This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy CARBON
4 8 ( 2 0 1 0 ) 7 0 6 –7 1 3
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/carbon
Acid properties of phosphoric acid activated carbons and their catalytic behavior in ethyl-tert-butyl ether synthesis A.M. Puziy a b
a,*
, O.I. Poddubnaya a, Yu.N. Kochkin b, N.V. Vlasenko b, M.M. Tsyba
a
Institute for Sorption and Problems of Endoecology, National Academy of Sciences of Ukraine, Naumov St. 13, 03164 Kyiv, Ukraine Institute of Physical Chemistry, National Academy of Sciences of Ukraine, Nauki Ave. 31, 03039 Kyiv, Ukraine
A R T I C L E I N F O
A B S T R A C T
Article history:
Carbon catalysts were prepared by phosphoric acid activation of styrene–divinylbenzene
Received 11 August 2009
copolymer followed by liquid phase oxidation with nitric acid. Their surface properties
Accepted 7 October 2009
were modified by heat treatment at 300–600 °C in an argon atmosphere, and their acid
Available online 13 October 2009
properties were characterized by quasi-equilibrium temperature desorption of ammonia and by acid–base titration. The pore structure was characterized by nitrogen adsorption at 196 °C. Their catalytic activity in the synthesis of ethyl-tert-butyl ether (ETBE) from isobutene and ethanol was investigated. It has been shown that the carbon catalysts are active and selective in ETBE synthesis. The main factor determining the catalytic activity of carbon catalysts is the total number of acid surface sites. A linear correlation between total amount of surface sites and catalytic activity was found. Ó 2009 Elsevier Ltd. All rights reserved.
1.
Introduction
Carbon materials have been used over many years in heterogeneous catalysis due to tailorable porosity, regulated surface chemistry and special electronic properties. Surface chemistry of carbon materials plays a crucial role in application of carbon materials as catalysts and catalysts supports, especially, in acid-catalyzed reactions [1–3]. The nature and concentration of surface sites on carbon materials is very important for describing mechanism of reactions and optimizing carbon catalyst for specific catalytic system. Among the reactions that can be catalyzed by carbon materials is heterogeneous synthesis of ethyl-tert-butyl ether (ETBE) from ethanol and isobutene (IB). ETBE is oxygenate additive to automotive gasoline for increasing octane number [4,5] and reducing exhaust CO and unburned hydrocarbons [6–9]. Commercially ETBE is produced by equilibrium-limited exothermic reaction of ethanol and IB over an acid ionexchange resin catalyst. Due to complex nature, the process
is a suitable model for revealing a role played by different acid sites of catalyst in reaction pathways. To investigate the effect of acid surface sites on reaction pathways the catalysts with similar structure and nature but having different acid characteristics are required. In present study, the catalysts obtained by heat treatment at different temperatures of oxidized form of phosphoruscontaining carbon have been used. The present study was aimed at elucidating the relationship between acid surface sites and catalytic characteristics of carbon catalysts in reaction of synthesis of ETBE from ethanol and IB.
2.
Experimental
2.1.
Carbons
Preparation of polymer-based phosphorus-containing carbon was described in previous studies [10,11]. Briefly, porous chloromethylated and sulfonated copolymer of styrene and
* Corresponding author: Fax: +38 044 4529327. E-mail address:
[email protected] (A.M. Puziy). 0008-6223/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2009.10.015
Author's personal copy CARBON
divinylbenzene was impregnated with 60% phosphoric acid to 0.75 acid/precursor ratio, dried in air and then carbonized in argon flow (1 L min1) at 800 °C for 30 min. After carbonization, the carbon was oxidized by boiling in 25% nitric acid for 4 h. To remove humic substances that could be formed during oxidation with HNO3 the carbon was soaked in 1% NaOH and then washed with hot water. The wash procedure was repeated several times until the colorless washing waters appeared. Then the carbon was transferred into an H-form with 1 M HCl and washed with distilled water until neutral pH was reached in wash waters. To alter acid properties obtained oxidized carbon was heat-treated at different temperatures (300–600 °C) for 30 min in argon flow (1 L min1). It was expected that heat treatment would cause gradual destruction of surface groups and thus will change acid properties of carbon catalyst.
2.2.
Quasi-equilibrium thermodesorption of ammonia
Acid properties of carbon catalysts were determined by quasi-equilibrium thermodesorption (QE-TD) of ammonia [12]. QE-TD of ammonia experiments were carried out by stepwise temperature rise under vacuum (constant pressure 0.133 Pa was maintained by vacuum pump) using thermogravimetric apparatus with quartz spring balance of McBain type. Prior to experiment, the sample was outgassed at its calcination temperature until constant weight was attained (typically 1–1.5 h). Then temperature was lowered to ambient conditions and ammonia was admitted into sample stepwise (10–20 Torr) until no uptake was observed. Finally, excess of ammonia was removed by evacuation at 50 °C. Then the temperature was raised stepwise (5 °C min1 between steps, step size 50 °C) with holding at constant temperature until constant weight was achieved (typically 10–15 min). Thus, experimental conditions corresponded to desorption at constant pressure, i.e. desorption isobar. Distribution of ammonia adsorption energy was calculated from QE-TD data by solving integral equation [12]: Z Emax Hloc ðP; T; EÞf ðEÞdE ð1Þ Hexp ðP; TÞ ¼ Emin
where Hexp is experimentally measured adsorption, Hloc is the kernel function describing local adsorption on surface site with energy E at pressure P and temperature T, f(E) is energy distribution function. Langmuir equation was used as kernel function describing adsorption in isobaric conditions: H¼
KP 1 þ KP
707
4 8 (2 0 1 0) 7 0 6–71 3
ð2Þ
with temperature dependence of constant K according to kinetic theory of gases [13,14]: NA r0 s0 E ð3Þ K ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi exp RT 2pMRT where NA is Avogadro’s number, r0 is area occupied by an adsorbed molecule, s0 is period of oscillation of the adsorbed molecule perpendicular to surface of adsorbent, M is molecular mass of adsorbed molecule, R is gas constant, T is absolute temperature and E is energy of interaction of adsorbed molecule with surface of adsorbent. Period of oscillation, s0, could be estimated as [13]:
s0 ¼
h kT
ð4Þ
where h is Planck’s constant and k is Boltzmann constant. The distribution function is a sought-for quantity that characterizes the catalyst’s surface. Adsorption integral Eq. (1) has been solved with respect to distribution function f(E) using CONTIN method [15–17].
2.3.
Acid–base titration
Acid–base titration experiments were performed in 0.1 M NaCl solution with 0.1 M NaOH or 0.1 M HCl using a 672 Titroprocessor combined with 655 Dosimat (Metrohm, Herisau, Switzerland). Proton concentration was monitored by means of an LL pH glass electrode (Metrohm, Herisau, Switzerland). Prior to experiments, the electrode electromotive force was calibrated to proton concentration by blank titration. The adsorbed amount of protons was calculated using the equation: Q¼
V0 þ Vt ð½Hþ i ½OH i ½Hþ e þ ½OH e Þ m
ð5Þ
where V0 and Vt are the volumes of background electrolyte and titrant added, and m is the mass of adsorbent. Subscripts i and e refer to initial and equilibrium concentrations. Proton binding constants were calculated by solving the integral adsorption equation using the CONTIN method [17,18]: Z Kmax Q loc ð½Hþ ; KH Þf ðKH ÞdKH þ Q 0 ð6Þ Q exp ð½Hþ Þ ¼ Kmin
where Qexp is the experimentally measured proton binding isotherm, Qloc is kernel function describing local adsorption on surface centre with binding constant KH, f(KH) is the proton affinity distribution (PAD) function, and Kmin and Kmax are the limits of integration. Q0 is a constant background term, which accounts for surface groups with binding constant outside the experimental window. The distribution function, f(KH), describes the site concentration as a function of the binding constant, and is a unique characteristic of the adsorbent material. As a local isotherm, which reflects the underlying mechanism of ion binding, the Langmuir equation was used: Ql ¼
KH ½Hþ 1 þ KH ½Hþ
ð7Þ
where KH is the binding constant and [H+] is the proton concentration.
2.4.
Nitrogen adsorption
Nitrogen adsorption–desorption was measured at 77 K using Autosorb-6 (Quantachrome, USA) volumetric adsorption analyzer. Prior to the measurements, all samples were degassed overnight at 200 °C under vacuum. The surface area of the samples under study was assessed by the standard BET method, using nitrogen adsorption data in the relative pressure range from 0.01 to 0.10. The total pore volume was calculated from the amount of nitrogen adsorbed at a relative pressure of 0.99. Micropore and mesopore
Author's personal copy CARBON
4 8 ( 2 0 1 0 ) 7 0 6 –7 1 3
volumes were calculated from pore size distribution for pores with size w < 2 nm and 2 < w < 50 nm, respectively. Pore size distributions (PSD) were calculated using the QSDFT model [19].
2.5.
Catalytic studies
6 5
200 300 400 500 600
4 3 2 1 0 0
100
200
3.
Results and discussion
3.1.
Characterization of carbon catalysts
3.1.1.
Acid characteristics from ammonia desorption
300
400
500
Temperature [°C]
(b) 0.25 200 300 400 500 600
0.2 0.15 F(E)
Catalytic activity of carbon catalysts was investigated in flow reactor with fixed bed [12]. About 1.5 cm3 of carbon catalyst with grain size 0.5–1 mm was used in all experiments. Molar ratio ethanol/isobutene was 1.5, LHSV (liquid hour space velocity) – 1 h1, carrier gas helium. Ethanol was used as azeotropic mixture (4.43 mass percentage of water). The pressure was 1.0 MPa, temperature range 80–180 °C. Reaction products were analyzed using gas chromatograph Agat (Mashpriborkomplekt, Russia) equipped with Chromaton N-AW column with 10% Carbowax 600 (3 mm id, 2 m length) and a thermal conductivity detector.
(a) Adsorption [mmol/g]
708
0.1 0.05 0
Previous studies have shown that phosphoric acid activation of polymer and lignocellulosic precursors leads to incorporation of significant amount of phosphorus (6–9%) [10,20–23]. Phosphorus heteroatoms impart acid properties to carbon surface and ability to adsorb cations [10,20–24]. It has been shown that phosphorus exists in carbon obtained at 800 °C as phosphate-like structure bound to carbon lattice via C–O– P bonding [10,25,26]. In this study the acidity of phosphoric acid activated carbon catalysts was characterized by QE-TD of ammonia from gas phase and by acid–base titration in aqueous solution. Fig. 1a shows thermodesorption of ammonia from investigated carbon catalysts. All curves show decreasing amount of adsorbed ammonia with increasing temperature. There is a correlation between complete ammonia desorption temperature and heat treatment temperature of samples. The curves are smooth with no obvious steps indicating heterogeneous character of surface with several types of surface groups. Modeling ammonia desorption with adsorption integral Eq. (1) gives excellent fit to experimental QE-TD data (Fig. 1a). Calculated adsorption energy distributions (Fig. 1b and Table 1) show the presence of five types of surface groups on carbon catalysts. These groups may be classed to very weak (62 kJ/mol), weak (82 kJ/mol), medium (107 kJ/mol), strong (136 kJ/mol) and very strong (165 kJ/mol) surface sites according to energy of adsorption. The distributions are dominated by very weak and weak surface sites whose contribution to the total amount of surface sites decreases from 100% to 55% with increasing heat treatment temperature from 200 to 600 °C. With increasing of heat treatment temperature the total amount of surface sites decreases mainly through decreasing the number of very weak and weak surface sites (Table 1) indicating gradual destruction and transformation of these surface groups. It is interesting that thermal destruction of surface groups results in formation
40
60
80
100
120
140
160
180
Energy [kJ/mol]
Fig. 1 – QE-TD of ammonia (a) and adsorption energy distributions (b) for carbon catalysts heat-treated at different temperatures.
of new sites with higher adsorption energy. Initially present two types of very weak and weak surface sites in parent carbon catalyst have been transformed upon heating up to five types with higher acidity. It should be noted that medium to very strong surface sites might also be due to mass loss through other reactions such as formation of amides from ammonia salt of carboxylic acids: RCOONH4 ! RCONH2 þ H2 O
ð8Þ
or through formation of lactones:
O
OH
O OH
O
Thus, we can draw conclusion that heat treatment of carbon catalyst resulted in decreasing total amount of surface sites while conclusion about increasing average surface acidity may be false.
3.1.2.
Acid characteristics from acid–base titration
Another method to estimate acid–base characteristics of solid adsorbent is acid–base titration [27,28]. This method gives
Author's personal copy CARBON
709
4 8 (2 0 1 0) 7 0 6–71 3
Table 1 – Site densities and their average energy of adsorption calculated from QE-TD of ammonia for carbon catalysts heattreated at different temperatures. Heat treatment temperature (°C)
Type of surface group I
II
a
200 300 400 500 600 Average energy
b
III
a
Q
E
Q
2.91 2.75 1.61 1.46 0.62
59.6 63.2 65.1 61.5 60.8 62 ± 2
3.01 1.34 0.75 0.40 0.62
E
b
Q
77.3 81.1 90.1 82.0 78.4 82 ± 5
a
0.61 0.68 0.38 0.36
IV b
E
97.3 109.1 112.7 106.9 107 ± 7
Q
a
0.63 0.48
Totala
V b
E
138.6 133.4 136 ± 4
a
Q
0.17
b
E
165.1 165
5.92 4.70 3.05 2.88 2.26
a Site density (mmol/g). b Average adsorption energy (kJ/mol).
(a)
0
Q [mmol/g]
-1 -2
200 300 400 500 600
-3 -4 -5 -6 2
(b)
3
4
5
6
7 pH
9
10
11
12
3.0 200 300 400 500 600
2.5 2.0 F (pK)
8
1.5 1.0 0.5 0.0 0
1
2
3
4
5
6
7
8
9
10
11
12
pK
Fig. 2 – Proton binding isotherms (a) and proton affinity distributions of acidic surface groups (b) for carbon catalysts heat-treated at different temperatures.
information about surface sites capable of dissociation or association of protons, i.e. acidic and basic sites on adsorbent surface. Surface sites on carbon materials can be Brønsted type (acid sites: carboxylic, enolic, phenolic and phosphate groups; basic sites: ethers, carbonyl groups) and Lewis type (basic site: protonation of p-electron system). Proton binding isotherms obtained from acid–base titration experiments are shown in Fig. 2a. The isotherms lie in negative region indicating that solely proton dissociation occurs in investigated pH range. Taking into account that Lewis type maybe only basic sites, we can conclude that all surface sites are of Brønsted type. The intersection of isotherms with horizontal axis
(point of zero charge where adsorption and desorption of protons are zero) was not observed for all isotherms. This fact implies very acidic surface of carbon catalysts obtained by phosphoric acid activation. With increasing heat treatment temperature, the amount of dissociated protons decreases (proton binding values become less negative) which means decreasing amount of acidic surface sites. Smooth curves without obvious steps tell in favor of some distribution of pK of surface groups. Proton affinity distributions calculated from proton binding isotherms show five types of surface groups for all carbon catalysts (Fig. 2b and Table 2). Commonly accepted interpretation of chemical structure of surface groups seen by acid–base titration method is based on comparison of dissociation constants of surface groups with those of simple compounds. However, it should be noted that this classification is conventional and does not give true information about the chemical structure of surface group because proposed types of simple compounds show pKs in broad and overlapping regions. For example, pK of carboxylic groups falls in the broad range between 0.7 and 7.5, pK of enol groups is within 5.8–10.7 and pK for phenols lies in the range 7.6–10.3 [29]. Thus, surface groups of I, II, III types may be carboxylic, surface groups of III, IV, V type may be enol and surface groups of IV, V types may be phenolic. In addition, carbons obtained by phosphoric acid activation method contain significant amount of phosphorus [10] in form of phosphate groups bound to carbon [25,26]. Taking into account dissociation constants of phosphate groups [29] the surface groups of type I and type III may also be due to the first and second dissociation constant of phosphate groups. With increasing heat treatment temperature, the content of surface groups of all types decreases. It is interesting to note that total amount of surface groups corresponds closely to the amount of surface groups obtained from QE-TD of ammonia (Tables 1 and 2). Taking into account that QE-TD of ammonia measures both Lewis and Brønsted types of surface groups and potentiometric titration gave information on only Brønsted sites it may be concluded that investigated carbon catalysts contain merely Brønsted sites.
3.1.3.
Porous structure
Porous structure of investigated carbon catalysts was characterized by nitrogen adsorption at 77 K (Fig. 3a). The isotherms
Author's personal copy 710
CARBON
4 8 ( 2 0 1 0 ) 7 0 6 –7 1 3
Table 2 – Site densities and their average pK calculated from proton binding isotherms for carbon catalysts heat-treated at different temperatures. Heat treatment temperature (°C) PZCa I
200 300 400 500 600 Average pK
Totala
Type of surface group
0.02 0.13 0.83 0.88 0.95
II
III
IV
V
Qb
pKc
Qb
pKc
Qb
pKc
Qb
pKc
Qb
pKc
1.46 1.24 0.93 0.71 0.52
2.14 2.11 2.61 2.64 2.54 2.4 ± 0.3
0.75 0.55 0.28 0.21 0.18
4.17 4.25 4.65 4.69 4.53 4.5 ± 0.2
0.97 0.97 0.96 0.68 0.55
5.69 5.91 6.87 6.96 6.89 6.5 ± 0.6
0.83 0.60 0.29 0.28 0.21
7.77 7.73 9.05 9.29 9.31 8.6 ± 0.8
1.85 1.79 1.16 0.77 0.66
10.18 10.14 10.59 10.70 10.76 10.5 ± 0.3
5.86 5.15 3.62 2.65 2.13
(a)
16
Adsorbed amount [mmol/g]
a Point of zero charge obtained by linear extrapolation of proton binding isotherm. b Site density (mmol/g). c Average pK.
14
200 300 400 500 600
12 10 8 6 4 2 0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Relative pressure
(b)
0.7 200 300 400 500 600
0.6 dV/dlog (w)
0.5 0.4 0.3 0.2 0.1 0 0.1
1
10
100
pressures they are intermediate between types II and IV with a hysteresis loop intermediate between types H3 and H4 (both being typical for materials containing slit-shaped pores). Definite hysteresis loop closure was observed at relative pressure 0.48 for all carbon catalysts. With increasing heat treatment temperature, nitrogen adsorption increases at all relative pressures indicating gradual development of porous structure. Parameters of porous structure for carbon catalysts heattreated at different temperatures are listed in Table 3 and pore size distributions calculated by QSDFT method [19] are shown in Fig. 3b. The data show gradual increase up to 50% of BET surface area and pore volumes with increasing heat treatment temperature. The development of porosity suggests destruction of surface groups and freeing up space, which is in line with data presented in Sections 3.1.1 and 3.1.2. Pore size distributions Fig. 3b show three types of pores present in all carbon catalysts. They are micropores with average size 0.7–0.8 nm and mesopores with average size 3.4 nm and about 13 nm. Micropores and mesopores of 3.4 nm arose due to chemical activation with phosphoric acid, while large mesopores were formed during copolymerization of polymer precursor.
w [nm]
3.1.4.
Fig. 3 – Nitrogen adsorption isotherms (a) and pore size distributions calculated by QSDFT method (b) for carbon catalysts heat-treated at different temperatures.
Catalytic activity
The activity of carbon catalysts was tested in ETBE synthesis from ethanol and IB C2 H5 OH þ ðCH3 Þ2 C@CH2 $ ðCH3 Þ3 C–O–C2 H5
belong to a mixed type of the IUPAC classification [30]. In their initial part, they are type I, with an important uptake at low relative pressures. At intermediate and high relative
ð10Þ
The main reaction (10) is accompanied by several side reactions, namely, dimerization of IB: 2ðCH3 Þ2 C@CH2 $ ðCH3 Þ2 CH–CH@CH–CHðCH3 Þ2
ð11Þ
Table 3 – BET surface area and pore volumes of carbon catalysts heat-treated at different temperatures. Temperature (°C) 100 300 400 500 600
ABET (m2/g)
Vtot (cm3/g)
Vmi (cm3/g)
Vme (cm3/g)
340 363 375 427 490
0.366 0.380 0.391 0.430 0.529
0.124 0.115 0.130 0.147 0.166
0.242 0.265 0.261 0.283 0.362
Author's personal copy CARBON
hydration of IB with formation of tert-butyl alcohol (TBA): ðCH3 Þ2 C@CH2 þ H2 O $ ðCH3 Þ3 C–O–H
ð12Þ
and dehydration of ethanol with formation of diethyl ether (DEE): 2C2 H5 OH $ C2 H5 –O–C2 H5 þ H2 O
ð13Þ
Our previous investigation has shown the importance of weak acid surface sites (<100 kJ/mol) in case of zeolite catalyst for ETBE synthesis [12]. Taking into account high concentration of weak acid surface groups on carbon catalysts (Table 1) it is anticipated that investigated carbons would also exhibit catalytic activity in this reaction. Fig. 4 indicates that all carbon catalysts are active in ETBE synthesis at all investigated reaction temperatures. With increasing reaction temperature, IB conversion increases up to maximum and goes down at higher temperatures (Fig. 4a). The decrease of IB conversion at higher temperatures is explained by thermodynamic limitations [31]. The highest IB conversion of 50–60% was obtained at 100–120 °C. With increasing heat treatment temperature of the carbon catalyst its catalytic activity decreases and temperature of maximum IB conversion shifts to higher reaction temperatures (Fig. 4a). It should be noted that development of porous structure with increasing heat treatment temperature (Section 3.1.3) would suggest increasing catalytic activity of carbon catalysts since textural properties of the catalyst as a rule influence mass transport. However, experimentally observed decreasing catalytic activity with increasing heat treatment temperature indicates that surface chemistry plays a dominant role in ETBE synthesis.
(a) 70 200 300 400 500 600
IB conversion [%]
60 50 40 30 20 10 0
60
80
100
120
140
160
180
Reaction temperature [°C]
IB was converted to ETBE with high selectivity (70–80%) at 100–160 °C (Fig. 4b). With increasing reaction temperature, the selectivity of ETBE formation increases from 60–65% to 85%. Small decrease of ETBE selectivity at highest reaction temperatures, observed for parent carbon catalyst, is attributed to increasing transformation of ethanol to DEE. Increasing heat treatment temperature of carbon catalyst does not affect the selectivity of ETBE synthesis. For all carbon catalysts ETBE was the main product of IB transformation according to reaction (10) (Fig. 4b), the second largest being TBA (Fig. 5) according to reaction (12). Dimerization of IB (reaction (11)) was not observed at all. Ethanol was also converted mainly to ETBE (reaction (10)). Small amount of DEE (reaction (13)) was observed only for parent carbon catalyst at highest reaction temperatures 140–180 °C (1% at 140 °C, 11% at 160 °C, 25% at 180 °C). It is apparent that the higher amount of active surface sites on carbon catalyst the higher catalytic activity should be observed. Fig. 6 shows IB conversion 120 °C as a function of concentration of total amount of acid surface groups obtained by QE-TD of ammonia and by acid–base titration. The figure clearly shows a good linear relationship between IB conversion and total site densities obtained by both methods. Supposing that all acid sites are available for ETBE synthesis reaction, carbon catalysts show turnover frequency values (TOF) at 120 °C from 0.00024 to 0.00044 s1 calculated from QE-TD of ammonia and from 0.00026 to 0.00040 s1 calculated from acid–base titration. The dependence of catalytic activity in ETBE synthesis on total amount of acid sites on carbon catalysts observed in this study is different from previously found dependence on concentration of weak acid sites on zeolite catalysts [12]. This discrepancy could be explained by different nature of acid surface sites on carbons and zeolites. Acid sites on zeolite surface measured by QE-TD of ammonia comprise of Brønsted silanol and bridging hydroxyl sites as well as Lewis sites. Each site forms adsorbed complexes of different composition and strength and thus shows different catalytic activity. It is suggested that active in ETBE synthesis weak acid groups of zeolite are associated with Brønsted silanol sites characterized by labile proton and thus exhibit weak acid properties [32]. For carbon catalysts of present investigation, despite the difference in acid strength, the surface active sites are
90
45
85
40
80
35
TBA selectivity [%]
ETBE selectivity [%]
(b)
75 70
200 300 400 500 600
65 60 55 80
100
120
140
200 300 400 500 600
30 25 20 15 10 5
50 60
711
4 8 (2 0 1 0) 7 0 6–71 3
160
180
Reaction temperature [°C]
Fig. 4 – Effect of reaction temperature on IB conversion (a) and ETBE selectivity (b) over carbon catalysts heat-treated at different temperatures.
0
60
80
100
120
140
160
180
200
Reaction temperature [°C]
Fig. 5 – Selectivity of TBA synthesis over carbon catalysts heat-treated at different temperatures versus reaction temperature.
Author's personal copy 712
CARBON
4 8 ( 2 0 1 0 ) 7 0 6 –7 1 3
[2]
[3]
[4] [5]
Fig. 6 – IB conversion at 120 °C versus very weak acid sites density obtained from QE-TD of ammonia and acid–base titration.
[6]
[7]
Brønsted type groups and hence they form the same type of surface complex and participate in ETBE synthesis. Consequently, the linear relationship is expected between catalytic activity and total amount of surface groups, which is clearly seen in Fig. 6. Thus, decreasing IB conversion with increasing heat treatment temperature of carbon catalyst (Fig. 4a) is caused by decreasing total amount of acid surface groups. Two mechanisms have been proposed for ETBE synthesis over zeolites and acid resin catalysts. The first is Langmuir– Hinshelwood that involves reaction of two adsorbed complexes of ethanol and isobutylene with formation of ETBE [33–35]. Second is Eley–Rideal mechanism describing reaction of adsorbed complex of ethanol with isobutylene from gas phase [36–38]. In some cases hybrid mechanism was also discussed [39,40]. The results of present investigation showed that the catalytic activity of carbon catalyst depends on the order in which catalyst contacted with reagents. When carbon catalyst contacted with isobutylene first and then with ethanol–isobutylene mixture, catalytic activity was very low. However, when carbon catalyst first contacted with ethanol and then with ethanol–isobutylene mixture, the activity was high. This fact suggests Eley–Rideal mechanism of ETBE synthesis over carbon catalysts involving reaction of adsorbed complex of ethanol with isobutylene from gas phase.
4.
Conclusions
Investigation of phosphoric acid activated carbons showed their catalytic activity and selectivity in ETBE synthesis. The main factor determining catalytic activity of carbon catalysts is total amount of acid surface sites. The linear correlation between total amount of surface site and catalytic activity was found. With increasing heat treatment temperature of carbon catalyst, its total amount of surface groups decreases with corresponding decrease in catalytic activity.
[8]
[9]
[10]
[11]
[12]
[13] [14] [15]
[16]
[17]
[18]
[19]
[20]
[21] R E F E R E N C E S
[22] [1] Radovic LR, Rodrı´guez-Reinoso F. Carbon materials in catalysis. In: Thrower PA, editor. Chemistry and physics of
carbon, vol. 25. New York: Dekker; 1997. p. 276–312, 329–34. Leon y Leon CA, Radovic LR. Interfacial chemistry and electrochemistry of carbon surfaces. In: Thrower PA, editor. Chemistry and physics of carbon, vol. 24. New York: Dekker; 1994. p. 276–85. Figueiredo JL, Pereira MFR. Carbon as catalyst. In: Serp P, Figueiredo JL, editors. Carbon materials for catalysis. Hoboken, NJ: Wiley; 2009. p. 177–218. Nadim F, Zack P, Haag GE, Liu S. United States experience with gasoline additives. Energy Policy 2001;29(1):1–5. Agarwal AK. Biofuels (alcohols and biodiesel) applications as fuels for internal combustion engines. Prog Energy Combust Sci 2007;33(3):233–71. Vinuesa J-F, Mirabel Ph, Ponche J-L. Air quality using reformulated and oxygenated gasoline fuel blends: application to the Strasburg area (F). Atmos Environ 2003;37(13):1757–74. Perry R, Gee IL. Vehicle emissions in relation to fuel composition. Sci Total Environ 1995;169:149–56. Taljaard HC, Jordaan CFP, Botha JJ. The effect of oxygen content in different oxygenate-gasoline blends on performance and emissions in a single cylinder, sparkignition engine. SAE Paper 910379; 1991. Pumphrey JA, Brand JI, Scheller WA. Vapour pressure measurements and predictions for alcohol-gasoline blends. Fuel 2000;79:1405–11. Puziy AM, Poddubnaya OI, Martı´nez-Alonso A, Sua´rez-Garcia F, Tasco´n JMD. Synthetic carbons activated with phosphoric acid. 1. Surface chemistry and ion binding properties. Carbon 2002;40(9):1493–505. Puziy AM, Poddubnaya OI, Zaitsev VN, Konoplitska OP. Modeling of heavy metal ion binding by phosphoric acid activated carbon. Appl Surf Sci 2004;221(1–4):421–9. Vlasenko NV, Kochkin YN, Puziy AM. Liquid phase synthesis of ethyl tert-butyl ether: the relationship between acid, adsorption and catalytic properties of zeolite catalysts. J Mol Catal A: Chem 2006;253(1–2):192–7. de Boer JH. The dynamic character of adsorption. Oxford: Clarendon Press; 1953. Adamson AW, Gast AP. Physical chemistry of surfaces. New York: Wiley; 1997. Provencher SW. A constrained regularization method for inverting data represented by linear algebraic or integral equations. Comput Phys Commun 1982;27:213–27. Provencher SW. CONTIN: a general purpose constrained regularization program for inverting noisy linear algebraic and integral equations. Comput Phys Commun 1982;27:229–42. Puziy AM, Matynia T, Gawdzik B, Poddubnaya OI. Use of CONTIN for calculation of adsorption energy distribution. Langmuir 1999;15:6016–25. Puziy AM, Poddubnaya OI, Ritter JA, Ebner AD, Holland CE. Elucidation of the ion binding mechanism in heterogeneous carbon-composite adsorbents. Carbon 2001;39(15):2313–24. Ravikovich PI, Neimark AV. Density functional theory model of adsorption on amorphous and microporous materials. Langmuir 2006;22:11171–9. Puziy AM, Poddubnaya OI, Martı´nez-Alonso A, Sua´rez-Garcia F, Tascon JMD. Synthetic carbons activated with phosphoric acid. 3. Carbons prepared in air. Carbon 2003;41(6):1181–91. Puziy AM, Poddubnaya OI, Gawdzik B, Sobiesiak M, Tsyba MM. Phosphoric acid activation – functionalization and porosity modification. Appl Surf Sci 2007;253(13):5736–40. Puziy AM, Poddubnaya OI, Martı´nez-Alonso A, Sua´rezGarcia F, Tascon JMD. Surface chemistry of phosphoruscontaining carbons of lignocellulosic origin. Carbon 2005;43(14):2857–68.
Author's personal copy CARBON
4 8 (2 0 1 0) 7 0 6–71 3
[23] Puziy AM, Poddubnaya OI, Martı´nez-Alonso A, Castro Mun˜iz A, Sua´rez Garcı´a F, Tasco´n JMD. Oxygen and phosphorus enriched carbons from lignocellulosic material. Carbon 2007;45(10):1941–50. [24] Puziy AM, Poddubnaya OI, Gawdzik B, Sobiesiak M, Dziadko D. Heterogeneity of synthetic carbons obtained from polyimides. Appl Surf Sci 2002;196(1–4):89–97. [25] Puziy AM, Poddubnaya OI, Ziatdinov AM. On the chemical structure of phosphorus compounds in phosphoric acid activated carbons. Appl Surf Sci 2006;252(23):8036–8. [26] Puziy AM, Poddubnaya OI, Socha RP, Gurgul J, Wisniewski M. XPS and NMR studies of phosphoric acid activated carbons. Carbon 2008;46(15):2113–23. [27] Szabo´ T, Tomba´cz E, Ille´s E, De´ka´ny I. Enhanced acidity and pH-dependent surface charge characterization of successively oxidized graphite oxides. Carbon 2006;44:537–45. [28] Seredych M, Bandosz T. Mechanism of ammonia retention on graphite oxides: role of surface chemistry and structure. J Phys Chem C 2007;111:15596–604. [29] Dean John A, editor. Lange’s handbook of chemistry. NY: McGraw-Hill, Inc.; 1999. p. 8.24–72 [section 8.2.1, table 8.8]. [30] Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti RA, Rouquerol J, et al. Reporting physisorption data for gas/solid systems with special reference to the determination of surface area and porosity. Pure Appl Chem 1985;57:603–19. [31] Oudshoorn OL, Janissen M, van Kooten WEJ, Jansen JC, van Bekkum H, van den Bleek CM. A novel structured catalyst packing for catalytic distillation of ETBE. Chem Eng Sci 1999;54:1413–8.
713
[32] Kondo JN, Yoda E, Ishikawa H, Wakabayashi F, Domen K. Acid property of silanol groups on zeolites assessed by reaction probe IR study. J Catal 2000;191:275–81. [33] Dogu T, Boz N, Aydin E, Oktar N, Murtezaoglu K, Dogu G. DRIFT studies for the reaction and adsorption of alcohols and isobutylene on acidic resin catalysts and the mechanism of ETBE and MTBE synthesis. Ind Eng Chem Res 2001;40:5044–51. [34] Chang KH, Kim GJ, Ahn WS. Methyl tert-butyl ether synthesis over titanium–silicalite I catalysts. Ind Eng Chem Res 1992;125:125–30. [35] Slomkiewicz PM. Determination of the Langmuir– Hinshelwood kinetic equation of synthesis of ethers. Appl Catal A: Gen 2004;269:33–42. [36] Tau L, Davis BH. Acid catalyzed formation of ethyl tertiary butyl ether (ETBE). Appl Catal 1989;53:263–71. [37] Kogelbauer A, Nikolopoulos AA, Goodwin Jr JG, Marcelin G. Reactant adsorption and its impact upon MTBE synthesis on zeolites. J Catal 1995;152:122–9. [38] Fite C, Iborra M, Tejero J, Izquierdo JF, Cunill F. Kinetics of the liquid-phase synthesis of ethyl tert-butyl ether (ETBE). Ind Eng Chem Res 1994;33:581–91. [39] Zhang T, Jensen K, Kitchaiya P, Phillips C, Datta R. Liquidphase synthesis of ethanol-derived mixed tertiary alkyl ethyl ethers in an isothermal integral packed-bed reactor. Ind Eng Chem Res 1997;36:4586–94. [40] Iborra M, Izquierdo JF, Tejero J, Cunill F. Application of the response surface methodology to the kinetic study of the gasphase addition of ethanol to isobutene on a sulfonated styrene–divinylbenzene resin. Ind Eng Chem Res 1992;31:1840–8.