Noncatalytic gasification of isooctane in supercritical water: A Strategy for high-yield hydrogen production

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 6

Avai lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ier . com/ loca te /he

Noncatalytic gasification of isooctane in supercritical water:A Strategy for high-yield hydrogen production

Ratna F. Susanti a,b, Agung Nugroho a,b, Jihye Lee a, Yunje Kim a, Jaehoon Kim a,b,*aClean Energy Center, Energy Division, Korea Institute of Science and Technology (KIST), 39-1 Hawolgok-dong, Seoungbuk-gu,

Seoul 136-791, Republic of KoreabDepartment of Clean Energy and Chemical Engineering, University of Science and Technology (UST), 113 Gwahangno, Yuseong-gu,

Daejeon 305-333, Republic of Korea

a r t i c l e i n f o

Article history:

Received 13 August 2010

Received in revised form

11 December 2010

Accepted 19 December 2010

Available online 22 January 2011

Keywords:

Hydrogen production

Supercritical water gasification

Haynes� 230� alloy

Isooctane

* Corresponding author. Clean Energy CenteSeoungbuk-gu, Seoul 136-791, Republic of K

E-mail address: [email protected] (J.0360-3199/$ e see front matter Crown Copyri

doi:10.1016/j.ijhydene.2010.12.095

a b s t r a c t

Continuous supercritical water gasification of isooctane, a model gasoline compound, is

investigated using an updraft gasification system. A new reactor material, Haynes� 230�

alloy, is employed to run gasification reactions at high temperature and pressure

(763 � 2 �C; 25 MPa). A large-volume reactor is used (170 mL) to enable the gasification to be

run at a long residence time, up to 120 s. Various gasification experiments are performed by

changing the residence time (60e120 s), the isooctane concentration (6.3e14.7 wt%), and

the oxidant concentration (equivalent oxidant ratio 0e0.3). The total gas yield and the

hydrogen gas yield increase with increasing residence time. At 106 s and an isooctane

concentration of 6.3 wt%, a very high hydrogen gas yield of 12.4 mol/mol isooctane, which

is 50% of the theoretical maximum hydrogen gas yield and 92% of the equilibrium

hydrogen gas yield under the given conditions, is achieved. Under these conditions,

supercritical water partial oxidation does not increase the hydrogen gas yield significantly.

The produced gases are hydrogen (68 mol%), carbon dioxide (20 mol%), methane (9.8 mol%),

carbon monoxide (1.3 mol%), and ethane (0.9 mol%). The carbon gasification efficiency is in

the range 75e91%, depending on the oxidant concentration. A comparison of supercritical

water gasification with other conventional methods, including steam reforming, auto-

thermal reforming, and partial oxidation, is also presented.

Crown Copyright ª 2010, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All

rights reserved.

1. Introduction and lowdegreeofhydrogenbondingofwater in its supercritical

Supercritical water gasification (SCWG) has recently received

much attention as a potential alternative to conventional

reformingmethods for hydrogen production; this is because of

the unique physical properties of supercritical water [1,2].

Noncatalytic reforming reactions are possible because of the

high reactivity of supercritical water. The low dielectric

constant (2e20,dependingonthetemperatureandpressure [3])

r, Energy Division, Koreaorea. Tel.: þ82 2 958 5874Kim).ght ª 2010, Hydrogen Ene

state can lead to high solubilities of hydrocarbon feeds. The

produced gases are also soluble in supercritical water. Thus,

a single-phase reforming reaction can be carried out in super-

criticalwater. Thehigh density, high thermal conductivity, and

highheat/mass transfer associatedwith supercriticalwater are

beneficial in developing a compact reformer system.

Over the last ten years, there has been considerable

interest in the use of biomass as a renewable energy source.

Institute of Science and Technology (KIST), 39-1 Hawolgok-dong,; fax: þ82 2 958 5205.

rgy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 63896

Many studies have demonstrated that supercritical water is

an excellent reaction medium for producing hydrogen from

biomass and its model compounds [4e8]. In the presence of

homogeneous or heterogeneous catalysts, high hydrogen gas

yields, up to the theoretical value, can be achieved. Hydrogen

produced from biomass may be used in the future in renew-

able energy sources. For a seamless transition from a fossil

fuel economy to a hydrogen economy, it may be necessary to

explore hydrogen production from fossil oil sources. In spite

of the advantages of SCWG, only a few studies have focused

on hydrogen production from long-chain hydrocarbons (>C6)

using SCWG [9e11]. It is more difficult to reform longer-chain

hydrocarbons because many competing reactions take place,

so, in some cases, catalysts are needed to achieve high

hydrogen gas yields [10,11]. In such cases, the catalyst deac-

tivation associated with the sulfur compounds present in

fossils fuels [12] and coke/tar formation [2] can cause major

problems. Development of a noncatalytic SCWG system that

can generate high hydrogen gas yields is therefore highly

desirable. Our group has worked on noncatalytic SCWG of

isooctane (C8H18) as a model gasoline compound [9,13]. The

first work was carried out using a downdraft tubular reactor

system, made from Hastelloy� C-276, at temperatures of

593e694 �C, residence times of 8e16 s, concentrations of

15e23 wt%, and a pressure of 25 MPa [13]. Hydrogen peroxide

(H2O2) was used as an oxidant source to enhance the hydrogen

gas yield by running partial oxidation in supercritical water.

The maximum hydrogen yield achieved using the downdraft

reactorconfigurationwasvery low:1.68molH2/mol isooctaneat

664 �C, 25 MPa, a residence time of 15 s, an isooctane concen-

tration of 22 wt%, and a H2O2 concentration of 1560.2 mmol/L.

The experimentally obtained maximum hydrogen gas yield

corresponds to only 7% of the theoretical maximum hydrogen

gas yield (24.2 mol/mol isooctane). In our second study of non-

catalytic SCWG of isooctane, a significant modification of the

gasification system was made to enhance the hydrogen gas

yield. This included tilting of the gasification reactor to 75� fromthe vertical position, positioning the reactor inlet at the bottom

and the reactor outlet at the top (updraft configuration), and

installing an insulation zone and a cooling zone [9]. A higher

hydrogengasyieldof6.13molH2/mol isooctanewasachievedat

637 �C, 25MPa, 18 s, an isooctane concentration of 9.9 wt%, and

a H2O2 concentration of 2701.1 mmol/L. Under similar gasifica-

tion conditions, the updraft configuration resulted in a higher

hydrogen gas yield, approximately four times higher than that

obtained with the downdraft gasification system. The gasifica-

tion efficiency also increased dramatically. However, the

maximum hydrogen yield of 6.13 mol H2/mol isooctane is still

low. This value only corresponds to 28% of the theoretical

maximum yield of hydrogen (21.9 mol/mol isooctane). Even

though previouswork suggests that higher hydrogen gas yields

can be achieved at higher reaction temperatures and/or longer

residence times, it was not possible to explore noncatalytic

SCWG further because the temperature was limited to 640 �C,and because the reactor volume was small.

In this paper, we demonstrate that a much higher

hydrogen gas yield of 12.6 mol H2/mol isooctane can be ach-

ieved, in the presence of a small amount of oxidant, using the

updraft gasification configuration. This value corresponds to

56% of the theoretical maximum (22.4 mol/mol isooctane), or

to 98% of the equilibrium hydrogen gas yield, under given

conditions. A new SCWG reactor was developed to enable the

investigation to be carried out at higher reaction temperatures

and longer residence times. The reactormaterial was changed

from Hastelloy� C-276 to Haynes� 230� alloy, to enable the

SCWG to be run at a temperature of 763 �C and a pressure of

25 MPa. In addition, the reactor volume was enlarged 5.7-fold

to examine long residence time conditions. The Haynes� 230�

alloy retains a higher strength and has a longer lifetime at

elevated temperatures and pressures than those achieved by

other alloy materials, including Hastelloy� C-276 [14e16]. The

use of the Haynes� 230� alloy for SCWG was pioneered by Lee

et al. [10,17]. The following sections describe the gasification

apparatus and process, and the effects of temperature, feed

concentration, residence time, and oxidant concentration on

the gas yields, gas compositions, and total content of polar

compounds in the liquid effluent. A comparison of the

hydrogen gas yields obtained by steam reforming, auto-

thermal reforming, and partial oxidation of isooctane with the

yields obtained by noncatalytic SCWG of isooctane will also be

presented.

2. Experimental method

2.1. Materials and analysis

Details of the materials and analytical tools (gas chromato-

graphy (GC) for gas analysis and total organic carbon (TOC) for

liquid analysis) were given in our previouswork in Ref. [9]. The

compositions of liquid products were analyzed using a GC

Agilent 6890N (Agilent, Palo Alto, CA, USA)/TOF-MS LECO

Pegasus III (LECO, St. Joseph, MI, USA) with high-purity helium

(>99.9999%) as the carrier gas. The column was an Ultra-2 (5%

phenylmethylsiloxane) of length of 25 m, inner diameter (ID)

0.2 mm, and thickness 0.25 mm. The temperature of the

column was first maintained at 100 �C for 1 min, then ramped

up to 300 �C at a rate of 20 �C/min, and held at 300 �C for 7min.

The injector port temperature was 280 �C and the injection

volume was 2 mL.

2.2. Apparatus and procedure

The supercritical water gasification was carried out using

a custom-built, continuous-flow tubular reactor system. A

schematic diagram of the gasification apparatus is shown in

Fig. 1. The apparatus consists of a tubular reactor (R); a water

pre-heater (WP); a feed pre-heater (FP); several heat furnaces,

including pre-heater furnaces (PF); a mixing part furnace (MF);

a reactor furnace (RF); an insulation part furnace (IF); three

high-pressure pumps (P-01, P-02, and P-03); cooling units (CT

and C-01); a metal filter (F); a back-pressure regulator (BPR);

a gaseliquid separator (S); a container for liquids (L); feed

tanks (T-01, T-02, T-03, and T-04); isolation and safety valves

(V1, V2, V3, and V4), pressure gauges (P); and thermocouples

(T). The details of the gasification apparatus and the gasifica-

tion procedure were described in our previous work in Ref. [9].

Only modifications of the previous system will be described

here. The reactor (R) was tubular and made from Haynes�

230� alloy. The size of the reactor was ID 2 cm and length

Fig. 1 e Schematic diagram of supercritical water gasification apparatus. R, reactor; M, mixing part; MF, mixing furnace; RF,

reactor furnace; WP, water pre-heater; FP, feed pre-heater; PF, pre-heater furnaces; IF, insulation furnace; RZ, reaction zone;

IZ, insulation zone; CT, cooling tube; C, condenser; F, metal filter; BPR, back-pressure regulator; P-01, high-pressure

isooctane pump; P-02 and P-03, high-pressure water pumps; S, gaseliquid separator; P, pressure gauge; T, thermocouple;

V1, needle valve; V2, relief valve; V3, three-way valve; V4, safety valve; T-01, isooctane feed tank; T-02 and T-03, DDI water

feed tanks; T-04, oxidant feed tank; L, container for liquid; GC, gas chromatograph; WG, wet gas meter.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 6 3897

54 cm, giving an effective volume of 170 mL. All the cooling

work was done by an air-cooled tube (CT) and C-01. The

condenser (C-01) was a shell and tube type heat exchanger

consisting of seven tubes enclosed in a cylindrical shell. The

shell had an outer diameter (OD) of 6 cm, an ID of 3.55 cm, and

a length of 24 cm. Each tube had an OD of 0.95 cm and an ID of

0.5 cm. The high-pressure pumps P-01 and P-02 were model

HKS-600 digital metering pumps, manufactured by Hanyang

Accuracy (Seoul, Korea). P-01 was used to deliver isooctane at

a maximum flow rate of 162 mL/h, and P-02 was used to

deliver water to replace isooctane feed during heating and

cooling of the system. The flow-rate range of the water pump

was 60e600 mL/h. Pump P-03 was a model HKS-3000 high-

pressure digital metering pump (Hanyang Accuracy) with

a maximum flow rate of 3000 mL/h; it was used as the main

distilled de-ionized (DDI) water pump.

In this paper, total gas yield is defined as the volume (L) of

total gaseous products at room temperature (25 �C) and

pressure (0.1 MPa) per weight (g) of feed, estimated by real-

time volumetric flow-rate measurements and feed concen-

trations. Carbon gasification efficiency (CE) is defined as the

total number of moles of carbon in the produced gases per

total number ofmoles of carbon in the feed. The individual gas

yields in the produced gases are defined as the number of

moles of each produced gas per mole of isooctane fed to the

reactor. The number of moles of each produced gas was

estimated by real-time volumetric flow-rate measurements

and the gas compositions as determined by GC. The dry gas

composition is defined as the mole per cent of each produced

gas in the sample. Oxygen (O2), which may be present in the

gaseous product when the oxidant effect is explored, and

nitrogen (N2), which was used to pressurize the feed tanks T-

01 and T-03, were not included in the number ofmoles of each

gas produced by the gasification. The oxygen equivalent ratio

(ER) is defined as the amount of oxidant added per oxidant

required for complete oxidation. For example, an ER of 1

indicates that the amount of oxidant is equivalent to complete

oxidation estimated by stoichiometry calculations. The stan-

dard error (SD of the mean) is used to describe statistically the

uncertainties in the measurements from four samplings. In

the experimental results, the standard errors are shown as

error bars.

3. Results and discussion

3.1. Chemical reactions

Supercritical water gasification of organic compounds is

a very complex process and various chemical reactions can

take place. The major chemical reactions can include steam

reforming, the wateregas shift reaction, pyrolysis, partial

oxidation if an oxidant is present, andmethanation. Themost

important reaction in SCWG is steam reforming. The steam

reforming reaction of isooctane is

C8H18 þ 8H2O 4 8CO þ 17H2 ΔH298 K ¼ 1274.47 kJ/mol (1)

The large and positive value of the heat of enthalpy indi-

cates that steamreformingof isooctane is highly endothermic.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 63898

A higher reaction temperature can therefore enhance steam

reforming, leading to higher yields of hydrogen gas.

Since the steam reforming reaction produces carbon

monoxide, and a large amount of water exists under typical

SCWG conditions, the wateregas shift reaction can take place.

CO þ H2O 4 CO2 þ H2 ΔH298 K ¼ �41.15 kJ/mol (2)

This is a highly desirable reaction because the forward

reaction can increase the hydrogen gas yield. The wateregas

shift reaction is exothermic and the forward reaction is

thermodynamically favored at temperatures lower than

815 �C [18].

When the steam reforming reaction and the wateregas

shift reaction are combined, the overall reaction is

C8H18 þ 16H2O 4 8CO2 þ 25H2 ΔH298 K ¼ 945.27 kJ/mol (3)

If all the isooctane is steam reformed, and all the CO

generated by steam reforming is consumed by the wateregas

shift reaction, the maximum hydrogen gas yield that can

theoretically be obtained from isooctane is 25 mol/mol

isooctane. Any of the other competing reactions, including

pyrolysis, partial oxidation, and methanation, that can occur

during SCWG would lead to a decrease in the hydrogen gas

yield. The combined reaction for the maximum hydrogen gas

yield is highly endothermic, so a high reaction temperature

may be favorable in enhancing the hydrogen gas yield.

Pyrolysis (or cracking) can take place during the SCWG of

isooctane. Complete pyrolysis produces hydrogen and solid

compounds such as coke and tar. In the course of pyrolysis,

cracked light gaseous compounds such as methane and

ethane can be formed.

C8H18 4 8C þ 9H2 ΔH298 K ¼ 224.1 kJ/mol (4)

C8H18 4 CxHy þ zH2 (5)

Pyrolysis is also endothermic, but the heat of enthalpy is

much smaller than that of steam reforming. When an oxidant

is used in SCWG, partial oxidation or total oxidation can take

place, depending on the amount of oxidant fed into the

system.

Partial oxidation:

C8H18 þ 4O2 4 8CO þ 9H2 ΔH298 K ¼ �748.13 kJ/mol (6)

Total oxidation:

C8H18 þ 25/2O2 4 8CO2 þ 9H2O

ΔH298 K ¼ �5100.53 kJ/mol (7)

The partial oxidation reaction is desirable because it can

generate hydrogen and carbon monoxide. The generated

carbon monoxide can react with water, yielding hydrogen by

the wateregas shift reaction. In addition, the exothermic

oxidation reaction can help to reduce coke or tar formation by

rapid internal heating of the reactants and reduce the external

energy required for the steam reforming reaction [19,20]. The

rapid internal heating by the partial oxidation can often lead to

an increase in thegasificationreaction.Note that themaximum

hydrogen yield that can be produced by the combined partial

oxidation and wateregas shift reactions is smaller than that of

the combined steam reforming and wateregas shift reactions.

In fact, the theoretical hydrogen gas yield in the presence of an

oxidant can be calculated as below:

C8H18 þ xO2 þ (16 � 2x)H2O 4 8CO2 þ (25 � 2x)H2 (8)

Thus the higher the amount of oxidant added, the lower

the theoretical hydrogen gas yield. For example, the

maximum hydrogen yield of isooctane is estimated to be 22.4

at ER 0.1, 19.8 at ER 0.2, and 17.5 at ER 0.3. At large values of ER,

the total oxidation can be a dominant reaction pathway,

producing the undesirable compounds carbon dioxide and

water.

Methane can be produced during SCWG by reaction of

carbon monoxide, carbon dioxide, or carbon with hydrogen.

Methanation of CO:

CO þ 3H2 4 CH4 þ H2O ΔH298 K ¼ �206.17 kJ/mol (9)

Methanation of CO2:

CO2 þ 4H2 4 CH4 þ 2H2O ΔH298 K ¼ �165.01 kJ/mol (10)

Methanation of C: C þ 2H2 4 CH4 ΔH298 K ¼ �74.87 kJ/mol

(11)

Methanation is undesirable because it consumes hydrogen

produced by SCWG and decreases the hydrogen gas yield.

Methane formation increases with decreasing temperature or

increasing amounts of reactants (H2, CO, CO2, or C). A high

reaction temperature is therefore favorable as it avoids

methanation. The methanation of CO and CO2 can be cata-

lyzed by ruthenium, iridium, rhodium, nickel, cobalt, osmium,

platinum, iron, palladium (group VIII), and molybdenum

(group VI), or silver (group I) [21,22]. Supercritical water reactor

materials such as Hastelloy� C-276, Inconel�, and Haynes�

230� contain nickel as the major component. It is well known

that nickel is a good catalyst for methanation [21] and for

steam reforming [23], suggesting that the presence of nickel in

the reactor wall can catalyze methanation or steam reform-

ing. Wall effects in SCWG have been studied in Refs. [24,25].

3.2. Equilibrium calculation

The equilibrium composition of isooctane gasification under

supercritical water conditions was calculated using Gibbs free

energy minimization and the PengeRobinson equation of

state. The calculation indicates that the formation of solid

carbon is not observed and that C2eC4 gases are formed in

negligible quantities (10�4e10�10 mol/mol isooctane), so these

are not discussed in the results. At feed concentrations of less

than 15 wt%, the isooctane is consumed completely. Trace

amounts of isooctane are found in the output at feed

concentrations of 15e30 wt%.

Fig. 2a shows theeffect of temperatureon thecalculatedgas

yield at a pressure of 25MPa and an isooctane concentration of

6.2wt%.An increase ingasification temperature leads toa large

Fig. 2 e Calculated equilibrium gas yield (a) as a function of

temperature at a fixed isooctane concentration of 6.2 wt%,

and (b) as a function of concentration at a fixed

temperature of 764 �C. The pressure is 25 MPa.

Fig. 3 e Calculated equilibrium (a) hydrogen gas yield,

(b) carbon dioxide gas yield, and (c) methane gas yield as

a function of ER at various temperatures, 25 MPa, and an

isooctane concentration of 6.2 wt%.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 6 3899

increase in hydrogen formation, a drastic decrease inmethane

formation, and a slight increase in carbon monoxide forma-

tion. The carbon dioxide gas yield increases with increasing

temperature, but the increase is not as large as that of

hydrogen. At a temperature of 763 �C and an isooctane

concentration of 6.2 wt% (the experimental gasification

conditions explored in this study), the equilibrium hydrogen

gas yield is estimated to be 13.6 mol/mol isooctane. The

concentration dependence of the effluent gas yield composi-

tionwas calculated at a temperature of 763 �Candapressureof

25 MPa, and the results are shown in Fig. 2b. The change in gas

composition is obvious at low feed concentrations. The

hydrogen yield decreases dramatically and the methane yield

increasesslightlyas the feedconcentration increases from1wt

% to 15 wt%. The theoretical hydrogen gas yield of 25 mol/mol

isooctane can be achieved when a dilute feed concentration of

w1 wt% isooctane is used at 763 �C and 25 MPa.

Fig. 3 shows the effects of oxygen concentration on the

equilibrium gas yield at 25 MPa and 6.2 wt% isooctane at

various temperatures. It can be seen that as the amount of

oxidant increases, the hydrogen and methane gas yields

decrease, but the carbon dioxide gas yield increases. The

effect of ER on the hydrogen gas yield is more obvious at high

temperatures.

3.3. Effects of residence time

Fig. 4 shows the effects of residence time on the total gas yield,

carbon gasification efficiency (CE), total organic carbon (TOC),

individual gas yield, and dry gas composition at a temperature

of 763 � 2 �C, a pressure of 25 MPa, and an isooctane

concentration of 6.3 � 0.3 wt%. A reaction temperature of

763� 2 �Cwas themaximum temperature achievable with the

current gasification system. The effects of residence time

Fig. 4 e Effects of residence time on (a) total gas yield, total

organic carbon (TOC), and carbon gasification efficiency

(CE); (b) individual gas yield composition; (c) dry gas

composition; and (d) liquid product color. Gasification was

conducted at 25 MPa, 763 ± 2 �C and isooctane

concentration of 6.3 ± 0.3 wt%. (d) shows photographs of

liquid product samples.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 63900

were studied by varying the total flow rate of the feeds that

were charged to the reactor. The residence time was esti-

mated from

s ¼ VRrR

y0r0

where VR is the effective reactor volume (cm3), y0 is the volu-

metric flow-rate of the liquid fed to the reactor (cm3/min), r0 is

the density of the liquid feed (g/cm3), and rR is the density of

the fluid at the experimental temperature and pressure

(g/cm3). As shown in Fig. 4a, as the residence time increased

from 60 s to 106 s, the gas yield increased from 3.1 L/g to 4.2 L/g

isooctane. A further increase in residence time to 120 s

resulted in a slight decrease in the gas yield to 3.9 L/g. A

decrease in the total organic carbon (TOC) value from 55 ppm

to 17 ppm was observed when the residence time was

increased from 60 s to 91 s. The TOC value at longer residence

times of 106e120 s showed negligible changes. The carbon

gasification efficiency (CE) was in the range 75e80% and did

not changemuch with residence time. It is noteworthy that in

this study, higher gas yields, much lower TOC values, and

much higher CE values were obtained than in previous gasi-

fication studies at shorter residence times (<40 s) and a lower

reaction temperature (632 �C) [9]. The total gas yields in the

previous work were 1.1e2.9 L/g isooctane, the TOC values

were 198e2790 ppm, and the CE values were 26.6e52.5%,

depending on the residence time.

As shown in Fig. 4d, the liquid was yellowish for a resi-

dence time of 60 s, but the liquid effluents were transparent at

longer residence times. This agrees well with the changes in

the TOC value with residence time. It was shown in our

previous work that the liquid product was yellow with a thin

layer of dark brown viscous oil on the surface when isooctane

was gasified in supercritical water at a lower temperature of

632 �C and shorter residence time of 6 s in Ref. [9]. The liquid

effluent produced at the residence time of 6 s in the previous

study, which consisted of a translucent water-dispersible

product and an oily product on the water surface, was not

observed in this study. This indicates that higher tempera-

tures and longer residence times are beneficial for higher

isooctane conversion. At least eight compoundswere found in

the liquid products by GC/TOF-MS analysis, including fluo-

ranthene, 2-phenylnaphthalene, 9H-fluorene, phenanthrene,

1-methylnaphthalene, hydroxybenzene, bicyclo[4,4,0]deca-

1,3,5,7,9-pentene and cyclohexylcyclohexane, for both of the

yellowish samples at a residence time of 60 s and for the

transparent liquid at 91 s. A representative GC/TOF-MS anal-

ysis profile of the liquid phase is provided in the

Supplementary data. The increased total gas yields and

reduced TOC values at longer residence times suggest that

isooctane was converted to gaseous products instead of being

converting to liquid products. It has been reported by other

researchers that longer residence times decreased the

concentration of organic acids and 2,5-hexanedione in the

liquid product in the gasification of glucose at 650 �C, sug-gesting conversion of the organic species to gaseous products

[26]. From carbon balance analysis, it was estimated that

approximately 17e24% of the carbon in isooctane was con-

verted to solid products. In fact, after each reaction, a very

small amount of coke was observed in the filter and in the

cooling tube.

The dependences of the individual gas yields on the resi-

dence time are shown in Fig. 4b, and the dry gas compositions

(mol%) are shown in Fig. 4c. When the residence time was

increased from 60 s to 91 s, a significant increase in hydrogen

gas yield, from 7.7 mol/mol isooctane to 12.4 mol/mol

Fig. 5 e Effects of isooctane concentration on (a) total gas

yield, total organic carbon (TOC), and carbon gasification

efficiency (CE); (b) individual gas yield composition; and

(c) dry gas composition. The gasification was conducted at

25 MPa, 764 ± 1 �C and a residence time of 106 s.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 6 3901

isooctane, was observed, and a slight decrease to 11.8mol/mol

isooctane was seen at the longer residence time of 120 s. The

higher hydrogen gas yields at longer residence times observed

in this study agree well with the results of previous work on

the supercritical water gasification of wood sawdust, glucose,

aviation fuel, andmicroalgae in Refs. [5,7,10,27]. The hydrogen

gas yield of 12.4 mol/mol isooctane almost reached the value

of 13.6 mol/mol isooctane predicted using the equilibrium

calculation described in Section 3.2. Note that this hydrogen

gas yield of 12.4 mol/mol isooctane is comparable to the yield

obtained by steam reforming of isooctane at 900 �C with

a molybdenum carbide catalyst (w12.6 mol/mol isooctane) in

Ref. [28]. This value is much higher than the hydrogen gas

yield from autothermal reforming (ATR) of isooctane at 800 �Cwith platinum/doped ceria (w9.5 mol/mol isooctane) in

Ref. [29] and that from partial oxidation (POX) of isooctane at

750 �C with noble metal and base metal/alumina (w8.4 mol/

mol isooctane) in Ref. [30]. Details of the hydrogen gas yield

comparison will be given later. Themethane gas yield showed

the opposite trend to the hydrogen gas yield; the methane gas

yield decreased from 2.33 mol/mol isooctane to 1.78 mol/mol

isooctane with an increment in residence time from 60 s to

106 s, and increased to 1.93 mol/mol isooctane at the longer

residence time of 120 s. The decrease in the hydrogen gas yield

and the increase in the methane yield at the longer residence

timemay indicate that the hydrogen producedwas consumed

by methanation reactions (Eqs. 9e11). Byrd et al. [31] also

reported an increase in methane and a decrease in hydrogen

at higher residence times (�2 s at 700 �C and �6 s 800 �C) at221 bar in the gasification of ethanol in supercritical water. As

shown in Fig. 4b, very small amounts of carbonmonoxide and

ethane were produced. The carbonmonoxide gas yield was in

the range 0.19e0.29 mol/mol isooctane and the ethane gas

yield was in the range 0.17e0.23 mol/mol isooctane.

The gas composition analysis (Fig. 4c) showed that the

amount of hydrogen in the produced gas increased with

increasing residence time. At 106 s, the composition of the

produced gases was H2 (68.0%), CO (1.3%), CO2 (20.0%), CH4

(9.8%), and C2H6 (0.9%). In contrast to the previous results at

shorter residence times (6e33 s) and a lower temperature

(632 �C) in Ref. [9], hydrocarbon gas products with a carbon

number higher than C2 (e.g., propane, propylene, butane,

butylenes) were not detected in this work. This clearly indi-

cates that the current gasification system and the reaction

conditions were favorable for cracking of longer-chain

hydrocarbons to shorter-chain species. The cracking reaction

is an endothermic process (Eqs. 4 and 5), so higher reaction

temperatures can enhance cracking. The longer residence

time can lead to better conversion to hydrogen. In fact, the

hydrogen composition of 68.0% is much higher than the

maximum hydrogen composition (56.5%) observed in our

previous work in Ref. [9]. Note that the carbon monoxide

concentration in the gaseous products was extremely low, in

the range 87e107 ppm. These values are comparable to the

allowable carbon monoxide concentration for proton

exchange membrane fuel cell applications (in the range

10e100 ppm) in Ref. [32]. The wateregas shift reaction under

supercritical water conditions, i.e. formation of hydrogen and

carbon dioxide with the consumption of carbon monoxide

(Eq. (2)), may be responsible for the low carbon monoxide

content. The kinetics of the noncatalytic wateregas shift

reaction under supercritical conditions has been reported by

other researchers in Refs. [33e35].

3.4. Effects of feed concentration

Fig. 5 shows the effects of isooctane concentration on the total

gas yield, CE, TOC, individual gas yields, and dry gas compo-

sition at a pressure of 25 MPa, a temperature of 764 � 1 �C and

a residence time of 106 s. When the concentration was

increased from 6.3 wt% to 14.7 wt%, the total gas yield

decreased from 4.2 L/g to 2.8 L/g, CE decreased from 75.1% to

65.5%, and TOC increased from 17 ppm to 38.5 ppm. Similar

trends were observed in the supercritical water gasification of

wood sawdust [36]. The lower total gas yield and the higher

TOC value at higher concentrations are caused by a shortage

of water as a reactant. Compared to previous results at a lower

Fig. 6 e Effects of oxidant concentration on (a) total gas

yield, total organic carbon (TOC), and carbon gasification

efficiency (CE); (b) individual gas yield composition; and

(c) dry gas composition. The gasification was conducted at

25 MPa, temperature of 764 ± 1 �C, residence time of

105 ± 1 s, and isooctane concentration of 6.2 ± 0.1 wt%.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 63902

temperature of 630 �C, a shorter residence time of 18 s in Ref.

[9], and the same feed concentration (10 wt%), the carbon

efficiency increased by w74% and the TOC values were two to

three orders of magnitude smaller.

As shown in Fig. 5b, the hydrogen gas yield decreased from

12.4 mol/mol isooctane to 8.6 mol/mol isooctane, the carbon

dioxide gas yield decreased from 3.65 mol/mol isooctane to

2.32 mol/mol isooctane, and the methane gas yield increased

from 1.78 mol/mol isooctane to 2.34 mol/mol isooctane when

the feed concentration was increased from 6.3 wt% to 14.7 wt

%. The dry gas composition in Fig. 5c clearly shows that

methane formation was suppressed at high water/isooctane

ratios. This is because the excess water will shift the equilib-

rium of the methanation reaction (Eqs. 9 and 10) to the left.

The amount of CO increased from 1.3% to 1.75% and that of

C2H6 increased from 0.93% to 1.22% as the isooctane concen-

tration increased from 6.3 wt% to 14.7 wt%. Compared to our

previous work in Ref. [9], the formation of hydrocarbon

species > C2 and formation of CO were greatly suppressed at

higher temperatures and longer residence times, using the

same concentration of isooctane.

3.5. Effects of oxidant

Partial oxidation in supercritical water gasification has typi-

cally been carried out to enhance the total gas and hydrogen

gas yields [10,20,37,38]. We performed a set of experiments to

investigate the oxidant role in the hydrogen gas yield, using

the current supercritical water gasification system. Fig. 6

shows the effects of hydrogen peroxide on the total gas

yield, CE, TOC, individual gas yield, and dry gas composition

when the oxygen equivalent ratio (ER) increased from 0 to 0.3

at 25 MPa, 764 � 1 �C, a residence time of 105 � 1 s, and an

isooctane concentration of 6.2 � 0.1 wt%. In sharp contrast to

previous results in Refs. [9,13], the total gas yield did not

increase on addition of hydrogen peroxide. In fact, the total

gas yield decreased from 4.22 L/g to 3.73 L/g when the ER

increased from 0 to 0.3, indicating that with the current gasi-

fication system and experimental conditions, partial oxida-

tion did not encourage gas production. The decrease in the gas

yield at higher ER values is due to the decrease in hydrogen gas

production. Since hydrogen is the major gas in the gaseous

products (58e68%, see Fig. 6c), a decrease in hydrogen gas

yield led to a decrease in the total gas yield. The dependence of

CE on ER showed a different trend. As shown in Fig. 6a, when

the ER increased from 0 to 0.3, the CE increased significantly

from 75.1% to 90.6%. Thus, at the high ER of 0.3, onlyw8.5% of

the carbon in isooctane is converted to a solid product and

w0.9% of the carbon is converted to a liquid product. The

increase in CE with increasing ER may occur because the

oxidant helps to enhance the rate of the gasification reaction

[20], resulting in an increased conversion of isooctane to

gaseous products. Since the oxidation reaction enhanced the

production of carbon dioxide, which is the second major

compound in the gaseous products (see Fig. 6b and c), the CE

increased as the ER increased. The TOC value increased from

14 ppm to 24 ppm as the ER increased from 0 to 0.3. Oxidation

reactions of solid product such as tar/cokemay be responsible

for the increase in TOC value at higher ER. A similar trend was

also previously observed in the supercritical water oxidation

of dog food, used as amodel municipal solid waste in Ref. [39],

and supercritical water oxidation of municipal excess sludge

and alcohol distillery waste water in Ref. [40].

As shown in Fig. 6b, the hydrogen gas yield increased

slightly from12.4mol/mol isooctane to 12.6mol/mol isooctane

as the ER increased from 0 to 0.1. A further increase in ER to 0.3

resulted in adecrease in thehydrogengas yield to 10.1mol/mol

isooctane. This clearly indicates that partial oxidation did not

play an important role in enhancing hydrogen production

when the reaction temperature was high (764 � 1 �C) and the

residence timewas long (105� 1 s). In our previous papers, we

showed that the hydrogen gas yield could be increased to the

maximum value by increasing the hydrogen peroxide

concentration in the SCWG of isooctane in Refs. [9,13]. For

example, the hydrogen gas yield increased from 4.0 mol/mol

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 6 3903

isooctane to 6.13 mol/mol isooctane (an increase of w38%) at

637 �C, 25 MPa, and a residence time of 18 s, with addition of

2701.1 mmol/L of H2O2 (or ER ¼ 0.12) as the oxidant in Ref. [9].

Other groups have observed similar trends. Picou et al. [10]

reported that the hydrogen gas yield increased to 26% as the

O2/fuel ratio increased from 0 to 5 at 767� 1 �C, 24.1� 0.1 MPa,

and a space time of 38 � 1 s in the SCWG of aviation fuel. Jin

et al. [20] reported a 26% increase in hydrogen gas yield as the

ER increased from 0 to 0.1 in the SCWG of lignin at 600 �C,25 MPa, and a feedstock flow rate of 25 g/min. Williams et al.

[38] reported an increase in hydrogen gas yield from 0.4 wt% to

1.26wt% at 380 �C and from0.2wt% to 1.27wt% at 374 �C as the

H2O2 concentration increased from 0 wt% to 4.5 wt%. Further

oxidant increases resulted in decreases in the hydrogen gas

yield. At least two possible reasonsmay be responsible for the

enhancements of the hydrogen gas yield as a result of adding

oxidants seen in the previous studies. Low reaction tempera-

tures (<600 �C) and/or short residence times often result in

incomplete gasification, leading to formation of unreacted

hydrocarbon species with carbon numbers higher than two

(>C2). In this case, partial oxidation can enhance the total

gasification rate by rapidheating of the>C2 species,which can

contribute to an increase in the hydrogen gas yield in Refs.

[20,38]. When the reaction temperature is sufficiently high,

partial oxidation can compete with pyrolysis (Eq. 4e6) and/or

can oxidize the coke produced by pyrolysis. Partial oxidation

would therefore produce more hydrogen and fewer hydrocar-

bons [10]. The decrease in the hydrogen gas yield obtained by

adding large amounts of oxidants can be attributed to more

complete oxidation to form carbon dioxide and water (Eq. (7)).

Based on the previous observations, the decrease in the

hydrogen gas yield obtained by addition of oxidant in the

current system may imply that there may be competition

betweensteamreforming (Eq. (1)) andpartial oxidation (Eq. (6)).

As discussed in the previous section, the theoretical hydrogen

yield of steam reforming is higher than that of partial oxida-

tion. Inaddition,asshowninFig. 3, thecalculatedhydrogengas

Table 1 e Comparison of hydrogen gas yields in noncatalytic sumodel compounds.

Feedstock Reactor type(Reactor material)

SCW

Methanol (CH3OH) Tubular reactor (Inconel� 600) 700 �C; 276 b

H2O/C ¼ 30

Tubular reactor (Inconel� 625) 600 �C; 250 b

Ethanol (C2H5OH) Fixed bed tubular

reactor (Inconel� 600)

800 �C; 221 b

Glycine (C2H5NO2) Tubular flow (Hastelloy� C-276) 500 �C; 25 M

Glycerol (C3H8O3) Tubular flow (Hastelloy� C-276) 500 �C; 25 M

Tubular reactor (Inconel� 600) 650 �C; 25 ba

Glucose (C6H12O6) Tubing reactor (Hastelloy� C-276) 740 �C; 28 M

Tubular reactor (Inconel� 600) 700 �C; 25 M

Isooctane (C8H18) Inclined tubular

reactor (Haynes� 230�)

764.7 �C; 25 M

neDecane (C10H22) Continuous flow

reactor (no information)

550 �C; 25 M

Aviation

fuel (�C12H26)

Tubular reactor (Haynes� 230�) 767 � 1 �C; 26.25 wt%; 15

yield decreases with increasing ER. Thus, when steam

reforming is dominant at a sufficiently high reaction temper-

ature and/or with a sufficiently long residence time, partial

oxidationmaysuppress thehydrogengasyield.A similar trend

wasobserved in theSCWGofaviationfuel inRef. [10].Whenthe

O2/fuel ratiowas increased, thehydrogengasyield increasedat

shorter residence times (38� 1 s and 75� 2 s), but thehydrogen

gas yield decreased at a longer residence time of 151 � 5 s at

767� 1 �C, 24.1� 0.1MPa, and an aviation fuel concentration of

6.25 wt%. It may therefore be unnecessary to use partial

oxidation reactions to enhance hydrogen gas yieldswhenhigh

temperatures, long residence times, and low feedstock

concentrations are used. However, note that partial oxidation

canreduce theenergyrequired for thegasificationreactionand

can reduce coke/tar formation. The energybalanceand reactor

plugging by coke/tar should therefore be taken into account in

determining the optimal amount of oxidant in SCWG.

The maximum hydrogen gas yield obtained in our study

was w12.6 mol/mol isooctane, which is w56% of the theo-

retical maximum and 98% of the equilibrium hydrogen gas

yield for an ER of 0.1. The maximum hydrogen gas yields

obtained by noncatalytic SCWG using various feedstocks are

compared in Table 1. When methanol was used as a low

carbon number feedstock, 85% of the theoretical maximum

hydrogen gas yield was achieved at 700 �C. When medium

carbon number species (C2eC6) such as ethanol, glycerol, and

glucosewere used, 11e73% of the theoretical maximumswere

obtained at temperatures of 500e740 �C.Much lower hydrogen

gas yields, 0.16e14.8% of the theoretical maximums, were

observed with higher carbon number feedstocks (C10eC12) at

temperatures of 550e767 �C. The yield obtained in this study,

>50% of the theoretical maximum hydrogen yield, was

therefore very high compared to the yields obtained in

previous studies with high carbon number feedstocks.

As shown in Fig. 6b, the methane gas yield decreased from

1.78 mol/mol isooctane to 1.49 mol/mol isooctane, and the

ethane gas yield decreased from 0.17 mol/mol isooctane to

percritical water gasification using hydrocarbon resources/

G conditions Maximum hydrogenyield in mol/mol feed

Ref.

Theoretical Experimental

ar; feed ¼ 1 mL/min, 3 w2.5 [25]

ar; 26 wt%; 4 s 3 w2.55 [41]

ar; 10 wt%; 4 s 6 3 [31]

Pa; 1 wt%; 0.98 min 4.5 w0.5 [42]

Pa; 1 wt%; 0.98 min 7 5.08 [42]

r; 10 wt%; 5 s 7 w2.6 [27]

Pa; 0.6 M glucose; 30 s 12 w7 [43]

Pa; 1 wt%; 4 s 12 w7.2 [5]

Pa; 6.3 wt%; 106 s 25 12.4 This

study

Pa; 20 vol%, 10 s 31 w0.05 [11]

4.1 � 0.1 MPa;

1 � 5 s

36 w5.33 [10]

Table 2 e Comparison of hydrogen gas yields by reforming of isooctane using various methods.

Reformingmethod

Reactor type Condition Catalyst Oxidant TheoreticalH2 yield

H2 yieldobtained

Ref.

Supercritical

water

gasification

Inclined

tubular reactor

764.7 �C, 25 MPa, s ¼ 106 s,

6.3 wt%

No No 25 12.4 This

study

763 �C, 25 MPa, s ¼ 106 s,

6.2 wt%

ER ¼ 0.1 22.4 12.6

Steam reforming Fixed bed tubular

(quartz) reactor

H2O/C ¼ 1.3; T ¼ 1000 �C Molybdenum

carbide (Mo2C)

No 25 w17b [44]

H2O/C ¼ 0.73; T ¼ 900 �C O2/C ¼ 0.12 24.97 w12.6b

Autothermal

reforming

NAa GHSV ¼ 15 000 h�1;

T ¼ 800 �CPlatinum/

doped ceria

O2/C8H18 ¼ 3.7 17.6 w9.5 [29]

Partial oxidation Fixed bed

single pass

T ¼ 750 �C Alumina/

noble metal

and base metal

O2/C ¼ 0.59 9 w8.4b [30]

a Not available.

b Calculated to give the same hydrogen gas yield unit using the definition provided in the literatures.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 63904

0.08 mol/mol isooctane, but the carbon dioxide gas yield

increased from 3.65 mol/mol isooctane to 5.2 mol/mol isooc-

tane, when the ER increased from 0 to 0.3. This may be

because the pyrolysis products (methane and ethane) react

with oxygen, producing carbon dioxide via the total oxidation

reaction (Eq. (7)). The oxidant can reactwith solid products (tar

and char), producing carbon dioxide via C þ O2 / CO2. The

compositions of the produced gases are illustrated in Fig. 6c.

The amount of hydrogen in the gaseous products decreased

from 68.0% to 58.4%, carbon dioxide increased from 20.0% to

30.0%, methane decreased from 9.8% to 8.6%, carbon

monoxide increased from 1.3% to 2.5%, and ethane decreased

from 0.9% to 0.5% when the ER was increased from 0 to 3. This

gas yield trendwith changing ER is similar to that predicted by

the equilibrium calculation in Fig. 3.

3.6. Comparison with other reforming methods

Steam reforming (SR), autothermal reforming (ATR), and

partial oxidation (POX) methods have been used to reform

isooctane to hydrogen in Refs. [28e30]. Each method has

different theoretical hydrogen gas yields. The addition of

oxygen in ATR or POX can drive the endothermic reforming

reaction; this is beneficial in lowering the reforming energy

consumption. In contrast, the presence of oxygen decreases

the theoretical hydrogen gas yield that can be achieved by

isooctane reforming. The SR reaction and the ATR reaction,

assuming all carbon monoxide is converted to hydrogen by

the wateregas shift reaction, and the POX reaction can be

described as shown below.

Combination of SR and WGS reactions:

C8H18 þ 16H2O / 8CO2 þ 25H2 (12)

ATR reaction:

CnHmOp þ x(O2 þ 3.7N2) þ (2n � 2x � p)

H2O / nCO2 þ [2n � 2x e p þ (1/2m)]H2 þ 3.7xN2 (13)

POX reaction:

C8H18 þ 4O2 / 8CO þ 9H2 (14)

A comparison of the hydrogen gas yields from isooctane

reforming using the various reforming methods are given in

Table 2. SR or SCWG give the highest theoretical hydrogen gas

yields, a value of 25 mol H2/mol isooctane. The theoretical

hydrogen gas yield using ATR is dependent on the amount of

oxygen used. The lower the amount of oxygen used, the

higher the theoretical hydrogen gas yield that can be ach-

ieved. The hydrogen gas yield by SR was w17 mol/mol isooc-

tane at a very high temperature (1000 �C) in the presence of

a molybdenum carbide (Mo2C) catalyst. ATR achieved a yield

of up to w9.5 mol/mol isooctane at 800 �C using a platinum/

doped ceria catalyst. POX gave a relatively lower hydrogen gas

yield of w8.4 mol/mol isooctane in the presence of noble

metal and basemetal/alumina catalysts. Therefore, compared

to the other reforming methods, our noncatalytic SCWG

system resulted in a relatively higher hydrogen gas yield of

12.4 mol H2/mol isooctane at a much lower temperature

(764.7 �C). Note that the typical residence time of a catalytic

reformer in operation is very small. For example, the resi-

dence time of a reactor of volume 170 mL, operating at a gas

hourly space velocity of 15 000 h�1, is 0.24 s.

4. Conclusion

Continuous noncatalytic supercritical water gasification of

isooctanewas investigated using an updraft configuration and

a 170-mL gasification reactor, with Haynes� 230� alloy as the

reactor material, to run gasification at a high temperature of

763 � 2 �C. The hydrogen gas yield increased with increasing

residence time and decreasing feed concentration. At a resi-

dence time of 106 s and an isooctane concentration of 6.3 wt%,

a very high hydrogen gas yield of 12.4 mol/mol isooctane,

which corresponds to 50% of the theoretical maximum

hydrogen gas yield and 92% of the equilibrium hydrogen gas

yield, was achieved without using oxidants. Under these

conditions, the gas yield was 4.2 L/g isooctane, the carbon

gasification efficiency was 75.1%, and the TOC value of the

liquid effluent was 17 ppm. The produced gas consisted of

hydrogen (68.0 mol%), carbon dioxide (20.0 mol%), methane

(9.8 mol%), carbon monoxide (1.3 mol%), and ethane (0.9 mol

%). Gaseous products with carbon numbers higher than three

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 6 3905

(>C3) were not produced. As the oxidant concentration

increased, methane/ethane decreased and carbon dioxide/

carbonmonoxide increased, as a result of oxidation reactions.

The carbon gasification efficiency increased to 91% at the

higher oxidant concentration of ER ¼ 0.3. The noncatalytic

SCWG of isooctane achieved higher hydrogen gas yields at

lower gasification temperatures than the yield obtained by

steam reforming, autothermal reforming, and partial oxida-

tion methods.

Acknowledgment

This project is supported by the Korea Ministry of the Envi-

ronment as a “Converging technology project”. Additional

support from the Korea Research Council of Fundamental

Science and Technology (KRCF) and the Korea Institute of

Science and Technology (KIST) for the “National Agenda

Program (NAP)” is appreciated.

Appendix. Supplementary data

Supplementary data associated with this article can be found

in the online version at doi:10.1016/j.ijhydene.2010.12.095

r e f e r e n c e s

[1] Andrea K. Supercritical water gasification. Biofuels BioprodBiorefin 2008;2:415e37.

[2] Guo Y, Wang SZ, Xu DH, Gong YM, Ma HH, Tang XY. Reviewof catalytic supercritical water gasification for hydrogenproduction from biomass. Renewable Sustainable Energy Rev2010;14:334e43.

[3] Fernandez DP, Goodwin ARH, Lemmon EW, Sengers JMHL,WilliamsRC.A Formulation for the Static Permittivity ofwaterand steamat temperatures from238K to 873K at pressures upto 1200 MPa, including derivatives and DebyeeHuckelCoefficients. J Phys Chem Ref Data 1997;26:1125e66.

[4] Yanik J, Ebale S, Kruse A, Saglam M, Yuksel M. Biomassgasification in supercritical water: Part 1. Effect of the natureof biomass. Fuel 2007;86:2410e5.

[5] Byrd AJ, Pant KK, Gupta RB. Hydrogen production fromglucose using Ru/Al2O3 catalyst in supercritical water. IndEng Chem Res 2007;46:3574e9.

[6] Williams PT, Onwudili J. Subcritical and supercritical watergasification of cellulose, starch, glucose, and biomass waste.Energy Fuels 2006;20:1259e65.

[7] Guo LJ, Lu YJ, Zhang XM, Ji CM, Guan Y, Pei AX. Hydrogenproduction by biomass gasification in supercritical water:a systematic experimental and analytical study. Catal Today2007;129:275e86.

[8] Antal MJ, Allen SG, Schulman D, Xu X, Divilio RJ. Biomassgasification in supercritical water. Ind Eng Chem Res 2000;39:4040e53.

[9] Susanti RF, Veriansyah B, Kim J-D, Kim J, Lee Y-W. Continuoussupercritical water gasification of isooctane: a promisingreactor design. Int J Hydrogen Energy 2010;35:1957e70.

[10] Picou JW,Wenzel JE,LantermanHB,LeeS.Hydrogenproductionby noncatalytic autothermal reformation of aviation fuel usingsupercritical water. Energy Fuels 2009;23:6089e94.

[11] Pinkwart K, Bayha T, Lutter W, Krausa M. Gasification ofdiesel oil in supercritical water for fuel cells. J Power Sources2004;136:211e4.

[12] Gadhe JB, Gupta RB. Hydrogen production by methanolreforming in supercritical water: Catalysis by in-situ-generated copper nanoparticles. Int J Hydrogen Energy 2007;32:2374e81.

[13] Veriansyah B, Kim J, Kim JD, Lee YW. Hydrogen productionby gasification of isooctane using supercritical water. Int JGreen Energy 2008;5:322e33.

[14] Safety in the operation of laboratory reactors and pressurevessels. Moline, Illinois, USA: Parr Instrument Company;2009.

[15] Factor MJ, Lanterman HB, Wenzel JE, Lee S. Use haynes alloy230 for supercritical water reactors. In: AIChE annualmeeting. Salt Lake City, Utah, USA: American Institute ofChemical Engineers; 2007.

[16] Haynes� 230� alloy. Kokomo, IN, USA: Haynes International,Inc.; 2007.

[17] Picou J, Wenzel J, Stever M, Bouquet J, Lee S. Noncatalyticreformation of sucrose in supercritical water. In: AIChEAnnual meeting. Philadelphia, PA, USA: American Instituteof Chemical Engineers; 2008.

[18] Lee S. Alternative fuels. Washington, D C: Taylor and Francis;1996.

[19] Matsumura Y, Minowa T, Potic B, Kersten SRA, Prins W, vanSwaaij WPM, et al. Biomass gasification in near- and super-critical water: Status and prospects. Biomass Bioenergy 2005;29:269e92.

[20] Jin H, Lu Y, Guo L, Cao C, Zhang X. Hydrogen production bypartial oxidative gasification of biomass and its modelcompounds in supercritical water. Int J Hydrogen Energy2010;35:3001e10.

[21] Fischer F, Tropsch H, Dilthey P. Brennst Chem 1925;6:265.[22] Seglin L, Geosits R, Franko BR, Gruber G. Survey of

methanation chemistry and processes. In: Methanation ofsynthesis gas. , Washington, DC: American Chemical Society;1975. p. 1e30.

[23] Rostrup-Nielsen JR. Activity of nickel catalysts for steamreforming of hydrocarbons. J Catal 1973;31:173e99.

[24] Yu D, Aihara M, Antal MJ. Hydrogen production by steamreforming glucose in supercritical water. Energy Fuels 1993;7:574e7.

[25] Gadhe JB, Gupta RB. Hydrogen production by methanolreforming in supercritical water: suppression of methaneformation. Ind Eng Chem Res 2005;44:4577e85.

[26] Goodwin AK, Rorrer GL. Conversion of glucose to hydrogen-rich gas by supercritical water in a microchannel reactor. IndEng Chem Res 2008;47:4106e14.

[27] Chakinala A, Brilman D, Swaaij W, Kersten S. Catalytic andnon-catalytic supercritical water gasification of microalgaeand glycerol. Ind Eng Chem Res 2010;49:1113e22.

[28] Marin Flores OG, Ha S. Study of the performance of Mo2C foriso-octane steam reforming. Catal Today 2008;136:235e42.

[29] Krumpelt M, Krause TR, Carter JD, Kopasz JP, Ahmed S. Fuelprocessing for fuel cell systems in transportation andportable power applications. Catal Today 2002;77:3e16.

[30] Cheekatamarla PK, Finnerty CM. Synthesis gas productionvia catalytic partial oxidation reforming of liquid fuels. Int JHydrogen Energy 2008;33:5012e9.

[31] Byrd AJ, Pant KK, Gupta RB. Hydrogen production fromethanol by reforming in supercritical water using Ru/Al2O3catalyst. Energy Fuels 2007;21:3541e7.

[32] Amphlett JC, Mann RF, Peppley BA. On board hydrogenpurification for steam reformation/PEM fuel cell vehiclepower plants. Int J Hydrogen Energy 1996;21:673e8.

[33] Holgate HR, Webley PA, Tester JW, Helling RK. Carbonmonoxide oxidation in supercritical water: the effects of heat

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 3 8 9 5e3 9 0 63906

transfer and the water-gas shift reaction on observedkinetics. Energy Fuels 1992;6:586e97.

[34] Rice SF, Steeper RR, Aiken JD. Water density effects onhomogeneous water-gas shift reaction kinetics. J Phys ChemA 1998;102:2673e8.

[35] Sato T, Kurosawa S, Smith RL, Adschiri T, Arai K. Water gasshift reaction kinetics under noncatalytic conditions insupercritical water. J Supercrit Fluids 2004;29:113e9.

[36] Lu YJ, Guo LJ, Ji CM, Zhang XM, Hao XH, Yan QH. Hydrogenproduction by biomass gasification in supercritical water:a parametric study. Int J Hydrogen Energy 2006;31:822e31.

[37] Youssef EA, Chowdhury MBI, Nakhla G, Charpentier P. Effectof nickel loading on hydrogen production and chemicaloxygen demand (COD) destruction from glucose oxidationand gasification in supercritical water. Int J Hydrogen Energy2010;35:5034e42.

[38] Williams PT, Onwudili J. Composition of products from thesupercritical water gasification of glucose: a model biomasscompound. Ind Eng Chem Res 2005;44:8739e49.

[39] Mizuno T, Goto M, Kodama A, Hirose T. Supercritical wateroxidation of a model municipal solid waste. Ind Eng ChemRes 2000;39:2807e10.

[40] Goto M, Nada T, Ogata A, Kodama A, Hirose T. Supercriticalwater oxidation for the destruction of municipal excesssludge and alcohol distillery wastewater of molasses.J Supercrit Fluids 1998;13:277e82.

[41] Boukis N, Diem V, Habicht W, Dinjus E. Methanolreforming in supercritical water. Ind Eng Chem Res 2003;42:728e35.

[42] Xu D, Wang S, Hu X, Chen C, Zhang Q, Gong Y. Catalyticgasification of glycine and glycerol in supercritical water. IntJ Hydrogen Energy 2009;34:5357e64.

[43] Lee I-G, Kim M-S, Ihm S-K. Gasification of glucose insupercritical water. Ind Eng Chem Res 2002;41:1182e8.

[44] Cheekatamarla PK, Thomson WJ. Hydrogen generation from2,2,4-trimethylpentane reforming over molybdenum carbideat low steam-to-carbon ratios. J Power Sources 2006;156:520e4.