Applied Energy 155 (2015) 294–307

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

A review on biomass gasification syngas cleanup

Nourredine Abdoulmoumine a, Sushil Adhikari b,⇑, Avanti Kulkarni b, Shyamsundar Chattanathan b
a
Biosystems Engineering and Soil Science Department, University of Tennessee, Knoxville, TN 37996, United States
b
Biosystems Engineering Department, Auburn University, Auburn, AL 36849, United States

h i g h l i g h t s

 The study reviews biomass syngas cleanup technologies and R&D status.
 In-depth synthesis of gas cleanup to remove major syngas contaminants is provided.
 Contaminants of focus are tar, NH3, HCN, H2S, COS, halides (HCl) and trace metals.
 Major technical barriers related to gas cleanup are discussed.

a r t i c l e i n f o a b s t r a c t

Article history: Energy, fuel and chemical production from biomass is increasingly attracting interest in the world.
Received 9 February 2015 Gasification of biomass can produce raw syngas which contains CO, CO2, H2 and CH4. In addition, raw syn-
Received in revised form 13 May 2015 gas contains minor but significant quantities of undesirable impurities – collectively known as syngas
Accepted 28 May 2015
contaminants. Syngas contaminants are composed of tars, nitrogen based compounds (NH3, HCN, etc.),
Available online 20 June 2015
sulfur based compounds (H2S, COS, etc.), hydrogen halides (HCl, HF, etc.) and trace metals (Na, K, etc.).
Raw syngas cleanup is an essential step prior to syngas utilization in downstream applications. In recent
Keywords:
years, significant research attention has been devoted to syngas cleanup to reduce contaminants below
Syngas cleanup
Cold gas cleanup
tolerable limits. The present paper is a comprehensive review of cold gas and hot gas syngas cleanup
Hot gas cleanup for major contaminants in syngas (tar, NH3, H2S, HCl and trace metals). This review organizes and dis-
Contaminants cusses investigations on syngas for all major contaminants, critically reviews important challenges in
Tar syngas cleanup and discusses recent advancements in hot and cold gas cleanup.
Ó 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
2. Cold gas cleanup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
2.1. Tar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
2.2. Nitrogen containing contaminants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
2.3. Sulfur containing contaminants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
2.4. Hydrogen halide and trace metal contaminants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
3. Hot gas cleanup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
3.1. Catalysts for tar removal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
3.1.1. Alkali and alkaline earth metal catalysts (dolomite, olivine, CaO, CaCO3 and MgO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
3.1.2. Transition metal based catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
3.1.3. Zeolite based catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
3.2. Catalysts for nitrogen contaminant removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
3.2.1. Alkaline metals catalysts (dolomite, olivine, CaO and MgO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
3.2.2. Transition metal based catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301

⇑ Corresponding author. Tel.: +1 334 844 3543; fax: +1 334 844 3530.
E-mail address: sushil.adhikari@auburn.edu (S. Adhikari).

http://dx.doi.org/10.1016/j.apenergy.2015.05.095
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307 295

3.3. Sorbents for sulfur contaminants removal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303

3.3.1. Alkaline earth metal based sorbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
3.3.2. Transition metal based catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
3.4. Sorbents for hydrogen halides contaminant removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
3.4.1. Alkali metal catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
3.4.2. Alkaline earth metal catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
3.4.3. Transition metal oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
3.5. Sorbents for trace metals removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
4. Technical barriers in syngas cleanup and future opportunities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
5. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306

1. Introduction halides, such as hydrogen chloride (HCl), and as their respective

In the context of gasification, the term ‘‘contaminant’’ is used to
refer to particulates, condensable organic compounds collectively
known as tar, nitrogenous compounds like ammonia (NH3) and
hydrogen cyanide (HCN), sulfur containing inorganic compounds
like hydrogen sulfide (H2S), carbonyl sulfide (COS) and carbon
disulfide (CS2), hydrogen halides and halogens such as hydrogen
chloride (HCl) and chlorine (Cl), and trace metals like sodium
(Na) and potassium (K). These contaminants emanate from volatile
organic and inorganic constituents in biomass and are conse-
quently present at widely varying concentrations in syngas.
In general, tars are considered as all organic compounds with
molecular weight greater than benzene as defined by a panel of
experts of EU/IEA/US-DOE in 1998 [1]. As such, the definition of
tar engulfs a wide range of compounds including oxygenates, phe-
nolic compounds and olefins, aromatic and polyaromatic hydrocar-
bons (PHA) [2]. Relative to other contaminants, tar is the most
abundant per unit weight of biomass gasified and has been the
focus of the majority of contaminant remediation studies [3]. Its
concentration in syngas depends on the feedstock, process condi-
tions and the type of gasifier with lower oxygen supply and tem-
perature resulting in higher tar concentration.
Nitrogen in biomass is converted to ammonia (NH3), hydrogen
cyanide (HCN), oxides of nitrogen (NO, NO2, N2O and other NOx)
and N2 [4]. However, among all deleterious nitrogenous species
in biomass gasification product gas, NH3 is the most abundant with
concentrations between 350 and 18,000 ppmv depending on the
nitrogen content in the feedstock [4–6]. The partition of biomass
nitrogen favors NH3 at temperatures below 800 °C with more than
60% by weight of the initial nitrogen in biomass converted to NH3
[4]. However, as temperature is increased without additional oxy-
gen supply, N2 becomes the most dominant species from biomass
nitrogen [4]. HCN and NOx are generally a smaller fraction of bio-
mass nitrogen [4,6–8].
Sulfur in the biomass is converted primarily to hydrogen sulfide
(H2S), carbonyl sulfide (COS), carbon disulfide (CS2) and other
minor sulfur containing compounds with H2S being the most dom-
inant [9]. H2S is typically present at concentrations less or near
100 ppmv for most woody and herbaceous biomass [10,11].
Halogens, like chlorine, and trace metals are released as hydrogen
salts by binding to metals in biomass such as alkali metals.
Among these species, HCl and alkali metals (Na, K) are the most
abundant hydrogen halides and metals, respectively for woody or
herbaceous biomass [12]. It is noteworthy to indicate that the rel-
ative abundance and prevalence of specific contaminants can
change drastically depending on the concentration of individual
precursors (i.e. N, S, halogens and metals) in biomass as well as
operating conditions.
The presence of these contaminants in syngas poses several
technical and operational issues ranging from equipment corrosion
(H2S) and fouling (tar), catalyst deactivation (tar, H2S, NH3, HCl and
trace metals) to environmental pollution (NH3) [5,6,12]. As shown
in Table 1, most downstream applications have very stringent
requirements for syngas cleanliness with varying levels of cleanup
demands depending on the application. Therefore, contaminant
levels must be decreased by gas cleanup to adhere to the require-
ments of downstream applications [13].
Few comprehensive reviews on gas cleanup have been pub-
lished in recent years. Woolcock and Brown presented an extensive
and comprehensive review of cleanup technologies for cold, warn
and hot gas cleanup for all major contaminants [13]. Others have
summarized hot gas catalytic or sorbent removal of tar [15–17],
hydrogen sulfide [17,18], ammonia [18] and catalytic conditioning
of biomass-derived syngas [19]. The present review updates and
complements those past reviews by extending the review to
halides and trace metals that have previously only received limited
attention and cold gas cleanup. However, this review does not
cover particulate removal technologies which have been covered
by a recent review [13]. Furthermore, this review critically dis-
cusses important challenges in syngas cleanup and discusses
recent advancements in hot gas cleanup. Finally, existing barriers
and the path forward to advance syngas cleanup are discussed.
2. Cold gas cleanup
Cold gas cleanup is commonly considered the conventional
approach in syngas cleanup due to its proven reliability and high
contaminant removal efficiency. The main distinguishing charac-
teristic of this approach is that it is carried out at low temperature,

Table 1
Upper limits of contaminants in gasification syngas for selected applications [14].

Applications Tars (mg/N m3) Sulfur contaminants (ppmv) Nitrogen contaminants (ppmv) Alkali (ppmv) Halides (ppmv)
Gas turbine n/a <20 <50 <0.02 <1
FT synthesis <0.1–1a 0.01 0.02 0.01 0.01
Methanol synthesis <1 <1b 0.1b n/a 0.1b

n/a means not available.

a
Units are in ppmv.
b
Units in mg/N m3.
296 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307

usually at or below room temperature. In terms of cost, this Table 3

approach is more amenable to larger scale applications due to Absorption efficiencies of tar components by different liquid scrubbing solvents [27].

the need to treat solvent effluents generated during gas cleanup. Absorbents % Absorption
As gasification is carried out at 800–900 °C, the main disadvantage Water Diesel Biodiesel Vegetable Engine
of this approach is the efficiency penalty paid due to syngas cooling fuel fuel oil oil
and the additional cost incurred to treat or dispose of contaminant Benzene 24.1 77 86.1 77.6 61.7
streams. Cold gas cleanup utilizes ‘‘wet’’ or ‘‘dry’’ processes, most Toluene 22.5 63.2 94.7 91.1 82.3
deployed and tested at the pilot scale level. Wet cold gas cleanup Xylene 22.1 n/a 97.8 96.4 90.7
processes employ spray and wash towers, impingement and ven- Styrene 23.5 57.7 98.1 97.1 91.1
Phenol 92.8 n/a 99.9 99.7 97.7
turi scrubbers and wet electrostatic precipitators or cyclones
Indene 28.2 97.9 97.2 97.6 88.7
[20]. These units remove contaminants by absorption and adsorp- Naphthalene 38.9 97.4 90.3 93.5 76.2
tion, filtration or a combination thereof. In contrast, dry cold gas
n/a represents values omitted from the primary reference.
cleanup processes employ mechanical, physical and electrostatic
separation including cyclones, adsorbing bed and other filters,
dry electrostatic precipitators, etc. Of the two types of cold gas for tar removal are increased in process complexity and cost of
cleanup approaches, wet cold gas cleanup processes are more com- oil-based absorbents ultimately resulting in an overall increase in
monly used because they easily allow for multiple-contaminants cost. For this reason, this approach is currently economically
removal. For example, NH3, HCl and H2S are readily soluble in attractive only in large-scale installations [28]. Consequently, in
water. Therefore, wet towers, scrubbers, ESP and cyclones using light of the effectiveness of non-polar scrubbing solvents as evi-
water as a solvent will remove all of these contaminants at varying denced in Tables 2 and 3, future research activities should aim to
removal efficiencies based on their solubility in water. (i) identify and evaluate cheaper non-polar alternative absorbents
and (ii) reduce the process complexity and (iii) develop affordable
2.1. Tar strategies for a wide range of applications.

Cold gas tar cleanup is carried out by wet scrubbing with a liq- 2.2. Nitrogen containing contaminants
uid absorbent. The selection of the solvent is critical in maximizing
the removal efficiency. Water has been used for scrubbing tars Cold gas cleanup of nitrogen containing contaminants is con-
from syngas primarily because it is cheaper [21]. However, while ventionally achieved through wet scrubbing although spray and
light and oxygenated tar compounds can be removed by water wash towers are suitable for this application. The wet scrubbing
scrubbing due to their inherent polarity, non-polar heavy and hete- approach presents practical advantages for solvent recovery as
rocyclic tar compounds have low removal efficiency as shown in absorbent can be regenerated by precipitation as an ammonia salt
Table 2. The challenges associated with water based tar scrubbing [24], in the case of NH3 absorption, or by solvent regeneration
are low removal efficiency and added cost treating the resulting through temperature swing [29]. Due to the high solubility of
waste water stream [22]. Recent advancements have focused on NH3 and HCN in water, water based scrubbing is regarded as the
oil based absorbents that afford high removal efficiency, particu- conventional approach for cold gas removal of nitrogen containing
larly for heavy and heterocyclic tar compounds. One of the earliest contaminants. In this approach, deleterious nitrogen species
pilot-scale deployment of oil-based tar scrubbers was demon- removal is often carried out simultaneously with tar further com-
strated at the Energy Research Centre of the Netherlands (ECN) plicating wastewater treatment due to the toxicity of many tar
through their OLGA technology [22,23]. As illustrated in Table 2,
the OLGA oil based scrubbers achieved higher efficiency that the
traditional water scrubbers, even when a wet electrostatic precip-
itator is used [23,24]. The OLGA technology consists of multiple
stage scrubbers where oil is the scrubbing solvent and can be
regenerated by hot air stripping [25].
Other recent cold gas tar cleanup advancements employ diesel
fuel, engine oil or fatty acid methyl esters derived from vegetable
oil as an absorbent, an approach demonstrated at the pilot-scale
level in Austria at the Güssing Renewable Energy gasification
installation [25,26]. As expected and shown in Table 3, hydrocar-
bon and fatty acid ester based absorbents outperformed water.
The benefits of oil-based scrubbing technologies are the possi-
bility to operate at higher temperatures thereby improving the
overall process efficiency and the ability to regenerate the oil based
absorbents by hot gas stripping to generate a concentrated tar
stream [23]. The major drawbacks of oil based scrubbing systems

Table 2
Comparison of wet scrubbing technologies for tar removal [24].

Absorbents % tar removal

Water Water and wet ESP Vegetable oil (OLGA) Fig. 1. Diagram of common sulfur removal solvents employed in conventional acid
Heavy tars 49 99 100 gas removal (AGR). MEA = methylethanolamine, DGA = diglycolamine,
Light tars 62 74 99 DEA = diethanolamine, DIPA = di-isopropanolamine, MDEA = methyl diethanola-
Heterocyclic tars 79 79 99 mine and TEA = trietahnolamine. Selexol, Coastal AGR, Genosorb, Rectisol, Ifpexol,
Dew point, °C 180 60 10 Purisol, Jeffsol, Sulfonol, Amisol and Flexsorb are commercial names of solvents
[36,37].
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307
Table 4
Conventional methods for sulfur removal from gasification syngas [44].
Solvent Mechanism Removal efficiency (%) Typical process conditions
MDEA Chemical H2S: 98–99 T: 30–35 °C
CO2: 630 P: 62.94 MPa
Selexol Physical H2S: 99 T: 7–4 °C
CO2: variable P: 6.87 MPa
Rectisol Physical H2S: 99.5–99.9 T: 35–60 °C
CO2: 98.5 P: 8.04 MPa
a
SR = stands for ‘‘Sulfur Recovery’’ using a Claus sulfur recovery system.
b
TGT = stands for ‘‘Tail gas treatment’’.
compounds to nitrifying organisms [30]. The benefits of conven-
tional wet scrubbing of syngas through water is the significant
reduction of NH3 concentrations with removal efficiency greater
than 99% observed when a spray tower is utilized as a contacting
device at ambient temperature [29]. In fact, it was reported that
water vapor in syngas can remove more than 90% of NH3 upon con-
densation [31]. Absorption in water is enabled by protonation of
NH3 according to the reaction below.
NH3 þ H2 O $ NHþ4 þ OH
+
In theory, NH3 protonation to NH4 can be achieved in various
dilute acid solutions. In fact, the EPA’s procedure for sampling
NH3 in gas stream from stationary sources such as coal-fired boil-
ers uses 0.1 N H2SO4 as a scrubbing absorbent (EPA Method
CTM-027). The on-going assessment of acid based scrubbing to
remove NH3 and other nitrogen species in syngas and flue gas rep-
resents an advancement over the conventional water-based scrub-
bing approach. Of the few investigations that have explored dilute
acid absorbents, NH3 removal efficiency varies widely from 50% at
500 ppmv to greater than 95% at 200 ppmv [32,33]. The benefits of
acid absorbents for nitrogen contaminant removal is that they
could co-absorb acid gases such as H2S to further enhance
co-absorption of NH3 as the acidity of the absorbent increases over
time. At present, major limitations of absorbents include their cor-
rosiveness when mineral acids (e.g., H2SO4) are used and their low
removal efficiency at higher NH3 concentrations (>500 ppmv) typ-
ically observed for herbaceous and other higher nitrogen contain-
ing biomass. As there are very limited systematic studies using
acid washing, future research activities could be carried out in this
area to systematically evaluate mineral and organic acids and ulti-
mately improve on the removal efficiency. Organic acids based
absorbent scrubbing would be very interesting in the context of
thermochemical conversion of biomass as pyrolysis and gasifica-
tion generate various organic acids that could find use as
absorbent.
2.3. Sulfur containing contaminants
Cold gas cleanup of sulfur contaminants in syngas can be
achieved by wet or dry cleaning processes. Thus far, the main
drawbacks of cold gas cleanup strategies for sulfur removal are
syngas cooling (40 °C for amine-based absorption processes and
to 40 °C for the refrigerated physical solvent processes) and
increase in capital cost due to refrigeration and solvent regenera-
tion requirements [34]. For wet gas cleanup, there are two modes
of sulfur removal in solvent: chemical reaction or physical absorp-
tion hence the classification of solvents as ‘‘chemical’’ or ‘‘physical’’
solvents. Absorptive wet gas cleanup technologies have been
enjoyed commercial success since the 1970s and are widely
regarded as the conventional approach in sulfur removal from syn-
gas. These technologies commonly rely on alkanolamines, alkaline
297
Exit gas quality Remarks
H2S: 10–20 ppmv Lowest capital cost
Moderate operating temp
Limited physical COS absorption
H2S: <30 ppmv Higher cost than MDEA
Good efficiency if coupled with SRa & TGTb
H2S: <0.1 ppmv Highest cost
CO2: down to few ppmv Highest removal of H2S
Good removal of COS
High selectivity for H2S over CO2
salts or aqueous ammonia and organic solvents with high affinity
for acid gases as chemical and physical solvents, respectively
[35]. Fig. 1 and Table 4 outline the most common physical, chem-
ical and hybrid solvents that are employed in acid gas removal as
well as the process conditions and expected outlet gas quality.
Solvent scrubbing has the added advantage of being regenerable
by heating and pressure letdown and of simultaneous
co-adsorbing COS and CO2 [36].
For integrated gasification combined cycle (IGCC), environmen-
ð1Þ tal regulations require syngas sulfur concentration less than
30 ppmv in order to meet stack SO2 emission limits [38]. In such
a case, chemical solvents such as amine-based solvents like
DMEA are more suitable due to the relatively high H2S tolerance
in the exit gas stream. For chemical and fuel synthesis, much lower
sulfur concentration is required due to severe catalyst poisoning
Ò Ò
and physical absorption solvents like Selexol and Rectisol are
preferred as they can achieve concentrations below 1 ppmv. In
addition to wet processes, adsorptive dry cold gas removal of sulfur
contaminants can be carried out on porous sorbents like activated
carbon or metal impregnated activated carbon [39,40]. For these
sorbents, surface chemistry is critical to achieve adequate removal
with basic surfaces performing better than acidic ones [39].
In recent years, ionic liquid based physical solvents are emerg-
ing as a new research area in wet cold gas cleanup of sulfur con-
taminants [41,42]. Often referred to as ‘‘designer’’ solvents due to
customizable physical properties depending on anion and cation
substituents, these solvents can be functionalized to provide selec-
tive separation of H2S and CO2 [43]. Furthermore, the use of ionic
liquid solvents can improve the overall thermal efficiency due to
their negligible volatility and thermal stability around warm gas
cleanup temperatures (200 °C) [42].
2.4. Hydrogen halide and trace metal contaminants
HCl and other hydrogen halides are commonly removed by wet
scrubbing usually through a caustic solution or, alternatively, sim-
ply water [45]. The use of caustic water, NaOH dissolved in water,
is logical as HCl will react with NaOH to produce NaCl which can be
precipitated out of solution.
HCl þ NaOH ! NaCl þ H2 O ð2Þ
The ability to remove NaCl by precipitation is an added benefit
to caustic based wet scrubbing particularly when compared to the
alternative water based scrubbing which will require costly ionic
exchanger to strip chlorine from the solution [46]. In addition,
caustic water can remove other contaminants such as COS
[24,47] as well as CO2 [24] according to the reactions below.
H2 S þ NaOH ! NaHS þ H2 O ð3Þ
H2 S þ 2NaOH ! Na2 S þ 2H2 O ð4Þ
298 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307

Fig. 2. Elements in the periodic table used as catalyst, catalyst promoter, catalyst support or sorbents in gas cleanup. References for promoters and supports [19,50–57].
Catalysts and sorbents compiled from over 20 gas cleanup studies.

COS þ 4NaOH ! Na2 S þ Na2 CO3 þ 2H2 O ð5Þ ingredients evaluated were summarized in the periodic table
shown in Fig. 2. The benefit of this approach is that it can potential
COS þ 2NaOH ! Na2 S þ H2 CO3 ð6Þ help future investigators to avoid duplicating existing information
and identify active ingredients that could be effective in the
CO2 þ 2NaOH ! Na2 CO3 þ H2 O ð7Þ removal of many species.
As NaOH reactions with HCl and H2S are fast, it is possible to
significantly reduce the concentrations of these two impurities 3.1. Catalysts for tar removal
when caustic water is used as an absorbent. While caustic water
is a potential scrubbing solvent, water remains the most com- Tar compounds can be removed by thermal cracking, hydroc-
monly used solvent [45]. racking, steam or dry reforming. Thermal cracking is the conver-
Gas cooling induces the condensation of most metals, particu- sion of tar to C and H2 at elevated temperature.
larly alkali metals which can be reduced by mild gas cooling below m
Cn Hm ! nC þ H2 ð8Þ
600 °C and subsequent adequate gas filtration [48,49]. However, 2
this approach only serves as a primary measure as some metals
Due to the refractory nature of gasification tar [58], extensive
will remain in gas phase and can be removed by wet scrubbing
tar removal by thermal cracking requires high temperatures
with water [49]. Most trace metals, like alkali metals, are suffi-
(>1100 °C) [59]. Tar can be decomposed in the presence of hydro-
ciently soluble in water and can consequently be co-absorbed in
gen through hydrocracking which produces methane as shown
water during wet scrubbing of tar and NH3.
below.
 
3. Hot gas cleanup 4n  m
Cn Hm þ H2 $ nCH4 ð9Þ
2
In addition to the conventional techniques described previ- Tar as well as light hydrocarbons can be converted to carbon
ously, warm and hot gas cleanup approaches are of great interest monoxide and hydrogen through stream reforming as shown
in contaminant removal from syngas. Warm and hot gas cleanup below.
focuses on contaminant removal from syngas at high temperature  m
(>300 °C) [13]. The advantage of hot gas cleanup is that it addresses Cn Hm þ nH2 O $ nCO þ n þ H2 ð10Þ
the loss in efficiency, a downside of cold gas cleanup. Additionally, 2
it reduces the waste streams with the potential of converting some Due to the high activation energy (250–350 kJ/mol) and its
contaminants into environmentally benign or even useful prod- endothermic nature, steam reforming of tars occurs appreciably
ucts. Over the years, group of catalysts have been tested in hot only above 900 °C in the absence of a catalyst [60]. When a suitable
gas cleanup of syngas. In an effort to distill the large volume of catalyst is employed, temperatures ranging between 650 and
hot gas cleanup catalytic and sorbents studies, the most active 900 °C are sufficient for high conversion. However, tar steam
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307 299

Table 5
Summary of catalytic tar reforming studies.

Catalysts Gas composition, vol%a Reaction conditions Conversion Ref.

3
Mg3Ca(CO3)4 Fluidized bed gasification tar: 1400 mg/m T: 900, s: 0.2–0.3 s 99 [64]
Fe2O3/Mg3Ca(CO3)4 CO: 9.8, CO2: 14.6, CH4: 3.6, H2: 9.8, H2O: 13.7, NH3: 0.41, N2: balance 99
Ni/Al2O3 100
Al2O3–SiO2 Updraft gasification tar: 9800 mg/m3, CO:17.2, CO2: 8.0, CH4: 1.6, H2: 13.2, H2O: 25.9, NH3: T: 900 °C, s: 0.2–0.3 77
Fe2O3 0.17, N2: balance 91
Mg3Ca(CO3)4 100
Fe2O3/Mg3Ca(CO3)4 100
Ni/Al2O3 100
Dolomite Biomass tar from power plant: 5000 mg/m3, CO, CO2, CH4, H2 concentration unknown T: 850 °C, GHSV: 95 [65]
12,000 h1
Fe2O3/Dolomite T: 850 °C, GHSV: 95
12,000 h1
Ni/dolomite T: 650 °C, GHSV: 85
12,000 h1
T: 850 °C, GHSV: 100
12,000 h1
T: 650 °C, GHSV: 85
12,000 h1
T: 700 °C, GHSV: 90
12,000 h1
T: 750 °C, GHSV: 95
12,000 h1
T: 850 °C, GHSV: 100
12,000 h1
T: 650 °C, GHSV: 85
12,000 h1
T: 850 °C, GHSV: 100
12,000 h1
Dolomite Biomass tar, CO, CO2, CH4, H2 concentration unknown T: 760 °C, GHSV: 77.8 [66]
9200 h1
T: 923 °C, GHSV: 96.8
10800 h1
T: 830 °C, GHSV: 91.3
12,200 h1
T:814 °C, GHSV: 94.4
13,700 h1
T: 805 °C, GHSV: 71.4
15,500 h1
T: 844 °C, GHSV: 91.4
9700 h1
Y-zeolite 1-Methyl naphthalene 3000 mg/m3, N2: 70, H2: 30 T: 550 °C, GHSV: 99 [67]
NiMo/Al2O3 30,000 h1 99
Lime 87
Al2O3 86
SiO2 81
North Carolina olivine Naphthalene 0.017, H2: 20.5, CO: 20.5, CO2: 28, H2O: 23.5, CH4: 5.6, Toluene: 0.28, N2 balance T: 850 °C, GHSV: 77.1 [68]
2500 h1
T: 900 °C, GHSV: 100
2500 h1
Toluene: 0.28, H2: 20.5, CO: 20.5, CO2: 28, H2O: 23.5, CH4: 5.6, Napthalene: 0.017, N2 balance T: 800 °C, GHSV: 70.1
2500 h1
T: 850 °C, GHSV: 88.3
2500 h1
T: 900 °C, GHSV: 89.4
2500 h1
Washington olivine Naphthalene: 0.04, H2: 16, CO: 8, CO2: 12, H2O: 16, CH4: 4, N2 balance T: 850 °C, GHSV: 80
1166 h1
Washington olivine T: 900 °C, GHSV: 97
1166 h1
North Carolina olivine T: 800 °C, GHSV: 75
1166 h1
North Carolina olivine T: 900 °C, GHSV: 98
1166 h1
Australian olivine T: 800 °C, GHSV: 90
1166 h1
Australian olivine T: 900 °C, GHSV: 100
1166 h1
Calcined Australian T: 900 °C, GHSV: 92
olivine 1166 h1
Preatreated olivine Naphthalene: 5000–7000 mg/m3, H2: 10, CO: 12, CO2: 12, H2O: 10, Ar balance T: 900 °C, s: 0.3 s 70 [69]
Preatreated olivine Naphthalene: 5000–7000 mg/m3, Ar balance 65
Ni/Olivine Methylnaphtalene: 0.0509, H2O: 1, N2 balance T: 900 °C 75 [70]

(continued on next page)

300 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307
Table 5 (continued)
Catalysts Gas composition, vol%a
Ni/CaO–Ca12Al14O33 Toluene: Unknown, H2O, N2 unknown
Ni–Fe–Ce/Al2O3 Naphthalene: 5000 mg/m3, H2O: 15, CO2: 10, N2 carrier
T = temperature, F = flow rate, GHSV = gas hourly space velocity.
reforming often suffers from catalyst deactivation through coke
although this can be minimized considerably by careful control
of the steam to carbon ratio [61]. Tar compounds could also be
reformed in the absence of steam with CO2 through dry reforming,
as shown below.
m
Cn Hm þ nCO2 ! 2nCO þ H2 ð11Þ
2
In practice, several condensable hydrocarbons, oxygenated aro-
matics and complex polyaromatic hydrocarbons (PAHs) fall within
the definition of tar proposed by a panel of experts in 1998 [1]. As a
direct result of the complex chemical composition of biomass tars,
researchers typically employ model compounds in catalytic tar
cleanup in order to obtain useful results. While benzene is techni-
cally outside the definition of tar, it is commonly quantified tar
content in gasification experiments as it is present in high concen-
trations. Benzene, toluene, phenol, and naphthalene together
reflect the majority of tar composition [62,63] and as such are com-
monly used as model compounds for tar removal studies. However,
naphthalene is the most common model compound because it is
considered the most stable and recalcitrant in catalytic tar cleanup.
Dolomite, olivine, limonite, calcium/magnesium carbonates as well
as nickel based catalysts have been widely used in tar removal in
bench-scale as well as pilot-scale setup.
Table 5 summarizes selected catalytic tar reforming studies. Of
these catalysts, mineral catalysts (dolomite, olivine), nickel and
iron based catalysts have been the most effective at temperatures
ranging from 500 to 900 °C.
3.1.1. Alkali and alkaline earth metal catalysts (dolomite, olivine, CaO,
CaCO3 and MgO)
Alkali metal catalysts have been investigated as primary cata-
lysts to improve the quality of the gasification syngas by several
researchers over the years. Their presence in biomass ash and their
catalytic effects in thermochemical conversion is well documented
[58]. These catalysts have however been less effective for carbon
conversion and often result in downstream recovery challenges
due to their volatility around typical gasification temperatures
[72].
K2CO3 supported on perovskite oxide (LaMn(1x)CuxO3) and
Al2O3 were investigated for in-situ tar removal in fluidized bed
gasifier. In the presence of 10 wt% K2CO3 supported on Al2O3 and
perovskite oxide, the tar concentration was reduced by 53% and
59%, respectively at low gasification temperature (700 °C) [73].
K2CO3 unsupported was less effective resulting only in 20% reduc-
tion of tar content.
Carbonate minerals of Ca and Mg, well known as dolomites, are
of great interest in gas cleanup. Several groups have studied dolo-
mites as in-situ and ex-situ catalysts for tar cleanup and have
showed that near complete conversion can be achieved under opti-
mum condition [64,65,74]. Corella et al. [74] showed that dolo-
mites are equally effective in-situ as ex-situ in tar removal.
Calcination of dolomites results in increase in activity because of
the increase in surface area and oxide content [74]. Improved dolo-
mite activity can be attained by the addition of other metal oxides
Reaction conditions Conversion Ref.
T: 650 °C 99 [71]
T: 550 °C, F: 500 l/min 97.2 [51]
T: 600 °C, F: 500 l/min 98
T: 650 °C, F: 500 l/min 97.7
T: 700 °C, F: 500 l/min 98.7
such as Fe2O3 [65,66], Ni [65]. Furthermore, bimetallic Ni–Pt sup-
ported on dolomite showed significantly better performance over
various Pt/Al2O3 and olivine with extensive cracking of tar based
on a GC–MS comparative analysis of uncatalyzed and catalyzed
samples [75]. Calcined dolomite catalysts seems to be more active
towards tar when the gasifying agent is steam rather than air sug-
gesting that dolomite removes tars via steam reforming primarily
[76,77]. The drawback of dolomites is that they can easily be
eroded limiting their use in fluidized bed reactors as primary or
secondary catalyst [69]. Olivine, like dolomite, is a natural mineral
catalyst that is composed of primarily of magnesium iron and sili-
con in the chemical formulas of Fe2SiO4 or Mg2SiO4. Like dolomite,
olivine can achieve high conversion (>90%) for naphthalene [68].
Kuhn et al. [68] compared the performance of several olivine cata-
lysts in tar reforming potential with naphthalene and toluene as
model compounds, the authors listed the origin, treatment and
phase of the Fe atom among factors affecting the activity.
Deactivation was observed and was attributed to carbon deposi-
tion as well as Fe dispersion. Like dolomites, greater activity was
reported for calcined olivine relative to the natural and uncalcined
olivine [69,78]. X-ray photoelectron spectroscopy analysis revealed
that calcination increased the Fe content on the surface as a poten-
tial reason for enhanced activity upon calcination [69]. Ni doping
on olivine showed enhancement of the support with conversion
increasing from 4% to near 30% at 900 °C [70]. Furthermore, steam
pretreatment of Ni/Olivine increased conversion near 75% at the
same temperature.
The effectiveness of dolomite and olivine in tar decomposition
has motivated researchers to look into the activity of individual
oxides (CaO, CaCO3 and MgO) in dolomites. Simell and coworkers
[64] investigated the activity towards tar decomposition of CaO
and MgO and reported activities in the following order:
CaO > CaO–MgO > MgO > NiMo/c–Al2O3 indicating that CaO has a
greater catalytic role than MgO in dolomite. Fe and Ni impregnated
scallop shell (CaO/CaCO3) as well as scallop shell were investigated
in steam reforming of biomass tar to enhance the production of
primary gases [79]. The authors reported that Ni loading outper-
formed Fe loading and CaO/CaCO3 showed good activity in steam
reforming of tar. In downdraft gasification of bagasse with CaO,
exit tar concentration was reduced by 80% in the presence of
6 wt% of CaO in the bed.
3.1.2. Transition metal based catalysts
3.1.2.1. Iron based catalysts. Limonite is a naturally occurring iron
ore with generic formula as FeOx(OH)y. They have been used in
tar cleanup studies in-situ and ex-situ. Indonesian limonite was
used for the decomposition of coal syngas tars resulting in a good
tar concentration reduction [80]. Investigations of the activity of
the catalyst revealed that transformation from FeOOH to Fe3O4 is
responsible for the increase in activity. Furthermore, there was evi-
dence that the Indonesian limonite catalyzed the steam reforming
reaction. Matsumura and coworkers [81] compared a catalyst con-
taining an Australian yellow yandi limonite against a conventional
NiO–MoO2/Al2O3 for the decomposition of asphaltene and reported
slightly higher conversion for the catalyst containing limonite.
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307
Brazilian limonite was also investigated for asphaltene decomposi-
tion with still greater activity than the conventional NiO–
MoO2/Al2O3 catalyst [82].
In addition to olivine and limonite, other iron containing cata-
lysts that have been studied for tar reduction include iron oxides
(FeO [83], Fe2O3 [64,83] and Fe3O4 [83]), ankerite (CaFe(CO3)2 as
a ferrous dolomite), sintered iron ore, and pelletized iron ore.
Nordgreen et al. [83] investigated the potential of FeO, Fe2O3 and
Fe3O4 and showed that FeO was more effective than Fe2O3 and
Fe3O4. FeO, Fe2O4, Fe3O4 and sintered ore have very little catalytic
activity in tar decomposition [72,84].
3.1.2.2. Nickel based catalysts. Nickel-based catalysts, commonly
supported on a metal oxide, have been extensively used in biomass
gasification tar cleanup [5,52,85]. They are not suitable for in-situ
tar destruction due to deactivation by coking [50,86]. However,
these catalysts have remarkable activity as secondary catalysts
[87]. Buchireddy et al. [88] investigated the impregnation effect
of nickel on zeolite catalysts and observed noticeable activity
improvement that was attributed to the steam reforming catalytic
activity of nickel. Dou and coworkers [87] investigated the poten-
tial of a NiMo catalyst containing 4% NiO in the conversion of
1-methyl naphthalene and reported nearly 100% conversion at
550 °C. Similar conversion was achieved when nickel catalysts
treated with various alkali promoters were used in a different
study [52]. Miyazawa et al. [89] assessed the activity of several
nickel supported catalysts and ranked them as follow:
Ni/Al2O3 > Ni/ZrO2 > Ni/TiO2 > Ni/CeO2 > Ni/MgO suggesting that
the support material can enhance the catalytic performance of
the nickel catalysts. Another group tested various bimetallic Ni
and Fe catalysts supported on alumina and found that the bimetal-
lic catalyst outperformed nickel supported on alumina [51]. Pt
appears to be a good promoter of tar reforming catalysts and mixed
oxides of Ni, Fe, Pt performed the best among several Ni, Fe and Ce
catalysts at 800 °C in naphthalene reforming. Similarly, Sutton
et al. [90] investigated several Ni-supported catalysts synthesized
via wet impregnation or co-precipitation and observed that the
co-precipitated catalysts performed slightly better than their coun-
terparts under similar conditions. Following the same logic, nickel
supported on dolomite and olivine catalysts have been evaluated
and a noticeable improvement of activity was observed upon
nickel addition [65,68,91].
3.1.3. Zeolite based catalysts
Zeolites are a group of crystalline aluminosilicate minerals
whose three-dimensional structures are derived from frameworks
of [SiO4]4 and [AlO4]5 coordination polyhedra. The catalysts have
been extensively used in the petrochemical industry to crack large
and complex hydrocarbons. These catalysts should be ideal for tar
cracking since they already catalyze the breakdown of hydrocar-
bons similar to those found in biomass syngas tars. Furthermore,
these catalysts are used for tar reforming of biomass pyrolysis with
good results [92,93] with increase in the concentration of primary
gases presumably due to steam reforming of pyrolytic tar [93]. Dou
et al. [87] satisfactorily used Y-zeolite to convert 1-methylnaphthalene
and achieved nearly complete conversion at 550 °C. Seven commercial
zeolites with different pore sizes and acidity were tested for naph-
thalene conversion. The authors stated that the naphthalene conver-
sion increased with increasing acidity [88]. Adversely, increasing
acidity also resulted in increasing coke formation. Other catalyst prop-
erties that affected naphthalene conversion are pore size and surface
area with in both properties resulting in higher conversion. The
same authors reported the following order for naphthalene
conversion at 750 °C all other reactions being identical: ZY-5.2
(SiO2/Al2O3 = 5.2) > ZY-30 (SiO2/Al2O3 = 30)  ZY-80 (SiO2/Al2O3 =
301
80) > Zb (SiO2/Al2O3 = 25) > Na-Chabazite (SiO2/Al2O3 = 4) > ZSM5
(SiO2/Al2O3 = 24) > Si/Al (SiO2/Al2O3 = 28).
3.2. Catalysts for nitrogen contaminant removal
Ammonia is not easy to decompose under typical gasification
temperature without the use of catalysts [54]. In the following sub-
sections, the various groups in ammonia decomposition are
reviewed. Table 6 summarizes experimental conditions and results
obtained in several catalytic ammonia decomposition studies. The
most effective active metals are Ni or Ni promoted with Ce or La, Co
or Co–Zn mixed oxide, Ru and Fe while Al2O3 and activated carbon
are among the most common and effective supports. In addition,
limonite is also very effective in NH3 decomposition. While the
experimental conditions vary widely depending on the studies,
NH3 decomposition in the context of hot gas cleanup is typically
investigated at temperatures ranging from 450 to 750 °C and
GHSV ranging from few thousands to as high as 45,000 h1.
3.2.1. Alkaline metals catalysts (dolomite, olivine, CaO and MgO)
In addition to its good activity in tar decomposition, dolomite
has also exhibited activity in ammonia decomposition [102,103].
Several investigators reported partial decomposition of ammonia
in the presence of CaO, MgO and dolomites at normal gasification
temperature [84,104]. Bjorkman and Sjoestroem [103] studied
ammonia decomposition catalyzed by calcined dolomite, CaO and
MgO and reported good decomposition activity for dolomite in inert
environment. They discovered that calcined dolomite was better
than CaO which was in turn better than MgO based on the activation
energy. However, their investigation further revealed that the activ-
ity of dolomite was not due to the additive effect of CaO and MgO,
both present in dolomite. Furthermore, the presence of other syngas
components (16.3, 12.4 and 4.7 vol% for CO, CO2 and H2, respec-
tively) resulted in lower conversion, possibly due to coking accord-
ing to the investigators. Corella et al. [102] investigated in-bed
dolomite and sand mixture (25–30/70–75 wt%) and established that
it was less effective for ammonia decomposition. Despite significant
investigation of dolomite on tar removal, fewer studies solely
focused on ammonia decomposition were carried out.
Like dolomite, ammonia decomposition activity was observed
for olivine though very few publications focused on olivine solely.
Corella et al. [102] investigated olivine and dolomite mixed with
sand (70–75 wt%) as bed materials. Under similar operating condi-
tions, they showed that the mixture with olivine resulted in lower
NH3 exit concentration than with dolomite in a bubbling fluidized
bed gasifier.
3.2.2. Transition metal based catalysts
3.2.2.1. Iron based catalysts. Limonite is very effective in ammonia
conversion with near complete conversion at 500 °C in inert envi-
ronment [105]. In the presence of primary gases (CO, H2 and H2O),
limonite’s activity in ammonia decomposition was seemingly
reduced and higher temperature (>750 °C) was required to achieve
conversions above 90% [105]. In addition to good catalytic activity,
Australian limonite showed resistance to sulfur deactivation in the
presence of 50–500 ppmv H2S [101].
Leppälahti et al. [84] investigated the effectiveness of iron sin-
ter, iron pellet, ferrous dolomite against dolomite limestone, inert
silicon carbide and a commercial nickel catalyst as the reference
in the decomposition of ammonia. Their results indicated that fer-
rous catalysts were more active than dolomite and limestone while
less active than nickel. An iron doped mesoporous activated car-
bon, prepared by H3PO4 activation of Canadian peat, was synthe-
sized and test by Donald and co-workers with great success
[106]. Under optimum conditions, NH3 conversions near 90% were
obtained for a gas mixture containing 2000 ppmv NH3 in a balance
302 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307

Table 6
Summary of catalytic ammonia decomposition studies.

Catalysts Gas composition, vol% Reaction conditions Conversion, % Ref.

1
Ni–Zn/Al2O3 NH3: 0.1, H2: 11.7, CO: 9.56, CO2: 5.20, N2 balance 650 °C, GHSV: 5000 h 95% [94]
Fe–Zn/Al2O3 20–30
Co–Zn/Al2O3 95%
5% Ru/Fly ash NH3: 100 T: 500 °C, F: 100 ml/min 15.3 [56]
Ru/CMK3-NaOH NH3: 100 T: 500 °C, WHSV: 30,000 ml g1 h1 50.8 [57]
Ru/CMK3-KOH 78.9
Ru/CMK3-Ca(OH)2 48.5
Ni–Ce/Al2O3 NH3: 100 T: 500 °C, WHSV: 30,000 ml g1 h1 100 [95]
Ce–Ni/SBA-15 NH3: 100 T: 650 °C, WHSV: 30,000 ml g1 h1 100 [50]
Ce–Ni/SBA-15 T: 550 °C, WHSV: 30,000 ml g1 h1 80–90
La–Ni/SBA-15 T: 650 °C, WHSV: 30,000 ml g1 h1 100
La–Ni/SBA-15 T: 550 °C, WHSV: 30,000 ml g1 h1 80–90
Fe/Activated carbon NH3: 0.20, Balance: He T: 750 °C, GHSV: 45,000 h1 90 [5]
Ni/activated carbon 75
Ni/activated carbon NH3: 100 T: 500 °C, GHSV: 6000 h1 60 [96]
Ru/LaZrO2 NH3: 100 T: 450, 700 °C, WHSV: 4000 ml g1 l1 80, 100 [97]
Ru/LaZrO2 40, 100
15% Ni/Al2O3 NH3: 0.21, CO: 16.5, CO2: 10.3, CH4: 4.6, H2: 16.7, H2O: 15.1, T: 850 °C, s: 0.6 s 80 [54]
Ni–Ru N2 balance 85
Ni/ZSM-5 NH3: 0.10, H2: 10.5, CO: 28.4, CO2: 3.6, H2O 3.1, O2: 0.008, T: 300 °C, GHSV: 20,000 h1 95 [98]
N2 balance
Ru/CNT NH3: 100 T: 500 °C, WHSV: 30,000 ml g1 h1 84.78 [53]
Ru/CNT
Rh/CNT 12.53
Pt/CNT 4.22
1 1
Pd/CNT T: 450 °C, WHSV: 30,000 ml g h 1.70
Pd/CNT T: 500 °C, WHSV: 30,000 ml g1 h1 4.28
Ni/CNT 8.67
Fe/CNT 1.94
Ru/MgO T: 500 °C, WHSV: 150,000 ml g1 h1 13.39
Ru/TiO2 11.30
Ru/Ac 10.49
Ru/Al2O3 10.79
Ru/ZrO2 5.10
k-Ru/CNT 28.58
k-Ru/ZrO2 18.89
Ni/Al2O3 NH3: 0.06, H2: 9–11, CO: 12–15, CO2: 13–18, CH4: 5–8, H2O: 7–15, T: 830 °C, P: 20 bar, s: 1.1 s 38.00 [54]
Ru–Al2O3 H2S: 0.0030–0.0070, N2 balance 28
Al2O3 23
Co3O4 NH3: 0.50, H2: 11.7, CO: 9.56, CO2: 5.20, N2 balance T: 650 °C, GHSV: 5000 h1 95 [55]
MoO3 96
NiO 97
ZrO2 30
WO3 20
ZnO 20
Co3O4 20
MoO3 20
NiO 20
Co/ZnO 95
Mo/ZnO 95
Ni/ZnO 95
Ru/Al2O3 NH3: 10, He balance T: 450 °C, F: 200 ml/min 90 [99]
Fe2O3 NH3: 0.17, CO: 17.2, CO2: 8.0, CH4: 1.6, H2: 13.2, H2O: 25.9, T: 900 °C, s: 0.2–0.3 s 87 [64]
Fe2O3/Mg3Ca(CO3)4 Tar: 0.98, N2 balance 75
Ni/Al2O3 100
Mg3Ca(CO3)4 53
Fe2O3/Mg3Ca(CO3)4 53
Ni/Al2O3 100
Fe/AC NH3: 0.20, He balance T: 750 °C, GHSV: 45,000 h1 100 [100]
Fe/AC 100
Ca/AC 100
Fe/AC 50
Fe/AC NH3: 0.20, CO/H2 = 2, CO2 95
Limonite NH3: 0.2, H2S: 0.0050–0.050, N2 balance T: 750 °C, GHSV: 45000 h1 100 [101]

T = temperature, F = flow rate, P = pressure, GHSV = gas hourly space velocity, WHSV = weight hourly space velocity.

of helium at 750 °C with conversion as high as 90%. Supported iron
catalysts have demonstrated greater activity than supported cal-
cium catalysts suggesting that iron is better than calcium in
ammonia decomposition [100].
3.2.2.2. Nickel catalysts. Nickel based catalysts, either supported or
unsupported, are very effective in NH3 decomposition, often with
conversions exceeding 95% [5,55]. The selection of the support
plays a key role in the activity of supported nickel catalysts: stron-
ger acidity tend to increase catalytic activity [98] and some com-
mon supports (ZrO2 [55], MoO3 [55], Al2O3 [107]) appear to be
active in NH3 decomposition even prior to Ni impregnation. The
effectiveness in NH3 decomposition increases with Ni loading as
shown by Mojtahedi et al. [54] who demonstrated that with
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307
increasing Ni loading from 2 to 15 wt%, NH3 conversion increased
from 54% to 92% at 850 °C. Nassos et al. [108] studied the effect of
nickel loading on three metal oxide supports, Ce0.9La0.1O2,
Ce0.9Zr0.1O2 and c-Al2O3 with nickel loading of 5 and 10 wt%. The
ammonia conversion increased with increasing nickel concentration
for Ce0.9La0.1O2 and Ce0.9Zr0.1O2. They also observed that c-Al2O3
was not as good of a support as were Ce0.9La0.1O2 and Ce0.9Zr0.1O2.
It has also been reported that Ce and La addition to Ni based cata-
lysts improved its performance in NH3 conversion [50,95].
Other metal oxides can also be combined with Ni to improve its
performance and stability. Simell et al. [64] reported 80% conver-
sion using a nickel monolith catalyst (NiMo/c–Al2O3). Upon
increasing the temperature to 900 °C, complete conversion was
achieved. Wang et al. [109] investigated hot gas ammonia decom-
position in the presence of a nickel based catalyst in an air blown
fluidized bed gasifier. A maximum ammonia conversion of 95%
was achieved under optimum conditions. Ni–Pt/Al2O3 catalyst
was used in the decomposition of ammonia and good activity
was observed for pure ammonia decomposition with 97.4% conver-
sion at only 560 °C [110].
While nickel based catalysts are very active in ammonia decom-
position, their major drawback is catalyst deactivation due to cok-
ing as observed by Wang et al. [111] and/or poisoning [112],
particularly by H2S present in syngas [55].
3.2.2.3. Ruthenium based catalysts. The use of ruthenium based cat-
alysts for NH3 decomposition is well documented [53,54,57].
Mojtahedi et al. [54] investigated Ru/Al2O3 in ammonia decompo-
sition and observed complete conversion from 800 to 950 °C for a
gas mixture containing 1 vol% NH3, 0.4 vol% H2 and N2 as balance.
However, the same catalyst was less effective with real syngas
achieving only 30% conversion at 830 °C and 20 bar. Other authors
observed that 10 wt% Ru supported on SiO2 achieved better ammo-
nia conversion than 10 wt% Ni/SiO2 under similar conditions [113].
In fact Ru catalysts with loading less than 10 wt% on various metal
oxide supports (Al2O3, SiO2, ZrO2, TiO2, MgO and carbon nan-
otubes) have been employed for hydrogen production through
ammonia decomposition [114–116] with better performance than
Ni supported on the same metal oxides under similar conditions.
When deposited on the above supports, the following order
of activity was observed: Ru/Carbon Nanotubes > Ru/MgO >
Ru/TiO2 ffi Ru/Al2O3 ffi Ru/ZrO2 [53]. Ru–Ni catalysts constitute a
powerful catalytic system in ammonia decomposition. Such cat-
alytic system have been successfully used in decomposition of
ammonia coal gasification syngas with conversion of 90% ammonia
conversion at 900 °C [117]. Despite its good catalytic activity in
ammonia decomposition, Ru suffers from catalyst deactivation by
coking and poisoning just like nickel based catalysts previously
discussed. Furthermore, Ru seems to promote the side reaction to
convert hydrogen to water instead of ammonia [118].
3.2.2.4. Other transition metal oxide catalysts. Juutilainen et al. [119]
investigated the removal of ammonia by ZrO2 and ZrO2/Al2O3. In a
comparative test with nickel and dolomite catalysts, they reported
better ammonia conversion below 650 °C for ZrO2 and ZrO2/Al2O3.
Yin et al. [53] compared Ni, Fe, Pt, Pd and Rh against Ru supported
on carbon nanotubes and reported significantly lower performance
for all catalysts relative to Ru. At 500 °C, the same authors reported
the following order in activity for ammonia decomposition:
Ru > Rh > Ni > Pt ffi Pd > Fe. Co3O4 and MoO3 achieved similar NH3
conversion (95%) at 650 °C [55]. It appears that Al2O3, as a sup-
port, contributes to NH3 decomposition as Ni impregnation has
been shown to only increase NH3 conversion on alumina from
23% to 38% [120]. Other metal oxides (ZrO2, WO3 and ZnO)
achieved NH3 conversion between 20% and 30% at 650 °C [55].
303
3.3. Sorbents for sulfur contaminants removal
Unlike tars and ammonia, sulfur removal is achieved with the
help of a sorbent, commonly a metal oxide, which chemically
reacts with sulfur contaminants to produce a metal sulfide accord-
ing the reaction below.
MO þ H2 S ! MS þ H2 O ð12Þ
This process is referred to, interchangeably as desulfurization,
from the point of view of sulfur removal from syngas or sulfidation,
from the point of view of metal oxide conversion to its correspond-
ing metal sulfide.
The most common catalysts employed in desulfurization are
metal oxides with ZnO and CuO often regarded as the reference.
As such, calcium, zinc and copper based sorbents are discussed in
separate sections. Calcium based sorbents are also discussed sepa-
rately as they are great of interest in gasification due to their rela-
tive lower cost when they come in the form minerals such
dolomite or limestone. Table 7 summarizes desulfurization studies
of sulfur contaminants on various catalysts.
3.3.1. Alkaline earth metal based sorbents
Calcium based catalysts, reported to be effective in hot gas
desulfurization, are CaO [9,126], CaCO3 [9] and dolomite [9]. The
gas–solid reaction of calcium based catalysts and hydrogen sulfide
proceeds according to the general metal oxide desulfurization
reaction discussed earlier and adapted to CaO and CaCO3 below.
CaO þ H2 S $ CaS þ H2 O ð13Þ
CaCO3 þ H2 S $ CaS þ CO2 þ H2 O ð14Þ
Calcination plays a crucial role in the activity of calcium based
catalysts. Uncalcined limestone (i.e. CaCO3) is less effective than
fully calcined limestone (i.e. CaO) [9]. A similar behavior is
observed for dolomite as well, although uncalcined dolomite per-
formed better than uncalcined limestone [9]. In dolomite, it was
reported that MgO does not bind sulfur although there is strong
evidence of a synergistic effect between Mg and Ca: the higher
the Mg to Ca ratio, the higher the conversion of Ca to CaS [9].
Both higher temperature and pressure increase H2S conversion
[127].
Conversion of H2S is inhibited by COS, H2 and CO, presumably
by reducing the number of free active sites [127] and likely by
reducing the CaO.
One benefit of calcium based sorbents, such as CaO, is that they
can absorb both sulfur and halogen species such as chlorine
according to the reaction outlined below [128].
CaO þ 2HCl $ CaCl2 þ H2 O ð15Þ
In addition, as other basic metal oxide catalysts of CaO and MgO
have shown activity in tar conversion as discussed above, it is
likely that Ca based sorbents can be of use in simultaneous removal
of multiple contaminants.
3.3.2. Transition metal based catalysts
3.3.2.1. Zinc based catalysts. It is well established that zinc oxide
and its derivatives are very effective sorbents for hydrogen sulfide
removal. Zinc oxide is one the best metal oxides catalyst for desul-
furization based on thermodynamic simulations [129] with the
ability to reduce hydrogen sulfide concentration to less than
1 ppmv in experimental studies [124]. However, the regeneration
of the spent catalyst suffers from several issues. In order to avoid
the formation of zinc sulfate, regeneration must be carried out
under a reducing gas stream. Under a reducing environment, zinc
oxide tends to be reduced to elemental zinc. Furthermore, zinc
oxide is prone to elemental zinc volatilization at temperatures
304 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307

Table 7
Catalytic desulfurization studies of syngas sulfur (mostly H2S) contaminants.

Catalysts Gas composition Reaction conditions Performance Ref.

Breakthrough Sorbent capacity (gS/gSorbent)
time (min)
Zn–Fe– H2S: 1, H2: 20, H2O: 7–20, N2: 72 T: 500–700 °C, GHSV: n/a 0.156–0.324 at 500 °C [121]
V–O 4000 h1 0.105–0.288 at 700 °C
LaFeO3/ H2S: 0.33, H2: 10.5, CO2 (or CO): 17.1, N2: 72.1 T: 450–550 °C, WHSV: n/a 0.0035 CO2 [122]
M41 9000 ml h1 g1 0.0324 with CO
La2ZnO4/ T: 450–550 °C, WHSV: n/a 0.35 CO2, 0.94 CO (g/100 [122]
M41 9000 ml h1 g1 sorbent)
La2CoO4/ T: 450–550 °C, WHSV: n/a 0.35 with CO2 [122]
M41 9000 ml h1 g1 0.0153 with CO
Ni–Zn/ H2S: 1, H2: 11.7, CO: 9.56, CO2: 5.20, NH3: 0.50, N2 balance T: 650 °C, F: 50 ml/min, 250–450 0.13 [94]
Al2O3 GHSV: 5000 h1
Fe–Zn/ T: 650 °C, F: 50 ml/min, 250–450 0.16 [94]
Al2O3 GHSV: 5000 h1
Co–Zn/ T: 650 °C, F: 50 ml/min, 250–450 Better than previous two [94]
Al2O3 GHSV: 5000 h1 catalyst but not indicated
Zn–Ti H2S: 1.5, H2: 11.7, CO: 9.6, CO2: 5.2, N2: balance T: 480 °C, 650 °C, F: 50 ml/ n/a <0.15 at 480 °C [123]
min, GHSV: 5000 h1 0.15–0.20 at 650 °C
Co–Zn–T n/a 0.05–0.17 at 480 °C [123]
0.20–0.25 at 650 °C
ZnO H2S: 0.01, He balance T: 250 °C, F: 400 ml/min 550 0.48 [124]
Cu–Ce H2S: 0.0394, CO: 10.5, H2: 10.5, CO2: 8.3, CH4: 6.3, C2H4: 1.7, T: 700 °C, GHSV: 56,000 h1 n/a 0.0004 [125]
Benzene: 0.0370, H2O: 0, 45 T: 700 °C, GHSV: 14,000 h1 0.0005
Mn/ H2S: 0.0394, CO: 10.5, H2: 10.5, CO2: 8.3, CH4: 6.3, C2H4: 1.7, T: 700 °C, GHSV: 3300 h1 n/a 0.004 (w/steam)
Al2O3 Benzene: 0.0370, H2O: 0, 25–65 0.0002 (w/steam)

n/a means not available, T = temperature, F = flow rate, GHSV = gas hourly space velocity, WHSV = weight hourly space velocity.

above 600 °C [130]. These drawbacks significantly limit the long
term performance of zinc oxide as a hydrogen sulfide sorbent. As
a remedy to these issues, iron oxide and titanium oxide are com-
monly incorporated with zinc oxide in the catalyst structure. Of
the two metal oxides, it appears that the addition of Fe2O3 is less
effective than that of TiO2 as a solution for the issues aforemen-
tioned [130]. While zinc ferrites are more active than zinc oxide
alone, they too are subjected to decomposition under gasification
condition [131]. On the other hand, zinc titanates have a similar
performance to zinc oxide: Fe2O3 synergistically participates in
the desulfurization reaction whereas TiO2 does not [131].
However, TiO2 improves thermal stability [129], prevents ZnO
reduction in the reducing environment of gasification [129,132]
if adequate dispersion is achieved and inhibits zinc sulfate forma-
tion [129] in the presence of excess oxygen. Other mixed metal
oxide catalysts of zinc oxide/titanates and other metal oxides such
ZrO2 [133], Co3O4 [123], CuO [134] and MnO [134] have been
investigated with promising results. Cobalt, in particular, appears
to be a promising promoter for zinc titanates as it has been
reported to enhance high (650 °C) and medium (480 °C) tempera-
ture desulfurization [123].
3.3.2.2. Copper based catalysts. Thermodynamically, copper, along
with zinc, is the most suitable transition metal for hydrogen sulfide
absorption. Unlike zinc, copper based sorbents have high sorption
capacity under reducing as well as oxidizing environments
although unsupported copper oxide is reduced to elemental cop-
per, a less effective sorbent, in gasification environment [130].
Besides reduction, copper oxide suffers from a loss of surface area
due to the formation of a sulfide layer [135]. This shortcoming can
be alleviated through dispersion on to a support in order to
improve its stability and increase the surface area. Dispersion on
a support can significantly improve catalytic activity to that of dis-
persed zinc oxide [136] with the hydrogen sulfide sorption capac-
ity being directly proportional to the quantity of copper oxide
dispersed on the support [137]. Copper oxide has been dispersed
on c-Al2O3 [138], SiO2 [137], ZrO2 [139], V2O5 [140], Ce [125]
and ZnO [134]. However, even supported copper based catalysts
can be affected by the presence of hydrogen as evidenced by the
decrease in breakthrough time in the presence of equimolar H2S
and H2 for copper supported on olivine [141].
3.3.2.3. Other transition metal oxides. Iron oxide has attracted atten-
tion as hydrogen sulfide sorbent due to its thermodynamically
favorable properties and relatively low cost [142]. However, in
the temperature ranges of interest, iron oxide catalyzes the
Boudouard reaction and thus leading to potential catalyst deactiva-
tion due to coke formation [129]. Nickel based catalysts, primarily
tested for tar and NH3 removal, can also remove H2S as well as
other sulfur based contaminants in syngas [5]. The reported H2S
removal capacity is consistent with the well-known poisoning
effect of sulfur compounds on Ni based catalysts.
Ceria based sorbents, either as active supports or as mixed
metal oxides, have been investigated as H2S sorbents [125,143].
It appears that ceria sorbent capacity for sulfur is significantly
diminished by reduction from H2 in syngas which can be alleviated
by steam in syngas [125].
3.4. Sorbents for hydrogen halides contaminant removal
The body of research on halide cleanup is not as exhaustive as
that of tar, ammonia or hydrogen sulfide in syngas, especially for
biomass derived syngas [144]. Due to its abundance, relative to
other hydrogen halides in syngas, HCl has been the target of most
research investigations.
Hydrogen halides are removed from syngas via dehydrohalo-
genation. In the specific case of HCl, this process is referred to as
dechlorination or dehydrochlorination and proceeds according to
the reactions given below:
MO þ 2HCl ! MCl2 þ H2 O ð16Þ
MCO3 þ 2HCl ! MCl2 þ H2 O þ CO2 ð17Þ
M2 CO3 þ 2HCl ! 2MCl þ H2 O þ CO2 ð18Þ
where M is metal elements. Thermodynamic simulations have indi-
cated that alkaline and alkali earth metals are promising halide sor-
bents for cleanup temperatures greater than 500 °C [145].
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307
3.4.1. Alkali metal catalysts
Alkali metals of interest in halides removal are Li, Na and K with
the following order established in terms of sorbent capability to
achieve high HCl removal in temperature ranges for hot gas
cleanup based on thermodynamic modeling [12].
Na2O  K2O  Li2O. Na2O and K2O are more suitable for high tem-
perature application with less than 5 ppmv of HCl up to 600 °C
whereas Li2O is only effective below 200 °C [144]. In syngas con-
taining high concentration of CO2, alkali metal oxides will be con-
verted to their respective carbonates (i.e. Li2CO3, Na2CO3 and
K2CO3). Na2CO3 tested in a simulated coal syngas with 200 ppmv
HCl was effective in achieving a concentration less than 1 ppmv
with a breakthrough time near 200 min [12]. Duo et al. [146]
reported the following order for Na carbonates sorbents:
Na2CO3.10H2O > Na2CO3 > NaHCO3.
In addition to HCl, alkali based sorbents can additionally
remove other halides. Na2CO3 can achieve complete removal below
600 °C for HI and HBr and below 400 °C for HF [144].
3.4.2. Alkaline earth metal catalysts
Alkaline earth metals of interest as HCl sorbents are Mg, Ca, Sr,
Ba. In their oxide form, these elements can be used as sorbent with
the following order established in terms of sorbent capability to
achieve high HCl removal over in hot gas cleanup based on thermo-
dynamic modeling: BaO > SrO > CaO > MgO. These sorbents are
more effective when they are not converted to their corresponding
carbonate. CaCO3 was reported to be less effective in HCl removal
than CaO [145,146]. Based on experimental studies on a simulated
syngas, the following order was observed for calcium based sor-
bents: CaO = Ca(OH)2 > CaCO3 [128,146].
3.4.3. Transition metal oxides
It appears that selected transition metal oxides (MnO, CoO, ZnO,
NiO, Y2O3, Cu2O, FeO) can be HCl sorbents albeit less effective than
alkali and alkaline earth metals based on a thermodynamic study.
These sorbents could be effective for low temperature range
dechlorination (<200 °C) [144].
3.5. Sorbents for trace metals removal
Granite et al. [147] investigated the removal of mercury (Hg)
from nitrogen as well as a gas mixture emulating syngas on solid
sorbents of metal oxides of Ir, Pt, Ru, Pd, Rh, Ti and Ag at 204,
288 and 371 °C. In nitrogen, Ir, Pt, Pd and Ru based sorbents per-
formed best at 288 °C with removal efficiencies of 93%, 100%,
65% and 65%, respectively. Low removal efficiencies were reported
at 204 and 371 °C for the sorbents as well as other (Rh, Ti and Ag)
sorbents. However, in an emulated syngas mixture (27% H2, 36%
CO, 4% CO2, 240 ppb mercury, balance N2), only Pt and Pd based
sorbents were effective at the same temperature with removal effi-
ciencies of 100% and 54%, respectively.
Metals can be conjointly removed along with halides in the
form of halide salts. For example, the removal of alkali metal
vapors often occurs in conjunction with halides as alkali metals
tend to combine to halides to form salts (NaCl, KCl, etc.) in syngas.
In a study focusing on alkali metals removal, fly ash, bentonite,
kaolin and bauxite were investigated as potential sorbents. The
authors reported that all sorbents reduced the release of alkali
metals although fly ash and bentonite recorded the best perfor-
mance [148].
Kaolinite as well as zeolite were also effective in alkali vapors
from syngas with the presence of H2O and H2S enhancing NaCl cap-
ture [149]. In another experimental study, several aluminosilicate
sorbents (bauxite, zeolites, bentonite, kaolinite) were effective in
reducing alkali metal concentrations with kaolinite, bauxite and
zeolites achieving the best performance.
305
4. Technical barriers in syngas cleanup and future opportunities
In 1998, Tom Reed nicely summed up the impact of syngas con-
taminants on the economic viability of gasification projects when
he stated that ‘‘The typical project starts with new ideas, announce-
ments at meetings, construction of the new gasifier. Then it is found
that the gas contains 0.1–10% ‘tars’. The rest of the time and money
is spent trying to solve this problem.’’ [2]. Unfortunately, Reed’s
statement is still very relevant over 15 years later with many gasi-
fication and thermochemical demonstration plants companies
struggling to overcome issues caused by these impurities.
According to a recent survey of a panel of experts, the control
and management of syngas contaminants is one of the biggest
deterrent to the commercial deployment of biomass gasification
[150].
The challenge of syngas contaminants is partially due to the
inherent variability of biomass. In the context of a biorefinery
where a wide portfolio of biomass is required to guarantee contin-
uous supply, the wide variability in biomass implies developing
customizable and flexible solutions to deal with varying levels of
contaminants. It is consequently necessary to develop economical
solutions that can handle a wide range of contaminant concentra-
tions and that can be applied at various application scales. While
effective cold gas cleanup technologies exist to remove virtually
all contaminants down to tolerable limits, this approach’s main
drawback is the reduction in the overall efficiency due to syngas
cooling and the increase in cost due to increase in complexity
and the need to treat effluent streams. Consequently, effective
existing cold gas cleanup solutions are more suitable to large scale
applications at the moment due to their higher cost [28]. As such,
activities in this area should aim to develop gas cleanup strategies
that will improve efficiency and decrease cost. Implicit efforts
should be made to develop feedstock and scale flexible solutions
so as to account for the wide variability in biomass composition
and application scales. In hot gas cleanup, despite extensive
research in recent years as illustrated in Tables 5–7, advancements
in studies have not translated to cost effective commercial solu-
tions and with limited pilot demonstration. The major limitation
of hot gas cleanup catalysts and sorbents is rapid catalyst deactiva-
tion [151]. The emphasis has been in most reported studies to
develop exotic and seemingly very active cleanup catalysts and
sorbents. However, this approach provides limited information as
most of these studies do not evaluate material lifetime, an essential
criterion when assessing the commercial potential of new catalysts
or sorbents. Consequently, serious attention must be devoted to
evaluating the activity and lifetime of hot gas cleanup catalysts
and sorbents in future studies. In both cold and hot gas cleanup,
solutions effective in multi-contaminant removal would be desir-
able to drive cost down.
5. Conclusions
Biomass has the potential to displace petroleum and other fossil
fuels via syngas generation through gasification and subsequent
conversion of syngas to fuel, energy and products. However, the
presence of deleterious contaminant species at concentrations
detrimental to most downstream applications hinders the utility
of biomass derived syngas. Cost effective and efficient remediation
technology of syngas contaminants is one of the biggest challenges
to commercial deployment of biomass gasification at large scale.
Effective cold gas cleanup technologies exist to remove virtually
all contaminants down to desirable lower concentration limits.
However, the main drawbacks of these technologies are that they
(i) reduce in the overall efficiency due to syngas cooling and (ii)
increase cost due to their complexity and the need to treat effluent
306 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307

streams. As such, activities in this area should aim to develop gas gasification Task Order 739656-00100, Revised Final Report, SFA Pacific,
Inc., Mountain View, CA, U.S. Department of Energy, National Energy
cleanup strategies that will improve efficiency and decrease cost.
Technology Laboratory, Pittsburgh, PA, 2002.
In addition to the cold gas cleanup approaches, warm and hot [35] Descamps C, Bouallou C, Kanniche M. Energy 2008;33:874–81.
gas cleanup approaches have also been investigated. Based on this [36] Burr B, Lyddon L. Bryan, Texas: Bryan Research & Engineering, Inc., 2008.
review, few catalysts or sorbents have demonstrated good activity [37] Kohl AL, Nielsen R. Gas purification. 5th ed. Houston, TX: Gulf Professional
Publishing; 1997.
in contaminant removal. The most effective catalysts and sorbents [38] Haslbeck JL, Kuehn NJ, Lewis EG, Pinkerton LL, Simpson J, Turner MJ, et al.
are: Ni, Fe and mineral based catalysts for tar, Ru and Ni based cat- Cost and performance baseline for fossil energy plants. Bituminous coal and
alysts for NH3, Zn and Cu based sorbents for H2S, alkali and alkaline natural gas to electricity DOE/NETL-2010/1397, vol. 1. Morgantown, WV:
National Energy Technology Laboratory, September 2013.
earth metal oxides and carbonates based (Na2CO3, K2CO3, CaO, [39] Bandosz TJ. J Colloid Interface Sci 2002;246:1–20.
CaCO3) and clay mineral sorbents (kaolinite and bentonite) for [40] Huang C-C, Chen C-H, Chu S-M. J Hazard Mater 2006;136:866–73.
halides and trace metals. The dominant persisting challenge with [41] Bara JE. In: Green solvents II. Springer; 2012. p. 155–67.
[42] Heintz YJ, Sehabiague L, Morsi BI, Jones KL, Luebke DR, Pennline HW. Energy
hot gas cleanup catalysts and sorbents is rapid catalyst deactiva- Fuels 2009;23:4822–30.
tion. This challenge is perhaps persisting due to the lack of serious [43] Hanioka S, Maruyama T, Sotani T, Teramoto M, Matsuyama H, Nakashima K,
attention to deactivation in many catalytic hot gas cleanup studies. et al. J Membr Sci 2008;314:1–4.
[44] Mondal P, Dang GS, Garg MO. Fuel Process Technol 2011;92:1395–410.
Consequently, in the future, it recommends that serious attention [45] Allegue LB, Hinge J. Biogas and bio-syngas upgrading. Aarhus,
be devoted to evaluate catalyst resistance to deactivation in hot Denmark: Danish Technological Institute; 2012.
gas cleanup studies. [46] Paasen SV, Cieplik M, Phokawat N. Gasification of non-woody biomass ECN-
C-06-032. Petten, Netherlands: Energy Research Centre of the Netherlands;
2006.
Acknowledgements [47] Kirk RE, Othmer DF. Encyclopedia of Chemical Technology, 5th ed., vol. 20.
New York (NY): Wiley-Interscience Publication; 2004.
[48] Salo K, Mojtahedi W. Biomass Bioenergy 1998;15:263–7.
The authors would like to acknowledge the Alabama [49] Stevens DJ. Hot gas conditioning: recent progress with larger-scale biomass
Agricultural Experiment Station (AAES) and the U.S. Department gasification systems NREL/SR-510-29952. Golden, Colorado: National
of Agriculture-National Institute of Food and Agriculture Renewable Energy Laboratory; 2001.
[50] Liu H, Wang H, Shen J, Sun Y, Liu Z. Catal Today 2008;131:444–9.
(USDA-NIFA) for funding this project.
[51] Ferella F, Stoehr J, Michelis ID, Hornung A. Fuel 2013;105:614–29.
[52] Zhang R, Brown RC, Suby A, Cummer K. Energy Convers Manage
References 2004;45:995–1014.
[53] Yin S-F, Zhang Q-H, Xu B-Q, Zhu W-X, Ng C-F, Au C-T. J Catal
2004;224:384–96.
[1] Devi L, Ptasinski KJ, Janssen FJJG. Biomass Bioenergy 2003;24:125–40.
[54] Mojtahedi W, Ylitalo M, Maunula T, Abbasian J. Fuel Process Technol
[2] Milne TA, Evans RJ, Abatzoglou N. Biomass gasifier ‘‘tars’’: their nature,
1995;45:221–36.
formation, and conversion NREL/TP-570-25357. Golden, Colorado: National
[55] Park JJ, Park CG, Jung SY, Lee SC, Ragupathy D, Kim JC. Res Chem Intermed
Renewable Energy Laboratory; 1998.
2011;37:1193–202.
[3] Abdoulmoumine N, Kulkarni A, Adhikari S. Ind Eng Chem Res
[56] Ng PF, Li, Wang S, Zhu Z, Lu G, Yan Z. Environ Sci Technol 2007;41:3758–62.
2014;53:5767–77.
[57] Li L, Wang S, Zhu Z, Yao X, Yan Z. Fuel Process Technol 2008;89:1106–12.
[4] Zhou J, Masutani SM, Ishimura DM, Turn SQ, Kinoshita CM. Ind Eng Chem Res
[58] Bridgwater A. Appl Catal A 1994;116:5–47.
2000;39:626–34.
[59] Sutton D, Kelleher B, Ross JRH. Fuel Process Technol 2001;73:155–73.
[5] Cui H, Turn SQ, Keffer V, Evans D, Tran T, Foley M. Energy Fuels
[60] Coll R, Salvado J, Farriol X, Montane D. Fuel Process Technol 2001;74:19–31.
2010;24:1222–33.
[61] Twigg MV, editor. Catalyst handbook. New York (NY): Oxford University
[6] Hurley S, Xu C, Preto F, Shao Y, Li H, Wang J, et al. Fuel 2012;91:170–6.
Press; 1997.
[7] Zhou J, Masutani SM, Ishimura DM, Turn SQ, Kinoshita CM. Ind Eng Chem Res
[62] Leppälahti J, Kurkela E. Fuel 1991;70:491–7.
2000;39:626–34.
[63] Link S, Arvelakis S, Paist A, Martin A, Liliedahl T, Sjöström K. Appl Energy
[8] Abdoulmoumine N, Kulkarni A, Adhikari S. Ind Eng Chem Res
2012;94:89–97.
2014;53:5767–77.
[64] Simell P, Kurkela E, Ståhlberg P, Hepola J. Catal Today 1996;27:55–62.
[9] Yrjas P, Iisa K, Hupa M. Fuel 1996;75:89–95.
[65] Wang T, Chang J, Lv P, Zhu J. Energy Fuels 2005;19:22–7.
[10] Van der Drift A, Van Doorn J, Vermeulen J. Biomass Bioenergy 2001;20:45–56.
[66] Orio A, Corella J, Narvaez I. Ind Eng Chem Res 1997;36:3800–8.
[11] Cui H, Turn SQ, Keffer V, Evans D, Tran T, Foley M. Fuel 2013;108:1–12.
[67] Dou B, Gao J, Sha X, Baek S. Appl Therm Eng 2003;23:2229–39.
[12] Ohtsuka Y, Tsubouchi N, Kikuchi T, Hashimoto H. Powder Technol
[68] Kuhn J, Zhao Z, Felix L, Slimane R, Choi C, Ozkan U. Appl Catal B
2009;190:340–7.
2008;81:14–26.
[13] Woolcock PJ, Brown RC. Biomass Bioenergy 2013;52:54–84.
[69] Devi L, Ptasinski K, Janssen F. Fuel Process Technol 2005;86:707–30.
[14] Woolcock PJ, Brown RC. Biomass Bioenergy 2013;52:54–84.
[70] Michel R, Łamacz A, Krzton A, Djéga-Mariadassou G, Burg P, Courson C, et al.
[15] Abu El-Rub Z, Bramer E, Brem G. Ind Eng Chem Res 2004;43:6911–9.
Fuel 2013;109:653–60.
[16] Dayton D. A review of the literature on catalytic biomass tar destruction
[71] D’Orazio A, Di Carlo A, Dionisi N, Dell’Era A, Orecchini F. Int J Hydrogen Energy
NREL/TP-510-32815. Golden, Colorado: National Renewable Energy
2013;38:13282–92.
Laboratory; 2002.
[72] Xu C, Donald J, Byambajav E, Ohtsuka Y. Fuel 2010;89:1784–95.
[17] Torres W, Pansare SS, Goodwin Jr JG. Catal Rev 2007;49:407–56.
[73] Kim Y-K, Park J-I, Jung D, Miyawaki J, Yoon S-H, Mochida I. J Ind Eng Chem
[18] Cheah S, Carpenter DL, Magrini-Bair KA. Energy Fuels 2009;23:5291–307.
2014;20:216–21.
[19] Yung MM, Jablonski WS, Magrini-Bair KA. Energy Fuels 2009;23:1874–87.
[74] Corella J, Aznar M-P, Gil J, Caballero MA. Energy Fuels 1999;13:1122–7.
[20] Bhave AG, Vyas DK, Patel JB. Renew Energy 2008;33:1716–20.
[75] Sun Y, Jiang J, Kantarelis E, Xu J, Li L, Zhao S, et al. Catal Commun
[21] Jansen JlC, Jönsson K, Hagman M. Water Sci Technol 2002;46:59–65.
2012;20:36–40.
[22] Bergman PC, van Paasen SV, Boerrigter H. Pyrolysis and gasification of
[76] Perez P, Aznar P, Caballero M, Gil J, Martin J, Corella J. Energy Fuels
biomass and waste, expert meeting, Strasbourg, France; 2002.
1997;11:1194–203.
[23] Zwart RWR, Van der Drift A, Bos A, Visser HJM, Cieplik MK, Könemann HWJ.
[77] Corella J, Orio A, Toledo J. Energy Fuels 1999;13:702–9.
Environ Prog Sustain Energy 2009;28:324–35.
[78] Hu G, Xu S, Li S, Xiao C, Liu S. Fuel Process Technol 2006;87:375–82.
[24] Zwart R. Gas cleaning downstream biomass gasification: status Report
[79] Guan G, Chen G, Kasai Y, Lim EWC, Hao X, Kaewpanha M, et al. Appl Catal B
2009. Petten, Netherlands: Energy research Centre of the Netherlands; 2009.
2012;115–116:159–68.
[25] Chen H, Namioka T, Yoshikawa K. Appl Energy 2011;88:5032–41.
[80] Li L, Morishita K, Takarada T. Fuel 2007;86:1570–6.
[26] Gil J, Caballero MA, Martin JA, Aznar M-P, Corella J. Ind Eng Chem Res
[81] Matsumura A, Sato S, Kondo T, Saito I, de Souza W. Fuel 2005;84:417–21.
1999;38:4226–35.
[82] Matsumura A, Kondo T, Sato S, Saito I, de Souza W. Fuel 2005;84:411–6.
[27] Phuphuakrat T, Namioka T, Yoshikawa K. Bioresour Technol 2011;102:543–9.
[83] Nordgreen T, Liliedahl T, Sjöström K. Fuel 2006;85:689–94.
[28] Neubauer Y. J Sustain Energy Environ Spec Issue 2011;67:71.
[84] Leppälahti J, Simell P, Kurkela E. Fuel Process Technol 1991;29:43–56.
[29] Resnik KP, Pennline HW. Fuel 2013;105:184–91.
[85] Dou B, Gao J, Sha X, Baek SW. Appl Therm Eng 2003;23:2229–39.
[30] Kim YM, Park D, Lee DS, Park JM. J Hazard Mater 2008;152:915–21.
[86] Huber GW, Iborra S, Corma A. Chem Rev 2006;106:4044–98.
[31] Pinto F, Lopes H, André RN, Dias M, Gulyurtlu I, Cabrita I. Energy Fuels
[87] Dou B, Gao J, Sha X, Baek S. Appl Therm Eng 2003;23:2229–39.
2007;21:2737–45.
[88] Buchireddy PR, Bricka RM, Rodriguez J, Holmes W. Energy Fuels
[32] Proll T, Siefert IG, Friedl A, Hofbauer H. Ind Eng Chem Res 2005;44:1576–84.
2010;24:2707–15.
[33] Khakharia P, Huizinga A, Jurado Lopez C, Sanchez Sanchez C, de Miguel
[89] Miyazawa T, Kimura T, Nishikawa J, Kado S, Kunimori K, Tomishige K. Catal
Mercader F, Vlugt TJH, et al. Ind Eng Chem Res 2014;53:13195–204.
Today 2006;115:254–62.
[34] Korens N, Simbeck DR, Wilhelm DJ, Longanbach JR, Stiegel GJ. Process
[90] Sutton D, Kelleher B, Doyle A, Ross J. Bioresour Technol 2001;80:111–6.
screening analysis of alternative gas treating and sulfur removal for
 N. Abdoulmoumine et al. / Applied Energy 155 (2015) 294–307
[91] Swierczynski D, Libs S, Courson C, Kiennemann A. Appl Catal B
2007;74:211–22.
[92] Agblevor F, Beis S, Mante O, Abdoulmoumine N. Ind Eng Chem Res
2010;49:3533–8.
[93] Kaewpanha M, Guan G, Hao X, Wang Z, Kasai Y, Kakuta S, et al. J Taiwan Inst
Chem E 2013;44:1022–6.
[94] Jung SY, Lee SJ, Park JJ, Lee SC, Jun HK, Lee TJ, et al. Sep Purif Technol
2008;63:297–302.
[95] Zheng W, Zhang J, Ge Q, Xu H, Li W. Appl Catal B 2008;80:98–105.
[96] Zhang H, Alhamed YA, Kojima Y, Al-Zahrani AA, Miyaoka H, Petrov LA. Int J
Hydrogen Energy 2014;39:277–87.
[97] Lorenzut B, Montini T, Pavel CC, Comotti M, Vizza F, Bianchini C, et al.
ChemCatChem 2010;2:1096–106.
[98] Ozawa Y, Tochihara Y. Catal Today 2011;164:528–32.
[99] Prasad V, Karim A, Arya A, Vlachos D. Ind Eng Chem Res 2009;48:5255–65.
[100] Ohtsuka Y, Xu C, Kong D, Tsubouchi N. Fuel 2004;83:685–92.
[101] Tsubouchi N, Hashimoto H, Ohtsuka Y. Powder Technol 2008;180:184–9.
[102] Corella J, Toledo JM, Padilla R. Energy Fuels 2004;18:713–20.
[103] Bjoerkman E, Sjoestroem K. Energy Fuels 1991;5:753–60.
[104] Simell PA, Hirvensalo EK, Smolander VT, Krause AOI. Ind Eng Chem Res
1999;38:1250–7.
[105] Tsubouchi N, Hashimoto H, Ohtsuka Y. Energy Fuels 2007;21:3063–9.
[106] Donald J, Xu C, Hashimoto H, Byambajav E, Ohtsuka Y. Appl Catal A
2010;375:124–33.
[107] Mojtahedi W, Abbasian J. Fuel 1995;74:1698–703.
[108] Nassos S, Elm Svensson E, Boutonnet M, Järås S. Appl Catal B
2007;74:92–102.
[109] Wang W, Padban N, Ye Z, Olofsson G, Andersson A, Bjerle I. Ind Eng Chem Res
2000;39:4075–81.
[110] Chellappa A, Fischer C, Thomson W. Appl Catal A 2002;227:231–40.
[111] Wang W, Padban N, Ye Z, Andersson A, Bjerle I. Ind Eng Chem Res
1999;38:4175–82.
[112] Depner H, Jess A. Fuel 1999;78:1369–77.
[113] Li XK, Ji WJ, Zhao J, Wang SJ, Au CT. J Catal 2005;236:181–9.
[114] Choudhary T, Sivadinarayana C, Goodman D. Chem Eng J 2003;93:69–80.
[115] Yin S, Xu B, Zhou X, Au C. Appl Catal A 2004;277:1–9.
[116] Zhang J, Xu H, Ge Q, Li W. Catal Commun 2006;7:148–52.
[117] Ozawa Y, Tochihara Y. Chem Eng Sci 2007;62:5364–7.
[118] Ozawa Y, Tochihara Y. Chem Eng Sci 2007;62:5364–7.
[119] Juutilainen SJ, Simell PA, Krause AOI. Appl Catal B 2006;62:86–92.
[120] Mojtahedi W, Ylitalo M, Maunula T, Abbasian J. Fuel Process Technol
1995;45:221–36.
[121] Akyurtlu JF, Akyurtlu A. Gas Sep Purif 1995;9:17–25.
[122] Liu B, Wan Z, Zhan Y, Au C. Fuel 2012;98:95–102.
307
[123] Jun HK, Lee TJ, Ryu SO, Kim JC. Ind Eng Chem Res 2001;40:3547–56.
[124] Rosso I, Galletti C, Bizzi M, Saracco G, Specchia V. Ind Eng Chem Res
2003;42:1688–97.
[125] Cheah S, Parent YO, Jablonski WS, Vinzant T, Olstad JL. Fuel 2012;97:612–20.
[126] Heesink ABM, Van Swaaij WPM. Chem Eng Sci 1995;50:2983–96.
[127] Heesink ABM, Van Swaaij WPM. Chem Eng Sci 1995;50:2983–96.
[128] Wang W, Ye Z, Bjerle I. Fuel 1996;75:207–12.
[129] Elseviers WF, Verelst H. Fuel 1999;78:601–12.
[130] Meng X, de Jong W, Pal R, Verkooijen AHM. Fuel Process Technol
2010;91:964–81.
[131] Lew S, Sarofim AF, Flytzani-Stephanopoulos M. Ind Eng Chem Res
1992;31:1890–9.
[132] Jothimurugesan K, Gangwal SK. Ind Eng Chem Res 1998;37:1929–33.
[133] Sasaoka E, Sada N, Manabe A, Uddin MA, Sakata Y. Ind Eng Chem Res
1999;38:958–63.
[134] Bu X, Ying Y, Zhang C, Peng W. Powder Technol 2008;180:253–8.
[135] Kyotani T, Kawashima H, Tomita A, Palmer A, Furimsky E. Fuel 1989;68:74–9.
[136] Ko T-H, Chu H, Chaung L-K. Chemosphere 2005;58:467–74.
[137] Karvan O, Atakül H. Fuel Process Technol 2008;89:908–15.
[138] Buelna G, Lin YS. Sep Purif Technol 2004;39:167–79.
[139] Baeza P, Aguila G, Gracia F, Araya P. Catal Commun 2008;9:751–5.
[140] Bineesh KV, Kim D-K, Kim M-IL, Park D-W. Appl Clay Sci 2011;53:204–11.
[141] Hachimi A, Vilcocq L, Courson C, Kiennemann A. Fuel Process Technol
2014;118:254–63.
[142] Wang J, Zhang Y, Han L, Chang L, Bao W. Fuel 2013;103:73–9.
[143] Zeng Y, Kaytakoglu S, Harrison D. Chem Eng Sci 2000;55:4893–900.
[144] Dolan MD, Ilyushechkin AY, McLennan KG, Sharma SD. Asia-Pac J Chem Eng
2011. n/a–n/a.
[145] Krishnan GN, Gupta R, Ayala R. Development of disposable sorbents for
chloride removal from high temperature coal-derived gases DE-AC21-
93MC30005. Morgantown, WV: U.S. Department of Energy; 1999.
[146] Duo W, Kirkby N, Seville J, Kiel J, Bos A, Den Uil H. Chem Eng Sci
1996;51:2541–6.
[147] Granite EJ, Myers CR, King WP, Stanko DC, Pennline HW. Ind Eng Chem Res
2006;45:4844–8.
[148] Bläsing M, Müller M. Combust Flame 2013;160:3015–20.
[149] Inai M, Ronhunshu KK. Fuel Energy Abstr 2000;41:313.
[150] Liliedahl T, Heyne S, Marklund Magnus. Biomass gasification – a synthesis of
technical barriers and current research issues for deployment at large
scale. Göteborg, Sweden: The Swedish Knowledge Centre for Renewable
Transportation Fuels; 2013.
[151] Bain RL, Dayton DC, Carpenter DL, Czernik SR, Feik CJ, French RJ, et al. Ind Eng
Chem Res 2005;44:7945–56.