A kinetic study of 2-propanol dehydration on carbon acid catalysts

Article abstract of DOI:10.1016/j.jcat.2010.01.023

Activated carbons with high surface acidity were obtained in a single step by chemical activation of hemp residues with H₃PO₄ and used as catalysts for the dehydration reaction of 2-propanol. Despite the washing process performed after the activation, the resulting carbons show a considerable amount of surface phosphorus as revealed by XPS. The surface acidity, predominantly of Broensted type, was dependent on the amount of phosphorus retained on the carbon surfaces. Conversion of 2-propanol yielded only dehydration products, mostly propylene with a very low amount of di-isopropyl ether. The effect of a thermal treatment performed to carbon catalyst on the surface chemistry, acidity and dehydration activity was analyzed. A kinetic study of the catalytic dehydration of 2-propanol on the best carbon catalyst was carried out, where two different Langmuir-Hinshelwood mechanisms, via surface elimination reactions, E1 and E2, were analyzed. The rate expressions derived from both models fitted properly the experimental results and activation energy values of about 98 kJ/mol were obtained. The results indicate that the 2-propanol dehydration reaction takes place by an E2 elimination mechanism with strong E1 character.

Full text of DOI:10.1016/j.jcat.2010.01.023

Journal of Catalysis 271 (2010) 33–42
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
A kinetic study of 2-propanol dehydration on carbon acid catalysts
*
J. Bedia, R. Ruiz-Rosas, J. Rodríguez-Mirasol , T. Cordero
Chemical Engineering Department, School of Industrial Engineering, University of Málaga, Campus de Teatinos s/n, 29071 Málaga, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Activated carbons with high surface acidity were obtained in a single step by chemical activation of hemp
residues with H3PO4 and used as catalysts for the dehydration reaction of 2-propanol. Despite the wash-
ing process performed after the activation, the resulting carbons show a considerable amount of surface
phosphorus as revealed by XPS. The surface acidity, predominantly of Brönsted type, was dependent on
the amount of phosphorus retained on the carbon surfaces. Conversion of 2-propanol yielded only dehy-
dration products, mostly propylene with a very low amount of di-isopropyl ether. The effect of a thermal
treatment performed to carbon catalyst on the surface chemistry, acidity and dehydration activity was
analyzed. A kinetic study of the catalytic dehydration of 2-propanol on the best carbon catalyst was car-
ried out, where two different Langmuir–Hinshelwood mechanisms, via surface elimination reactions, E1
and E2, were analyzed. The rate expressions derived from both models fitted properly the experimental
results and activation energy values of about 98 kJ/mol were obtained. The results indicate that the 2-
propanol dehydration reaction takes place by an E2 elimination mechanism with strong E1 character.
Ó 2010 Elsevier Inc. All rights reserved.
Received 4 August 2009
Revised 20 January 2010
Accepted 26 January 2010
Available online 2 March 2010
Keywords:
Activated carbon
Chemical activation
Surface acidity
2-Propanol dehydration
1. Introduction
69.0 million tons in 2006, and the demand is expected to increase
faster than supply up to 350 million tons in 2020 [8]. Nowadays,
Flash pyrolysis of biomass produces liquid oil (bio-oil) consti-
tuted mostly by oxygenated compounds which, after the proper
treatments, are of great interest as fuel [1,2]. Like the biomass pre-
cursor, this bio-oil contains almost negligible amount of nitrogen,
sulfur and metals. However, transformation of bio-oil is necessary
to reach a quality similar to that of the conventional fuel. Studies
have been conducted to improve the quality of bio-oil as fuel
through catalytic processes with inorganic acid solids [3–5]. Gay-
ubo et al. [4] studied the catalytic transformation of several model
components of biomass pyrolysis oil, such us phenols and alcohols
(1 and 2-propanol, 1 and 2-butanol, . . .), over HZSM-5 zeolite
obtaining mainly light olefins and aromatics.
Alcohol decomposition is also used as test reaction for the eval-
uation of the acid and base properties of catalysts. Alcohol dehy-
drogenation products (aldehydes and ketones) are preferentially
formed on basic catalysts, while dehydration products (olefins
and ethers) are favored when acidic sites are present [6]. It is gen-
erally accepted that 2-propanol decomposition over basic sites
proceeds through an elimination reaction yielding acetone. Over
acid sites, 2-propanol dehydrates to propylene and to di-isopropyl
ether [7]. Propylene is a key building block for the petrochemical
industry used for the production of polypropylene, acrylonitrile,
propylene oxide, cumene, oxo-alcohols and other intermediates.
Propylene demand has grown from 16.4 million tons in 1980 to
propylene is obtained from non-renewable sources mainly as by-
product of ethylene production in stream crackers and from refin-
ery in FCC streams. Therefore, the finding and developing of new
processes for propylene production from renewable sources is of
great interest from economical and environmental point of views.
Activated carbons are mainly used as adsorbent and catalyst
supports. However, the special characteristics of these materials
related to their porous structure and surface chemistry make them
suitable to be used also as catalysts per se [9]. Activated carbons
show some important advantages to be used as catalysts [10], like
a high specific surface area, high thermal and chemical stability
and the possibility to be obtained from many diverse materials
including different types of lignocellulosic waste [11–14]. Carbon
materials can be used as catalysts for acid/base reactions due to
the presence of surface oxygen groups of acidic and/or basic char-
acter [15,16]. To increase the surface acidity, carbons are usually
oxidized with different chemical compounds in order to introduce
oxygen surface complexes [16,17]. The increase in the surface acid-
ity is associated to the formation of carboxylic and lactonic groups,
which show a low to moderate stability [18]. Chemical activation
with phosphoric acid results in carbons with high thermal stability
surface groups that provide high surface acidity. These groups
decompose at temperatures higher than 700 °C producing CO and
CO2, as a consequence of the decomposition of oxygen–phosphorus
surface groups formed during the activation process [19].
In this study, acid carbons were studied as catalysts for the
selective catalytic dehydration of 2-propanol (as model compound
* Corresponding author. Fax: +34 951952385.
E-mail address: mirasol@uma.es (J. Rodríguez-Mirasol).
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.01.023

34
J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
Nomenclature
ANBE2T
apparent surface area (m2 gꢀ1
)
PoIPA
2-propanol partial pressure at the reactor inlet (atm)
PIPA
PPROPY
PROPY propylene
R
Rep
rCARB
rPROPY
rIPA
DS0ad
Sg0
2-propanol partial pressure (atm)
propylene partial pressure (atm)
CO2
A
narrow micropore surface area (m2 gꢀ1
)
DR
AtN
external surface area (m2 gꢀ1
Damköhler number
)
2
universal gas constant (R = 8.31 J molꢀ1 Kꢀ1
particle Reynolds number
rate of cationic species formation (mol sꢀ1 gꢀ1
rate of propylene formation (mol sꢀ1 gꢀ1
2-propanol conversion rate (mol sꢀ1 gꢀ1
standard entropy of adsorption (J molꢀ1 Kꢀ1
)
Da
CARB
dp
cationic species (carbocation or isopropoxide)
particle diameter (cm)
)
EaCARB
activation energy of the cationic species formation reac-
)
)
tion (kJ molꢀ1
)
EaPROPY activation energy of the propylene formation reaction
(kJ molꢀ1
standard enthalpy of adsorption (kJ molꢀ1
enthalpy of adsorption of 2-propanol (kJ molꢀ1
HPROPY enthalpy of adsorption of propylene (kJ molꢀ1
)
)
standard entropy in gas phase (J molꢀ1 Kꢀ1
)
DH0ad
HIPA
)
D
)
Sdii
Sp
di-isopropyl ether selectivity
propylene selectivity
temperature (K)
D
IPA
kCARB
KIPA
)
2-propanol
cationic species formation rate constant (mol gꢀ1 sꢀ1
2-propanol adsorption constant (atmꢀ1
preexponential factor of the cationic species formation
(mol gꢀ1 sꢀ1
preexponential factor of the 2-propanol adsorption
(atmꢀ1
T
V
CO2
DR
narrow micropore volume (cm3 gꢀ1
)
)
)
VNm2es
mesopore volume (cm3 gꢀ1
micropore volume (cm3 gꢀ1
)
koCARB
VtN
)
2
)
W=FoIPA 2-propanol space time (g s molꢀ1
)
KoIPA
X
conversion
)
Xexp
XIPA
Xsim
experimental conversion
2-propanol conversion
simulated conversion
KoPROPY preexponential factor of the propylene adsorption
(atmꢀ1
koPROPY preexponential factor of the propylene formation
(mol gꢀ1 sꢀ1
kPROPY
KPROPY
Lb
)
)
propylene formation rate constant (mol gꢀ1 sꢀ1
)
Greek letters
u
g
gext
hi
propylene adsorption constant (atmꢀ1
bed length (cm)
)
Thiele modulus
interphase internal effectiveness factor
interphase external effectiveness factor
fractional coverage of species i
n
reaction order
m
Pep
molecular weight (g molꢀ1
particle Peclet number
)
hV
fractional vacant acid sites
of bio-oil) to propylene. Acid-activated carbons were obtained by
chemical activation with H3PO4 of hemp residues in a single step,
without the need for incorporating acid surface groups in a com-
plementary stage by chemical oxidation of the carbon surface.
The effect of the activation conditions and the surface oxygen
groups on the catalytic dehydration of 2-propanol over the acid-
activated carbons was investigated. A kinetic study of 2-propanol
dehydration reaction on the carbon catalyst was carried out.
2.2. Catalyst characterization
The surface chemistry of the samples was analyzed by temper-
ature-programmed desorption (TPD), X-ray photoelectron spec-
troscopy (XPS) and Fourier transform infrared (FTIR)
spectroscopy. TPD profiles were obtained in a custom quartz tubu-
lar reactor placed inside an electrical furnace. The samples were
heated from room temperature up to 900 °C at a heating rate of
10 °C/min in a helium flow (200 cm3 STP/min). The amounts of
CO and CO2 desorbed from the samples were monitored with
non-dispersive infrared (NDIR) gas analyzers (Siemens ULTRAMAT
22), and the H2 concentration was measured by a mass spectrom-
eter (Pfeiffer Vacuum, OmniStar model). X-ray photoelectron spec-
troscopy (XPS) analyses of the samples were obtained using a
2. Experimental
2.1. Catalyst synthesis methods
Hemp stems supplied by Alsativa (Sociedad Cooperativa Agraria
Andaluza del Cáñamo, Pórtugos, Granada) were used as precursor
of the activated carbons. The preparation of the activated carbons
is reported elsewhere [19]. The precursor was impregnated with
85% (w/w) H3PO4 aqueous solution at room temperature and dried
for 24 h at 60 °C in a vacuum dryer. The impregnated hemp stems
were activated under continuous N2 flow (150 cm3 STP/min) in a
conventional tubular furnace. The activation temperature was
reached at a heating rate of 10 °C/min and maintained for 2 h. Dif-
ferent activation temperatures (350–550 °C) and impregnation ra-
tios (1,2), defined as H3PO4/precursor weight ratio, were studied.
The activated samples were cooled inside the furnace under N2
flow, then washed with distilled water at 60 °C until neutral pH
and negative phosphate analysis in the eluate [20] and dried at
110 °C. The activated carbons obtained are denoted by the letter
C followed by a number corresponding to the impregnation ratio
and by a second number representing the activation temperature
in degree Celsius.
5700C model Physical Electronics apparatus, with Mg Ka radiation
(1253.6 eV). For the analysis of the XPS peaks, the C1s peak posi-
tion was set at 284.5 eV and used as reference to locate the other
peaks. The fitting of the XPS peaks was done by least squares using
Gaussian–Lorentzian peak shapes. Fourier transform infrared
(FTIR) spectra were obtained using a Bruker Optics Tensor 27 FTIR
spectrometer by adding 256 scans in the 4000–400 cmꢀ1 spectral
range at 4 cmꢀ1 resolution. Pressed KBr pellets at a sample/KBr ra-
tio of around 1:250 were used.
The total acidity and acid strength distribution of the catalysts
were determined by temperature-programmed desorption of
ammonia. The NH3–TPD was performed using 80 mg of catalysts
saturated with pure NH3 at 100 °C. After saturation, the NH3
weakly adsorbed was desorbed in a He flow at the adsorption tem-
perature, until no NH3 was detected in the outlet gas. The TPD was
performed by raising the temperature up to 500 °C at a heating rate
of 10 °C/min. Outlet NH3 concentrations were measured by gas

J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
35
chromatography using a thermal conductivity detector. The type of
6
5
4
3
2
1
0
2.0
1.5
1.0
0.5
0.0
CO
CO2
H2
surface acidity (Brönsted or Lewis) was studied by adsorption and
desorption of pyridine (Py) and 2,6-dimetilpyridine (DMPy) carried
out in a thermogravimetric system (CI Electronics) at 100 °C. The
inlet partial pressure of the organic bases was 0.02 atm, and it
was established saturating He with the corresponding organic base
in a saturator at controlled temperature. After saturation of the
carbon, desorption is carried out at the adsorption temperature
in Helium flow. FTIR analysis of the carbons with adsorbed Py
and DMPy were performed.
(mmol/g) (mmol/g) (mmol/g)
1st cycle
2nd cycle
3rd cycle
3.96
1.18
1.04
0.66
0.07
0.06
1.00
0.39
0.39
H2
CO
CO2
CO2
1000
2.3. 2-Propanol dehydration
0
200
400
600
800
Temperature (ºC)
The decomposition of 2-propanol (IPA) was carried out at atmo-
spheric pressure in a quartz fixed bed microreactor (4 mm i.d.)
placed inside a vertical furnace with temperature control, using
Fig. 1. CO (black closed symbols), CO2 (gray closed symbols) and H2 (open symbols)
evolved with temperature during three consecutive TPDs (1st cycle: squares; 2nd
cycle: triangles; 3rd cycle: circles) performed to C1-450 carbon.
65 mg of catalyst (100–300 lm particle size) dispersed in 1 g of
SiC. Nitrogen was saturated with 2-propanol (Sigma–Aldrich,
99.5%, HPLC grade) vapor by contact in a saturator. To avoid the
condensation of 2-propanol or any reaction product, all the lines
from the saturator to the chromatograph were heated above
120 °C. The concentrations of the reactant and the products were
measured by online gas chromatography (Perkin–Elmer Autosys-
tem GC, 50 m HP-1 methyl silicone capillary column, flame ioniza-
tion detector) and by mass spectrometry (Pfeiffer Vacuum,
OmniStar model). Standard conditions for 2-propanol dehydration
on the catalysts were a 2-propanol partial pressure of 0.022 atm
CAOAP groups, respectively [19]. An alternative explanation for
the CO2 desorbed at high temperatures is that secondary reactions
between the CO and the oxygen surface groups take place [19,22].
The second and third TPDs showed a little but constant pres-
ence of CO desorbed at high temperatures (850 °C) with a small
peaks of CO2 and H2. The presence of CAOAPO3 groups might ex-
plain these results. The oxygen bonded to C and P evolves as CO
at high temperatures during the first TPD. The CO desorption gen-
erates active surface sites susceptible to be attacked probably by
the P and the hydroxyl group linked to phosphorus, producing a
CAPO2AOAC type surface groups and H2. This new group may pro-
duce, during the subsequent TPDs, a new desorption of CO at high
temperature. The CO2 evolution may be attributed to the second-
ary reactions previously commented. It is noteworthy the amount
of gases evolved during the second and the third TPDs were similar
suggesting a stabilization of the groups that desorbed at high tem-
peratures. The high amount of desorbed H2 may be a consequence
of both the aromatic condensation and the dehydrogenation of the
acid phosphates on the carbon surface.
and a space time of 0.043 g s/
2-propanol partial pressure varied from 0.012 to 0.081 atm and the
space time from 0.012 to 0.170 g s/ mol.
lmol. For the kinetic study, the inlet
l
The conversion was defined as the molar ratio of 2-propanol
converted to the 2-propanol supplied to the reactor. The selectivity
for each product was defined as the molar ratio of each product to
2-propanol converted.
3. Results and discussion
Infrared spectroscopy (IR) provides information about the
chemical structure of carbon catalysts. Fig. 2 shows the FTIR spec-
trum of the C1-450 carbon and of a sample obtained by carboniza-
tion of hemp cane at 450 °C denoted as C-450 [19]. The spectra of
both carbons show bands at around 1570 cmꢀ1, characteristic of
the aromatic ring stretching mode. The presence of these bands
suggests the existence of single or multiple aromatic rings in the
structure of the carbons. The presence of shoulders at 890, 820
and 760 cmꢀ1, due to CAH out-of-plane bend vibrations, supports
the aromatic character of the carbons. The presence of aliphatic
structures is suggested by the peaks at 2960, 2920 and
3.1. Preparation and characterization of the catalysts
The preparation and characterization of the activated carbons
are reported elsewhere [19]. Four different activated carbons were
prepared, namely, C1-450, C2-350, C2-450 and C2-550. The iso-
therms of the carbons are modified type I, characteristic of carbons
with a well-developed microporous structure and a significant
contribution of mesoporosity. The carbons obtained present BET
surface areas (ABET) between 700 and almost 1500 m2/g and show
a significant amount of surface phosphorus, as a consequence of
the activation procedure [14,19], which increases with the activa-
tion temperature and decreases with the impregnation ratio.
Carbon C1-450, which shows the highest phosphorus surface
concentration [19], was submitted to successive TPD. Fig. 1 shows
the evolution with temperature of CO, CO2 and H2 during three
successive TPDs. A table with the total amount of CO, CO2 and H2
evolved for each TPD has been included in the figure. The CO pro-
file of the first TPD is mainly characterized by a considerable
amount of CO desorbed at high temperatures (T > 700 °C). This
behavior, reported in the literature [21], could be associated to
the decomposition of phosphorus groups of high thermal stability
formed during the activation process. The amount of carboxylic,
lactonic and anhydride groups, which decompose as CO2 during
the thermal treatment, is much lower, as indicates the lower
amount of CO2 than that of CO desorbed. It is noticeable the pres-
ence of high stable groups that decompose as CO2 at temperatures
higher than 700 °C. The evolution of CO2 and CO at high tempera-
tures might be due to the decomposition of stable O@CAOAP and
C1-450
C450
4000
3500
3000
2500
2000
1500
1000
500
0
-1
Wavenumber (cm )
Fig. 2. FTIR spectrum of C1-450 carbon catalyst and C-450 carbon.

36
J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
2850 cmꢀ1 characteristic of aliphatic CAH stretching vibration, at
2960 cmꢀ1 due to methyl CAH asymmetric stretch and at 2920
and 2850 cmꢀ1 as a consequence of methylene CAH asymmetric/
symmetric stretch [23].
Chemisorption of basic molecules such as pyridine (Py) and 2,6-
dimethylpyridine (DMPy) is often used to probe the acidity of sol-
ids [34–37]. Pyridine interacts with acidic sites because it has a
lone electron pair at the nitrogen atom available for donation to
a Lewis acidic site and because it can accept a proton from Brön-
sted sites. In contrast, 2,6-dimethylpyridine is selectively adsorbed
on Brönsted acid sites because of its higher basicity and steric hin-
drance of the methyl groups [38,39]. The difference between the
total amount of adsorbed pyridine and 2,6-dimethylpyridine corre-
sponds to the amount of Lewis acid sites. Fig. 4 represents the
adsorption and desorption kinetics of Py and DMPy on C1-450 car-
bon at 100 °C. The amounts of Py and DMPy retained after the
desorption process (irreversible chemisorbed Py and DMPy) are
very similar, suggesting that most of the acid sites of the carbon
are Brönsted sites. Therefore, the chemical activation of biomasic
residues with H3PO4 yields carbons with a predominantly Brönsted
acidity. The amount chemisorbed of both organic bases is very sim-
ilar to the amount of ammonia desorbed during the TPD (Fig. 3).
Fig. 5 represents the room temperature FTIR spectra of C1-450
activated carbon with adsorbed Py and DMPy. Interaction of pyri-
dine, via the nitrogen lone-pair electrons, with aprotic (Lewis) and
protonic (Brönsted) acid sites can be analyzed by following the ring
vibration modes 8a and 19b, respectively [40]. The interaction of
Py with Lewis type acid sites produces 8a vibration mode usually
observed at around 1624 cmꢀ1 and 19b vibration mode at around
1452 cmꢀ1. The former was not observed, and the latter appeared
in form of a low intensity shoulder suggesting a low presence of Le-
wis acid sites on the C1-450 carbon surface. The shoulder at
1574 cmꢀ1 is characteristic of weakly adsorbed (physisorbed) pyr-
idine. The pyridine interaction with Brönsted acid sites is observed
at 1587 cmꢀ1 for 8a vibration mode and at 1442 cmꢀ1 for 19b
vibration mode. Both peaks are clearly observed in the Py-adsorbed
spectra, supporting the results previously obtained in the TG anal-
ysis that the acid sites of the carbon surface are predominantly of
Brönsted type. The band at 1483 cmꢀ1 could be associated to 19a
vibration of pyridine ions. Further characterization of Brönsted
acidity was performed by FTIR spectroscopy of adsorbed DMPy.
This probe molecule is more sensitive to Brönsted acidity than it
is pyridine. The FTIR spectrum of C1-450 carbon with DMPy ad-
sorbed shows four adsorption bands at 1605, 1589, 1477 and
1464 cmꢀ1. The last two bands can be assigned to weakly adsorbed
(physisorbed) DMPy [35,41,42]. The bands at 1605 and 1589 cmꢀ1
were attributed to DMPy molecules hydrogen-bonded to carbon
surface [39,43,44], indicating the presence of Brönsted type acid
sites on the carbon surface.
The presence of phenol groups is supported by the peak at
1400 cmꢀ1, characteristic of OAH bending vibrations [24]. The
peak at around 1700 cmꢀ1 is characteristic of C@O absorption in
carboxylic acids. The low intensity of these peaks indicates a lower
amount of phenol and carboxylic groups compared with other oxy-
gen groups of the carbon catalyst. The spectrum of C1-450 carbon
displays a wide band between 1300 and 900 cmꢀ1 with the maxi-
mum at 1160 cmꢀ1. A shoulder is observed at 1085 cmꢀ1. The
adsorption in this region is generally found in oxidized carbons
and is attributed to CAO stretching vibrations in acids, alcohols,
phenol, ethers and esters [25,26].
It is interesting to point out that C1-450 carbon, with a signifi-
cant P surface content [19], shows clearly a broad band between
1300 and 900 cmꢀ1, almost not observed in the spectrum of C-
450 carbon that contains a negligible amount of surface P [19]. This
broad band between 1300 and 900 cmꢀ1 is also assigned to several
types of phosphorus groups formed during the chemical activation
of carbons with phosphoric acid [23,27,28]. The strong intensity
peak at 1160 cmꢀ1 is characteristic of the stretching vibrations of
hydrogen-bonded P@O groups [29–31] from phosphates and poly-
phosphates of the OAC stretching vibrations in the PAOAC (aro-
matic) linkage [28,29] and of O@PAOH groups [28]. This
supports the results previously discussed in the XPS analysis. The
shoulder at 1085 cmꢀ1 is assigned to PAOꢀ in acid phosphate es-
ters and to the symmetrical vibration in polyphosphate chain
PAOAP [32,33]. The presence of phosphate groups is supported
by the peak at 620 cmꢀ1 and by the shoulder at 985 cmꢀ1, both
characteristic of stretching vibrations of phosphate groups [28].
The shoulder at 985 cmꢀ1 is also associated to the PAOAC stretch-
ing vibration for pentavalent phosphorus [28]. The presence of CAP
groups (795–650 cmꢀ1) [31] is supported by the peaks located at
760 and 680 cmꢀ1. The spectrum does not show bands’ character-
istic of PAH (2500–2225 cmꢀ1) [31], which rule out the presence of
phosphine derivates.
Fig. 3 shows the NH3–TPD profiles for different carbon catalysts
after adsorption and desorption of NH3 at 100 °C. The desorption
temperature indicates the acid strength of the sites, with weaker
sites desorbing at lower temperatures. The C1-450 carbon shows
the highest amount of NH3 desorbed, confirming the higher acidity
of this carbon. The total acidity, expressed as the total amount of
NH3 desorbed during the TPD, is reported in Fig. 3. The amount
of desorbed NH3 increases with the amount of surface phosphorus
[19] revealed by XPS analyses, suggesting that the acidity of the
carbons is related to the phosphorus retained on the carbon sur-
face, probably associated to the OH groups in phosphates.
In order to analyze the role of the surface oxygenated groups in
the acidity of the carbons and in the alcohol decomposition reac-
tion, the C1-450 carbon was submitted to thermal treatments in
Py
ads.
3.0
0.25
C1-450
Amount of
NH3 desorbed
(mmol/g)
2.5
DMPy desorption
in He flow
Carbon
0.20
2.0
C1-450
C2-350
C2-450
C2-550
1.85
0.82
1.02
1.33
0.15
0.10
0.05
0.00
Py desorption
1.5
in He flow
DMPy
1.0
adsorption
0.5
0.0
0
100
200
300
400
500
100
200
300
Temperature (ºC)
400
500
Time (min)
Fig. 4. Adsorption and desorption kinetics of pyridine (Py) and 2,6-dimethylpyr-
idine (DMPy) on C1-450 carbon at 100 °C.
Fig. 3. NH3–TPD profiles for the carbon catalysts.

J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
37
bonded to a carbon site throughout an O atom (CAOAPO3 and/or
(CAO)3PO)) [21], the binding energy of about 133.2 eV is character-
istic of CAP bonding as in CAPO3 and/or C2PO2 [21], and the value
132.0 eV of the main peak of the doublet is associated to C3PO
groups [45]. Finally, a low intensity doublet peak with the main
peak centered at around 131.0 eV is ascribed to C3P groups [45].
The thermal treatment results in a reduction in the binding energy
of the maximum of the peaks and a displacement to lower binding
energies of the entire signal, being this displacement greater at the
highest thermal treatment temperature analyzed. This result sug-
gests that the thermal treatment decreases the proportion of oxy-
gen in the surface phosphorus groups, with a reduction in the most
oxygenated phosphorus groups ((CAOAPO3 and/or (CAO)3PO))
and an increase in the C3PO and C3P groups.
1442
1587
1589
1574
1452
1483
Py
1624
1605
1464
1477
DMPy
1700
1650
1600
1550
1500
1450
1400
Wavenumber (cm-1)
Fig. 5. FTIR spectra of C1-450 carbon catalyst after adsorption of pyridine (Py) and
2,6-dimethyl pyridine (DMPy).
3.2. 2-Propanol catalytic dehydration
Table 1
The catalytic decomposition of 2-propanol has been studied on
different carbon catalysts in order to analyze the effect of phospho-
rus and oxygen surface groups. Also, a kinetic study has been car-
ried out. To verify that axial dispersion problems are negligible, the
following criterion is often used [46–48]:
Mass surface concentration (%) determined by XPS quantitative analysis and total
amount of NH3 desorbed in the NH3–TPD for C1-450-TT700 and C1-450-TT900.
XPS
C 1s
NH3–TPD
NH3 desorbed (mmol/g)
O 1s
P 2p
C1-450-TT700
C1-450-TT900
63.2
66.4
23.8
19.1
13.0
14.0
n.m.
1.23
ꢀ
ꢁ
Lb 20 ꢁ n
1
>
ꢁ Ln
ð1Þ
n.m. = not measured.
dp
Pep
1 ꢀ X
where Lb is the bed length, n is the reaction order, X is the conver-
sion, dp is the particle diameter and Pep is the particle Peclet num-
ber. For large particle Reynolds number (Rep) values, Pep is about 2,
whereas in the conditions used for laboratory reactors Rep is often
low and therefore Pep is lower than 2. For Rep < 1, a value for Pep
of 0.5 has been suggested [49]. It has also been stated that criterion
(1) is too conservative and that the factor 20 can be replaced by 8
[49]. In the conditions used in this study, the Rep number was
0.25 and so a value of 0.50 was chosen for the Pep number. The va-
lue of Lb/dp was 100.0 and the values of the right-hand side of the
Eq. (1) were 92.1 and 36.8 when using a factor of 20 and 8, respec-
tively. Therefore, the effects of axial dispersion could be neglected.
The absence of external and internal mass transfer limitations
was checked theoretically. The effect of the external mass transfer
limitations was evaluated using the Damköler number, Da, i.e. the
ratio of the chemical reaction rate to the mass transfer rate. An
inert atmosphere up to 700 and 900 °C. The carbons obtained are
denoted as C1-450-TT700 and C1-450-TT900, respectively. Table 1
summarizes the mass surface concentration of C, O and P deter-
mined by XPS quantitative analysis of these carbons. The thermal
treatment increases progressively the amount of surface phospho-
rus from 11%, for the original carbon [19], up to 14%, for the carbon
treated at 900 °C. This might be due to the loss of oxygen surface
groups during the thermal treatment that decreases the surface
oxygen concentration and increases the carbon and phosphorus
content at the external surface of the carbon. The thermal treat-
ment also produces a significant decrease in the surface acidity,
as reveals the lower amount of NH3 desorbed for the C1-450-
TT900 carbon (1.23 mmol/g) than that observed for the carbon
C1-450 (1.85 mmol/g). Fig. 6 represents the P 2p zone of the car-
bons C1-450, C1-450-TT700 and C1-450-TT900. The phosphorus
spectra were deconvoluted using two doublet peaks with an area
ratio of 0.5 and a separation between peaks of 0.84 eV [45]. For
the fitting, the position of the center of the peaks was allowed to
vary 0.2 eV, while the full width at half-maximum (FWHM) was
varied between 0.80 and 1.5 eV. The P 2p zone is the result of
the contribution of different phosphorus states. A main peak of
doublet at a value around 134.2 eV can be assigned to the P groups
interphase external effectiveness factor, gext, was used to evaluate
the influence of external mass transport on the global reaction rate.
A value of 0.955 for gext and a value of 4.66 ꢂ 10ꢀ2 for Da number
were obtained, respectively. These values confirmed the negligible
external mass transfer effects. The absence of internal mass trans-
fer limitation was evaluated using the isothermal intraphase inter-
nal effectiveness factor,
g, which is a function of the Thiele
modulus, u. A value of 0.255 was obtained for the Thiele modulus
and a value of 0.996 was obtained for the intraphase effectiveness
factor. Therefore, internal mass diffusion effects were also negligi-
ble in the experimental conditions studied in this work.
(C-O)3PO
C-O-PO3
C2PO2
C-PO3
C3PO
The 2-propanol steady-state conversion as a function of the
reaction temperature for different carbon catalysts is represented
in Fig. 7. Conversion increases with the reaction temperature.
There is no relation between the apparent surface area [19] and
the 2-propanol steady-state conversions for the different carbon
catalysts. The carbon C1-450, which shows the lowest surface area,
displays the highest steady-state conversion, probably because of
its higher acidity as measured with NH3–TPD (Fig. 3 and Table 1).
The increase in the activation temperature leads to carbons with
higher conversions (C2-550 > C2-450 > C2-350). This suggests that
the increase in the amount of phosphorus with the activation tem-
perature produces carbons with higher surface acidity and result-
ing in higher 2-propanol steady-state conversions. On the
C3P
C1-450-TT900
C1-450-TT700
C1-450
138
137
136
135
134
133
132
131
130
Binding energy (eV)
Fig. 6. XPS P 2p zone of carbons C1-450, C1-450-TT700 and C1-450-TT900.

38
J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
propylene (Sp) and to di-isopropyl ether (Sdii) with the reaction
temperature for carbon C1-450 at an inlet 2-propanol partial pres-
sure of 0.022 atm and a space time of 0.043 g s/ mol. The main
reaction product is propylene with lower amounts of di-isopropyl
ether. Only at low temperatures and low conversion values the
selectivity to the ether is significant. An increase in the reaction
temperature results in higher olefin selectivity. At temperatures
higher than 175 °C, the 2-propanol dehydration on C1-450 is
100% selective to propylene. The other carbon catalysts show a
similar behavior.
Fig. 9 represents the 2-propanol conversion and selectivities to
propylene and di-isopropyl ether with the reaction time at 170 °C
for C1-450 carbon. The 2-propanol conversion does not decrease
with the reaction time, but increases slightly from around 0.5 up
to 0.6. The selectivities to both propylene and di-isopropyl ether
does not change with the reaction time, suggesting that the cata-
lyst obtained does not suffer deactivation during the reaction, at
the conditions analyzed in this work.
1
0.8
0.6
0.4
0.2
0
l
C1-450
C2-350
C2-450
C2-550
C1-450-TT700
C1-450-TT900
100
150
200
250
300
350
400
Temperature (ºC)
Fig. 7. Conversion of 2-propanol as a function of temperature for the C1-450
catalyst ðPoIPA ¼ 0:022 atm; W=FIoPA ¼ 0:043 g s=
lmolÞ.
contrary, the thermal treatments performed to carbon C1-450 after
the activation process results in a decrease in the carbon activity,
being this decrease more pronounced at higher thermal treatment
temperatures. The thermal treatment reduces the amount of sur-
face oxygen and the surface acidity, and so, decreases the activity
of the carbons for 2-propanol decomposition. The sample obtained
by carbonization at 450 °C without phosphoric acid, C-450, shows
a 2-propanol conversion lower than 10% at 350 °C (not repre-
sented), much lower than those observed for the carbons activated
with phosphoric acid, as a consequence of a much lower surface
acidity. The lower activity of this char may also be associated to
its lower surface area (256 m2/g). However, we have previously re-
ported [7] that a commercial activated carbon with a similar sur-
face area, but significantly lower surface acidity needs much
higher temperatures for reaching similar 2-propanol conversion
than carbons activated with phosphoric acid. Furthermore, acti-
vated carbons prepared by chemical activation with phosphoric
acid at different conditions that presented higher surface areas,
but retained lower amount of surface phosphorus (lower acidity),
showed lower activity for alcohol dehydration than those contain-
ing higher amount of surface phosphorus groups. This fact and the
fact that carbonization of biomass does not produce carbons with
relevant acidity suggest that the lower activity of C-450 char is
mainly due to its lower surface acidity, beside its lower surface
area.
Fig. 10 shows the effect of the inlet water vapor partial pressure
in the conversion values for the C1-450 carbon at 155 °C. The var-
iation of the water vapor partial pressure showed little effect on
the 2-propanol conversion and on the product distribution, which
probably indicates a negligible adsorption of water vapor on the
active sites during 2-propanol dehydration reaction. In this sense,
adsorption of water vapor was not considered in the development
of the kinetic model.
Fig. 11 represents 2-propanol conversions at different partial
pressures
of
2-propanol
and
constant
space
time
ðW=FoIPA ¼ 0:043 g s=
lmolÞ, at reaction temperatures of 125, 140,
1
1
X
Sp
0.8
0.6
0.4
0.2
0
0.8
Sdii
0.6
0.4
0.2
0
The 2-propanol decomposition yields only dehydration prod-
ucts, propylene and di-isopropyl ether, for all the catalysts studied.
Fig. 8 represents the evolution of the conversion and selectivity to
0
100
200
300
40 0
500
Time (min)
Fig. 9. Evolution of the 2-propanol conversion and selectivity to propylene (Sp) and
di-isopropyl ether (Sdii) with the reaction time at 170 °C on the C1-450 catalysts
ðPoIPA ¼ 0:022 atm; W=FIoPA ¼ 0:043 g s=
lmolÞ.
1
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
0.8
0.6
0.4
0.2
0
X
Sp
Sdii
100
150
200
250
300
0.00
0.02
0.04
0.06
0.08
0.10
Temperature (ºC)
Water vapor pressure (atm)
Fig. 8. Conversion of 2-propanol and selectivity to propylene (Sp) and to di-
Fig. 10. Evolution of the 2-propanol conversion with the inlet water vapor partial
pressure for the C1-450 catalysts at 155 °C ðPoIPA ¼ 0:022 atm; W=FoIPA ¼ 0:043 g s=
isopropyl ether (Sdii) as
a function temperature for the C1-450 catalyst
ðPoIPA ¼ 0:022 atm; W=FIoPA ¼ 0:043 g s=
lmolÞ.
lmolÞ.

J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
39
1
0.8
0.6
0.4
0.2
0
3.2.1. E1 model
125 ºC
140 ºC
155 ºC
175 ºC
E1 mechanism involves the initial protonation of the OH group
in the molecule of 2-propanol, followed by the cleavage of the CO
bond producing a carbocation (CARB) and a water molecule as
leaving group. This step is unimolecular and limits the rate of the
reaction. The second step is a fast deprotonation of the carbocation
by the loss of one b hydrogen and the formation of the double
bond, C@C, and thus the propylene molecule (PROPY). The elemen-
tary steps for the proposed model are shown in the following equa-
tions, in which ꢃ represents an acid active site:
2-Propanol adsorption:
0
0.0 2
0.0 4
0.0 6
0.0 8
0.1
IPAðgÞ þ ꢃ K$IPA IPAꢃ
ð3Þ
PoIPA (atm)
Surface reaction:
Fig. 11. Evolution of 2-propanol conversion with inlet partial pressure of 2-
propanol
at
different
temperatures
and
constant
space
time
kCARB
IPAꢃ ! CARBꢃ þ H2OðgÞ
ð4Þ
ð5Þ
ðW=FoIPA ¼ 0:043 g s=
lmolÞ.
kPROPY
CARBꢃ ! PROPYꢃ
1
Propylene desorption:
125 ºC
140 ºC
155 ºC
175 ºC
1=KPROPY
PROPYꢃ
$
þ ꢃ
ð6Þ
0.8
0.6
0.4
0.2
0
The adsorption of 2-propanol is assumed to take place molecu-
larly and in quasi-equilibrium between the adsorbed and vapor-
phase 2-propanol. The formation of the propylene molecule is sup-
posed to occur in two sequential elementary steps. In the first step,
the formation of the adsorbed carbocation and the gas-phase water
molecule is assumed to be the rate-determining step (rds),
whereas in the second step, the formation of adsorbed propylene
molecule takes place. In the last step, a fast desorption of propyl-
ene molecules occurs. With these considerations, the global reac-
tion rate is
0
0.05
0.1
0.15
0.2
W/Fo
(g·s/µmol)
IPA
Fig. 12. Evolution of 2-propanol conversion with space time at different temper-
atures and constant 2-propanol inlet partial pressure ðPoIPA ¼ 0:022 atmÞ.
ꢀrIPA ¼ rCARB ¼ kCARB ꢁ hIPA
ð7Þ
where kCARB is the forward rate constant of the reaction (4) and hIPA
is the fractional coverage of 2-propanol. From the consideration of
quasi-equilibrium for adsorption of 2-propanol and desorption of
propylene it can be obtained
155 and 170 °C for the C1-450 carbon. At PoIPA < 0:02, the conver-
sion increases with the 2-propanol partial pressure. This increase
is higher at high temperatures. At partial pressures of 2-propanol
higher than 0.02 atm, the conversion remains almost constant.
Fig. 12 depicts the evolution of the conversion of 2-propanol with
hIPA ¼ KIPA ꢁ PIPA ꢁ hV
hPROPY ¼ KPROPY ꢁ PPROPY ꢁ hV
ð8Þ
ð9Þ
the space time (up to values of around 0.170 g s/lmol) at a 2-pro-
where hV and hPROPY are the fraction of vacant acid sites and propyl-
ene occupied sites, respectively, KIPA and KPROPY are the adsorption
constants and PIPA and PPROPY are the partial pressures of 2-propanol
and propylene, respectively. Assuming a pseudo-steady state for the
adsorbed carbocation formation, the next expression is obtained:
panol partial pressure of 0.022 atm and temperatures between 125
and 170 °C. Alcohol conversion increases with the space time at all
the temperatures studied.
Integral reactor behavior was used for the interpretation of the
experimental data. To this purpose, the reactor continuity, Eq. (2),
was integrated numerically to calculate the exit conversion of 2-
propanol dehydration reaction.
dhCARB
dt
kCARB ꢁ hIPA
¼ kCARB ꢁ hIPA ꢀ kPROPY ꢁ hCARB ꢄ 0 ) hCARB
¼
ð10Þ
ð11Þ
ð12Þ
kPROPY
The site balance is shown in the following equation,
dXIPA
ꢂ
ꢃ
¼ rIPA
ð2Þ
W
1 ¼ hV þ hIPA þ hCARB þ hPROPY
d
FoIPA
and rearranging the Eqs. (8)–(11) yields,
1
In Eq. (2), rIPA represents the conversion rate of 2-propanol, XIPA is
ꢄ
ꢅ
the total conversion of 2-propanol and W=FoIPA is the 2-propanol
space time. The following suppositions were assumed: homoge-
neous distribution of active sites on the catalyst surface, catalyst
is assumed to operate at steady-state conditions, diffusional con-
straints and transport limitations are negligible (as proven theoret-
ically) and changes in temperature and pressure within the reactor
were neglected. As the total conversion to di-isopropyl ether is al-
ways lower than 1%, the intermolecular dehydration to the ether
was not considered in the kinetic model resolution.
hV
¼
kCARBꢁKIPAꢁPIPA
1 þ KIPA ꢁ PIPA
þ
þ KPROPY ꢁ PPROPY
kPROPY
and the final rate expression results in
kCARB ꢁ KIPA ꢁ PIPA
ꢀrIPA ¼ rCARB
¼
ð13Þ
kCARBꢁKIPAꢁPIPA
1 þ KIPA ꢁ PIPA
þ
þ KPROPY ꢁ PPROPY
kPROPY
3.2.2. E2 model
Two Langmuir–Hinshellwood mechanisms were considered,
assuming that the surface reaction was an E1 (two-step) or an E2
(one-step) elimination mechanism.
This model assumes a Langmuir–Hinshelwood mechanism in
which the surface reaction is an E2 elimination in a single step.
In this case, both the leaving group, OH and the b proton depart

40
J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
simultaneously in a single concerted step. The elementary steps for
this mechanism are shown below.
2-Propanol adsorption:
Both models fit properly the experimental data with low error
values, although slightly lower values were obtained for E1 model
(error = 0.031) than for E2 model (error = 0.037). This could be due
to that both mechanisms are similar and, probably, both mecha-
nisms occur simultaneously over the carbon surface. The result of
the model E1 application to the experimental data is shown in
Figs. 11 and 12 as solid lines (symbols are the experimental val-
ues). Fig. 13 represents the simulated conversion, Xsim, assuming
the E1 model, versus the conversion obtained experimentally, Xexp
for dehydration of 2-propanol on C1-450 catalysts. The result
shows that E1 model represents properly the experimental data.
The values of the parameters obtained by numerical simulation
to both models, E1 and E2, are summarized in Table 2. The activa-
tion energy for the rate-determining step of both models is very
similar, nearly 98 kJ/mol. This fact indicates that formation of the
carbocation through and E1 mechanism and the formation of the
propylene by an E2 mechanism occur at a very similar rate, and,
therefore, it is probably that both mechanisms take place simulta-
neously. The literature shows a broad spectrum of apparent activa-
tion energy values for this reaction, ranging from 12 to 133 kJ/mol
[6, 52-54]. This large variation in the activation energy values
could be due to the different type of interaction between 2-propa-
nol molecules and the catalyst surfaces. Also, diffusional limita-
tions may reduce the value of the apparent activation energy.
In order to corroborate the validity of the proposed reaction
models, some requirements have to be fulfilled [55]. The enthalpies
of adsorption must be negative, as adsorption is always exother-
IPAðgÞ þ ꢃ K$IPA IPAꢃ
ð14Þ
ð15Þ
ð16Þ
Surface reaction:
kPROPY
IPAꢃ ! PROPYꢃ þ H2OðgÞ
Propylene desorption:
1=KPROPY
$
PROPYꢃ
PROPYðgÞ þ ꢃ
In this model, as for E1 model, the 2-propanol adsorption is as-
sumed to take place molecularly and in quasi-equilibrium between
the adsorbed and vapor-phase 2-propanol. The formation of the
propylene molecule is supposed to occur in a single concerted step.
Again, desorption of the molecules of propylene is assumed to be
faster than propylene adsorbed formation. Considering the forma-
tion of adsorbed propylene molecule like the rate-determining
step, the global reaction rate can be written as
ꢀrIPA ¼ rPROPY ¼ kPROPY ꢁ hIPA
ð17Þ
where kPROPY is the forward rate constant of the reaction (15) and
hIPA is the fractional coverage of 2-propanol. The expressions for
adsorption of 2-propanol and desorption of propylene are Eqs. (8)
and (9), respectively. In this model, only molecules of 2-propanol
and propylene could adsorb on the active sites of the carbon sur-
face, and so, the site balance is
mic. This condition is satisfied for both models, since
D
D
HIPA and
HPROPY are negative. The standard entropy of adsorption,
S0ad
S0ad must be negative
D
,
must fulfill two additional requirements: (i)
D
1 ¼ hV þ hIPA þ hPROPY
ð18Þ
and (ii) D
S0ad must have an absolute value smaller than S0g, the stan-
dard entropy in gas phase [56,57]. The values of S0g for 2-propanol
and propylene are 181.1 and 266.8 J/mol K, respectively [58]. The
standard enthalpy and entropy in gas phase are related by the
expression:
with the Eqs. (8) and (9) leads to
1
hV
¼
ð19Þ
1 þ KIPA ꢁ PIPA þ KPROPY ꢁ PPROPY
and the final rate expression results in
ꢀ
D
R ꢁ T
H0ad
D
S0ad
R
LnðKÞ ¼
þ
ð26Þ
kPROPY ꢁ KIPA ꢁ PIPA
ꢀrIPA ¼ rPROPY
¼
ð20Þ
1 þ KIPA ꢁ PIPA þ KPROPY ꢁ PPROPY
The values of the standard entropy of adsorption could be obtained
For both models, the kinetic parameters were estimated by a
Runge–Kutta method combined with an optimization routine
based in the Levenberg–Marquardt [50,51] algorithm, minimizing
the error function:
taking into account that at 1/T = 0, Ln(K0i) = S/R. The values of
D
D
S0ad
for 2-propanol and propylene for model E1 are ꢀ135.5 and ꢀ38.3 J/
mol K, respectively, and these values for model E2 are ꢀ128.8 and
ꢀ29.5 J/mol K, respectively. Both models yield standard adsorption
entropies that fulfill the aforementioned requirements.
X
ꢄ
ꢅ
2
error ¼
Xexp ;i ꢀ Xsim;i
ð21Þ
Fig. 14 represents the FTIR spectrum obtained at room temper-
i
ature of C1-450 activated carbon after reaction for 2 h at 150 °C
where Xexp is the value of the steady-state conversion obtained
experimentally and Xsim is the simulated value. The dependence
of the kinetic parameters with the temperature was considered to
follow an Arrhenius law for the kinetic constant (Eqs. (23) and
(24)) or a Van’t Hoff law for the adsorption constants (Eqs. (22)
and (25)).
with 2-propanol ðPoIPA ¼ 0:022 atm; W=FoIPA ¼ 0:043 g s=
molÞ. The
l
spectrum of the carbon that was not subject to reaction has been
1
0.8
0.6
0.4
0.2
0
ꢀ
ꢁ
ꢀ
DHIPA
KIPA ¼ KoIPA ꢁ exp
ð22Þ
ð23Þ
R ꢁ T
ꢀ
ꢁ
ꢀEaCARB
kCARB ¼ koCARB ꢁ exp
R ꢁ T
ꢀ
ꢁ
ꢀEaPROPY
R ꢁ T
kPROPY ¼ koPROPY ꢁ exp
KPROPY ¼ KoPROPY ꢁ exp
ð24Þ
ꢀ
ꢁ
ꢀ
DHPROPY
ð25Þ
R ꢁ T
0
0.2
0.4
0.6
0.8
1
where koi, Koi are the preexponential factors,
D
HIPA and
DHPROPY are
Xexp
the enthalpies of adsorption of 2-propanol and propylene, respec-
tively, and EaCARB and EaPROPY are the activation energies for the
reactions represented in Eqs. (4) and (5) and (15), respectively.
Fig. 13. Conversion simulated by E1 model versus conversion obtained
experimentally.

J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
41
Table 2
Optimized kinetic parameters for mechanisms E1 and E2.
Model
KIPA
kCARB
kPROPY
KPROPY
KoIPA (atmꢀ1
)
D
HIPA (kJ/mol)
koCARB (mol/g s)
EaCARB (kJ/mol)
koPROPY (mol/g s)
EaPROPY (kJ/mol)
KoPROPY (atmꢀ1
)
DHPROPY (kJ/mol)
E1
E2
8.3 ꢂ 10ꢀ8
ꢀ78.3
ꢀ75.6
8.7 ꢂ 10+6
98.5
–
1.0 ꢂ 10+7
65.0
98.3
1.0 ꢂ 10ꢀ2
ꢀ35.8
ꢀ33.4
1.8 ꢂ 10ꢀ7
–
8.9 ꢂ 10+6
2.9 ꢂ 10ꢀ2
reaction with 2-propanol suggests the coexistence of at least two
adsorbed species, one undissociated as 2-propanol molecules and
the other one dissociated forming isopropoxide or carbocation spe-
cies [59–61]. These results, although not conclusive, may support
that both mechanisms take place simultaneously during dehydra-
tion reaction, an E1 mechanism in which the reaction goes through
intermediate isopropoxide–carbocation species and the E2 mecha-
nism in which the propylene is formed from 2-propanol adsorbed
molecules.
2920
2850
1085
1160
Fig. 15 outlines the surface reaction mechanisms proposed, E1
and E2. The 2-propanol molecule is adsorbed on the OH group of
the phosphate, probably by hydrogen bonding. The adsorbed 2-
propanol could coexist with the protonated species, as shown by
FTIR analysis, formed by the transfer of a proton from the Brönsted
acid site that yields isopropoxide–carbocation ions in equilibrium,
which could be considered as only one species in the reaction
mechanism. Both species produce the olefin by surface elimination
mechanisms. Via E1 mechanism, the adsorbed carbocation is
formed, and subsequently a b hydrogen departs and the propylene
is obtained. Via E2, the adsorbed olefin is produced in a single con-
certed step, with the loss of the water molecule and the deprotona-
tion in b happen simultaneously.
4000 3500 3000 2500 2000 1500 1000
500
0
-1
Wavenumber (cm )
Fig. 14. FTIR spectrum of C1-450 carbon catalyst after 2-propanol adsorption.
subtracted. It should be noted that it can be difficult to discern be-
tween surface isopropoxide species and molecularly adsorbed iso-
propanol without other additional characterization techniques.
The reaction of 2-propanol with the carbon surface produces an in-
crease in the two bands clearly observed at 2920 and 2850 cmꢀ1
probably as a consequence of the CAH stretching vibrations of
undissociated 2-propanol molecules. The absence of a defined
band at around 3663 cmꢀ1 suggested a weak interaction between
the adsorbed 2-propanol and the carbon surface [55]. The peaks
at 1160 and 1085 cmꢀ1 are assigned to the vibration of P@O groups
and of PAOꢀ, respectively. Both peaks may be increased by the
adsorption of cationic species (isopropoxide or carbocation) on
the carbon surface. The FTIR spectrum of the carbon C1-450 after
4. Conclusions
The chemical activation of hemp residues with H3PO4 results in
carbons with a high surface acidity because of the residual phos-
Fig. 15. Scheme of the proposed surface reaction mechanisms from a 2-propanol molecule adsorbed on a phosphate group.

42
J. Bedia et al. / Journal of Catalysis 271 (2010) 33–42
[13] J. Rodríguez-Mirasol, J. Bedia, T. Cordero, J.J. Rodríguez, Sep. Sci. Technol. 40
(2005) 3113.
[14] J.M. Rosas, J. Bedia, J. Rodríguez-Mirasol, T. Cordero, Fuel 88 (2009) 19.
[15] C.A. Leon y Leon, L.R. Radovic, in: P.A. Thrower (Ed.), Chemistry and Physics of
Carbon, M. Dekker, New York, 1992, pp. 213–310.
[16] G.S. Szyman´ ski, G. Rychlicki, Carbon 31 (1993) 247.
[17] C. Moreno-Castilla, F. Carrasco-Marín, C. Parejo-Pérez, M.V. López Ramón,
Carbon 39 (2001) 869.
[18] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Órfao, Carbon 37 (1999)
1379.
phorus groups with a high thermal stability that remain in the car-
bon surface.
The TPD–CO profiles are mainly characterized by a considerable
amount of CO desorbed at high temperatures (T > 700 °C) associ-
ated to the decomposition of the CAOAPO3 and CAPO3 groups
formed during the activation process. The surface acidity of the
carbons, determined by NH3–TPD, seems to be related to the phos-
phorus retained on the carbon surface. Irreversible adsorption of
organic bases, such as pyridine and 2,6-dimethyl pyridine, suggests
the presence of Brönsted acid sites.
[19] J.M. Rosas, J. Bedia, J. Rodríguez-Mirasol, T. Cordero, Ind. Eng. Chem. Res. 47
(2008) 1288.
[20] E. González-Serrano, T. Cordero, J. Rodríguez-Mirasol, L. Cotoruelo, J.J.
Rodríguez, Water Res. 38 (2004) 3043.
Conversion of 2-propanol yields only dehydration products and
is highly selective to propylene, with very low amounts of di-iso-
propyl ether observed only at low conversion values. Kinetic inter-
pretation of the experimental data was performed using two
elimination mechanisms; an E1 mechanism (two-step mechanism)
and an E2 mechanism (one-step mechanism). The rate expressions
derived from both models represent properly the experimental re-
sults, suggesting that probably the two mechanisms occur simulta-
neously. This supposition is supported by the similar rate constant
obtained for the rate-determining step of both models, formation
of the carbocation in E1 mechanism and formation of the propyl-
ene in the E2 mechanism. The activation energy of the rds of both
models is very similar, nearly 98 kJ/mol. This fact indicates that
formation of the carbocation through and E1 mechanism and for-
mation of the propylene by an E2 mechanism occur at a very sim-
ilar rate, and, therefore, it is probably that both mechanisms take
place simultaneously. Furthermore, the FTIR spectrum of the car-
bon C1-450 after reaction with 2-propanol suggests the coexis-
tence of at least two adsorbed species, one undissociated as 2-
propanol molecules and the other one dissociated forming isoprop-
oxide–carbocation species.
[21] X. Wu, L.R. Radovic, Carbon 44 (2006) 151.
[22] P.J. Hall, J.M. Calo, Energy Fuels 3 (1989) 370.
[23] A.M. Puziy, O.I. Poddubnaya, A. Martínez-Alonso, F. Suárez-García, J.M.D.
Tascón, Carbon 43 (2005) 2857.
[24] J. Coates, in: R.A. Meyers (Ed.), Interpretation of Infrared Spectra, A Practical
Approach, John Wiley & Sons Ltd., Chichester, 2000, pp. 10815–10837.
[25] J. Zawadzki, in: P.A. Thrower (Ed.), Infrared Spectroscopy in Surface Chemistry
of Carbons, Marcel Dekker, New York, 1989, pp. 147–386.
[26] P. Vinke, M. van der Eijk, M. Verbree, A.F. Voskamp, H. van Bekkum, Carbon 32
(1994) 675.
[27] M.S. Solum, R.J. Pugmire, M. Jagtoyen, F. Derbyshire, Carbon 33 (1995) 1247.
[28] Spectroscopic Tools. <http://www.chem.unipotsdam.de/tools/index.html>.
[29] L.J. Bellamy, The Infra-red Spectra of Complex Molecules, Wiley, New York,
1954.
[30] D.E.C. Corbridge, J. Appl. Chem. 6 (1956) 456.
[31] G. Socrates, Infrared Characteristic Group Frequencies, Wiley, New York, 1994.
[32] R. Xie, B. Qu, K. Hu, Polym. Degrad. Stab. 72 (2001) 313.
[33] S. Bourbigot, M. Le Bras, R. Delobel, Carbon 33 (1995) 283.
[34] F.M. Bautista, J.M. Campelo, A. Garcia, D. Luna, J.M. Marinas, A.A. Romero, M.R.
Urbano, J. Catal. 172 (1997) 103.
[35] H.A. Benesi, J. Catal. 28 (1973) 176.
[36] H. Knözinger, H. Krietenbrink, P. Ratnasamy, J. Catal. 48 (1977) 436.
[37] A. Corma, C. Rodellas, V. Fornes, J. Catal. 88 (1984) 374.
[38] M.H. Healy, L.F. Wieserman, E.A. Arnett, K. Wefers, Langmuir 5 (1989) 114.
[39] A. Travert, O.V. Manoilova, A.A. Tsyganenko, F. Maugé, J.C. Lavalley, J. Phys.
Chem. B 106 (2002) 1350.
[40] G. Turnes Palomino, J.J. Cuart Pascual, M. Rodríguez Delgado, J. Bernardo Parra,
C. Otero Areán, Mater. Chem. Phys. 85 (2004) 145.
[41] P.A. Jacobs, C.F. Heylen, J. Catal. 34 (1974) 267.
Acknowledgments
[42] C. Petit, F. Maugé, J.C. Lavalley, Stud. Surf. Sci. Catal. 106 (1997) 157.
[43] A.A. Tsyganenko, E.N. Storozheva, O.V. Manoilova, T. Lesage, M. Daturi, J.C.
Lavalley, Catal. Lett. 70 (2000) 159.
[44] C. Morterra, G. Cerrato, G. Meligrana, Langmuir 17 (2001) 7053.
[45] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, in: J. Chastain, R.C. King Jr.
(Eds.), Handbook of X-ray Photoelectron Spectroscopy, Physical Electronics
Inc., Eden Prairie, MN, 1995, pp. 4872–4875.
The authors thank Ministry of Science and Education of Spain
(DGICYT, Projects CTQ2006/11322 and CTQ2009-14262). J.B.
acknowledges the assistance of the Ministry of Science and Educa-
tion of Science and Education of Spain for the award of a FPI Grant.
[46] D.E. Meras, Chem. Eng. Sci. 26 (1971) 1361.
[47] C.N. Satterfield, Heterogeneous Catalysis in Practice, McGraw-Hill, New York,
1991.
[48] J.A. Moulijn, A. Tarfaoui, F. Kapteijn, Catal. Today 11 (1991) 1.
[49] H. Gierman, Appl. Catal. 43 (1988) 277.
References
[1] C. Branca, P. Giudicianni, C . Di Blasi, Ind. Eng. Chem. Res. 43 (2003) 3190.
[2] A.E. Pütün, A. Özcan, H.F. Gerçel, E. Pütün, Fuel 8 (2001) 1371.
[3] J.D. Adjaye, S.P.R. Katikaneni, N.N. Bakhshi, Fuel Process. Technol. 48 (1996)
115.
[4] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Ind. Eng. Chem. Res.
43 (2004) 2610.
[5] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, M. Olazar, J. Bilbao, Ind. Eng.
Chem. Res. 43 (2004) 2619.
[6] J.M. Campelo, A. Garcia, J.F. Herencia, D. Luna, J.M. Marinas, A.A. Romero, J.
Catal. 151 (1995) 307.
[50] K. Levenberg, Q. Appl. Math. 2 (1944) 164.
[51] D. Marquardt, SIAM J. Appl. Math. 11 (1963) 431.
[52] M.A. Aramendia, V. Borau, C. Jiménez, J.M. Marinas, A. Porras, F.J. Urbano, J.
Catal. 161 (1996) 829.
[53] W. Turek, J. Haber, A. Krowiak, Appl. Surf. Sci. 252 (2005) 823.
[54] P. Trens, V. Stathopoulos, M.J. Hudson, P. Pomonis, Appl. Catal. A 263 (2004)
103.
[55] R.M. Rioux, M.A. Vannice, J. Catal. 216 (2003) 362.
[56] M. Boudart, D.E. Meras, M.A. Vannice, Ind. Chem. Belge 32 (1967) 281.
[57] M.A. Vannice, S.H. Hyun, B. Kalpalci, W.C. Liauh, J. Catal. 56 (1979) 358.
[58] J.A. Deon, Lange’s Handbook of Chemistry, 15th ed., McGraw-Hill, New York,
1999. p. 6.42.
[7] J. Bedia, J. Márquez, J. Rodríguez-Mirasol, T. Cordero, Carbon 47 (2009) 286.
[8] H. Koempel, W. Liebner, Stud. Surf. Sci. Catal. 167 (2007) 261–267.
[9] L.R. Radovic, F. Rodríguez-Reinoso, in: P.A. Thrower (Ed.), Chemistry and
Physics of Carbon, vol. 25, Marcel Dekker, New York, 1997, pp. 243–358.
[10] R.W. Coughlin, Ind. Eng. Chem. Prod. Res. Develop. 8 (1964) 23.
[59] J. Zawadzki, M. Wisniewski, J. Weber, O. Heintz, B. Azambre, Carbon 39 (2001)
187.
[60] J.L. Davis, M.A. Barteau, Surf. Sci. 235 (1990) 235.
[61] R.F. Rossi, G. Busca, V. Lorenzelli, O. Saur, J.C. Lavalley, Langmuir 3 (1987) 52.
´
[11] F. Rodriguez-Reinoso, M. Molina-Sabio, Carbon 30 (1992) 1111.
[12] O. Ioannidou, A. Zabaniotou, Renew Sustain Energy Rev 11 (2007) 1966.