination after 120 min
reaction (dimensionless);
Fractional conversion of
2,4-dichlorophenol, defined by Eq. (1)
(dimensionless);
(
(
X
X
Cl
2
,4-DCP
YC=O
Yield of cyclohexanone after 120 min
reaction (dimensionless).
3) A higher S2-CP at higher [NaOH]0 (pH > 10) is observed
for both catalysts and can be linked to a suppression of
the 2-CP to phenol HDC step as a result of the stronger
electrostatic repulsion between the 2-chlorophenolate
anion and the negatively charge surface. S2-CP differs for
reaction over Pd/Al2O3 and Pd/C at pH < 10 where, in
the absence of base, a higher S2-CP is delivered by Pd/C
due to the weaker dispersion attraction between the 2-
CP and the carbon surface while at pH 8–10 the lower
S2-CP associated with Pd/C is the result of a more effec-
tive electrostatic interaction with the 2-chorophenolate
anion.
Acknowledgment
This work was supported in part by the National Science
Foundation through Grant CTS-0218591.
References
[1] P.M. Armenante, D. Kafkewitz, G.A. Lewandowski, C.-J. Jou, Water
Res. 33 (1999) 681.
[
2] US EPA, The Inventory of Sources of Dioxin in the United States,
EPA/600/P-98/002Aa, 1998.
(
(
4) In the case of Pd/C, (RHDC)0 at higher [AMOH]0
−3
+
(
0.074 mol dm ) increases in the order of Li
≈
[3] US EPA, Chemical Advisory and Notice of Potential Risk: Skin Expo-
sure to Molten 2,4-Dichlorophenol (2,4-DCP) can Cause Rapid Death,
Advisory No. 8EHQ-14302, 2000.
+
+
+
+
Na < K ≈ Rb > Cs which coincides with the ad-
sorption affinity sequence on a negatively charged car-
bon surface where interaction with Pd/C through charge
transfer must serve to raise the HDC activity of the Pd
sites; the Pd/Al2O3 HDC rate is insensitive to the nature
of the AM cation.
[4] C.C. Wang, C.M. Lee, C.H. Kuan, Chemosphere 41 (2000) 447.
[5] S. Sabhi, J. Kiwi, Water Res. 35 (2001) 1994.
[6] M.D. Erickson, S.E. Swanson, J.D. Flora, G.D. Hinshaw, Environ. Sci.
Technol. 23 (1989) 462.
[7] E.-J. Shin, M.A. Keane, Chem. Eng. Sci. 54 (1999) 1109.
[8] P.N. Rylander, Catalytic Hydrogenation over Platinum Metals, Acad-
emic Press, New York, 1967, p. 405.
5) A significantly higher S2-CP at higher [CsOH]0 is ob-
served for both catalysts and can be attributed to
[9] J.B. Hoke, G.A. Gramiccioni, E.N. Balko, Appl. Catal. B 1 (1992) 285.
+
chlorophenolate anion/Cs ion pair formation that hin-
[
10] V. Felis, P. Fouilloux, C.D. Bellefon, D. Schweich, Rec. Prog. Gen.
ders ortho-Cl dechlorination.
Proc. 13 (1999) 303.

522
G. Yuan, M.A. Keane / Journal of Catalysis 225 (2004) 510–522
[
11] F.H. Ribeiro, C.A. Gerken, G.A. Somorjai, C.S. Kellner, Catal. Lett. 45
[28] L.R. Radovic, I.F. Silva, J.I. Ume, J.A. Menendez, C.A. Leon y Leon,
A.W. Scaroni, Carbon 35 (1997) 1339.
(1997) 149.
[
12] B. Coq, G. Ferrat, F. Figueras, J. Catal. 101 (1986) 434.
[29] J.C. Liu, C.P. Huang, J. Colloid Interface Sci. 153 (1992) 167.
[30] G. Yuan, M.A. Keane, Catal. Commun. 4 (2003) 195.
[31] J. Dolfing, B.K. Harrison, Environ. Sci. Technol. 26 (1992) 2213.
[32] E.N. Balko, E. Przybylski, F. Von Trentini, Appl. Catal. B 2 (1993) 1.
[33] N. Sahai, Geochim. Cosmochim. Acta 64 (2000) 3629.
[34] R.G. Pearson, Inorg. Chem. 27 (1988) 734.
[35] K. Pirkanniemi, M. Sillanpaa, Chemosphere 48 (2002) 1047.
[36] W. Piasecki, Phys. Chem. Chem. Phys. 5 (2003) 713.
[37] E. Laiti, P. Persson, L.-O. Ohman, Langmuir 12 (1996) 2969.
[38] J. Bandara, J.A. Mielczarski, J. Kiwi, Appl. Catal. B 34 (2001) 307.
[39] C. Namasivayam, D. Kavitha, J. Harzard. Mater. B 98 (2003) 257.
[40] Yu. Matatov-Meytal, Yu. Shindler, M. Sheintuch, Appl. Catal. B 45
(2003) 127.
[
13] M.A. Aramendia, R. Burch, I.M. Garcia, A. Marinas, J.M. Marinas,
B.W.L. Southward, F.J. Urbano, Appl. Catal. B 31 (2001) 163.
14] V. Felis, C.D. Bellefon, P. Fouilloux, D. Schweich, Appl. Catal. B 20
[
[
(1999) 91.
15] M.A. Aramendia, V. Borau, I.M. Garcia, J.C. Jimenez, F. Lafont, A.
Marinas, J.M. Marinas, F.J. Urbano, J. Mol. Catal. A: Chem. 184
(2002) 237.
[16] W.S. Stoesser, E.H. Sommerfield, A.H. Widiger, Ind. Eng. Prod. Res.
Dev. 10 (1971) 76.
[17] G. Yuan, M.A. Keane, Chem. Eng. Sci. 58 (2003) 257.
[18] G. Yuan, M.A. Keane, Catal. Today 88 (2003) 27.
[19] Yu. Shindler, Yu. Matatov-Meytal, M. Sheintuch, Ind. Eng. Chem.
Res. 40 (2001) 3301.
[41] C. Moreno-Castilla, Carbon 42 (2004) 83.
[
[
20] N. Krishnankutty, M.A. Vannice, J. Catal. 155 (1995) 312.
21] J. Vakros, C. Kordulis, A. Lycourghiotis, J. Chem. Soc., Chem. Com-
mun. (2002) 1980.
[42] K. Laszlo, A. Szucs, Carbon 39 (2001) 1945.
[43] A.P. Terzyk, J. Colloid Interface Sci. 268 (2003) 301.
[44] T.M. Grant, C.J. King, Ind. Eng. Chem. Res. 29 (1990) 264.
[45] E. Knettig, B.M. Thomson, S.E. Hrudey, Environ. Poll. Ser. B, Chem.
Phys. 12 (1986) 281.
[46] A.A.M. Daifullah, B.S. Girgis, Water Res. 32 (1998) 1169.
[47] S. Scire, C. Crisafulli, R. Maggiore, S. Minico, S. Galvagno, Appl.
Surf. Sci. 93 (1996) 309.
[
[
[
[
22] R.T.K. Baker, E.B. Prestridge, R.L. Garten, J. Catal. 59 (1979) 293.
23] B. Ealet, E. Gillet, Surf. Sci. 281 (1993) 91.
24] H. Dropsch, M. Baerns, Appl. Catal. A 158 (1997) 163.
25] C.A. Leon y Leon, J.M. Solar, V. Calemma, L.R. Radovic, Carbon 30
(1992) 797.
[
[
26] J.S. Noh, J.A. Schwarz, J. Colloid Interface Sci. 130 (1989) 157.
27] F. Rodriguez-Reinoso, Carbon 36 (1998) 159.
[48] V. Vishwanathan, N. Mahata, M.A. Keane, React. Kinet. Catal.
Lett. 72 (2001) 297.