close

Вход

Забыли?

вход по аккаунту

?

Pyrolysed powdered mussel shells for eutrophication control effect of particle size and powder concentration on the mechanism and extent of phosphate removal.

код для вставкиСкачать
ASIA-PACIFIC JOURNAL OF CHEMICAL ENGINEERING
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
Published online 12 February 2010 in Wiley Online Library
(wileyonlinelibrary.com) DOI:10.1002/apj.426
Research Article
Pyrolysed powdered mussel shells for eutrophication
control: effect of particle size and powder concentration
on the mechanism and extent of phosphate removal
Arjan Abeynaike, Luyao Wang, Mark I. Jones and Darrell A. Patterson*
Department of Chemical and Materials Engineering, University of Auckland, Auckland, New Zealand
Received 18 February 2009; Revised 29 November 2009; Accepted 6 December 2009
ABSTRACT: The international shellfish farming industry has a growing problem with respect to sustainability: the
shells, a by-product, are currently being mostly wasted to landfill. Instead, this calcium-rich resource can be used to
produce lime [calcium oxide (CaO)] and then used to remove phosphate from rural wastewaters. Powdered mussel shell
was heat treated to form lime. Two different shell particle sizes (fine, 53–106 µm; coarse, 212–250 µm) as well as
various pyrolysis times, heating rates, pyrolysis temperatures and shell concentrations were used to determine the effects
of these parameters on the lime formation and subsequent phosphate removal from a synthetic wastewater. Furthermore,
the mechanisms of phosphate removal were determined by quantifying the phosphate content in all components before
and after reaction with the synthetic wastewater. It was found that with excess of partially calcined pyrolysed shells, at
a concentration of 5 g l−1 , more than 95% phosphate removal was achieved, irrespective of particle size or pyrolysis
conditions. When using optimally heat-treated shells (particle size: 53–106 µm, pyrolysed for 1 h at 750 ◦ C), it was
possible to achieve over 90% phosphate removal using just 196 mg l−1 of shell. For the pyrolysed shells, the main
mechanisms of phosphate removal were homogeneous nucleation to form a suspended precipitate, as well as adsorption
and heterogeneous precipitation on the surface of the remaining calcite shell particles.  2010 Curtin University of
Technology and John Wiley & Sons, Ltd.
KEYWORDS: wastewater treatment; phosphate precipitation; mussel shell; calcination; eutrophication
INTRODUCTION
Phosphates are common pollutants in domestic and
industrial wastewaters. High levels of phosphates and
nitrates in natural waterways lead to eutrophication,
causing the excessive growth of algae. These algae
deplete the water of oxygen, subsequently reducing the
ability to sustain other aquatic life.[1,2] This is a major
problem in many countries, particularly those where
agriculture is a major component of their economy, such
as New Zealand (NZ). One NZ survey found that up to
40% of the 700 or more shallow lakes were eutrophic,
mainly in rural areas where there is a high concentration
of agricultural activity (where the phosphates mainly
come from fertilisers). Most of these lakes no longer
supported fish life.[3] The latest report[4] shows that the
phosphate levels are now increasing in rivers, signalling
a long-term trend that is likely to trigger undesirable
changes in the river ecosystems. In NZ, as well as many
*Correspondence to: Darrell A. Patterson, Department of Chemical
and Materials Engineering, University of Auckland, Private Bag
92019, Auckland, New Zealand.
E-mail: darrell.patterson@auckland.ac.nz
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
Curtin University is a trademark of Curtin University of Technology
other countries where agricultural-based industries are
important to the economy, it is therefore essential to
develop a cost-effective method of removing phosphates
from rural and agricultural wastewaters before they
enter rivers and lakes in order to prevent further
eutrophication, as well as to help recover these aquatic
environments from the impact of phosphates.
Lime [calcium oxide (CaO)] treatment of wastewater has been commonly used as a relatively inexpensive method to remove dissolved phosphates by
precipitation.[5] The precipitate forms a sludge that is
removed from the water and landfilled. Lime is traditionally produced from limestone, whereby pyrolysis at
temperatures greater than 600 ◦ C[6] results in the decomposition of the limestone according to:
(1)
Limestone is a finite non-renewable resource, so is
not a sustainable source of lime. However, in NZ and
the rest of the world, there is a rich, renewable and sustainable source of calcium carbonate: the shells from
the aquaculture industry. These shells are however currently mainly going to waste. In NZ for instance, the
232
A. ABEYNAIKE et al.
annual production of mussel shells alone is currently
estimated at 12 000 t (Mandeno M, personal communication) and has been estimated up to 100 800 t.[7]
Furthermore, the NZ aquaculture industry is expected
to grow from a $300 million to a billion-dollar p.a.
industry by 2025,[8] indicating that this wastage is set to
significantly increase. Currently, shells go to landfill or
are crushed and used as chicken feed and burley,[9] or as
a low-value material substitute for road fill.[7] However,
in NZ, shell waste volume exceeds the demand in such
applications, and landfill disposal costs are predicted
to increase.[9,10] This rich primary production resource
needs to be, and should be, utilised. Previous work has
shown that these shells can be calcined into lime.[11 – 14]
In this paper, one application route for lime is further
explored: applying the lime to the treatment of phosphate containing wastewaters.
This work concentrates on utilising one specific type
of mussel shell, the NZ Greenshell Mussel (Perna
canaliculus). The results can however be generalised
to other types of mussel and aquaculture shells. P.
canaliculus is unique to NZ, and is a class of bivalve
mollusc with a shell comprised three layers. The outer
periostracum is a thin layer of organic material that provides protection from dissolution in water. The middle
prismatic layer is composed of calcium carbonate in
the form of thin vertical crystals in a protein matrix.
The innermost nacreous layer is also calcareous but
in the form of thin sheets separated by an organic
matrix.[15] The cumulative work by Refs [11–14] has
shown that on heating the shell to above 500 ◦ C, the
organic material can be burnt out and the crystal structure of the calcium carbonate changes from the ambient
temperature aragonite polymorph, with an orthorhombic
crystal structure, to the calcite polymorph with a trigonal–rhombohedral structure. Furthermore, heating in a
carrier gas allows dissociation of the calcite into lime
with the evolution of CO2 , as per reaction 1. The temperature of the polymorphic transformation, typically
above 500 ◦ C, is dependent on the composition, purity
and origin of the aragonite material, whereas the dissociation temperature depends on mineral content, crystallinity, impurity levels, particle size and carbon dioxide partial pressure,[14] and occurs typically between
600 and 800 ◦ C. There is some evidence that the conversion occurs via a shrinking core mechanism,[14] indicating that insufficient heat treatment of the shells may
leave an unreacted calcite core beneath a lime outer
shell.
Studies by Kwon et al .,[11] Lee et al .,[12,13] Currie
et al .[14] and Namasivayam et al .[16] have demonstrated
that raw (non-pyrolysed) and pyrolysed shell powder
(both mussel and oyster) are capable of removing dissolved phosphate. Kwon et al .[11] and Currie et al .[14]
performed trials in stirred batch reactors, achieving 90%
and 98% phosphate removal, respectively. Lee et al .[12]
performed trials in a fluidised bed reactor, achieving
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pacific Journal of Chemical Engineering
92% reduction in phosphate within 5 min and in other
trials[13] using a stirred batch reactor achieved 98%
reductions within 10 min from a 11.9-mg l−1 phosphate solution with 6 g l−1 of pyrolysed oyster shells.
In a related study, Kwon and Lee[17] report using lime
derived from pyrolysed aquaculture shells in the same
type of reaction, though aimed at producing a pure
precipitate of hydroxyapatite. These studies have also
shown that the raw oyster and mussel shell powder also
remove phosphate, though to a much lesser extent than
the pyrolysed shells. However, despite all this work, the
dominant mechanism (or mechanisms) through which
the pyrolysed shells remove phosphate has not yet been
definitively identified. There are three possible mechanisms, which may all remove phosphate to a greater or
lesser extent:
1. Reaction through homogeneous precipitation in the
reaction solution;
2. reaction through heterogeneous precipitation on the
pyrolysed shell surface;
3. absorption or adsorption onto the pyrolysed shell
surface.
Both reaction mechanisms 1 and 2 are well understood. Firstly, the solid lime reacts with water to form
calcium hydroxide, which then dissociates in the water
to calcium and hydroxide ions[6] :
(2)
The phosphates in a wastewater (or orthophosphates)
may exist in several different forms depending on the
pH of the solution, as shown by Eqn (3)[18] :
(3)
Thus, orthophosphate ions can react with the calcium
and hydroxide ions (from reaction 2) to form a range
of different calcium phosphates. However, in order to
ensure that the phosphate is irreversibly and therefore
stably removed from the wastewater, the ultimate goal,
is to form one specific calcium phosphate, the insoluble
precipitate hydroxyapatite[5] :
10Ca2+ + 6PO4 3− + 2OH− −−→ Ca10 (PO4 )6 (OH)2
(4)
Hydroxyapatite is the most thermodynamically stable
form of calcium phosphate.[18] Precipitation to hydroxyapatite takes place once the pH has risen above 10.[5]
The reaction is favoured by a high pH, as sufficient OH−
ions must be supplied to form the precipitate. If the initial solution is acidic, the reaction may take longer, as
the hydroxide ions are consumed by neutralisation. A
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
PYROLYSED SHELLS: MECHANISM AND EXTENT OF PHOSPHATE REMOVAL
high pH will also improve the supply of PO4 3− ions, as
shown in Eqn (3).
Hydroxyapatite precipitation can occur either homogenously or heterogeneously, giving the two phosphate removal mechanisms as previously discussed. Lee
et al .[13] demonstrated that the precipitation with pyrolysed oyster shells occurs homogenously and with no
adsorption of phosphates. In contrast, Song et al .[19]
have shown that the precipitation of hydroxyapatite is
much faster when it occurs on the surface of calcium
carbonate (as in unpyrolysed shells), which acts as a
seed for the nucleation of the hydroxyapatite crystal.
The heterogeneous nucleation has a lower energy barrier to overcome than homogeneous nucleation, causing
an increased rate of crystallisation. This could therefore indicate that heterogeneous precipitation may be
a dominant phosphate removal mechanism for partially
converted pyrolysed mussel shells (which due to the
shrinking core mechanism in the pyrolysis have an outer
surface of lime and an inner unreacted calcite core),
where dissolved calcium ions from the lime reprecipitate onto the surface of the undissolved calcite core.
The absorption/adsorption mechanism onto raw and
pyrolysed mussel shells is less well understood than
the reaction steps. For raw shells, because the calcium carbonate is relatively insoluble compared to lime,
it is likely that the main phosphate removal mechanism is by absorption and/or adsorption. Freundlich
and Langmuir models have been applied to phosphate
adsorption based on empirical data.[20,21] These models, however, have been criticised as being too simplistic to describe the adsorption mechanisms.[22] For
pyrolysed shells, adsorption is the less preferred reaction mechanism for a wastewater treatment process,
because phosphate removal capacity is limited to the
surface area of the particles, whilst precipitation with
lime can remove the phosphate in much larger stoichiometric amounts. However, if the pyrolysis does
not achieve full conversion to lime, there is the potential of a dual mechanism, where the phosphate could
both adsorb and/or precipitate onto the surface of the
unreacted core. This competition of processes therefore
complicates the phosphate removal process, potentially
resulting in impurities and rate limitations. For example, surface absorption/adsorption may slow down the
rate of heterogeneous precipitation by blocking the surface and pores. House and Donaldson[21] reported that
the adsorption is rapid enough to be independent of
surface precipitation. The extent to which they compete for surface sites in the pyrolysed shells is however
unknown.
All these indicate that the reaction mechanisms need
to be resolved in order to optimise the particle sizing
and heat treatment of the shells to give the maximum
removal of phosphate. For instance, if most of the
phosphate is removed via homogeneous precipitation
from the solution, then the heat treatment of the
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
minimum practical particle size (i.e. which avoids
caking during heat treatment) to give the maximum
amount of lime formation is optimal. If heterogeneous
surface precipitation is the dominant phosphate removal
mechanism, then less lime formation to retain an
undissolved solid for the precipitation to occur on
is optimal. A mixed mechanism may require a heat
treatment intermediate to these two cases. If adsorption
is a dominant phosphate removal mechanism, however,
favourable conditions for homogenous precipitation (i.e.
conversion of the shells to 100%) may need to be
engineered.
Consequently, the overall objective of this work is
to establish an optimised process in which the shellfish industry’s shell waste can be converted into lime
and then used for phosphate removal from wastewaters. To do this, the effect of varying the pyrolysis
conditions, shell particle size and shell concentration
on the phosphate removal from a synthetic wastewater was determined. This information was then used
to establish the dominant mechanisms of this phosphate removal. The phosphate removal mechanisms for
the equivalent shell particle size and concentration of
raw (unpyrolysed) shells were used as a benchmark for
comparison.
EXPERIMENTAL
Shell preparation and treatment
Shells from green-lipped mussels (P. canaliculus) were
kindly donated by the Sanford Limited shellfish processing plant, Coromandel, New Zealand. All shells
were stored in a freezer to prevent spoilage of the
remaining organic material. For processing, the shells
were first thawed, and then cleaned by scrubbing with
steel wool, since although they were received free of
meat, some strongly attached organic matter remained
on the shells. The outer protein layer (the periostracum)
was not removed however. Thereafter, the shells were
oven dried at 80 ◦ C for 2 h. They were then crushed
into pieces with a hammer before being further ground
in a ring mill for 10 s. The shell powder was sieved
into size ranges of 53–106 µm (referred to as fine
powder) and 212–250 µm (coarse powder) – both the
particle sizes are smaller than the 300–600 µm size
range used by Currie et al .[14] in order to increase
the amount of lime formed in each shell particle
compared to that study. Brunauer, Emmet and Teller
(BET) surface area analysis was performed on the
crushed and sieved powders using a Micromeritics Tristar 3000.
Pyrolysis was performed in an in-house built horizontal tube furnace with a nitrogen atmosphere replenished at 1.5 l min−1 to ensure air was purged from the
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
233
234
A. ABEYNAIKE et al.
Asia-Pacific Journal of Chemical Engineering
furnace. Samples were placed in the furnace and heating began at room temperature and was increased at
20 ◦ C min−1 until the desired temperature of 750 or
800 ◦ C was reached. The temperature was held constant for a specified time before the furnace was turned
off and allowed to cool overnight. The shell powder
was weighed before and after pyrolysis. Lime content
in the pyrolysed particles was calculated by weighing
the shell samples before and after heating, assuming
that all weight loss corresponds to carbon dioxide evolution as a result of the limestone calcination reaction
[Eqn (1)] minus the carbon from the organic matter in
the shell (mainly from the periostracum). The number of moles of carbon dioxide evolved must correspond to the number of moles of lime formed. The
weight percentage of lime can then be calculated from
the final weight of the sample. To determine the carbon attributable to the organic matter in the shell,
the percentage of organic matter in the raw samples
was calculated from the weight loss of the raw particles pyrolysed at 500 ◦ C for 1 h. This temperature
ensured complete decomposition of organic matter without forming lime on the surface of shell particles. It
is to be noted that in this work, shells were not necessarily fully converted into lime, but instead to a
mixture of lime and calcite, in order to determine the
effect of partial calcination on the phosphate removal
mechanisms.
Batch phosphate removal trials
The details of the main trials conducted are summarised
in Table 1. Shell powder that had undergone the pyrolysis treatment (pyrolysed shell ) and shell powder that
had no pyrolysis (raw shell ) were initially reacted in
a stirred batch reactor, a 2-l cylindrical perspex vessel with a diameter of 235 mm, with shell powder in
excess in the reactor (Trials 1–7). Shell samples varied in particle size and pyrolysis treatments. Further
trials (Trials 8–16) were conducted in a 1-l conical
flask, where various shell concentrations were used,
but the shell particle size and pyrolysis treatment were
maintained constant. All trials used an orthophosphate
solution of 10 mg PO4 2− l−1 , which was prepared by
dissolving potassium di-hydrogen phosphate (Biolab,
NZ) in deionised water. It is to be noted that additional
trials to those detailed in Table 1 were conducted using
a smaller particle size (0–53 mm) and longer pyrolysis
times. The phosphate removal results for these will not
be reported; however, Environmental Scanning Electron Microscopy (ESEM) images from these runs are
presented, where they are representative of the trends
in all the results.
The reaction method used for Trials 1–7 and 8–16
(Table 1) was the same. A weighed amount of powdered
mussel shell was added to the phosphate solution.
Samples of the supernatant were taken over a period
of 90 min. The pH was measured at the beginning
and at the end of each run using a Sartorius PB10 digital pH meter. These samples were filtered
through 0.45 µm filter paper to remove all solids. A
sample of the solids remaining in the solution at the
end of each trial was collected by vacuum filtration
and then dried. Two phases of solid were collected
separately: white suspended solids and the remaining
mussel shell powder that had settled at the bottom of
the vessel. The solids were oven dried at 120 ◦ C for
analysis.
Table 1. Summary of the main batch trials, including pyrolysis treatment, mussel shell concentration and measured
pH at beginning and end of batch runs.
Trial #
Particle
Size (µm)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
53–106
53–106
53–106
53–106
53–106
212–250
212–250
53–106
53–106
53–106
53–106
53–106
53–106
53–106
53–106
53–106
Pyrolysis
Treatment
1
1
2
2
1
1
1
1
1
1
1
1
1
1
No treatment
hour at 750 ◦ C
hour at 750 ◦ C
hours at 750 ◦ C
hours at 800 ◦ C
No treatment
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
hour at 750 ◦ C
Reaction
vessel
Mussel shell
concentration (g L−1 )
pH at
start
pH at
end
2L perspex
2L perspex
2L perspex
2L perspex
2L perspex
2L perspex
2L perspex
1L glass
1L glass
1L glass
1L glass
1L glass
1L glass
1L glass
1L glass
1L glass
5.00
5.00
5.00
5.00
5.00
5.00
5.00
0.033
0.033
0.033
0.049
0.065
0.081
0.098
0.196
3.71
6.25
6.35
6.57
6.66
6.38
6.50
6.43
–
–
–
–
–
–
–
–
–
7.91
12.03
12.20
12.59
12.48
7.67
11.42
–
–
–
–
–
–
–
–
–
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
PYROLYSED SHELLS: MECHANISM AND EXTENT OF PHOSPHATE REMOVAL
Sample and data analysis
Chemical analytical procedures
The concentration of phosphate in the initial liquid
solution and the reaction supernatant was determined
using a standard vanadomolybdophosphoric acid colorimetric method.[23] Spectrophotometric analysis was
performed using a Perkin Elmer Lamda 35 UV/Vis
spectrophotometer (Trials 1–7) and an Agilent 8453
UV/Vis spectrophotometer (Trials 8–16). Wavelengths
of 420 and 400 nm were used depending on the sensitivity required. Calibration tests were performed using
five standard solutions before each set of liquid samples
was analysed.
The solids were analysed for their phosphate content using a modified vanadomolybdophosphoric acid
method.[23] The vanadomolybdate reagent solution is
highly acidic (4.0 N) and is assumed to have hydrolysed all solid phosphate into liberated orthophosphate.
The suspended solids were analysed by adding a 10-mg
sample to 10.0 ml deionised water and 3.0 ml vanadomolybdate reagent. The resulting solution was diluted
to 100 ml with deionised water. The sample was filtered
through 0.45 µm filter paper before spectrophotometric
analysis. The mussel shell was analysed by adding a
100-mg sample to 10.0 ml deionised water and 3.0 ml
vanadomolybdate reagent. The resulting solution was
not diluted but was filtered and analysed in the spectrophotometer.
ESEM and electron dispersive spectroscopy
(EDS)
The raw and pyrolysed shells, as well as the solid
reaction products, were imaged by ESEM using a FEI
Quanta 200 FEG ESEM (5 kV accelerating voltage).
Samples were sputter-coated (Polaron SC 7640) with
platinum to prevent charging. EDS was performed on
selected samples using an EDAX Phoenix EDS system
at an accelerating voltage of 20 kV. For these samples,
the particles were coated with carbon to mitigate
charging, because the platinum signal from platinum
coating can interfere with the phosphorus signal.
X-ray diffraction (XRD)
XRD was performed using a Bruker D8 Advance
diffractometer with a 40-kV Cu Kα X-ray source,
scanning within a 2θ range of 20–70◦ with a 0.02degree step size. Spectra were measured from the
pyrolysed mussel shell powder before and after use
in the batch reactor. All XRD spectra are plotted as
intensity (arbitrary units) vs diffraction angle, 2θ .
Statistical analysis of data
The data for Trials 8–16 at 30, 45 and 60 min were
analysed to determine if there were significant differences between the data points for the various shell
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
concentrations. The spread for each data point was estimated from the standard deviation of the three measurements at 0.33 mg l−1 excess. It was assumed that
the standard deviation was the same for all trials, and
that the mean of each trial was at the measured data
point. With this basis, a one-way analysis of variance
(ANOVA) was performed and Tukey intervals were calculated at 95% confidence intervals. These calculations
were done using the statistical analysis programme R,
which is available freely from http://www.r-project.org.
RESULTS AND DISCUSSION
Conversion of shells to lime by pyrolysis
Pyrolysis resulted in the partial calcination of the shell
particles into lime, as per equation 1. This was proved
by XRD (Fig. 1). Figure 1 shows that the raw shells
consist of the aragonite calcium carbonate polymorph,
which is transformed into both the calcite polymorph
and lime by the calcination, as expected. The weight
fraction of lime in the pyrolysed shell is given in
Table 2, as calculated from the weight loss during pyrolysis as described in Section ‘Shell preparation and
treatment’. Both these results confirm the findings of
Currie et al .[14] who also achieved partial lime formation. ESEM images of the raw shell (Fig. 2a and b;
before pyrolysis) (Fig. 2c and d; after pyrolysis) for fine
particles (sieve size 0–53 µm) showed that the surface
of the shells becomes a more open, porous morphology, indicative of the pyrolysis reaction (reaction 1),
which releases carbon dioxide. The pyrolysed morphology is different for larger particle sizes. For example,
for 106–150 µm sieve size shell pyrolysed at 750 ◦ C
for 1 h (Fig. 2e and f), there is evidence that there is an
outer layer on top of a more closed, dense packed sublayer (which is clearly evident in Fig. 2f). This gives
further evidence that reaction 1 starts at the surface
and converts into lime and moves towards the centre of the shell particle in the classic shrinking core
mechanism.[24] These changes in morphology are consistent at larger particle sizes. A more detailed study
of the pyrolysis has been conducted, but will not be
presented here.
Treatment of the phosphate wastewater:
nature and morphology of mussel shells
XRD of the partially transformed shells before and
after use in the batch reactor are shown in Fig. 3. It
is to be noted that the amount of calcite in the original
mixed sample for this experiment (as indicated by peak
intensity) is higher than that shown in Fig. 1 due to
the lower pyrolysis temperature. This figure shows that
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
235
A. ABEYNAIKE et al.
Asia-Pacific Journal of Chemical Engineering
(a)
Intensity (arbitrary units)
Aragonite
20
30
40
50
60
70
2θ (degrees)
(b)
Calcite
Lime
Intensity (arbitrary units)
236
20
30
40
50
2θ (degrees)
60
70
Figure 1. XRD spectrum demonstrating the change in
crystal structure and chemical composition of a typical
pyrolysed mussel shell powder sample: (a) raw, which
consists predominantly of aragonite (b) 106–150 µm
mussel shell pyrolysed at 800 ◦ C, showing the transformation to calcite and lime. This figure is available in
colour online at www.apjChemEng.com.
Table 2. Weight percent of lime in each pyrolysed shell
sample.
Particle size
(µm)
53–106
53–106
53–106
212–250
Weight percentage
lime
Pyrolysis
750 ◦ C,
750 ◦ C,
800 ◦ C,
750 ◦ C,
1
2
2
1
h
h
h
h
30
24
84
23
Comparing ESEM images of raw and pyrolysed shells
before (Fig. 2) and after reaction in the batch reactor
(Figs 4 and 5) show that the structures have changed
after exposure to aqueous phosphate, indicating either
reactions or a physical process such as adsorption has
occurred. The raw (unpyrolysed) shells (Figs 4a and
b and 5a and b) do not appear to have significantly
changed from the types of surfaces as shown in Fig. 2,
indicating that they have not significantly dissolved.
Instead, they have a new crystalline phase precipitated
onto the shell surface. This indicates that heterogeneous adsorption and/or precipitation of the phosphate
is occurring on these shells, confirming the findings of
Currie et al .[14] In contrast, the surface structure of the
pyrolysed shells has all changed significantly after exposure to aqueous phosphate. The surfaces appear to have
both dissolved to become either more porous and irregular for fine powder (Fig. 4c and d) or flatter for the
coarse powder (Fig. 5c and d) than the surfaces were
prior to reaction. The difference between the different
size fractions is possibly because for the fine powder,
the particle size is smaller (53–106 µm shell particle
size), so more of the calcium carbonate throughout the
particle was transformed into CaO (lime) during pyrolysis – a smaller core is left during the shrinking core process due to the smaller original particle sizes – whereas
a larger core was left and only a surface shell was converted into the larger coarse powder (212–250 µm shell
particle size). Thus, when the fine powder reacted, lime
throughout the particles was dissolved, leaving an irregular inert skeleton, whereas a surface layer on the larger
coarse powder was dissolved, leaving a smoother surface inert particle. Furthermore, crystalline precipitates
are again present on the surfaces; however, the surface coverage is much lower than for the unpyrolysed
shells. All these indicate that at least some of the lime
formed during pyrolysis has dissolved and reacted with
the phosphate, some of it precipitating and/or adsorbing
on the surface of the remaining shell.
pH and extent of phosphate removal
following the reaction experiment the only mineral
present was calcite, indicating that all the lime had
dissolved into solution to leave behind the uncalcined
core of the particles. This further confirms the shrinking
core mechanism occurs during calcination (reaction 1).
It is to be noted that no calcium phosphate peaks
were detected in this sample, perhaps indicating that
for this size fraction there is little surface phosphate
due to precipitate and adsorption and/or the surface
concentration was too small to be detected by the X-ray
diffractometer used. Consequently, ESEM and EDS and
a wet phosphate analysis method as outlined in Section
Experimental were used to analyse for phosphate on the
reacted shells (results presented later).
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
pH measurements (Table 1) show that the solution was
initially slightly acidic. Initial pH ranges from 6.25
to 6.66. This is due to the potassium di-hydrogen
phosphate. After reaction, the pH increases for both raw
and pyrolysed shells. The solution becomes very basic
after reaction with pyrolysed shells compared to the raw
shells (the final pH ranges from 11.4 to 12.6 compared
to 7.7 to 7.9 for raw shells). This indicates a high
concentration of hydroxide ions due to the dissolution
of lime from the shells. The raw shells only became
slightly basic due to the smaller amount of calcium ion
dissolution. This further confirms that lime has been
formed from the pyrolysed shells, where there was none
in the raw shells. These results are comparable to those
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
PYROLYSED SHELLS: MECHANISM AND EXTENT OF PHOSPHATE REMOVAL
(a)
(b)
50.0µm
(c)
20.0µm
(d)
50.0µm
(e)
20.0µm
(f)
100.0µm
20.0µm
Figure 2. ESEM images of the shells before reaction with phosphate, showing the
effect of pyrolysis and particle size. Top: 0–53 µm sieve size, raw (unpyrolysed) shell
(a) 3000× magnification, (b) 6000× magnification. Middle: 0–53 µm sieve size shell
pyrolysed at 750 ◦ C for 1 h, (c) 2000× magnification, (d) 8000× magnification.
Bottom: 106–150 µm sieve size shell pyrolysed at 750 ◦ C for 1 h. (e) 1600×
magnification, (f) 6000× magnification.
found by Lee et al .,[13] who found a pH of more than
11 with pyrolysed oyster shells and 8.5 with raw oyster
shells.
The phosphate removal results using a mussel shell
concentration of 5 g l−1 (Trials 1–7 in Table 1) are
shown in Fig. 6. A very rapid and thorough removal of
phosphates was achieved by all the pyrolysed shell samples, regardless of the different particle sizes or pyrolysis conditions, with phosphate concentration reduced
to less than 0.5 mg l−1 within 5 min. This represents
a 95%-reduction in phosphate, comparable with previous results.[11,14] The exact reduction is unclear as
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
0.5 mg l−1 is the limit of detection for the colorimetric
analysis method.
The high concentration of mussel shell used in
these trials provides calcium and hydroxide ions well
in excess of the stoichiometric amount required for
hydroxyapatite precipitation. Based on the weight
fraction of lime (Table 2), the calcium ion concentration in the solution is at least 20 mmol l−1 .
However, the stoichiometric concentration of calcium
ions for hydroxyapatite formation is much less at
0.18 mmol l−1 . With such a high excess of calcium, the
equilibrium of the reactions 2 and 4 is pushed towards
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
237
A. ABEYNAIKE et al.
Asia-Pacific Journal of Chemical Engineering
Table 3. BET surface area of the raw mussel shells.
Intensity (arbitrary units)
(a)
Sieve size range (µm)
Calcite
Lime
20
30
40
50
Surface area (m2 g−1 )
53–106
212–250
60
70
2θ (degrees)
1.22
0.43
solution is to use pyrolysed shell powder to get stoichiometric reaction and removal of the phosphate with
the powder.
(b)
Intensity (arbitrary units)
238
Pyrolysed shell phosphate removal
mechanisms
Calcite
20
30
40
50
2θ (degrees)
60
70
Figure 3. XRD spectrum of mussel shell powder, sieve size
53–106 µm, pyrolysed at 750 ◦ C for 2 h, (a) after pyrolysis
at 750 ◦ C for 1 h, (b) after reaction with phosphate in the
batch reactor. This figure is available in colour online at
www.apjChemEng.com.
the formation of hydroxyapatite, as well as increasing
the reaction rate. This explains the high conversion in
such a short time for all the pyrolysed shell samples.
Raw shell phosphate removal mechanism
The raw mussel shells removed far less phosphate than
the pyrolysed shells and so are not well suited for phosphate removal applications. The coarse shell resulted
in a 22% decrease in phosphate, whereas the fine shell
resulted in a 45% decrease. BET analysis shows that the
fine powder has three times the surface area available
than that of the coarse powder (Table 3). The phosphate removal increases with surface area, confirming
that adsorption must be the mechanism of phosphate
removal for the raw powders. The results are similar to
those found by Currie et al .[14] Kwon et al .[11] reported
less phosphate removal, however, the particle size used
in that study is not clear, so the use of larger particles
may have caused the lower result. Potentially, the phosphate removal of raw shell powder can be improved
by increasing the shell concentration or using a finer
powder (increasing surface area). However, in practice, this would introduce more material handling costs,
more sludge to dispose of and difficulty in keeping finer
powders dispersed in the solution. Therefore, a better
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
For the pyrolysed mussel shells, a mass balance showing the final location of the total phosphate has been
based on the measurements of phosphate in each phase
(phosphate content in the liquid, the suspended solids
and on the non-suspended remaining shell particles after
reaction) and is presented in Fig. 7. When pyrolysed
shell was used in the batch reactor, significant levels of
phosphate were detected in both the suspended precipitate and the shell particles at the end of the trials. The
phosphate in the precipitate is formed by the homogeneous nucleation of calcium phosphates. The phosphate
on the calcite shell particles is likely the result of both
heterogeneous nucleation of a calcium phosphate precipitate and adsorption of phosphate.
The suspended solids are also not pure calcium
phosphate (such as hydroxyapatite or brushite), as the
amount of phosphate detected does not exceed 7% of the
mass of precipitate. This result is supported by an EDS
spectrum that was measured in the suspended solids
(results not presented), which also shows a relatively
small amount of phosphorus relative concentration
to calcium, carbon and oxygen. For a pure calcium
phosphate precipitate, e.g. hydroxyapatite, the mass
fraction of phosphate should be higher (e.g. 57% for
hydroxyapatite). The suspended solids are therefore
likely to be mainly calcium carbonate – possibly some
from the unreacted shell cores and some formed by
the reaction of calcium ions with dissolved carbon
dioxide (reactions 5 and 6) – with some hydroxyapatite
and other calcium phosphates formed before the pH
exceeded 10.
(5)
(6)
The crystal phases in the suspended precipitate could
not be determined by XRD as the sample was too
small to obtain a definitive spectrum. However, because
a pH of 12 was achieved during the reaction, it is
likely that at least some of the precipitate formed
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
PYROLYSED SHELLS: MECHANISM AND EXTENT OF PHOSPHATE REMOVAL
(a)
(b)
100.0µm
(c)
10.0µm
(d)
100.0µm
20.0µm
Figure 4. ESEM images showing the effect of pyrolysis and particle size on the surface
of the fine mussel shell powder (sieve size 53–106 µm) after reaction with aqueous
potassium di-hydrogen phosphate in the batch reactor. Surfaces shown both at low and at
high magnifications. Top: raw (unpyrolysed) shells, (a) 1504× magnification, (b) 10 000×
magnification. Bottom: shells pyrolysed for 2 h 750 ◦ C, (c) 1500× magnification, (d) 6000×
magnification.
was hydroxyapatite, which is the desired end-product
because it is the most thermodynamically stable form
of calcium phosphate[18] and therefore will enable the
most irreversible phosphate removal. ESEM images
(Fig. 8) of the precipitates from Trials 4 and 5 in
Table 1 show that the suspended solids are crystalline
and quite different to the shells, both before (Fig. 2)
and after reaction (Figs 4 and 5) with phosphate,
indicating that the suspended solids should mainly
contain precipitates that originate from reactions within
the solution rather than calcite from unreacted shell
cores. Therefore, a proportion of phosphate in the
suspended solids is indicative of a greater propensity for
homogenous precipitation of this phosphate, rather than
heterogeneous precipitation on the surface of shells.
Figure 7 also shows that the amount of phosphate
in the suspended solids was doubled when the coarse
shell powder was used compared with the fine powder.
Coarse powders have a smaller surface area than fine
powders, so there is less area available for heterogeneous (surface) precipitation. The coarse powder still
achieves the same overall phosphate removal, but most
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
of the removal is by homogeneous precipitation, which
forms the precipitate in the suspended solids. Adsorption alone cannot account for this (compared to raw
shells – Fig. 6), thus heterogeneous precipitation of calcium phosphates must also be occurring.
Therefore, more heterogeneous precipitation can be
encouraged for a finer particle size as long as the pyrolysed shell has not been fully converted (i.e. it has an
insoluble calcium carbonate core). In contrast to these
results, when Lee et al .[13] investigated the reaction
mechanisms with pyrolysed oyster shells, they found
that the phosphorous removal was by homogenous precipitation with little adsorption or heterogeneous precipitation. The difference in mechanism can possibly
be explained by a higher percentage conversion to lime
of their oyster shells, therefore minimising the calcium
carbonate surfaces that may be seeding the heterogeneous precipitation in this work. In terms of the practical
implications of this result, the precipitation of calcium
phosphates, such as hydroxyapatite, has been shown
to be much faster when it occurs on the surface of
calcite,[19] therefore, for a faster treatment time, fine
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
239
A. ABEYNAIKE et al.
Asia-Pacific Journal of Chemical Engineering
(a)
(b)
10.0µm
200.0µm
(c)
(d)
200.0µm
20.0µm
Figure 5. ESEM images showing the effect of pyrolysis and particle size on the surface
of the coarse mussel shell powder (sieve size 212–250 µm) after reaction with aqueous
potassium di-hydrogen phosphate in the batch reactor. Top: raw (unpyrolysed) shells,
(a) 600× magnification, (b) 10 000× magnification. Bottom: shells pyrolysed for 2 h at
750 ◦ C, (c) 800× magnification, (d) 6000× magnification.
Solution
Shell - measured
100
Suspended solids
Shell - calcium interference
90
Fine, 800°C, 2hr
80
Phosphate Removal (%)
240
70
Fine, 750°C, 2hr
60
Fine, 750°C, 1hr
50
40
Coarse, 750°C, 1hr
30
Coarse, no pyrolysis
20
10
Fine, no pyrolysis
0
0
10
20
30
40
50
60
70
80
90
Time (min)
Figure 6. Phosphate removal over time showing the effect
of particle size, shell pyrolysis time and shell pyrolysis
temperature, Trials 1–7. , fine 1 h 750 ◦ C; , coarse 1 h
750 ◦ C; , fine 2 h 750 ◦ C; ◊, fine 2 h 800 ◦ C; , fine, not
pyrolysed; , coarse, not pyrolysed. This figure is available
in colour online at www.apjChemEng.com.
°
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
0%
20%
40%
60%
80%
100%
Percentage of total initial phosphate
Figure 7. Average percent phosphate in the three different
phases in the reaction vessel at the end of the batch run (in
solution, in the suspended solids, and on the undissolved
shell), for Trials 2–7. The vertical axis details the type of shell
used for each trial. The phosphate attributed to the shell
was quantified in two ways – that which could be directly
measured and that which was estimated because it could
not be directly measured because of calcium interference in
an assay.
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
PYROLYSED SHELLS: MECHANISM AND EXTENT OF PHOSPHATE REMOVAL
(a)
(b)
20.0µm
20.0µm
Figure 8. ESEM images of the crystalline suspended solids resulting from reacting
pyrolysed shells (sieve size 53–106 µm) with aqueous potassium di-hydrogen
phosphate, (a) Run 4 (Table 1), 6000× magnification, (b) Run 5 (Table 1), 5000×
magnification.
particles (i.e. particle size: 53–106 µm) with a partial
conversion of the shells to lime like those used in this
work is recommended, or alternatively some seeding in
the reactor of fine raw shell particles mixed with fine
pyrolysed shell particles that have been fully converted
into lime, in order to allow heterogeneous precipitation
to occur. Further work is continuing to optimise this
type of treatment system.
Having measured the phosphate content of the suspended solids and remaining shells, there is still a proportion of the phosphate removed from the solution that
is unaccounted for (Fig. 7). This is most likely due to
interference of the colorimetric test by calcium ions that
dissolved from the solids into the acidic reagent solution. These ions do not interfere at concentrations of
less than 1000 mg l−1 .[23] This limit was well exceeded
when the shell powder samples were tested because of
the large amounts of calcium carbonate dissolving in
the reagent. The suspended solid samples had much
less mass than the shell samples, so interference is
less likely. Therefore, this unaccounted phosphate was
treated as being on the shells in the above analysis.
Optimising mussel shell concentration
100
90
Phosphate Removal (%)
80
70
60
50
40
30
20
10
0
−10
−20
0
10
20
30
Time (min)
40
50
60
Figure 9. Phosphate removal over time using various
concentrations of heat-treated mussel shell powder, Trials
8–14 (particle size: 53–106 µm, pyrolysed for 1 h at
750 ◦ C). , 33 mg l−1 (0% excess); , 49 mg l−1 (50%
excess); ◊, 65 mg l−1 (100% excess); ×, 81 mg l−1 (150%
excess); , 98 mg l−1 (200% excess); , 196 mg l−1 (500%
excess); ♦, 5000 mg l−1 (11 250% excess). This figure is
available in colour online at www.apjChemEng.com.
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
Using mussel shell in excess can potentially lead to
an expensive wastage of both material (i.e. the large
amount of excess shell increases the equipment and processing costs, also increasing the amount of used shell
that must be removed and disposed of) and energy (i.e.
to process, handle and transport the shells). Therefore,
experiments to minimise the amount of shell added were
conducted. Varying amounts of mussel shell was added
to 10 mg PO4 l−1 synthetic wastewaters in a 1-l reactor to determine the effect of changing the mussel shell
concentration on the extent of phosphate removal and
identify the optimal mussel shell loading (Trials 8–15 in
Table 1). The stoichiometric (0% excess) mussel shell
concentration for hydroxyapatite formation (used as it is
the most stable and preferred precipitate for phosphate
removal) was calculated as being 33 mg l−1 , based on
a lime content of 30% by weight and reaction 4. The
percentage of the shells added in excess was then determined based on this. Results (Fig. 9) show that phosphate removal mainly occurs within the first 20 min of
reaction and decreases after this. This may be because
of the phosphate redissolving as the system settles at
equilibrium. Therefore, in practice, the reacted solids
will need to be removed from the wastewater just before
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
241
242
A. ABEYNAIKE et al.
20 min of treatment time has expired, to obtain the maximum removal of phosphate.
In terms of optimising the mass of shell added, the
results (Fig. 9) show that when using the stoichiometric amount of pyrolysed shell (0% excess), no significant removal of phosphate was observed for more than
1 h. As excess shell above the stoichiometric amount
is increased, the percentage phosphate removal also
increased, though this effect was not as apparent at concentrations greater than 150% excess shell (81 mg l−1 ).
This result is confirmed by a statistical analysis of the
experimental uncertainty in the data. This experimental
uncertainty was determined from the spread between
the data points from the three repeated trials conducted
at 0% excess. Using this spread in a statistical analysis
of the data, it was found that at 30 and 45 min there is
a clear difference between the data points up to 150%
excess shell, confirming that up to this shell loading,
increasing the shell concentration has a significant effect
on the phosphate removal. However, for 150% excess
shell and higher, there is no statistically significant difference between the data points, indicating that there is
no statistically discernable effect (within the calculated
error in the data) of increasing the concentration.
The data, however, indicates that an excess of 500%
can more readily yield a higher phosphate removal value
than at 150% excess, which is comparable phosphate
removal to that with 5 g l−1 shell (11 250% excess) for
the first 20 min of experiment, but using the 25 times
less pyrolysed shell. At this point, a 90%-reduction
in phosphate is achievable, giving a final phosphate
concentration of 1 mg l−1 . Therefore, based on these
results, it can be concluded that 196 mg l−1 shell (500%
excess) is a practical dosage to achieve the maximum
phosphate removal with these pyrolysed shells.
CONCLUSIONS
This work has established an optimised process in
which the shellfish industry’s shell waste can be converted into lime and then used for phosphate removal
from wastewaters. Specifically, this work has shown
that:
1. Phosphate removal of more than 95% can be
achieved by partially calcined pyrolysed mussel
shells at a concentration of 5 g l−1 within 5 min,
regardless of particle size or pyrolysis conditions.
The large amount of excess calcium and hydroxide
ions available under these conditions caused the high
conversion and rapid reaction rate.
2. There is a mixed mechanism for the phosphate
removal with partially calcined pyrolysed mussel
shells: phosphate is removed from solution by the
homogeneous nucleation of a calcium phosphate precipitate, forming part of the suspended solids, and by
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pacific Journal of Chemical Engineering
3.
4.
5.
6.
the heterogeneous nucleation of the precipitate on the
surface of the calcite shell powder. Heterogeneous
nucleation dominated on the fine shell particles (original shell size: 53–106 µm) and homogeneous nucleation dominated on coarse particles (original shell
size: 212–250 µm).
Therefore for a faster treatment time heterogeneous
precipitation should be encouraged. This could be
achieved by using fine particles partially calcined
to lime, or alternatively by seeding the reactor with
fine raw shell particles mixed in fine pyrolysed shell
particles that have been fully converted to lime.
Raw shells are unlikely to be suitable for phosphate
removal applications as they remove much less
phosphate than the pyrolysed shells. The amount of
phosphate removed was dependent on the surface
area of the shell particles, indicating that adsorption
was the phosphate removal mechanism.
For concentrations of partially calcined pyrolysed
mussel shell between 33 and 196 mg l−1 (0% and
500% excess of the stoichiometric concentration for
complete phosphate removal in hydroxyapatite), the
maximum removal of phosphate is reached within
the first 20 min of the treatment time after which
phosphate redissolves. Therefore in practice, the
solids must be removed before 20 min to obtain the
maximum removal of phosphate.
For 10 mg L−1 synthetic phosphate wastewaters in
a 1 l- reactor, 196 mg l−1 shell (500% excess) is the
practical dosage to achieve the maximum phosphate
removal with partially calcined pyrolysed mussel
shells.
Overall, these results demonstrated that pyrolysed
mussel shells are a viable alternative raw material
source for producing lime for wastewater treatment.
Because the dominant mechanism of removal is precipitation (whether it be homogeneous or heterogeneous),
this lime can be used to permanently remove phosphates
as a calcium phosphate precipitate from wastewaters
before they enter rivers and lakes, in order to prevent
eutrophication. Such a treatment process is ideal for
countries such as NZ, where there is a rich resource
of shells from the aquaculture industry, as well as
a need to prevent further eutrophication in lakes and
rivers.
Acknowledgements
The authors would like to thank Laura Liang, Jeffrey
Ang, Allan Clendinning, Catherine Hobbis and Bryony
James from the Department of Chemical and Materials
Engineering, University of Auckland for their help
with this work. The authors would also like to thank
Su-Ling Brooks, associated with the Department of
Process Engineering & Applied Science at Dalhousie
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
PYROLYSED SHELLS: MECHANISM AND EXTENT OF PHOSPHATE REMOVAL
University, Canada, who instigated this project area at
the University of Auckland. Finally, the authors would
like to thank Sanford Limited for donating the mussel
shells used in this work.
REFERENCES
[1] P.M. Haygarth, L.M. Condron, A.L. Heathwaite, B.L. Turner,
G.P. Harris. Sci. Total Environ., 2005; 344, 5–14. DOI:
10.1016/j.scitotenv.2005.02.001.
[2] J. Hilton, M. O’Hare, M.J. Bowes, J.I. Jones. Sci. Total Environ., 2006; 365, 66–83. DOI: 10.1016/j.scitotenv.2006.02.055.
[3] R. Taylor, I. Smith, P. Cochrane, B. Stephenson, N. Gibbs.
The State of New Zealand’s Environment 1997, The Ministry
for the Environment and GP Publications: Wellington, 1997.
[4] Ministry for the Environment. Environment New Zealand 2007,
Publication ME847, Ministry for the Environment: Wellington,
2007.
[5] G. Tchobanoglas, F.L. Burton. Wastewater Engineering: Treatment, Disposal and Reuse, 3rd ed., McGraw-Hill: New York,
1991.
[6] R.S. Boynton. Chemistry and Technology of Lime and
Limestone, John Wiley and Sons: New York, 1980.
[7] C. Barnaby, An Investigation into the Reuse of Organic Waste
Produced by the New Zealand Mussel Industry. Masters of
Applied Science Thesis, AUT University, Auckland, 2004.
[8] M. Burrell, L. Meehan, S. Munro, The New Zealand Aquaculture Strategy. Commissioned by the New Zealand Aquaculture
Council with assistance of the New Zealand Seafood Industry
Council and the Ministry of Economic Development. Creative
Design Advertising Limited: Hastings, 2006.
[9] Sanford Limited. Sustainable Development Report 2006.
Sanford Limited: Auckland, 2006.
[10] G. Robertson. Shellfish Waste Arisings in Scotland: Strategic
Baseline Analysis. A report prepared for the UK Scallop
Association. Business Environment Partnership North East:
Bucksburn, 2004.
 2010 Curtin University of Technology and John Wiley & Sons, Ltd.
[11] H.B. Kwon, C.W. Lee, B.S. Jun, J.D. Yun, S.Y. Weon,
B. Koopman. Resour. Conserv. Recycl., 2004; 41, 75–82,
DOI: 10.1016/j.resconrec.2003.08.005.
[12] C.W. Lee, H.C. Kwon, H.P. Jeon, B. Koopman. Mater. Sci.
Forum., 2005; 486-487, 177–180.
[13] C.W. Lee, H.B. Kwon, H.P. Jeon, B. Koopman. J Cleaner
Prod., 2009; 17, 683–687. DOI: 10.1016/j.jclepro.2008.11.
019.
[14] J.A. Currie, N.R. Harrison, L. Wang, M.I. Jones, M.S. Brooks.
Asia-Pac. J. Chem. Eng., 2007; 2, 460–467, DOI:
10.1002/apj.82.
[15] E.E. Ruppert, R.S. Fox, R.D. Barnes. Invertebrate Zoology,
Brooks/Cole-Thomson Learning: Belmont, California, 2004;
pp.283–412.
[16] C. Namasivayam, A. Sakoda, M. Suzuki. J. Chem. Technol.
Biotechnol., 2005; 80, 356–358. DOI: 10.1002/jctb.1175.
[17] H.B. Kwon, C.W. Lee. Method for producing hydroxyapatite complex using fluidised bed reactor and pyrolysed oyster shell and producing device therefore. Korean Patent
KR20040046179, 2004.
[18] G.H. Nancollas, M.B. Tomson, G. Battaglia, H. Wawrousek,
M. Zuckerman. In Chemistry of Wastewater Technology (Ed.:
A.J. Rubin), Ann Arbor Science: Ann Arbor, 1978; pp.17–30.
[19] Y. Song, P.G. Weidler, U. Berg, R. Nüesch, D. Donnert.
Chemosphere, 2006; 63, 236–243. DOI: 10.1016/j.chemo
sphere.2005.08.021.
[20] T. Suzuki, S. Inomata, K. Sawada, J. Chem. Soc., Faraday
Trans. 1, 1986; 82, 1733–1743. DOI: 10.1039/F19868201733.
[21] W.A. House, L. Donaldson. J. Colloid Interface Sci., 1986;
112, 309–324. DOI: 10.1016/0021-9797(86)90101-3.
[22] Z.R. Hinedi, S. Goldberg, A.C. Chang, J.P. Yesinowski. J.
Colloid Interface Sci., 1992; 152, 141–160. DOI: 10.1016/
0021-9797(92)90015-E.
[23] A.D. Eaton, L.S. Clesceri, E.W. Rice, A.E. Greenberg (eds).
Standard Methods for the Examination of Water & Wastewater,
21st ed., American Public Health Association: Washington,
DC, 2005; pp.4-146–4-162.
[24] O. Levenspiel, Chemical Reaction Engineering, 3rd ed., John
Wiley & Sons: New York, 1998; pp.566–588.
Asia-Pac. J. Chem. Eng. 2011; 6: 231–243
DOI: 10.1002/apj
243
Документ
Категория
Без категории
Просмотров
1
Размер файла
591 Кб
Теги
powder, powdered, phosphate, mechanism, mussel, extent, eutrophication, control, effect, removal, concentrations, pyrolysed, size, shell, particles
1/--страниц
Пожаловаться на содержимое документа