Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34

Contents lists available at ScienceDirect

Journal of Analytical and Applied Pyrolysis

journal homepage: www.elsevier.com/locate/jaap

Catalytic effects of magnesium on the characteristics of fast pyrolysis

products – Bio-oil, bio-char, and non-condensed pyrolytic gas
fractions夽
Hyewon Hwang, Shinyoung Oh, In-Gyu Choi, Joon Weon Choi ∗
Department of Forest Sciences and Research Institute for Agriculture and Life Science, Seoul National University, 599 Gwanak-ro, Gwanak-gu, Seoul
151-921, Republic of Korea

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

Article history: Fast pyrolysis of yellow poplar impregnated with different concentrations of MgCl2 was carried out
Available online 13 October 2014 at 450, 500, and 550 ◦ C to investigate the effects of magnesium on the characteristics of the pyrolytic
products. Analysis of the physicochemical properties of bio-oil, bio-char, and non-condensed pyrolytic
Keywords: compounds was carried out accordingly. The results indicate that the yield of char as well as some physi-
Fast pyrolysis cochemical properties of the bio-oil was influenced by the magnesium concentration. The water content
Bio-oil
increased due to a dehydration reaction induced by magnesium. Additionally, the viscosity significantly
Magnesium
increased from 45 cSt to 216 cSt as the magnesium content increased. Magnesium promoted a repoly-
Recombination
Inorganic distribution
merization reaction leading to increases of the average molecular weight and solid content of the bio-oil
Aromatic hydrocarbons from 950 Da to 1670 Da and 0.37 wt% to 0.73 wt%, respectively. Also, the amounts of levoglucosan and
other small molecules detected by gas chromatography-mass spectroscopy analysis decreased as large
fractions such as oligomers or char fines were formed. Meanwhile, the yield and elemental composi-
tion of the bio-oil barely changed. Following the pyrolysis, most of the inorganic metals remained in the
bio-char. Furthermore, the various forms of aromatic hydrocarbons obtained from the non-condensed
pyrolytic gas fractions were identified and quantified in this study.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction It has been reported that inherently present inorganic elements

The development of alternative energy processes has become
imperative due to concerns about the depletion of petroleum
resources. Lignocellulosic biomass is an abundant and renewable
source that can be converted into fuel and chemical feedstocks.
Fast pyrolysis is a promising process for the production of liquid
fuels from biomass. This process is simpler and more inexpen-
sive than other conversion technologies such as bio-fermentation
and gasification [1,2]. In addition, liquid bio-oil fuel can be used
in industries for power generation and has potential as a trans-
portation fuel after being subjected to an upgrading process
[3].
Biomass, however, contains inorganic elements (mainly K,
Mg, and Ca) and their amounts vary depending on the source.
Determination of the effects of inorganic elements on the pyrol-
ysis of biomass is important because of their unique impacts.
夽 Selected Paper from Pyrolysis 2014, Birmingham, U.K. 19–23 May 2014.
∗ Corresponding author. Tel.: +82 2 880 4788; fax: +82 2 873 2318.
E-mail address: cjw@snu.ac.kr (J.W. Choi).
in biomass can cause several problems such as corrosion and
deposition during processes [4]. In addition, they also influence
the decomposition behavior of biomass and char formation [5,6].
With the removal of inorganic metals, the maximum decompo-
sition temperature of biomass increases while char formation is
deteriorated. With regard to bio-oil properties, the production of
sugar and water in bio-oil was enhanced in the presence of inor-
ganic metals [7]. It is known that the metals can play a catalytic
role during pyrolysis although the mechanism is not completely
understood.
Meanwhile, there is a lack of studies concerning the effects of
individual inorganic elements as well as their concentrations on
the fast pyrolysis process and properties of the pyrolytic products.
On the other hand, alkaline earth metals, which are essential inor-
ganic elements in biomass, have been reported to have conflicting
influences and little is known concerning their effects on bio-oil
[8,9]. Furthermore, bio-oil and bio-char contain significant amounts
of inorganic elements originating from biomass which are related
to further applications in diverse fields [4,10]. Thus, understand-
ing of the distribution of inorganic elements remaining in pyrolytic
products is also required.

http://dx.doi.org/10.1016/j.jaap.2014.09.028
0165-2370/© 2014 Elsevier B.V. All rights reserved.
28 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34

In our previous study, the effects of potassium were investigated

in which the decomposition of organic components was induced
to form small molecules [10]. Also, magnesium and calcium consti-
tute a half of total inorganic contents in yellow poplar, but calcium
was rarely removed by demineralization treatment due to its strong
combination with biomass. In this study, we investigated the effects
of magnesium, not only one of major elements in biomass but also
easy one to control, on the characteristics of pyrolytic products,
especially bio-oil, produced by fast pyrolysis at various tempera-
tures (450, 500, and 550 ◦ C). Different concentrations of magnesium
were impregnated in the biomass samples and characterization
of the fast pyrolysis process, bio-oil, bio-char, and non-condensed
pyrolytic compounds was conducted for each sample. The retained
amount of inorganic metals in the pyrolytic products was examined
to confirm their distribution after the pyrolysis process.
Fig. 1. Contents of inorganic elements in Mg-impregnated biomass samples.

2. Materials and methods

an ICPS-1000IV instrument (Shimadzu, Japan). The properties of
2.1. Preparation of Mg-impregnated biomass samples the samples are shown in Table 1 and Fig. 1.
In order to examine the thermal decomposition behaviors of the
Biomass samples with different magnesium contents were pre- biomass samples, thermogravimetric (TG) and differential thermo-
pared through a demineralization and impregnation process which gravimetric (DTG) analyses of the Mg-impregnated samples were
was demonstrated in a previous study [8]. carried out using a Q-5000 IR instrument (TA Instruments, USA).
Yellow poplar wood (Liriodendron tulipifera: YP) provided by the In this analysis, the samples less than 0.1 mm in size were heated
Korea Forest Research Institute (Seoul, Korea) was used as the raw from 40 ◦ C to 800 ◦ C at a heating rate of 10 ◦ C/min under a nitrogen
biomass and the size of the biomass sample was less than 0.5 mm. atmosphere with the flow rate of 25 ml/min.
All of the samples were dried overnight in an oven at 75 ◦ C prior
to pyrolysis and the final moisture contents of the samples were
2.3. Fast pyrolysis
∼2 wt%.
YP, a control sample, was treated with 3 wt% hydrofluoric acid
Fast pyrolysis of the Mg-impregnated samples was conducted
(HF; Duksan, Korea, extra pure grade) for 1 h to demineralize the
using a fluidized bed-type pyrolyzer which was described in a pre-
inorganic elements in the raw biomass.
vious study [8]. The operating temperatures of the pyrolysis process
The HF-treated sample was then placed into an aqueous solu-
were 450, 500, and 550 ◦ C and the residence time was fixed at 1.3 s.
tion of magnesium chloride (MgCl2 ; Sigma–Aldrich, Korea, ≥98%)
The sample was fed into the reactor at a rate of 2.5 g/min and the
with different concentrations of 0, 0.5, 1.0, and 2.0 wt% based on
reactor was maintained under inert conditions by nitrogen gas flow.
the biomass sample. The mixture was stirred for 48 h at room tem-
The temperature of the cooler was adjusted to 0 ◦ C by using an
perature in order to impregnate the sample with magnesium. After
ethanol-circulating system and the final liquid product condensed
filtering, the Mg-impregnated sample was also dried overnight in
by an electric precipitator was obtained as bio-oil. The yields of bio-
an oven at 75 ◦ C. The samples are referred to Mg-0.5, Mg-1.0, and
char and bio-oil were calculated on a wet basis and that of gas was
Mg-2.0 according to concentration of the solution and the sample
determined by difference. After that, part of the gas was absorbed
treated with a 0 wt% solution is referred to as demineralized yel-
using bottles filled with acetone connected to the exit of an electric
low poplar (D-YP). In this study, YP and D-YP related data from a
precipitator. The compounds dissolved in the acetone are referred
previous study [8] was utilized as the same biomass sample was
to ‘non-condensed pyrolytic compounds’ and they were evaluated
used and to provide connectivity to the previous study concerning
by gas chromatography/mass spectrometry (GC/MS).
inorganic metals.

2.4. Characterization of pyrolytic products

2.2. Characterization of the samples
The ratio of C, H, and N was calculated using a US/CHNS-932
(LECO Corp., USA) and the ratio of oxygen was calculated by dif-
ference. The holocellulose, lignin, and ash contents of the samples
were obtained using National Renewable Energy Laboratory (NREL)
standard procedures [11,12]. Inductively coupled plasma-emission
spectroscopy (ICP-ES) analysis was performed to determine the
content of inorganic elements in the biomass samples. Weighed
samples (0.5 g) were digested in 10 mL of a mixture of HNO3 : HCl:
H2 O2 (8:1:1, v/v; all reagents were purchased from Duksan, Korea,
extra pure grade) using a microwave (Mutiwave 3000, Anton Paar).
The heating program is ramped at 5 ◦ C/min to 550 ◦ C and held for
5 min, then ramped to 1200 ◦ C with the same heating rate, and
held 10 min at final temperature. The pressure rate was 0.3 bar/s
and the reaction pressure was 55 bar. After digestion, the samples
were diluted with 50 mL of purified water and filtered with filter
paper (Whatman No. 42). The samples were then analyzed using
2.4.1. Elemental analysis, higher heating value, and inorganic
content
Elemental and inorganic content analysis of the bio-oil and bio-
char were implemented using the same method for the biomass
sample (Section 2.2). The higher heating values (HHV) were esti-
mated utilizing Sheng and Azevedo’s [13] correlation.
2.4.2. Water content, viscosity, and acidity
The water content and viscosity of the bio-oil was determined by
a 870 KF Titrino Plus (Radiometer, Switzerland) and by a capillary-
type viscometer at 40 ◦ C, respectively. The acidity of bio-oil was
measured by the total acid number (TAN) method suggested by
Shao et al. [14]. Bio-oil (0.025 g) and phenolphthalein were diluted
with acetone and then, aqueous potassium hydroxide (0.1 M KOH)
was added until the color of the solution changed. The TAN was
calculated by an equation detailed in a prior study [15].
 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34 29

Table 1
Chemical composition of raw and Mg-impregnated biomass.

Elemental analysis (wt.%) Holocellulose (wt.%) Lignin (wt.%) Ash (wt.%)

C H N Oa

YPb 45.8 5.8 0.2 48.2 74.0 (0.4) 30.0 (0.6) 0.45 (0.0)
(1.8) (0.5) (0.0) (2.3)
D-YPc 47.1 6.1 0.1 46.7 72.1 (0.5) 29.2 (0.4) 0.17 (0.0)
(0.1) (0.1) (0.0) (0.3)
d
Mg-0.5 46.6 6.2 0.2 47.1 73.2 (0.1) 27.9 (0.1) 0.48 (0.0)
(0.2) (0.4) (0.0) (0.1)
Mg-1.0d 46.0 6.1 0.2 47.6 72.8 (0.4) 28.1 (0.0) 0.55 (0.0)
(0.7) (0.1) (0.0) (0.3)
Mg-2.0d 46.5 6.0 0.2 47.3 73.2 (0.6) 28.0 (0.5) 0.62 (0.0)
(0.3) (0.2) (0.0) (0.5)
a
By difference.
b ,c
Yellow poplar and demineralized yellow poplar, respectively and the data was referred from the previous study [8].
d
Inorganic metal-impregnated samples with concentrations of MgCl2 0.5, 1.0, 2.0 wt.%, respectively.

2.4.3. Gel permeation chromatography
The molecular weight distribution of the components in bio-
oil was measured using a gel permeation chromatography (GPC)
max instrument (Viscotek RImax, Viscotek, UK) coupled with a
UV–Vis detector (VE3210, Viscotek) equipped with PL gel 5 ␮m
MIXED-C columns (300 mm × 7.5 mm, Varian, Inc.). Each sample
(ca. 2 ± 0.3 mg) was dissolved in tetrahydrofuran and filtered to
remove the insoluble solids. Polystyrenes were used as the stan-
dards to create a calibration curve.
2.4.4. Compositional analysis
The organic components in bio-oil and non-condensed pyrolytic
compounds dissolved in the acetone were identified by GC/MS
analysis. The detailed experimental conditions for bio-oil were
previously described [8]. Additionally, quantitative analysis of non-
condensed pyrolytic compounds was performed in this study. The
response factors between the internal standard, phenol (25 mg/ml
acetone), and major compounds were determined. The amount of
each compound was calculated by using the following equation.
Amount of compound (mg/g biomass)
Rf × Apy /AI.S × QI.S /Qg
= × Vtotal
Wb
Here, Rf is the response factor, Apy and AI.S are the peak areas
of the pyrolytic compound and internal standard, respectively, QI.S
and Qg are the amounts of the internal standard and acetone con-
taining gas fraction injected into the GC/MS, respectively, Wb is
the weight of the biomass in an experiment, and Vtotal is the total
volume of acetone containing gas fraction after the reaction.
rate of Mg-2.0 was 1.63 wt%/◦ C while that of D-YP was 1.24 wt%/◦ C.
In addition, as the magnesium content of the biomass increased,
the overall decomposition range shifted to lower temperatures due
to catalytic effects of magnesium [17]. The maximum decomposi-
tion temperature of D-YP was 373.9 ◦ C and it gradually decreased
to 337.6 ◦ C for Mg-2.0. This change of the thermal stability of the
biomass can be attributed to the fact that the degree of polymeriza-
tion (DP) of cellulose was reduced due to the altered reactivity of
the surface molecules by magnesium [18]. Magnesium in biomass
would combine to the oxygen containing functional groups and
linkages at surface molecules, and this could facilitate the decom-
position of biomass constituent at low temperature by relatively
weakening of the intramolecular bond strength.
3.2. Yields of pyrolytic products
Table 2 shows the yield of pyrolytic products as functions of the
magnesium content in the biomass and the pyrolysis temperature.
The yield of bio-oil seems to be influenced more by temperature
than the magnesium content. The yields obtained at a tempera-
ture of 550 ◦ C were low in all samples, where the lowest yield was
44.6 wt% from Mg-2.0 at 550 ◦ C and the highest yield was 59.7 wt%
from YP at 450 ◦ C. Meanwhile, both the presence of magnesium and
a low temperature resulted in a high yield of bio-char. The high-
est yield of bio-char was 8.5 wt% obtained from Mg-2.0 at 450 ◦ C
while the lowest yield was 3.3 wt% from D-YP at 550 ◦ C. Char for-
mation was specifically promoted by magnesium, which could be

3. Results and discussion

3.1. Thermal decomposition behaviors of Mg-impregnated

samples

TG and DTG analyses were implemented to determine the ther-

mal behaviors of the Mg-impregnated samples and the resulting
DTG curves are provided in Fig. 2. Two degradation points were
observed as the sample was heated at a rate of 10 ◦ C/min. The first
peak was revealed at approximately 300 ◦ C followed by the sec-
ond peak at 370 ◦ C, which indicate the decomposition ranges of
hemicelluloses and cellulose, respectively [16]. Biomass with a high
magnesium content showed a diminished first peak and a narrower
decomposition range than D-YP, which suggests that magnesium
significantly influenced the decomposition of hemicelluloses and
the initial decomposition of cellulose. The maximum degradation Fig. 2. DTG curves of raw and Mg-impregnated biomass (10 ◦ C/min).
30 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34

Table 2
Yield of pyrolytic products from raw and Mg-impregnated biomass under various pyrolysis temperatures.

(wt%) Bio-oil Bio-char Gas

450 ◦ C 500 ◦ C 550 ◦ C 450 ◦ C 500 ◦ C 550 ◦ C 450 ◦ C 500 ◦ C 550 ◦ C

YPa 59.7 (0.6) b 52.6 (0.3) 46.8 (0.5) 9.6 (0.4) 8.4 (0.1) 5.8 (0.0) 30.7 (1.0) 39.0 (0.4) 47.5 (0.5)
D-YPa 59.6 (1.5) 55.5 (1.7) 46.7 (3.6) 6.2 (0.3) 4.7 (0.3) 3.3 (0.7) 34.2 (1.7) 39.8 (2.0) 50.0 (2.9)
Mg-0.5 53.3 (1.5) 56.7 (1.7) 45.9 (2.4) 8.1 (0.5) 5.1 (0.3) 5.1 (0.0) 38.6 (2.0) 38.2 (1.5) 49.0 (2.4)
Mg-1.0 54.0 (1.8) 52.3 (2.2) 47.2 (1.0) 8.2 (0.5) 6.7 (0.9) 6.2 (0.4) 37.8 (1.3) 41.0 (1.3) 46.6 (0.6)
Mg-2.0 55.8 (1.4) 54.8 (0.2) 44.6 (0.6) 8.5 (0.2) 7.7 (0.9) 7.0 (0.1) 35.7 (1.6) 37.5 (1.0) 48.4 (0.7)
a
The data was referred from the previous study [8].
b
Standard deviation

attributed to the recombination of decomposed volatile fractions
favored by magnesium ions via an ionic mechanism, which will
be discussed further in Section 3.3.3. Moreover, agglomeration of
char induced by inorganic metal during pyrolysis has been reported
in prior studies [8,19]. It was revealed that magnesium also had
an effect on the characteristics of char. The agglomeration of char
proceeded with high magnesium contents and this phenomenon
was intensified at the low temperature. The agglomerated bio-
char, however, did not prevent the flow of volatiles or gaseous
compounds during pyrolysis.
3.3. Influence of magnesium on the characteristics of the
pyrolytic products
3.3.1. Elemental composition and HHV of bio-oil and bio-char
The effects of magnesium on the characteristics of bio-oil and
bio-char are shown in Table 3. As the magnesium content increased,
the ratio of carbon in the bio-oil slightly decreased while the ratio
of oxygen increased. The tendency was observed more clearly at
the high temperature and it could be a result of the dehydration
reaction promoted by magnesium. Meanwhile, the oxygen ratio in
the amounts of bio-char obtained from the Mg-impregnated sam-
ples ranged from 20.0–24.4%, which was higher than that obtained
from D-YP, which was in the range of 17.1–19.8%. The magnesium
in biomass would be converted to magnesium oxide during pyroly-
sis which finally remained in the bio-char containing oxygen atoms.
In addition, the combination between carbonyl groups on the sur-
face of bio-char and magnesium could also influence the high ratio
of oxygen for Mg samples [20–22]. As the pyrolysis temperature
increased, however, functional groups such as carbonyl groups in
bio-char were released by bond cleavage and the ratio of oxygen
slightly decreased.
The HHVs of bio-oil and bio-char were estimated to be in the
ranges of 18.8–19.5 MJ/kg and 24.1–25.3 MJ/kg, respectively. The
values decreased in the presence of magnesium due to the low
carbon ratio and high oxygen ratio, as mentioned above, but the
differences were not significant.
3.3.2. Comparison of the physicochemical properties of bio-oil
The acidity of bio-oil was determined by the TAN analysis. The
bio-oil typically contains volatile acids and phenols which can
impact the TAN values [15]. As shown in Table 3, the TAN of bio-oil
varied from 187 mg KOH/g to 214 mg KOH/g and it was revealed
that magnesium did not have much of an effect on the acidity of
bio-oil.
The water content and viscosity of the bio-oils with different
concentrations of magnesium at the various pyrolysis tempera-
tures are shown in Table 3. In comparison to the water content
of the bio-oil from derived D-YP, that resulting from Mg-2.0 was
higher because of the dehydration reaction where magnesium acts
as a Lewis acid and through the cleavage of the glycosidic bonds of
the biomass [23]. In addition, the dehydration reaction was accel-
erated at 550 ◦ C. For instance, the increment of the water content
between D-YP and Mg-2.0 was 7.7% at 550 ◦ C whereas values of
2.0% and 2.1% were obtained at 450 ◦ C and 500 ◦ C, respectively.
This result could be attributed to the strong hydration ability of
Mg2+ [24]. The combination of water and Mg2+ is difficult to break
down at temperatures higher than the boiling point of water and

Table 3
Physicochemical properties of bio-oil and bio-char obtained from Mg-impregnated biomass.

450 ◦ C 500 ◦ C 550 ◦ C

Sample YPa D-YPa Mg-0.5 Mg-1.0 Mg-2.0 YPa D-YPa Mg-0.5 Mg-1.0 Mg-2.0 YPa D-YPa Mg-0.5 Mg-1.0 Mg-2.0

Bio-oil
Elemental analysis (%)
Carbon 46.9 46.5 46.1 45.6 45.3 46.2 46.7 45.2 44.1 44.0 46.1 47.0 46.5 44.3 43.3
Hydrogen 6.9 6.9 6.4 6.8 6.8 7.3 7.0 6.6 6.9 7.0 6.7 7.0 6.8 7.0 7.1
Nitrogen 0.2 0.2 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.1
Oxygen 46.0 46.4 47.4 47.5 47.8 46.3 46.1 48.1 48.9 48.9 47.0 45.8 46.6 48.6 49.5
HHV (MJ/kg) 19.6 19.5 19.1 19.2 19.1 19.7 19.7 19.0 18.9 18.9 19.3 19.7 19.5 19.0 18.8
Water content (%) 14.9 14.4 13.2 15.7 16.4 15.8 14.5 13.5 15.4 16.6 17.2 14.4 15.0 20.0 22.1
Viscosity (cSt) 29 40 31 54 148 29 45 67 101 216 18 34 31 44 124
TAN (mg KOH/g) 214 206 189 210 202 200 201 187 188 187 210 211 196 188 187
Bio-char
Elemental analysis (%)
Carbon 74.6 77.5 75.5 74.7 72.3 75.4 75.8 76.3 76.6 75.5 80.3 79.2 77.1 76.3 77.1
Hydrogen 3.6 2.1 2.4 2.3 2.8 3.2 3.2 2.3 2.8 2.7 3.0 2.8 1.4 1.7 2.3
Nitrogen 0.5 0.7 0.6 0.6 0.5 0.5 1.2 0.7 0.6 0.6 0.5 0.9 0.5 0.5 0.5
Oxygen 21.3 19.7 21.5 22.4 24.4 20.9 19.8 20.7 20.0 21.2 16.2 17.1 21.0 21.5 20.1
HHV (MJ/kg) 25.2 25.0 24.7 24.4 24.1 25.2 25.3 24.8 25.3 24.9 26.4 26.0 24.5 24.4 25.1
a
The data was referred from the previous study [8].
 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34
Fig. 3. GPC curves of bio-oil from Mg-impregnated biomass.
dehydration of hydrated magnesium ions has been observed even
above 500 ◦ C [18,25]. Therefore, more water would be produced in
the bio-oil at high temperatures such as 550 ◦ C in the presence of
magnesium.
Magnesium also had an effect on the viscosity of the bio-oil.
Interestingly, the higher the magnesium content in the biomass,
the higher the viscosity of bio-oil despite the higher water con-
tent. Specifically, the viscosity dramatically increased from 45 cSt
for D-YP to 216 cSt for Mg-2.0 at 500 ◦ C. According to previous
studies, there are some factors which affect the viscosity of bio-
oil such as submicron char particles and the molecular weight of
constituent compounds as well as the water content [26,27]. In
order to verify this result, the solid content of bio-oil pyrolyzed
at 500 ◦ C was analyzed as a supplementary experiment using a
method suggested by Oasmaa et al. [28]. As a result, the amount of
solids increased from 0.37 wt% for D-YP to 0.73 wt% for Mg-2.0, indi-
cating that magnesium enhanced the formation of char fines in the
bio-oil.
Furthermore, GPC analysis was performed to examine the
changes of the distribution of the molecular weight of the con-
stituents in the bio-oil. Fig. 3 shows the GPC curves of bio-oil
obtained from the Mg-impregnated samples at different temper-
atures. Regardless of the temperature conditions, the molecular
weight distribution changed, as clearly observed for Mg-1.0 and
Mg-2.0. With a high content of magnesium, the left side of the
31
curves soared along with a decrease of the middle curves while the
right side slightly changed. This result demonstrates that the high
molecular weight compounds could be produced by the repoly-
merization of middle molecular weight substances rather than
low molecular weight compounds. In addition, numerical data
of the weight average molecular weight (Mw) and the number
average molecular weight (Mn) was obtained. The Mw of bio-
oil from Mg-2.0 at 500 ◦ C was estimated to be 1670 Da, which
was much higher than that of YP (640 Da). The polydispersity
index (PDI) also increased with increasing magnesium content.
For instance, the PDI range of bio-oil obtained from Mg-2.0 was
in the range of 2.9–3.4 whereas that from YP was in the range of
1.6–2.0.
Therefore, the increase of the solid content and high molecular
weight compounds in bio-oil would result in a high viscos-
ity. Additionally, these properties could be attributed to the
recombination reactions induced by magnesium, as discussed in
Section 3.3.3.
3.3.3. Changes of the organic components of bio-oil
The organic components detected by GC-MS analysis are listed
in Table 4 and the amounts of each group are shown in Fig. 4.
A total of 15 compounds derived from carbohydrate and 29
compounds derived from lignin were identified, and all of the
compounds were grouped as described in a previous study [8].
As shown in Fig. 4, magnesium affected the composition of
the bio-oil. The amount of each chemical group decreased as
the magnesium content increased, which could be due to the
recombination reaction promoted by magnesium, as described
below.
In general, inorganic metals in biomass act as a catalyst for
the decomposition of glycosidic units by a heterolytic mechanism
during pyrolysis [29,30]. Meanwhile, as magnesium is a divalent
ion, it would also act as a catalyst like a bridge to connect each
volatile fraction. Thus, a repolymerization reaction could be pro-
moted with a high concentration of chemicals by simultaneously
forming two bonds. Moreover, it has been reported that the affinity
of Mg2+ for the ring oxygen of levoglucosan can result in the ring-
opening polymerization of each levoglucosan [31,32]. The catalytic
activities of magnesium toward the production of higher molecu-
lar weight compounds rather than levoglucosan were estimated to
be about 5.0 while those of the other inorganic species, potassium,
were estimated to be about 0.3 [10].
Therefore, the amount of levoglucosan (Fig. 4 (a)) may have been
reduced due to the initial decomposition of the biomass or the
secondary repolymerization by magnesium. The amounts of lev-
oglucosan for Mg-2.0 were 93.35 mg/g, 87.52 mg/g, and 68.53 mg/g,
which correspond to decreases of 42.6%, 49.7%, and 62.1% com-
pared to D-YP as the temperature was increased from 450 ◦ C to
500 ◦ C and 550 ◦ C, respectively. With a high content of magne-
sium, the amount of lignin groups (Fig. 4 (b)) especially catechols
and guaiacol-related groups also decreased. This could be due
to the formation of large molecules such as oligomers, usually
called pyrolytic lignin, and char fines via reactions of the functional
sites. Accordingly, a recombination reaction would be preferred
rather than decomposition of lignin polymer in the presence of
magnesium.
3.3.4. Analysis of non-condensed pyrolytic compounds
The non-condensed pyrolytic compounds from the gas fractions
were analyzed by GC/MS. The qualitative analysis performed in
a previous study [8] was expanded by the quantitative analysis
carried out in this study. Table 5 shows the gas fraction results
and specifically, the benzene, toluene, ethylbenzene,xylene, and
32 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34

Table 4
Comparison of concentration of organic components in bio-oil pyrolyzed from Mg-impregnated biomass at 450 ◦ C.

No. Source Compound Concentration (mg/g biomass)

YPa D-YPa Mg-0.5 Mg-1.0 Mg-2.0

1 Cb Acetic acid 35.78 36.71 32.23 37.07 38.15

2 C 1-hydroxy-2-Propanone 1.90 2.42 4.94 2.69 0.95
3 C 3-hydroxy-2-Butanone 0.20 1.74 0.12 0.09 0.05
4 C 3-Furaldehyde 0.34 0.49 0.16 0.27 0.11
5 C Furfural 5.54 4.97 3.88 5.29 3.25
6 C 2-Cyclopentene-1,4-dione 0.60 0.21 0.29 0.23 0.19
7 C 2-metyl-2-Cyclopenten-1-one 0.50 0.65 0.25 0.29 0.23
8 C 2(5H)-Furanone 9.94 2.60 4.37 4.90 3.67
9 C 1,2-Cyclopentanedione 1.81 2.06 1.46 1.64 0.51
10 C 5-metyl-2(5H)-Furanone 0.71 0.51 3.24 3.62 4.23
11 C 5-methyl-2-Furancarboxaldehyde 0.82 0.99 0.87 1.03 0.43
12 C 2H-Pyran-2-one 0.29 0.55 0.81 0.82 0.85
13 Lc Phenol 0.21 0.35 0.38 0.39 0.44
14 C 3-methyl-1,2-Cyclopentanedione 1.61 0.86 0.66 0.73 0.73
15 C 4-methyl-5H-Furan-2-one 0.91 0.39 0.65 0.69 0.40
16 L o-Cresol 0.15 0.10 0.26 0.23 0.34
17 L m-Cresol 0.24 0.31 0.36 0.28 0.23
18 L Guaiacol 0.85 0.20 0.00 0.00 0.20
19 L 2-ethyl-Phenol 0.42 0.10 0.06 0.08 0.00
20 L 4-methyl Guaiacol 0.69 0.24 2.13 1.74 1.75
21 L Catechol 1.66 0.84 0.18 0.30 0.29
22 L 3-ethyl-5-methyl-Phenol 0.18 0.19 0.18 0.16 0.12
23 L 3-methyl Catechol 0.68 0.42 0.65 0.87 0.41
24 L 4-ethyl Guaiacol 0.50 0.23 0.13 0.16 0.13
25 L 4-methyl Catechol 0.21 0.10 0.51 0.51 0.23
26 L 4-vinyl Guaiacol 0.95 0.42 0.16 0.26 0.11
27 L Syringol 0.99 0.47 0.78 0.64 0.26
28 L Eugenol 0.34 0.49 0.12 0.06 0.07
29 L 3-methoxy Guaiacol 0.37 0.10 0.11 0.03 0.06
30 L 4-ethyl Catechol 0.69 0.34 1.27 1.00 0.60
31 L Vanillin 0.47 0.18 0.13 0.10 0.05
32 L 4-methyl Syringol 1.39 1.33 2.47 2.32 1.60
33 L Isoeugenol (trans) 1.20 0.39 0.25 0.20 0.07
34 L Acetoguaiacone 0.28 0.00 0.11 0.13 0.07
35 C Levoglucosan 17.68 119.38 93.35 87.52 68.53
36 L 4-vinyl Syringol 2.35 1.49 0.17 0.37 0.12
37 L cis-4-propenyl Syringol 0.50 0.00 0.26 0.05 0.04
38 L Coniferyl alcohol 1.22 0.00 0.16 0.14 0.13
39 L trans-4-propenyl Syringol 0.37 0.00 0.06 0.04 0.02
40 L Syringaldehyde 0.97 0.00 0.33 0.42 0.19
41 L methoxy Eugenol 1.88 0.00 0.02 0.00 0.00
42 L Acetosyringone 0.60 0.00 0.01 0.03 0.00
43 L Coniferyl aldehyde 0.48 0.00 0.00 0.00 0.00
44 L Sinapaldehyde 0.34 0.00 0.11 0.07 0.08
Total 99.81 182.84 158.64 157.46 129.89
a
The data was referred from the previous study [8].
b
Carbohydrate derivatives.
c
Lignin derivatives.

styrene (BTEXS) components. These compounds were detected not
in the bio-oil but in the collector filled with acetone, probably due
to absence of functional groups [8]. The main compounds were
benzene and toluene which comprised 60–86% of the total amount.
The maximum amount of BTEXS was 3.90 mg/g for D-YP at 450 ◦ C
and it decreased in the presence of inorganic metals. This may be
explained to be due to two reasons. Firstly, lignin in biomass is likely
to be converted to char after fast pyrolysis although BTEXS are pri-
marily derived from lignin [33,34]. Inorganic metal could enhance
the char formation reaction and the release of vapors from lignin
would decrease. Secondly, inorganic metal could be combined with
oxygen containing fractions from the cleavage of ether linkages in
the lignin complex by pyrolysis to form a transition state, leading
to reactions between the fractions. Thus, considering the results
discussed in Section 3.3.3 where a decrease in the amount of lignin
derived monomer occurred, recombination could be induced. Both
possibilities result in a decrease of volatiles from lignin, a precursor
of BTEXS, such that the amount could decrease in the presence of
magnesium.
3.4. Distribution of inorganic metals in pyrolytic products
The recovery rate of main inorganic metals (Mg, K, and Ca)
based on the original biomass is shown in Fig. 5. The rate of the
gas was estimated by difference. Most of the inorganic metals,
about 85–92%, were left in the bio-char while only part of them
were condensed in the bio-oil or evaporated to gas. This result is
in agreement with previous studies [4,8]. Meanwhile, the distribu-
tion changed slightly with increasing pyrolysis temperature. The
amount which remained in the bio-char gradually decreased. On
the contrary, the amounts from bio-oil and gas slightly increased
due to the different release behaviors of inorganic metals depend-
ing on the pyrolysis temperature of the biomass [8]. Additionally,
as the temperature increased, the carbon ratio of the bio-char
increased (Table 3), which indicates that a condensed structure was
formed so that fewer functional groups to combine with inorganic
metals would exist. The bio-char, however, still contained lots of
inorganic metals as metal oxides or in an ionized state, which could
be utilized for gas storage or soil applications [23,35].
 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34 33

Fig. 4. Effect of magnesium on the amount of different groups of organic components derived from carbohydrates (a) and lignin (b).

Table 5
Quantitative analysis of aromatic hydrocarbons from non-condensed pyrolytic gas fractions.

Compound Benzene Toluene Ethylbenzene p-Xylene Styrene Total amount of BTEXSa

(mg/g biomass)

450 ◦ C
YP 0.82 0.33 0.04 0.08 0.10 1.37
D-YP 1.72 1.39 0.23 0.30 0.25 3.90
Mg-0.5 0.79 0.57 0.05 0.11 0.14 1.66
Mg-1.0 0.95 0.38 0.09 0.03 0.09 1.54
Mg-2.0 0.78 0.57 0.07 0.02 0.17 1.62
500 ◦ C
YP 1.03 0.63 0.13 0.13 0.24 2.16
D-YP 1.48 1.31 0.20 0.23 0.17 3.40
Mg-0.5 0.44 0.34 0.05 0.06 0.13 1.02
Mg-1.0 0.33 0.49 0.07 0.10 0.19 1.19
Mg-2.0 0.57 0.52 0.04 0.10 0.35 1.58
550 ◦ C
YP 0.38 0.81 0.16 0.19 0.25 1.80
D-YP 1.42 1.19 0.21 0.22 0.45 3.48
Mg-0.5 1.05 0.71 0.25 0.24 0.66 2.92
Mg-1.0 0.69 0.92 0.15 0.23 0.46 2.44
Mg-2.0 1.16 1.00 0.15 0.20 0.43 2.93
a
Acronym for Benzene, Toluene, Ethylbenzene, p-Xylene, and Styrene.
34 H. Hwang et al. / Journal of Analytical and Applied Pyrolysis 113 (2015) 27–34
Fig. 5. Inorganic distributions in bio-oil and bio-char produced from Mg-
impregnated biomass.
4. Conclusions
The catalytic effects of magnesium in biomass in a fast pyrol-
ysis process and the characteristics of pyrolytic products were
investigated with various concentrations of magnesium at three
different temperatures (450, 500, and 550 ◦ C). The TGA results
showed that magnesium lowered the thermal decomposition tem-
perature of the biomass. Char formation and its agglomeration
were facilitated by magnesium at the low temperature. The proper-
ties and chemical composition of the bio-oil were largely affected
even though the yield and elemental composition changed only
slightly. Specifically, the water content increased due to the dehy-
dration reaction catalyzed by magnesium. In addition, the viscosity
of bio-oil dramatically increased along with an increase of the mag-
nesium content while the amounts of compounds detected by the
GC/MS analysis decreased. Consequently, magnesium induced the
recombination of levoglucosan and small molecules to form large
fractions such as oligomers and char fines. This result was con-
firmed by GPC analysis with an increase of high molecular weight
compounds in the bio-oil. Furthermore, aromatic hydrocarbons
were identified in the non-condensed gas fractions and the total
amount decreased in the presence of inorganic metals. The dis-
tribution of inorganic metals after the pyrolysis revealed that the
bio-char contained most of them and a higher temperature led to
a decrease of the amount remaining.
Acknowledgements
This research was carried out with the support of Forest Sci-
ence & Technology Projects, provided by Korea Forest Service
(Project No. S111313L010110) and Basic Science Research Pro-
gram through the National Research Foundation (NRF), funded by
the Ministry of Education, Science and Technology (Project No.
2013R1A2A2A01011112), Republic of Korea.
References
[1] S. Czernik, A. Bridgwater, Overview of applications of biomass fast pyrolysis oil,
Energy & Fuels 18 (2) (2004) 590–598.
[2] D. Mohan, C.U. Pittman Jr., P.H. Steele, Pyrolysis of wood/biomass for bio-oil: a
critical review, Energy & Fuels 20 (3) (2006) 848–889.
[3] R. Thamburaj, Dynamotive Engineering. Fast Pyrolysis of Biomass for Green
Power Generation, 2000.
[4] F. Agblevor, S. Besler, Inorganic compounds in biomass feedstocks. 1.
Effect on the quality of fast pyrolysis oils, Energy & Fuels 10 (2) (1996)
293–298.
[5] I.Y. Eom, J.Y. Kim, T.S. Kim, S.M. Lee, D. Choi, I.G. Choi, J.W. Choi, Effect of essential
inorganic metals on primary thermal degradation of lignocellulosic biomass,
Bioresour. Technol. 104 (2012) 687–694.
[6] P.R. Patwardhan, J.A. Satrio, R.C. Brown, B.H. Shanks, Influence of inorganic salts
on the primary pyrolysis products of cellulose, Bioresour. Technol. 101 (12)
(2010) 4646–4655.
[7] D. Mourant, Z. Wang, M. He, X.S. Wang, M. Garcia-Perez, K. Ling, C.-Z. Li, Mallee
wood fast pyrolysis: effects of alkali and alkaline earth metallic species on the
yield and composition of bio-oil, Fuel 90 (9) (2011) 2915–2922.
[8] G. Varhegyi, M.J. Antal Jr., T. Szekely, F. Till, E. Jakab, Simultaneous
thermogravimetric-mass spectrometric studies of the thermal decomposition
of biopolymers. 1. Avicel cellulose in the presence and absence of catalysts,
Energy & Fuels 2 (3) (1988) 267–272.
[9] H. Kawamoto, D. Yamamoto, S. Saka, Influence of neutral inorganic chlorides
on primary and secondary char formation from cellulose, J. Wood Sci. 54 (3)
(2008) 242–246.
[10] H. Hwang, S. Oh, T.-S. Cho, I.-G. Choi, J.W. Choi, Fast pyrolysis of potassium
impregnated poplar wood and characterization of its influence on the forma-
tion as well as properties of pyrolytic products, Bioresour. Technol. 150 (2013)
359–366.
[11] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D. Templeton, Determination
of Ash in Biomass Standard Biomass Analytical Procedures, National Renewable
Energy Laboratory, 2005.
[12] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D. Templeton, D. Crocker,
Determination of Structural Carbohydrates and Lignin in Biomass, Laboratory
Analytical Procedure, 2008.
[13] C. Sheng, J. Azevedo, Estimating the higher heating value of biomass fuels from
basic analysis data, Biomass Bioenergy 28 (5) (2005) 499–507.
[14] J. Shao, F. Agblevor, N.O. Mante, N. Abdoulmoumine, New rapid method for
determination of the total acid number (TAN) of fast pyrolysis oils, in: The
32nd Symposium on Biotechnology for Fuels and Chemicals, Florida, 2010.
[15] A. Oasmaa, D.C. Elliott, J. Korhonen, Acidity of biomass fast pyrolysis bio-oils,
Energy & Fuels 24 (12) (2010) 6548–6554.
[16] D.J. Nowakowski, J.M. Jones, Uncatalysed and potassium-catalysed pyrolysis of
the cell-wall constituents of biomass and their model compounds, J. Anal. Appl.
Pyrolysis 83 (1) (2008) 12–25.
[17] D. Nowakowski, J. Jones, R. Brydson, A. Ross, Potassium catalysis in the pyrolysis
behaviour of short rotation willow coppice, Fuel 86 (15) (2007) 2389–2402.
[18] N. Shimada, H. Kawamoto, S. Saka, Different action of alkali/alkaline earth metal
chlorides on cellulose pyrolysis, J. Anal. Appl. Pyrolysis 81 (1) (2008) 80–87.
[19] B.-J. Skrifvars, R. Backman, M. Hupa, Characterization of the sintering tendency
of ten biomass ashes in FBC conditions by a laboratory test and by phase equi-
librium calculations, Fuel Process. Technol. 56 (1) (1998) 55–67.
[20] H. Wu, D.M. Quyn, C.-Z. Li, Volatilisation and catalytic effects of alkali and alka-
line earth metallic species during the pyrolysis and gasification of Victorian
brown coal. Part III. The importance of the interactions between volatiles and
char at high temperature, Fuel 81 (8) (2002) 1033–1039.
[21] D.M. Keown, G. Favas, J.-I. Hayashi, C.-Z. Li, Volatilisation of alkali and
alkaline earth metallic species during the pyrolysis of biomass: differences
between sugar cane bagasse and cane trash, Bioresour. Technol. 96 (14) (2005)
1570–1577.
[22] W.-J. Liu, H. Jiang, K. Tian, Y.-W. Ding, H.-Q. Yu, Mesoporous carbon stabi-
lized MgO nanoparticles synthesized by pyrolysis of MgCl2 preloaded waste
biomass for highly efficient CO2 capture, Environ. Sci. Technol. 47 (16) (2013)
9397–9403.
[23] Z. Wang, F. Wang, J. Cao, J. Wang, Pyrolysis of pine wood in a slowly heating
fixed-bed reactor: potassium carbonate versus calcium hydroxide as a catalyst,
Fuel Process. Technol. 91 (8) (2010) 942–950.
[24] J. Bart, W. Roovers, Magnesium chloride—ethanol adducts, J. Mater. Sci. 30 (11)
(1995) 2809–2820.
[25] N. Shimada, H. Kawamoto, S. Saka, Solid-state hydrolysis of cellulose and
methyl ␣-and ␤-D-glucopyranosides in presence of magnesium chloride, Car-
bohydr. Res. 342 (10) (2007) 1373–1377.
[26] A.V. Bridgwater, An overview of fastpyrolysis of biomass, Org. Geochem. 30
(1999) 1479–1493.
[27] Q. Lu, W.-Z. Li, X.-F. Zhu, Overview of fuel properties of biomass fast pyrolysis
oils, Energy Convers. Manage. 50 (5) (2009) 1376–1383.
[28] A. Oasmaa, P. Koponen, Physical Characterisation of Biomass-based Pyrolysis
Liquids, Espoo, 1997.
[29] G.N. Richards, Glycolaldehyde from pyrolysis of cellulose, J. Anal. Appl. Pyrolysis
10 (3) (1987) 251–255.
[30] S. Julien, E. Chornet, P. Tiwari, R. Overend, Vacuum pyrolysis of cellulose: Fourier
transform infrared characterization of solid residues, product distribution and
correlations, J. Anal. Appl. Pyrolysis 19 (1991) 81–104.
[31] H. Kawamoto, M. Murayama, S. Saka, Pyrolysis behavior of levoglucosan as an
intermediate in cellulose pyrolysis: polymerization into polysaccharide as a key
reaction to carbonized product formation, J. Wood Sci. 49 (5) (2003) 469–473.
[32] H. Kawamoto, W. Hatanaka, S. Saka, Thermochemical conversion of cellulose
in polar solvent (sulfolane) into levoglucosan and other low molecular-weight
substances, J. Anal. Appl. Pyrolysis 70 (2) (2003) 303–313.
[33] D.J. Nowakowski, A.V. Bridgwater, D. Elliott, D. Meier, P. De Wild, Lignin fast
pyrolysis: results from an international collaboration, J. Anal. Appl. Pyrolysis
88 (1) (2010) 53–72.
[34] O. Faix, D. Meier, I. Fortmann, Thermal degradation products of wood, Holz Roh
Werkst. 48 (7–8) (1990) 281–285.
[35] H. Wu, K. Yip, Z. Kong, C.-Z. Li, D. Liu, Y. Yu, X. Gao, Removal and recycling of
inherent inorganic nutrient species in mallee biomass and derived biochars by
water leaching, Ind. Eng. Chem. Res. 50 (21) (2011) 12143–12151.