Biodiesel is an alternative fuel produced from renewable resources such as
palm, canola, soya bean and rape seed oil (Hass, 2005).
It is produced via transesterification reaction of fatty acid and alcohols.
Approximately 10% of glycerol, an alcohol with three hydroxyl groups is produced
as a byproduct in this reaction (Karinen and Krause, 2006;
Chew and Bhatia, 2008; Huber et
al., 2007). The crude glycerol obtained is about 50% pure and is a huge
potential as raw material to produce other valuable products (Chitra
et al., 2005; Pathak et al., 2010)
in manufacturing of drugs, cosmetics, synthetic resins, ester gums and gas production
(Wolfson et al., 2007). As more and more crude
glycerol is generated as biodiesel by-product, efforts to seek for economical
ways to convert glycerol to useful chemicals continue and this can indirectly
minimize the cost of biodiesel production. One of the valuable chemicals that
can be converted from glycerol is olefin (Zhou et al.,
2008; Pagliaro and Rossi, 2008), a petrochemical derivative
conventionally produced by thermal cracking of natural gas and crude oil. Lower
olefins have short chains with only two to four carbon atoms and the simplest
one is ethylene. The higher olefins have chains of up to twenty or more carbon
atoms. Main olefins products like ethylene, propylene, butadiene and C4
derivatives are used to produce plastics, chemical intermediates, industrial
solvents and others (Li et al., 2007).
The objective of this study is to investigate the most suitable metal addition in zeolite HZSM-5 catalyst to increase the production of olefins via catalytic conversion of glycerol. In order to achieve the objective, a set of HZSM-5 zeolite based catalysts were impregnated with potential metals and tested. The structure of the modified catalysts was observed and the effect of the modified structure towards production of olefins from glycerol was studied. Parameters for this study include types of metal used to enhance active acid sites on the surface of the zeolite catalyst. Catalysts characterizations employed were X-ray Diffraction (XRD) and Fourier Transform Infra Red (FTIR). Products from the reaction were analyzed using gas chromatography (GC) from which the best catalyst was identified.
MATERIALS AND METHOD
Materials: The main reagent used was HZSM-5 (nSiO2/nAl2O3 = 30, molar ratio) which was purchased from Zeolist International and glycerol of 87.5% pure. The unmodified HZSM-5 catalyst was used as reference.
Catalyst preparation: The HZSM-5 catalyst was impregnated with different
metals, namely chromium, aluminium, calcium, copper and nickel. The zeolite
was added into 100 mL distilled water. Then the metals were added into the solution.
For chromium metal, the zeolite was impregnated by 0.3 g of Cr(NO3)3.9H2O
with incipient-wetness impregnation method. For aluminum, calcium, copper and
nickel, the zeolite was impregnated with 0.3 g of aluminum nitrate, calcium
chloride, copper (III) hydrate and nickel nitrate, respectively.
||Table of catalysts
The impregnation period lasted for half an hour at a temperature range of 30-40°C
and then dried at 100°C and finally calcined at 550°C for 5 h (Murata
et al., 2008).
The weight of all catalysts was measured before and after impregnation in order to observe the weight loss during the process. The weight loss of the catalyst is listed in Table 1. One of the factors contributing to weight loss is the decomposition of metal precursors throughout the process.
Then, the catalysts were transferred into vials and marked as HZSM-5, CrHZSM-5, CuHZSM-5, AlHZSM-5, NiHZSM-5 and CaHZSM-5. Visually, catalysts with copper, calcium, nickel and aluminium metal were white in colour while slightly greenish powder was formed for catalyst with chromium.
X-Ray diffraction was performed using Siemens Diffractometer D5000 (Software Diffract Plus) with CuK α radiation, λ = 1.54056A° at 40 kV and 30 mA in the range of 2θ = 5° to 80° at a screening speed of 0.05° per second, with a vertical ganiometer at room temperature (20°C). The sample was held on mountain sample holder and then grounded before mounting on glass.
The existence of impregnated metal in zeolite catalysts was analyzed by Fourier Transform Infra Red (FTIR). Spectrum one FTIR Spectrometer and spectrum 20 explorer with instrumentation model spectrum one was used to characterize the catalysts. The sample was grounded with spectrograde KBr to form a mixture, which was pelletized using a hydraulic press. This pellet was used to record the infrared spectrum in a range of 4000-375 cm-1.
Catalyst evaluation: The catalysts performance testing was carried out
in a fixed bed stainless steel reactor, with outside diameter of 13 mm and 35
cm length, positioned inside a Carbolite tubular furnace as illustrated in Fig.
1. Glycerol solution with 50% dilution was prepared as feed for the reaction.
The prepared catalyst was sandwiched between quartz wools inside the reactor.
Prior to starting the test, the catalyst was activated in a flow of helium gas
for 1 h at 500°C to remove moisture and prepare the catalyst for reaction.
||The reactor is heated inside this tubular furnace which is
part of the experimental set-up of the reaction
This was immediately followed by the reaction where temperature was set at
700°C, pressure 1 atm and lower space velocity with long residence times
(20-80 sec) in order to obtain high olefins (Corma et
al., 2008). The products were then analyzed with gas chromatography
equipped with both Thermal Conductive Detector (TCD) and Flame Ionization Detector
(FID). The HP 6890N GC-TCD/FID is equipped with HP plot/Q, HP-Mole sieve, DB-1
and Hayasep-Q columns which are able to detect important compounds in this research
such as olefins, paraffin and oxygenates.
RESULTS AND DISCUSSION
Catalyst characterization findings: Figure 2 presents
XRD patterns of HZSM-5 and the modified zeolite with different metals of nickel,
aluminium, copper, chromium and calcium. The XRD patterns were analysed between
5° and 45° with 2θ scale. The X-ray diffraction pattern of the
modified HZSM-5 completely matched with that of the parent HZSM-5, which indicates
that the combined-modification measurement has no obvious effect on parent zeolite
structure (Peiqing et al., 2004). There was also
no new phase formation during heat treatment and zeolite modification. The XRD
pattern of the modified zeolite indicates that the structure of the zeolites
remains intact after doping different metals over HZSM-5. No peak related to
the metals was found on the diffractogram (Zaidi and Pant,
2005; Corma et al., 2008).
This pattern reveals that the metals were highly dispersed on the HZSM-5. This
finding is corroborated by the suggestions made by Jiangyin
et al. (2006), where he stated that there are no diffraction peaks
corresponding to the metal observed up to 40% loading in the XRD analysis. However
the obtained intensities differ depending upon metals content. The highest intensity
was observed for the CrHZSM-5, followed by AlHZSM-5, CaHZSM-5, NiHZSM-5, CuHZSM-5
and HZSM-5. The decrease in the intensities may due to the higher absorption
coefficient of copper metals.
Figure 3 illustrates the FTIR spectra when different metals
were impregnated to the HZSM-5 catalysts. The spectra shows that the impregnated
metals do not change the characteristic of zeolite catalysts. The broad peak
was observed at 3600 cm-1 that consist of O-H bond for phenols functional
groups. Another peak was present at 1637 cm-1 for all catalyst, corresponding
to the H-O-H bonding in the structure. All catalysts also show three other significant
broad bands at 1230, 1100 and 796 cm-1 which are associated with
asymmetric and symmetric stretching vibrations of the Si-O-Si. These bands are
also related to the internal linkages in SiO2 or Al2O3
of zeolite lattice (Jiangyin et al., 2006). The
FTIR spectra for all modified catalysts were similar to each other except for
AlZSM-5 that has a slightly different pattern. This may be due to the metal
addition that alters the compositions of aluminium compound in zeolite structure,
since the zeolite itself has certain aluminium ratio.
||XRD spectra for various impregnated metal ZSM-5 catalyst
||FTIR spectra for various impregnated metal ZSM-5 catalyst
||The reactivity and selectivity (%) of M-ZSM-5 for the reaction
of glycerol to olefin at T = 700°C
Reactivity and selectivity: The effect of different metals on the reactivity
and selectivity of MZSM-5 (M = metal) catalysts for the reaction of glycerol
at high temperature (700°C) with catalyst load mass 0.2 g is shown in Table
2. From the table, it is shown that the highest conversion of glycerol to
olefins is with CuZSM-5 with 17.72% conversion. It is followed by the reaction
of glycerol with NiZSM-5, CaHZSM-5, HZSM-5, CrHZSM-5 and AlHZSM-5 with percentage
conversion of 16.07%, 15.02%, 11.62%, 10.85 and 10.44%, respectively. From the
experiments it was found that the reaction produced methanol, olefins (3-methyl-butenes
and 2-methyl-1, 2-butadiene) and other products. The highest selectivity of
olefins is 100% with aluminium impregnated zeolites, evident by a GC report
that shows only olefins peak is available for the reaction. Besides that, about
51.27% selectivity of olefins was recorded when glycerol was cracked with the
assistance of CrHZSM-5 at 700°C. This agrees well with Lu
et al. (2006) who has reported that the selectivity to butene smoothly
increased with raising the loading amount of the Cr (III), when CrZSM-5 reacts
in catalytic cracking of isobutane to produce light olefins. The analysis of
the gaseous product was carried out using gas chromatography (GC-TCD) having
a 10 ft column 80/100 Porapak N. From the chromatogram obtained no peaks were
detected. This might be due to the small volume of the gaseous product. Besides,
the low concentration of the gaseous product also makes it difficult for the
GC to detect and identify the compound.
Effect of metal impregnated ZSM-5 catalyst on product yield: Compared
with unmodified HZSM-5 zeolite, more olefins were comparatively produced from
CaHZSM-5, CuHZSM-5 and NiHZSM-5. However, the yield of olefins decreased when
chromium was used as active metal as shown in Fig. 4. Highest
olefin yield was observed for reactions with copper impregnated zeolite. As
stated by Zaidi and Pant (2005), copper oxide doped
HZSM-5 catalyst result in an increased activity for conversion of methanol to
liquid hydrocarbons of gasoline range. Compared to HZSM-5 catalyst there was
a significant increase in conversion and hydrocarbon yield over CuO impregnated
catalyst. The percentage of olefins yields using copper, nickel and calcium
impregnated ZSM-5 is high compared to unmodified zeolite.
||Olefin yield for various active metal ZSM-5 catalysts
These results demonstrate that a small amount of metal inclusion like copper,
calcium and nickel plays an important role of modification, not only did it
enhance the reactivity for the glycerol conversion, but also increase the selectivity
of olefins, especially the selectivity to 3-methyl-1,2-butadiene. CrZSM-5 and
AlZSM-5 gave the lowest olefin yield. This might due to the metals altering
the active sites of zeolite catalyst, hence resulting the catalyst to deactivate
and not reactive. This is consistent with work reported by Jiangyin
et al. (2006) and Lu et al. (2006)
where one reason for the ineffective of chromium metal zeolite to yield high
olefins is because the trace amount of Cr alters the acidity of the HZSM-5 catalysts.
This shows that suitable strength and amount of acidity of the catalyst are
important factors in favouring high olefin yields.
Catalysts screening of various metal impregnated ZSM-5 for catalytic conversion of glycerol to olefins has been carried out by testing its performance in a fixed bed reactor. Evidently, metal doped ZSM-5 resulted in an increase of activity and improved conversion of glycerol as well as the olefins yields. The products obtained were methanol, olefins and other unidentified compounds. CuZSM-5 has shown the best glycerol conversion and yield with 17.76% and 3.55%, respectively. The high reaction temperature makes the catalyst more active and favours the yield of light olefins.
The authors would like to acknowledge Ministry of Higher Education, Malaysia (MOHE) for the financial supports received for this work through the project number (Vot 78401) under Fundamental Research Grant Scheme (FRGS).