kinetics of steam reforming of acetol over pt/c...

CHAPTER V

Kinetics of steam reforming

of acetol over Pt/C catalyst

Kinetics of steam reforming of acetol over Pt/C catalyst

Study of catalytic steam reforming of model bio-oil compounds Page 43

5.1. Introduction

In the present work, hydroxyacetone (or acetol) was selected as the model

compound. This hydroxyl group-bearing compound is a major constituent of the

ketonic fraction of biomass pyrolysis liquids (Vispute and Huber 2009; Oasmaa and

Meier, 2005). There is just scarce information in the published literature on this

reaction system available (Ramos et al., 2007; Bimbela et al., 2009; Medrano et al.,

2009). Reaction kinetics was investigated in a fixed-bed reactor over the ranges in

temperature, 623-773 K, acetol concentration, 5-20 wt%, and W/FAO ratio, 0.93 to

4.66 g h/mol using a commercial Pt/C catalyst. We selected Pt due to its high activity

for C-C bond cleavage and water gas shift, and high thermal stability (Vaidya and

Rodrigues, 2006). Because Pt/C has high efficacy for H2 production from other

hydrocarbons such as glycerol (Soares et al., 2006; Sutar et al., 2010) and ethylene

glycol (Davda et al., 2005), we anticipated the stable conversion of acetol into H2

can

be achieved using Pt/C.

5.2.

5.2.1 Experimental Procedure

Experimental

An aqueous acetol solution was charged from a feed vessel by using an HPLC

pump, vaporized in a pre-heater (573 K) and passed over the catalyst placed inside

the reactor at 773 K. Before each experiment, the catalyst was reduced at 773 K for 1

h in the presence of pure H2 employed at a flow rate of 50 cm3/min. Since a steady

state condition was achieved within 1 h, the results reported here are after 1 h of

reaction. Because of the uncertainties in the estimation of gas-phase conversion, we

considered the residual acetol content in the liquid product for evaluating acetol

conversion. Acetol conversion was calculated by Eq. 4.1 and H2

Its maximum stoichiometric value equals 1 mol/mol. W/F

yield was calculated

by Eq. 4.2 (see Chapter-IV).

AO (g h/mol) was

defined as the ratio of the mass of the catalyst (Pt/C) to the molar flow rate of acetol

at the inlet; the space time W/Q0 is much lower. The amount of catalyst (W) used in

all experiments was kept constant at 0.1 g. unless stated otherwise, an initial acetol

concentration of 10 wt % was used. The steam-to-carbon molar ratio was equal to

12.33. The liquid feeding rate was varied from 0.25 to 1.25 cm3/min. The total flow

of the aqueous acetol solution was varied from 338 to 1690 cm3/min. A nitrogen flow

rate of 100 cm3/min was used in all experiments. Thus, the total volumetric gas flow



rate (303 K, 0.1 MPa) was varied in the range 438 < Qo < 1790 cm3/min. All

experiments were performed at diluted conditions and hence the volume change due

to reaction was negligible. Using a W/FA0 ratio equal to 1.16 g h/mol, one experiment

at 773 K was repeated. From the variation in the values of % acetol conversion (59.5%

and 56.5%0 and H2 yield (0.37 and 0.35 mol/mol), It was concluded that the

experimental error was within the limits of engineering accuracy. At 773 K, these

results reported that the values of acetol conversion (11%) and H2

yield (0.07

mol/mol) in the absence of the catalyst. Previous studies suggest that the non-catalytic

steam reforming of acetol is noticeable (Ramos et al., 2007; Bimbela et al., 2009;

Medrano et al., 2009).

5.2.2 Product Analysis

A MS 13X column was used for the detection of H2 in the gaseous product

stream, whereas a silica gel column was used for the analysis of CH4 formed during

reaction. CO2 was detected by using H2 as the carrier gas in a Porapak N column.

From a comparison with the analysis of calibration gas mixtures, we found that

concentrations of C2 and C3 hydrocarbons in the reformed gas were negligible.

Similarly, CO concentration in the product stream was below the detection level. CO2

was detected by using H2 as the carrier gas in a Porapak N column. Although CO is

one of the major products of acetol steam reforming (Ramos et al., 2007, Bimbela et

al. 2009, Medrano et al. 2009). We found that CO concentration in the product stream

was below the detection level. This was mainly because of two reasons: first, N2 was

used as a carrier gas during experimentation, and second, all kinetic experiments were

performed at diluted conditions. Interestingly, when no carrier gas was used during

one of the experiments at 773 K (steam/carbon ratio = 37), CO was detected by using

Porapak N; the concentrations of CH4, CO2 and CO in the exit stream (on water-free

basis) were 17.6%, 23.9% and 11.8%, respectively. We checked the gas balance in

GC analysis and found that the balance closure was ~100%. The residual acetol

content of the liquid samples was estimated by using 3% SP2100 + 2% SP2300

column with a chromosorb WHP stationery phase coating. Acetone was the major by-

product detected in the liquid samples; few others (e.g., acetic anhydride, acetic acid,

acetaldehyde and 2-propanol) were detected in trace amounts, too, and their identity

was confirmed by using GC-MS technique (QP-2010 Shimadzu).


Study of catalytic steam reforming of model bio-oil compounds 5

5.2.3 Catalyst Characterization

A surface area analyzer (Micromeritics Model ASAP 2010) was used for

investigation of various catalytic features. Using N2 adsorption technique, it was

found that the fresh catalyst had the following properties: BET surface area, 948 m2/g,

micropore volume, 0.43 cm3

The surface morphology of the catalyst was studied by using the scanning

electron microscopy technique (JEOL-JSM 6380 LA Scanning Electron Microscope).

Two SEM image of the unused catalyst is shown in Fig. 5.1. Irregular particles with

different sizes are seen. Certainly, the solid has a porous structure, and there is a

presence of granules distributed on the solid surface. The granular characteristics of

the catalyst are distinctly seen. Besides, a predominantly porous region with a solid

phase distributed on its top is evident.

/g, and average pore diameter, 6.5 nm. Using a Coulter

LS 230 particle size analyzer, we found that the catalyst particle size varied in the

range, 23-69 μm.

Transmission electron microscopy was performed on PHILIPS CM-200

microscope instrumentation. Fig. 5.2 shows the TEM image of the unused catalyst

which illustrates black coloured, small-sized nearly spherical shapes well dispersed

on the carbon support. Thus, it could be said that the Pt metal were uniformly and

finely loaded on the carbon support. The average particle size of Pt/C catalyst was

determined as 6.3 nm. The average particle size of the catalyst was in full agreement

with the BET analysis executed. Powder X-ray diffraction patterns of the Pt/C

catalysts were obtained using Rigaku Miniflex D500 diffractometer and

monochromic Cu Kα radiation. The used and unused catalyst exhibits similar XRD

patterns (see Fig. 5.3). The peak at 26.6o is attributed to C, whereas the peaks at

39.79o, 46.24o and 67.45o

are characteristic of Pt (Dipti et al. 2007) (Luo et al. 2008).

The used catalyst does not exhibit new diffraction lines, thereby suggesting that no

new phase is formed in the catalyst.

5.3. Results and Discussion

Kinetic data in the temperature range, 623-773 K, are represented in Table 5.1.

As the W/FAO ratio increases from 0.93 to 2.33 g h/mol, there is considerable increase

in acetol conversion; thereafter, this effect is less pronounced at all temperatures.

Maximum acetol conversion (81.1 %) and H2 yield (0.66 mol/mol) were achieved at

773 K at W/FAO equal to 4.66 g h/mol. A comparison with the aforementioned non-



catalytic experiment at 773 K highlights the efficacy of the Pt/C catalyst. The effect

of the space time on acetol conversion at 623, 673, 723 and 773 K is represented in

Fig. 5.4. Because of the low reaction temperature and space time used for this study,

complete conversion of acetol was not achieved; even so, the increase in the values of

these reaction variables caused the expected increase in % acetol conversion. Using

Ni coprecipitated catalyst (33% Ni), Bimbela et al., (2009) reported 36% carbon

conversion at T=823 K and W/FAO= 0.1 g h/mol. The dependence of the H2 yield on

temperature is shown in Fig. 5.5. It is, thus, obvious that H2 production is favored at

high temperatures; a similar trend was also observed in previous works (Ramos et al.,

2007;Bimbela et al., 2009). In Fig. 5.6, a plot of H2 yield vs. acetol conversion at 773

K is shown. The higher the extent of reaction (acetol conversion), the higher is the

relative yield of H2. The effect of temperature on the CH4/H2 ratio was studied; these

results are shown in Fig. 5.7. As the temperature increased from 623-773 K, the

CH4/H2 ratio decreased from 0.54 to 0.23. It is worthy of note that Ramos et al., 2007

reported a value of 0.05 at 773 K for equilibrium. Finally, the effect of acetol

concentration (5-20 wt% acetol) in the feed solution on the conversion and yield was

studied (see Table 5.2). The liquid feeding rate was kept constant, whereas the W/FAO

ratio was varied. The overall tendency observed when the steam/carbon molar ratio

was increased is that the H2 yield was improved; the results of Ramos et al. (2007)

using Ni–Al catalyst at 923 K are similar. It is, thus, obvious that high water content

in the feed facilitates H2

The main reactions describing the reforming process are acetol steam reforming

(Ramos et al., 2007; Bimbela et al., 2009) and water gas shift (WGS):

production, possibly suggesting the shift towards products in

the water gas shift (WGS) reaction.

C3H6O2 + H2O → 3 CO + 4 H2 (5.3)

CO + H2O ⇔ CO2 + H2 (5.4)

The overall reaction, which is the sum of reactions represented by Eq. 5.3 and

5.4, is represented as:

C3H6O2 + 4 H2O → 3 CO2 + 7 H2 (5.5)

The thermodynamic equilibrium yields for acetol steam reforming at 773 K are

0.154, 0.095, 1.479 and 0.056 (g/g acetol) for H2, CO, CO2 and CH4, respectively

(Ramos et al., 2007). Methane steam reforming, which is thermodynamically limited,

may occur, too:



CH4 + H2O ⇔ CO + 3 H2 (5.6)

Besides, cracking reactions which result in the formation of oxygenated organic

compounds, gases and coke Hu and Lu, 2009 and steam reforming of the oxygenated

intermediates Ramos et al., 2007; Bimbela et al., 2009; Medrano et al., 2009) may

occur. These reactions are represented as:

CnHm Ok → CxHyOz + gas(CO, CO2, CH4, H2) (5.7)

CnHm Ok → CxHyOz + coke (5.8)

C2Hn + 2H2O → 2CO + (2 + n 2⁄ )H2 (5.9)

Acetol steam reforming occurs even without any catalyst; the reaction is further

accelerated in the presence of a catalyst. Our results comprise contributions from the

non-catalytic as well as the catalytic reaction. The latter is a heterogeneous gas–solid

catalyzed reaction system, which involves the following transfer processes: diffusion

of the reactants from the bulk gas phase to the catalyst surface, intra-particle diffusion

followed by chemical reaction at the active centers and diffusion of the products. Any

of these mass transfer processes (external or internal) can influence the rates of

reaction. To determine the kinetic parameters, it is essential to ensure the absence of

mass transfer limitations. In the kinetically controlled reaction regime, the conversion

of the reactant should not depend on the total gas flow rate for a fixed value of the

W/FAo ratio. This effect was experimentally studied and the results are represented in

Table 5.3. Acetol concentration in feed was 10 wt%, while the catalyst loading was

0.1 g. It was found that there was practically no change in the acetol conversion at

773 K, while varying the N2 flow rate from 50 to 200 cm3/min, and hence, the total

volumetric flow rate (Q0) from 1402 to 1552 cm3/min at a W/FAo ratio equal to 1.16

g h/mol. Thus, it was established that the external mass transfer resistance was absent

over the entire temperature range studied. To ascertain the absence of pore diffusion

limitation, the effect of catalyst particle size on conversion was studied at 773 K

using W/FAO ratio equal to 1.16 g h/mol and N2 flow rate equal to 100 cm3/min. The

total gas flow rate was 1452 cm3

The acetol disappearance rate was described by using the following correlation:

/min. It was found that the conversion obtained while

using larger particles in the size range 150–210 μm (57.4%) was almost identical to

that obtained using particles in the range 23–69 μm (59.5%). Therefore, it was

concluded that the intra-particle diffusion resistance was negligible.



r = dXd(W FA o⁄ )

(5.10)

Using values of the slopes at various points of the X vs. W/FAO

To gain further insight into the reaction mechanism and kinetics, a heterogeneous

kinetic model was proposed. A single-site mechanism comprising the following steps

was considered: reversible adsorption of acetol (here denoted as A) on the active site

(S) of the catalyst, reaction between adsorbed acetol and water (denoted as B) to form

the complex ABS, decomposition of ABS into intermediates and further reaction to

give products, CO

curves, the

experimental rates were found. Analysis of the experimental data using a power law

model suggested that the reaction order with respect to acetol was equal to 0.85, 0.82,

0.78 and 0.88 at 623, 673, 723 and 773 K (see Fig. 5.7). Power law kinetics of

methanol, ethanol and glycerol was earlier considered by several other researchers

(Akhande et al., 2006; Adhikari et al., 2009; Morgenstern et al., 2005; Sutar et al.,

2007; Idem and Bakhshi, 2009; Vaidya and Rodrigues, 2006).

2 and H2

A + S k1,k−1�⎯⎯� AS (5.11)

. The above-mentioned elementary steps in this

mechanism are represented as:

AS + Bk2→ ABS (5.12)

ABS ks→ intermediates

k4→ CO2 + H2 (5.13)

when surface reaction between A and B (viz. Eq. 5.12) is rate controlling and

product adsorption is weak, the reaction rate is given by:

r = k2KA pA pB1+KA pA

= kpA1+KA pA

(5.14)

where k = 2KApB . Since water was in large excess and pressure was constant, pB

was assumed to be nearly equal to pB0. Eq. (5.14) was linearized by plotting 1/r vs.

1/pA and this plot yielded a satisfactory relationship at all temperatures studied (see

Fig. 5.9). From the slopes and intercepts, the values of k and KA at various

temperatures were estimated (see Table 5.4). A comparison of the predicted and

experimental rates is presented in Fig. 5.10. Thus, the proposed model could

adequately describe reaction kinetics. There are few examples where the above-

mentioned single-site mechanism has been satisfactorily used, e.g., for describing H2

production from ethanol Byrd et al., 2007; Vaidya and Rodrigues, 2006 and glycerol



Byrd et al., 2008 using Ru/Al2O3

r = k2KA KB pA pB(1+KA pA +KB pB )2 =

k′ p A

�1+Kp A �2 (5.15)

. Contrary to our proposition that Eq. 5.12 is the

rate-limiting step, these researchers considered decomposition of the complex ABS

(see Eq. 5.13) as rate determining and developed a different kinetic model; even so,

the rate expressions in their works could be simplified to a form similar to that of Eq.

(5.14). Therefore, it is necessary to exercise caution while deducing the rate-limiting

step in the acetol reforming process. When a similar mechanism suggesting the

adsorption of both A and B was considered and surface reaction was assumed to be

the rate-determining step, the following rate expression was obtained:

where k’ and K are two new parameters introduced in Eq. (5.15), and pB = pB0

. Eq.

(5.15) was rearranged to the form, �𝑘𝑘′𝑝𝑝𝐴𝐴 𝑟𝑟⁄ = (1 + 𝐾𝐾𝑝𝑝𝐴𝐴 ) . We found that the

plots �𝑝𝑝𝐴𝐴 𝑟𝑟⁄ vs 𝑝𝑝𝐴𝐴were nonlinear and did not yield any satisfactory relationship at all

temperatures; therefore, this mechanism was rejected.

5.4. Conclusions

Acetol is a major component of the ketonic fraction in pyrolysis oil. In this work,

the kinetics of the acetol steam reforming reaction was investigated in a fixed-bed

reactor over wide ranges in temperature (623-773 K), acetol concentration (5–20 wt%)

and W/FAo ratio (0.93–4.66 g h/mol) using a commercial 5% Pt/C catalyst. It was

found that high temperature and space time facilitate H2 production from acetol.

Maximum acetol conversion (81%) and H2 yield (0.66 mol/mol) were achieved at

773 K at a W/FAO

ratio equal to 4.66 g h/mol. The absence of mass transfer

resistances (external and internal) was ascertained. Kinetic data were described by a

single-site heterogeneous kinetic model, which assumes that the surface reaction

between adsorbed acetol and water in the gas phase is rate-controlling.



Figure 5.1: SEM image of the unused Pt/C catalyst.



Figure 5.2: TEM image of the unused Pt/C catalyst.



Figure 5.3: XRD patterns of the unused and used Pt/C catalysts.



Figure 5.4: Effect of the W/Qo

ratio on acetol conversion at 623, 673, 723 and 773 K.



Figure 5.5: Effect of temperature on the H2

yield.



Figure 5.6: A plot of H2

yield vs. acetol conversion at 773 K.

0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1

H2

yiel

d (m

ol/m

ol)

Fractional acetol conversion



Figure 5.7: Effect of temperature on the CH4/H2

ratio.



Figure 5.8: Plots of ln r vs. ln pA

at various temperatures.



Figure 5.9: Plots of 1/r vs. 1/pA

at various temperatures.



Figure 5.10: A comparison of the predicted (Eq. 8) and experimental rates.

0.1

0.2

0.3

0.4

0.1 0.2 0.3 0.4

Pred

icte

d ra

tes (

mol

/(h g

cat))

Experimental rates (mol/(h gcat))



Table 5.1: Kinetic data at various temperatures range from 623-773 K

over Pt/C catalyst

Temp.

(K)

Acetol

conversion

(%)

H2

(mol/mol)

yield Q

(cmo

3

W/F

/min) (g h/mol) AO Mole

fraction of

acetol in

feed

Mole

fraction of

water in

feed

623 18.9 0.14 1790.2 0.93 0.025 0.919

623 27.3 0.25 1452.1 1.16 0.024 0.906

623 39.2 0.27 1114.1 1.55 0.023 0.886

623 55.3 0.37 776.1 2.33 0.022 0.848

623 62.3 0.46 438.0 4.66 0.020 0.751

673 27.3 0.20 1790.2 0.93 0.025 0.919

673 36.4 0.27 1452.1 1.16 0.024 0.906

673 45.5 0.34 1114.1 1.55 0.023 0.886

673 63.0 0.43 776.1 2.33 0.022 0.848

673 72.1 0.52 438.0 4.66 0.020 0.751

723 32.2 0.23 1790.2 0.93 0.025 0.919

723 41.3 0.29 1452.1 1.16 0.024 0.906

723 53.2 0.36 1114.1 1.55 0.023 0.886

723 67.9 0.46 776.1 2.33 0.022 0.848

723 76.9 0.56 438.0 4.66 0.020 0.751

773 43.4 0.29 1790.2 0.93 0.025 0.919

773 59.5 0.37 1452.1 1.16 0.024 0.906

773 61.6 0.46 1114.1 1.55 0.023 0.886

773 75.5 0.54 776.1 2.33 0.022 0.848

773 81.1 0.66 438.0 4.66 0.020 0.751



Table 5.2: Effect of acetol concentration in feed on acetol conversion and H2

yield

Temp.

(K)

S/C W/FAo

(g h/mol)

Acetol conversion

(%)

H2

(mol/mol)

yield

623 26 2.39 35.5 0.24

623 12.33 1.16 27.3 0.25

623 7.67 0.77 21.6 0.18

623 5.33 0.58 19.3 0.14

673 26 2.39 41.2 0.27

673 12.33 1.16 36.4 0.27

673 7.67 0.77 29.9 0.23

673 5.33 0.58 24.8 0.17

723 26 2.39 48.4 0.35

723 12.33 1.16 41.3 0.29

723 7.67 0.77 37.3 0.25

723 5.33 0.58 35.2 0.22

773 26 2.39 61.4 0.41

773 12.33 1.16 59.5 0.37

773 7.67 0.77 47.7 0.36

773 5.33 0.58 40.0 0.34



Table 5.3: Effect of total gas flowrate on acetol conversion at 773 K

N2

(cm

flowrate

3

Q

/min)

o

(cm

3

Acetol Conversion

/min) (%)

50 1402 58.5

100 1452 59.5

150 1502 60.9

200 1552 61.9



Table 5.4: Model parameters at various temperatures (cf Eq. 5.14)

Temperature

(K)

k

(mol/(g h atm))

KA

(1/atm)

623 15.9 11.1

673 21.7 17.1

723 27.8 23.1

773 32.3 6.0

kinetics of steam reforming of acetol over pt/c...

Documents