1
5
Journal of Materials Chemistry
PAPER
b807920j
1
5
1
10
Synthesis and characterisation of magnesium substituted
calcium phosphate bioceramic nanoparticles made via
continuous hydrothermal flow synthesis
10
2 Aqif Anwar Chaudhry, Josie Goodall, Martin Vickers,
Jeremy Karl Cockcroft, Ihtesham Rehman,
Jonathan Knowles and Jawwad Arshad Darr*
15
15
20
20
25
25
30
30
35
35
40
40
45
45
50
50
55
55
ART B807920J_GRABS
PAPER
1
5
1
www.rsc.org/materials | Journal of Materials Chemistry
Synthesis and characterisation of magnesium substituted
calcium phosphate bioceramic nanoparticles made via continuous
hydrothermal flow synthesis†
1
5
Aqif Anwar Chaudhry,abc Josie Goodall,a Martin Vickers,a Jeremy Karl Cockcroft,a Ihtesham Rehman,b
Jonathan Knowlesd and Jawwad Arshad Darr*a
10
15
20
25
Received 9th May 2008, Accepted 24th September 2008
First published as an Advance Article on the web ?????
DOI: 10.1039/b807920j
10
Continuous hydrothermal flow synthesis (CHFS) technology has been used as an efficient and direct
route to produce a range of largely crystalline magnesium substituted calcium phosphate bioceramics.
Initially, magnesium substituted hydroxyapatite, Mg-HA, according to the formula
[Ca10xMgx(PO4)6(OH)2] was prepared in the CHFS system for x ¼ 0.2 [where x:(10 x) is the Mg:Ca
ratio used in the reagents]. Biphasic mixtures of Mg-HA and Mg-whitlockite were obtained
corresponding to x values in the range x ¼ 0.4–1.6. The direct synthesis of phase pure crystalline Mgwhitlockite [based on the formula (Ca3yMgy(HPO4)z(PO4)22z/3] was also achieved using the CHFS
system for the range y ¼ 0.7–1.6 (this corresponds to the range x ¼ 1.6–5.3). With increasing
substitution of magnesium for calcium, the material became ever more amorphous and the BET surface
area generally increased. All the as-precipitated powders (without any additional heat treatments) were
analyzed using techniques including X-ray powder diffraction, Raman spectroscopy and Fourier
transform infra-red spectroscopy. Transmission electron microscopy (TEM) images revealed that in the
case of y ¼ 1.2, the Mg-whitlockite material was comprised of ca. 28 nm sized spheres. The use of the
CHFS system in this context facilitated rapid production of combinations of particle properties
(crystallinity, size, shape) that were hitherto unobtainable in a single step process.
1. Introduction
40
45
Synthetic hydroxyapatite, HA, [Ca10(PO4)6(OH)2], is a bioceramic that is chemically similar to the mineral component of
hard tissues (biological apatite) and resorbs slowly. HA is widely
used in hard tissue augmentation and replacement.1 In contrast,
beta-tricalcium phosphate, b-TCP, [Ca3(PO4)2], another bioceramic, is resorbable in vivo.2,3 Consequently, the use of
biphasic HA and b-TCP mixtures for bone grafts can facilitate
rapid bone formation around the implant site compared to HA
alone.4,5 It has been reported previously that when biphasic
calcium phosphate (BCP) ceramic (55% HA, 45% b-TCP) was
implanted in femoral cortical bones of dogs, new bone formed
more quickly for BCP compared to HA alone.6,7
Ionic substitutions are often introduced into synthetic bioceramics in order to more closely match the chemical compositions found in biological apatites. Indeed, biological apatites
differ from synthetic HA in that the former also contain
50
a
55
20
25
30
30
35
15
Clean Materials Technology Group, Department of Chemistry, University
College London, Christopher Ingold Laboratories, 20 Gordon Street,
London WC1H 0AJ, UK. E-mail: j.a.darr@ucl.ac.uk
b
Department of Materials and IRC in Biomedical Materials, Queen Mary
University of London, Mile End Rd, London E1 4NS, UK
c
IRC in Biomedical Materials, Comsats Institute of Information
Technology, Lahore Campus, Defence Road, Lahore, Pakistan
d
Division of Biomaterials and Tissue Engineering, UCL Eastman Dental
Institute, 256 Gray’s Inn Road, London WC1X 8LD, UK
† Electronic supplementary information (ESI) available: Tables, XRD,
DTA. See DOI: 10.1039/b807920j
carbonate (CO32), magnesium (Mg2+), zinc (Zn2+), sodium
(Na+), silicate (SiO44) and fluoride (F) ions, to name a few.8–11
These ions can play an important role in the biological behaviour
of biological apatites.12
In Mg2+ containing biological apatites13 such as those found in
pathological calcifications, human dental calculus or carious
lesions, a magnesium stabilized b-TCP (Mg-bTCP) or magnesium whitlockite [Mg-whitlockite, Ca18Mg2H2(PO4)14] is often
formed.14–16 Dentine, enamel and bone typically contain up to
1.1, 0.4 and 1.0 wt% Mg2+ ions, respectively.17,18 In calcified
tissues, the Mg2+ ion content is higher at the beginning of the
calcification process and decreases with greater calcification.13,18,19 Mg depletion in people results in decreased osteoblastic and osteoclastic activities and bone fragility.13,20 Indeed,
there are studies showing that younger bone in mammals
contains higher Mg levels than older bone.21
Mg2+ substitution for Ca2+ in HA reduces crystallinity,
increases solubility, and lowers the temperature at which
conversion of HA into b-TCP occurs.12,17,19,22,23 As the amount of
substituted magnesium affects thermal phase stabilities, this has
implications for sintering behaviour.3 For example, the b-TCP to
a-TCP phase transformation normally occurs at ca. 1180 C.
However Mg2+ substitution for Ca2+ can increase the transformation temperature to ca. 1500 C.3,24 This enables improved
sintering of b-TCP at elevated temperatures without deleterious
formation of a-TCP (the latter is a less bioactive polymorph).
Magnesium substituted HA (Mg-HA) can be made via
ambient
temperature
precipitation
reactions
from
This journal is ª The Royal Society of Chemistry 2008
J. Mater. Chem., 2008, 18, 1–10 | 1
ART B807920J
35
40
45
50
55
1
5
10
15
20
25
30
35
40
45
50
solution2,4,12,25–27 or by solid-state reactions at elevated temperatures.5,20 Magnesium substitution levels of up to 1.6 wt% have
been reported using such precipitation reactions.13 Surprisingly,
as much as 28.4 wt% substitution of Mg2+ in HA has also been
claimed in the literature using mechanochemical synthesis
routes.28 However, an excess of Mg2+ can also be undesirable as it
is known to reduce bioactivity in certain biomaterials.29
As well as Mg-HA, other phases such as Mg-whitlockite (or
Mg-bTCP), can also be made from precipitation reactions at
relatively low temperatures (range 37 to 95 C) and acidic or
neutral pH.22,26,30 For example, Mg2+ ions can stabilise the
formation of Ca2.7Mg0.3(PO4)2.H2O.31
BCP mixtures are of interest for use in repair of periodontal
and bone defects,32,33 as bone graft substitutes in mastoid cavity
obliteration34 and scoliosis surgery.35 Within these applications,
the particle size, presence of substituting elements and the crystallinity strongly affect resorbability. Thus, the ability to prepare
biphasic mixtures with controlled crystallinity, size and phase
composition is of great interest.
Crystalline whitlockite is usually made from a wet precipitated
amorphous phase that may require a further heat treatment
step.22,23,31 Similarly, there do not appear to be any reports of bTCP being precipitated directly from solution.36 Rather, it is
routinely obtained via solid-state reactions or via thermal
decomposition of calcium-deficient hydroxyapatite (CDHA)
near or above 750 C.2,4,26,37 Interestingly, when carbonate and
magnesium ions are simultaneously substituted into HA at
a basic (rather than acidic or neutral) pH of ca. 9, the resulting
apatite is much more thermally stable.12,18,24
To date, there have been no reports of direct and rapid
syntheses of crystalline Mg-HA or Mg-whitlockite. The
majority of the reported synthesis methods tend to be multistep, energy intensive or time consuming processes. In our
previous report, the authors previously described a novel
method for the instant synthesis of crystalline nano-sized HA
using a continuous hydrothermal flow synthesis system (CHFS)
which utilises a flow of superheated water to effectively nucleate
and crystallise the HA in flow without the need for any further
processing or ageing.38 This method can be compared to more
conventional synthesis methods conducted at room temperature
which require ageing steps and/or heat-treatment to obtain
crystalline HA.39–42 CHFS methods have been used previously
by the authurs and others for the synthesis of a range of
nanoceramic oxides.38,43–51
In this work, a CHFS reactor was used as a direct and efficient
method to controllably produce a range of magnesium
substituted calcium phosphates and biphasic nano-bioceramic
mixtures, with differing crystallinities. In particular, the first
rapid and CHFS of crystalline Mg-HA and crystalline Mgwhitlockite, respectively, are reported.
2. Experimental
55
2.1.
Materials and analytical equipment
Diammonium hydrogen phosphate, [(NH4)2HPO4, 98.3%],
calcium nitrate tetrahydrate [Ca(NO3)2$4H2O, 99%] and
magnesium nitrate hexahydrate [Mg(NO3)2$6H2O, 97%] were
supplied by Sigma-Aldrich Chemical Company (Dorset, U.K.)
and used as obtained. Aqueous ammonia (ca. 16.9 mol L1)
supplied by VWR International (UK) was used to adjust the
solutions’ pH (sold as 30% w/w). 10 mega-ohms deionised water
was used in all reactions.
Samples were freeze-dried using a Virtis Advantage Freeze
Dryer, Model 2.0 ES, supplied by BioPharma. A JEOL 2010 A
TEM (200 kV accelerating voltage) was used for generating images
of particles. Elemental analyses for the powders were performed
using a calibrated energy dispersive spectrometer (EDS) attached
to the TEM. A calibrated Oxford Instruments Inca 400 EDX
detector connected to a scanning electron microscope (JEOL 5410
LZSEM) was also used to carry out the elemental analysis for all
the samples. Averages of 10 area scans were used to calculate
average elemental compositions. The SEM was operated at 25 kV
and all samples were placed onto conductive carbon paper and
then carbon coated. Image Tool UTHSCSA version 3.0 software
was used for estimating particle size.
For batch sample work, PXRD (powder X-ray diffraction)
data were collected on a Siemens D5000 X-ray diffractometer
using Cu-Ka radiation (l ¼ 1.5418 Å, i.e. 15.415 nm) over the
2q range 5–70 with a step size of 0.02 and a count time of 1
s. Samples were placed into a 2.5 cm wide, 2 mm well in
a holder as flat pressed powders (no further sample preparation). For Rietveld refinement work, PXRD data were
collected on a Stoe StadiP transmission-geometry diffractometer using Ge <111> monochromated Cu Ka1 radiation (l ¼
1.54056 Å). Diffraction patterns were obtained from samples
sealed in 0.3 mm diameter borosilicate glass capillaries and
1
5
10
15
20
25
30
Table 1 Sample identification, corresponding wt% (using EDS attached
to a SEM) and x and y values according to the formulae used. The x
values assume a formula of Ca10-xMgx(PO4)6(OH)2 as observed for
samples 2–5 and the y values are based on the formula
Ca3yMgy(HPO4)z(PO4)22z/3 as assigned for samples 6–11 (for selected
samples where y is quoted, the equivalent value of x is also given in
brackets for comparison). The sample labelled as CaP is HA reported
from results published elsewhere.38 The sample labelled as MgP represents a powder made using magnesium nitrate and no calcium nitrate.
Key: Mg-HA ¼ magnesium substituted hydroxyapatite, Mg-W ¼ Mgwhitlockite, BCP ¼ Mg-HA + Mg-W.
40
Magnesium (wt%)
1
2
3
4
5
6
7
8
9
10
11
12
Sample ID
Added
Measured
x
y
XRDa
CaP
0.5Mg-CaP
1Mg-CaP
1.5Mg-CaP
2Mg-CaP
4Mg-CaP
6Mg-CaP
8Mg-CaP
10Mg-CaP
12Mg-CaP
14Mg-CaP
MgP
0
0.5
1
1.5
2
4
6
8
10
12
14
—
0
0.7
1.0
1.6
2.1
3.9
5.8
8.0
9.1
11.4
13.1
—
—
0.2
0.4
0.6
0.8
1.6
(2.4)
(3.1)
(3.9)
(4.6)
(5.3)
—
—
—
—
—
—
—
0.7
0.9
1.2
1.4
1.6
—
HA
Mg-HAb
BCPbc
BCPbc
BCPbc
BCPbc
Mg-Wc
Mg-Wc
Mg-Wc
Mg-Wcd
Mg-Wcd
Highly
amorphous
a
Structural assignment based on PXRD data gave patterns that were
a good match to the following: b ICDD pattern 09-0432c
ICDD
pattern
70-2064Hydroxyapatite-Ca10(PO4)6(OH)2;
d
whitlockite-Ca18Mg2H2(PO4)14; the PXRD data show very broad
humps which give a good approximation to ICDD pattern 70-2064
whitlockite.
This journal is ª The Royal Society of Chemistry 2008
2 | J. Mater. Chem., 2008, 18, 1–10
ART B807920J
35
45
50
55
1
5
10
15
20
25
30
measured with a linear position sensitive detector (nominal
aperture 4.5 2q), which was scanned from 2 to 95 2q in steps
of 0.1 2q with a count time of 30 s per step. Samples were
prepared by gentle crushing of the powders with a mortar and
pestle, filling the capillaries with powder by tapping to fill the
tube and then sealing the ends under a flame. Each scan was
repeated 8 times, compared and checked for consistency, and
then they were added together to create a single summed data
set with data binned in steps of 0.02 2q. The long scan times
were necessary to provide data of suitable statistical quality for
Rietveld refinement purposes. Four high-resolution data sets
were collected from samples with differing percentages of
added Mg during the synthesis stage, namely CaP,38 0.5MgCaP, 1.5Mg-CaP and 4Mg-CaP (CaP represents HA as
reported elsewhere by us38 and the number before Mg-CaP is
the nominal Mg wt% value expected to be substituted into HA
(in place of calcium) on the basis of the reagent composition
(see Table 1). The XRD data were analysed with the PROFIL
Rietveld refinement suite52 using known single-crystal models
for the structures of HA53 and whitlockite.54
A Nicolet Almega Dispersive Raman Spectrometer (785 nm
laser) was used in the wavenumber range 1267–91 cm1 averaging
20 scans for 2 s each. Fourier-transform infrared (FTIR) spectra
of the samples were collected using a Nicolet FTIR 800 spectrometer fitted with a photoacoustic sampler (MTech PAS Cell).
Samples were presented as flat powders in the cell with no special
sample preparation. Spectra were obtained in the range 4000–
400 cm1, at 8 cm1 resolution averaging 256 scans. BET surface
area measurements (using N2 gas adsorption method) were
performed on a Micromeritics Gemini analyser; powders were
first degassed at 110 C for 3 h prior to analyses.
2.2.
35
40
45
50
55
Apparatus and syntheses
All samples (as shown in Table 1) were made using a CHFS
system (see Fig. 1), which consists of the following basic
components. The system consists of three Gilson 305 HPLC
pumps fitted with 25 mL pump-heads, 316SS Swagelok 1/800
stainless steel fittings and tubing (except for the 3/800 countercurrent mixer, the design of which has been published elsewhere46) and a 2.5 kW electrically powered water preheater
which were built from 1/400 fittings. A small 5 cm long and 2.5cm
diameter, 200 W band heater (Watlow) was placed around
a duraluminum interface jacket which surrounds a countercurrent mixer pipe.46 The superheated water feed was created by
pumping water through the preheater (via pump 1). This feed was
then mixed with calcium phosphate mixture at a 3/800 countercurrent mixer and then the resultant nanoparticles pass through
a ¼00 and 97 cm long vertical water cooled stainless steel pipe
(shown as C in Fig. 1). Each high pressure line for each pump
also incorporates a pressure gauge (RS model no. 540-255),
Nupro SS-4R3A pressure release safety valves (set to 29 MPa)
and Swagelok non-return valves (not shown in Fig. 1 for clarity).
Other parts of the system include a stainless steel mesh 7 micron
filter (model SS-2TF-7) to remove any large aggregates and
a Tescom back-pressure regulator (model 26-1762-24) to maintain system pressure (shown as F and B in Fig. 1).
The CHFS system (see Fig. 1) allows pumping and then mixing
of the Ca (with Mg) salt and a basic phosphate source at a ¼00
1
5
10
15
Fig. 1 Scheme of the three-pump continuous hydrothermal flow
synthesis system used for the preparation of Mg-substituted calcium
phosphates. Key: P ¼ HPLC pump, H ¼ heater, C ¼ cooler, F ¼ filter, B
¼ back-pressure regulator, R ¼ counter-current reactor, T ¼ stainless
steel T-piece mixer.
stainless steel Swagelok T-piece mixer (whereupon initial
precipitation at ambient temperature occurs in flow). This slurry
then meets the flow of superheated water in a stainless steel
counter-current mixer (R in Fig. 1),42–46 whereupon the bioceramic material is crystallised in a continuous fashion. The
products are then cooled, filtered to remove large agglomerates
and pass out of a back-pressure regulator which maintains system
pressure. Pump rates of 20, 10 and 10 mL min1 were used for
superheated water, calcium (plus magnesium) nitrate solution and
phosphate source solutions, respectively. A (Ca + Mg)/P molar
ratio of 1.67 (0.0835 M/0.05 M) was hence maintained in the
starting solutions. All reactions were carried out using a superheated water feed at 400 C and 24 MPa. Additionally, a small 5
cm long, 200 W, 2.5cm diameter band heater set to 400 C (unless
stated otherwise) was used on the counter-current mixer (where
the superheated water and metal salts mix and react) to maintain
a high reaction temperature (this gave a mixture temperature of
ca. 275 C as measured via a thermocouple which was placed 7 cm
upstream of the duraluminum block).
For a standard reaction, magnesium nitrate and calcium
nitrate were accurately weighed and added to 100 mL deionised
water so that the total metal ion concentration was 0.0835 M (see
ESI† Table S1). The pH of a stock solution of 0.05 M diammonium hydrogen phosphate (13.2 g in 2000 mL deionised
water) was adjusted to pH 10 by adding 60 mL of neat aqueous
ammonia solution. All calculations were initially based on the
formula for Mg-substituted HA, i.e., [Ca10xMgx(PO4)6(OH)2]
(see Table 1). However, as a HA-like phase was not always
obtained, it can be assumed that the x:(10 x) ratio is simply the
Mg:Ca ratio used in the reagents (see Table 1).
For the reactions where Mg-substituted HA was formed, the
reaction proceeded via the following equation:
xMg(NO3)2 + (10 x)Ca(NO3)2 + 6(NH4)2HPO4 + H2O /
[Ca10xMgx(PO4)6(OH)2] + 12NH4NO3 + 8HNO3
This journal is ª The Royal Society of Chemistry 2008
J. Mater. Chem., 2008, 18, 1–10 | 3
ART B807920J
20
25
30
35
40
45
50
55
1
5
10
15
In order to investigate the effect of water temperature, the
reaction for sample 10Mg-CaP was also repeated using a superheated water feed at 450 C and band heater at 450 C.
The respective slurries were collected from exit of the backpressure regulator (at the end of the CHFS system) in 50 mL
falcon tubes. Between consecutive collections, the CHFS system
was purged with clean water for 10 minutes. Each slurry was
centrifuged at 4500 rpm for one minute (per 50 mL suspension),
ca. 45 mL of liquid was removed from each sample and 45 mL
de-ionized water was added with shaking. Each sample was then
re-centrifuged at 4500 rpm for 1 minute and 45 mL liquid was
removed for a final time to give concentrated slurry. The slurries
were placed in a freezer at 5 C for 60 minutes and then freezedried for 18 h at 1 104 mbar. The yields for the reactions
carried out in this study are summarized in ESI† Table S2.
1
5
10
15
3. Results and discussion
20
25
30
35
40
45
50
55
The authors previously reported the synthesis of ‘‘instant
hydroxyapatite’’, in which highly crystalline HA nanorods were
synthesized in the CHFS system using a superheated water feed
at 400 C and 24 MPa.38 In the same work, the authors reported
that not only could the crystallinity be controlled by changing the
superheated water temperature, but also the Ca:P ratio could be
affected by changing the pH or the temperature (this ratio affects
resorbability in vivo). As in the aforementioned case, the rationale behind the synthesis of Mg-substituted calcium phosphates
was to see if we could generate bioceramics with unique physical
or compositional attributes (particle size, ion substitution level,
crystallinity and phase composition) that may possess novel
physical or other properties. In this work we used solutions of
calcium nitrate and magnesium nitrate as the metal ion feeds in
the CHFS system. By controlling the Mg:Ca ratios in the starting
solutions, it was possible to affect the Mg-substitution levels in
the resulting bioceramics or mixtures. The products were
obtained as soft white free flowing powders in high yield after
clean up and freeze drying of the individual slurries. All powders
were analysed without further processing or heat treatments.
For selected samples, transmission electron microscopy
(TEM) images were used to investigate the particle morphology
with ever increasing magnesium content or with changes in
phase. Fig. 2a and b reveal distinct nano-rods of size ca. 234
(93) 57 (16) nm, for sample 0.5Mg-CaP. TEM images of
sample 4Mg-CaP (subsequently identified from PXRD data to
be largely Mg-whitlockite) are shown in Fig. 2c and d. A mixture
of both lozenge shaped particles and semi-spherical agglomerates
of ca. 30–60 nm [average particle size is 43 (12) nm; 50 particles
sampled] sized particles were observed, suggesting a deviation
from the rod-like morphology observed at lower Mg-substitution
levels and for phase-pure HA.38 Fig. 2e and f show TEM images
of sample 6Mg-CaP in which agglomerates as large as ca. 600 nm
in diameter are comprised of smaller primary particles (average
particle size is 47 17 nm; 50 particles sampled). The TEM
image in Fig. 2f suggests that the spherical particles are possibly
hollow. TEM images of sample 10Mg-CaP in Fig. 2g and h show
very small particles, ca. 37 nm 13 nm in diameter (range ¼ 20–
60 nm; 50 particles sampled). These particles also appear to be
hollow, as observed for sample 6Mg-CaP.
20
25
Fig. 2 Transmission electron microscope images of sample (a) 0.5MgCaP (15000 magnification; bar ¼ 500 nm), (b) 0.5Mg-CaP (150000;
bar ¼ 50 nm), (c) 4Mg-CaP (15000; bar ¼ 500 nm) and (d) 4Mg-CaP
(100000; bar ¼ 50 nm), (e) 6Mg-CaP (10000; bar ¼ 500 nm), (f) 6MgCaP (10000; bar ¼ 50 nm), (g) 10Mg-CaP (10000; bar ¼ 500 nm), (h)
10Mg-CaP (100000; bar ¼ 50 nm).
30
35
Elemental analyses (by an EDS detector attached to a TEM)
for selected samples (three areas selected per sample) suggested
the magnesium substitution levels in powders were 0.5 wt% 0.2 wt% for the rods in 0.5Mg-CaP, 4.0 wt% 0.6 wt% for the
agglomerates in 4Mg-CaP, 6.8 wt% 0.3 wt% for 6Mg-CaP and
8.3 wt% 0.9 wt% for 10Mg-CaP. In sample 4Mg-CaP the EDS
was focused on three lozenge shaped particles which form
a slightly different shape to the majority of the sample in the
frame (the vast majority are rounded agglomerates) and these
suggested a Mg content of 2.7 wt% 0.8 wt%. This suggests that
the lozenge shaped particles in sample 4Mg-CaP may be slightly
lower in Mg compared to the more rounded agglomerates. The
lozenge shape can essentially be viewed as a rounded rod like
shape (such as the rods observed at lower Mg substitution levels
for Mg-HA). Average area scans measurements of the Mg
content using an EDS detector attached to a SEM are shown in
Table 1.
Powder X-ray diffraction data were collected for all samples to
investigate how Mg substitution affected phase composition and
phase purity. The PXRD data for sample 0.5Mg-CaP, shown in
Fig. 3a, gave a good match to the line pattern for crystalline HA
[ICDD pattern 09-0432, Ca5(PO4)3(OH)]. In contrast, the PXRD
data for sample 4Mg-CaP gave a good match to the reference line
This journal is ª The Royal Society of Chemistry 2008
4 | J. Mater. Chem., 2008, 18, 1–10
ART B807920J
40
45
50
55
1
1
5
5
10
10
15
Fig. 4 BET surface areas of ‘‘as precipitated’’ and freeze dried magnesium substituted calcium phosphates (and biphasic mixtures) with variation in magnesium content (measured using EDS) in the solids. The data
points from left to right correspond to samples CaP, 0.5Mg-CaP, 1MgCaP, 1.5Mg-CaP, 2Mg-CaP, 4Mg-CaP, 6Mg-CaP, 8Mg-CaP, 10MgCaP, 12Mg-CaP and 14Mg-CaP.
20
25
Fig. 3 Powder X-ray diffraction patterns of ‘‘as precipitated’’ Mgsubstituted calcium phosphate powders for samples (a) 0.5Mg-CaP, (b)
1Mg-CaP, (c) 1.5Mg-CaP, (d) 2Mg-CaP, (e) 4Mg-CaP, (f) 6Mg-CaP, (g)
8Mg-CaP, (h) 10Mg-CaP, (i) 12Mg-CaP, (j) 14Mg-CaP and (k) MgP.
Note: phase separation from pure Mg-substituted HA (Mg-HA) into
a mixture of Mg-HA and Mg-whitlockite occurs from PXRD pattern (b)
and for several higher Mg loadings.
30
35
40
45
50
55
pattern for crystalline Mg-whitlockite [ICDD pattern 70-2064,
Ca18Mg2H2(PO4)14].55 Samples 1Mg-CaP, 1.5Mg-CaP and 2MgCaP were identified as biphasic mixtures, with a good match to
the aforementioned patterns for HA and whitlockite, respectively
(Fig. 3b–d). Increasing the magnesium content further (Fig. 3f–j)
resulted in significantly broader PXRD peaks assigned as phase
pure Mg-whitlockite [Ca3yMgy(HPO4)z(PO4)22z/3] (where y ¼
0.7–1.2). This was not surprising given that magnesium is known
to retard the crystallisation and growth of calcium phosphates in
solution and to stabilise whitlockite.24 Of relevance to Mgwhitlockite, the batch hydrothermal synthesis of an analogous
material, Mg3.5H2(PO4)3, was recently reported.56 It crystallised
in a triclinic system with space group P1, Z ¼ 2, with unit-cell
parameters: a ¼ 6.438(1) Å, b ¼ 7.856(1) Å, c ¼ 9.438(1), a ¼
104.57(1) , b ¼ 108.61(1) , g ¼ 101.28(1) , V ¼ 739.99Å3).
Previously, Mg-whitlockite crystals, Ca18Mg2H2(PO4)14, synthesised in a batch hydrothermal reactor, were reported in the
rhombohedral space group R3c with ZR ¼ 1, a ¼ 13.765(8) Å and
a ¼ 44.25(5) with the equivalent hexagonal parameters suggested to be a ¼ 10.350(5), c ¼ 37.085(12) Å and ZH ¼ 3.57
The PXRD data for our sample MgP (powder made using
magnesium nitrate and no calcium nitrate) showed extremely
broad peaks, suggesting the material is possibly amorphous. The
broad peak positions roughly approximated to several standard
line patterns, particularly ICDD patterns 11-0041-magnesium
phosphate–Mg(PO4)3 and 08-0038–magnesium phosphate–
Mg2P2O7. PXRD data of the sample obtained from the control
experiment (re-run of sample 10Mg-CaP using superheated
water at 450 C and band heater at 450 C) suggested the
material was more crystalline (narrower peaks) than the similar
sample made at 400 C (Fig. 3h) and gave a good match to the
pattern for whitlockite (this PXRD pattern is reported in the
ESI†).
Fig. 4 shows the trends in BET surface areas for the magnesium substituted calcium phosphate samples. For sample 0.5MgCaP (phase pure Mg-HA) the BET surface area was 20.7 m2 g1.
Samples 6Mg-CaP and 8Mg-CaP (phase-pure Mg-whitlockite)
had lower surface areas of 10.3 and 11.7 m2 g1, respectively. This
is unsurprising given the extent of agglomeration observed in
Fig. 2c–f. The surface area shows a ca. five fold increase for
samples 10Mg-CaP (50.2 m2 g1), 12Mg-CaP (54.3 m2 g1) and
14Mg-CaP (62.3 m2 g1). This suggests that there is possibly
a threshold above which Mg substitution severely retards crystallisation and growth of particles in solution under these
conditions. This hypothesis is supported by TEM images shown
in Fig. 2c–h.
Raman spectroscopy and FTIR spectroscopy were used to
analyze the samples and aid identification of different calcium
phosphates.12,58,59 Fig. 5 shows Raman spectra for magnesium
substituted calcium phosphates. Peaks in Fig. 5a–d for samples
0.5Mg-CaP to 2Mg-CaP are very similar to those observed for
phase pure HA. Peaks at ca. 1083 and 1054 cm1 in the Raman
spectra, for these samples, correspond to asymmetric stretching
(y3) of the P–O bond in phosphate. The peak around
970 cm1corresponds to the symmetric stretching mode (y1) of
the P–O bond of the phosphate group. This peak is generally
observed around 963 cm1 in HA.60 The shift could be due to
magnesium incorporation into the HA lattice. Indeed, Rietveld
refinement of the XRD data (discussed later) revealed a decrease
in unit cell volume with increase in magnesium content (due to
substitution of the larger calcium ion with the smaller magnesium ion). Therefore, the resultant compressive stresses in the
lattice may lead to this difference in peak position.61
When the vibrational behaviour of the phosphate ions is discussed (with Mg substitution), the influence of two factors may
also be considered: (i) the weakening of the intramolecular P–O
This journal is ª The Royal Society of Chemistry 2008
J. Mater. Chem., 2008, 18, 1–10 | 5
ART B807920J
15
20
25
30
35
40
45
50
55
1
1
5
5
10
10
15
15
20
25
30
35
40
45
50
55
Fig. 5 Raman spectra in the range 1250–350 cm1 for samples (a)
0.5Mg-CaP, (b) 1Mg-CaP, (c) 1.5Mg-CaP, (d) 2Mg-CaP, (e) 4Mg-CaP,
(f) 6Mg-CaP, (g) 8Mg-CaP, (h) 10Mg-CaP, (i) 12Mg-CaP, (j) 14Mg-CaP
and (k) MgP. Note: the peak corresponding to O–P–O bending of the
HPO42 group can be seen in (c) and increases for samples with higher
magnesium loading.
bonds owing to interionic Mgd+–Od interactions, and (ii) the
repulsion potential of the lattice, which is inversely proportional
to the unit-cell volume in the case of an isostructural series. As
a result, a shift in the P–O vibrational modes is seen in the spectra.
Peaks at 615 cm1 and 437 cm1 correspond to the y4 and y2
bending, respectively, of the O–P–O linkage in phosphate in HA.
For the samples in the range y ¼ 0.5–1.6 (formula of
[Ca3yMgy(HPO4)z(PO4)22z/3]), the Raman spectra reveal
a peak corresponding to symmetric stretching of the P–O bond in
the range 970 to 976 cm1 (see Fig. 5e–i). The Raman spectra for
samples 4Mg-CaP, 6Mg-CaP and 8Mg-CaP (Fig. 5e–g) each
reveal a peak at ca. 633 cm1, corresponding to bending modes of
the O–P–O linkage in phosphate (of whitlockite). Increase of
magnesium content to 1.5 wt% (and above) results in the
appearance of new peaks in the Raman spectra at 415 and 555
cm1, which correspond to the respective y2 and y4 bending
modes of the O–P–O linkage in HPO42 ions62 (of whitlockite).
This is supported by the appearance of whitlockite peaks in the
analogous PXRD data in Fig. 3.
The FTIR spectrum for sample 0.5Mg-CaP (Fig. 6a) revealed
peaks similar to those normally seen for HA; these include peaks
at 3570 cm1 (strong) and 636 cm1 due to the stretching mode
(ys) and the librational mode (yL), respectively, of hydroxyl
groups.62,63 A band in the range 1150–990 cm1 corresponds to
asymmetric P–O stretching (y3) of the phosphate group, whilst
a peak at 964 cm1, corresponds to symmetric P–O stretching (n1)
of phosphate. The peaks at 605 and 567 cm1 correspond to the
y4 O–P–O bending mode, whilst the peak at 478 cm1 corresponds to the y2 O–P–O bending mode of phosphate.
The weak peak centred at ca. 872 cm1 may be due to the
bending mode (y2) of a minute amount of carbonate which is
present. With an increase in magnesium content from 0.5 to
Fig. 6 FTIR spectra in ranges (i) 3700–3400 cm1 and (ii) 1800–400
cm1, respectively, for samples (a) 0.5Mg-CaP, (b) 1Mg-CaP, (c) 1.5MgCaP, (d) 2Mg-CaP, (e) 4Mg-CaP, (f) 6Mg-CaP, (g) 8Mg-CaP, (h) 10MgCaP, (i) 12Mg-CaP, (j) 14Mg-CaP and (k) MgP. Note: the phosphate
bands in the range 1150–950 cm1, become broader with increasing Mg
substitution.
2 wt% (x ¼ 0.2 to 0.8), there was a general decrease in the
intensity of hydroxyl peaks at 3570 and 633 cm1 in the corresponding FTIR spectra. This coincides with onset of Mg-whitlockite formation (as seen in the PXRD data) at greater THAN
or equal to 1 wt% magnesium substitution (x ¼ 0.4). The FTIR
spectra for samples with Mg substitution of 6 wt% (x ¼ 2.4) and
higher (as seen in fig. 6f–j) revealed a phosphate peak (range
1550–1450 cm1) that became broader with increasing magnesium substitution. This coincided with a decrease in crystallinity
of Mg-whitlockite (as seen in the PXRD data shown in Fig. 3).
For biphasic samples and phase pure Mg-whitlockite (i.e.
samples 1Mg-CaP to 14Mg-CaP, fig. 6b–j), the peak at 964 cm1
may be due to the presence of HPO42. Fig. 6k shows the FTIR
spectrum for sample MgP. The weak bands centred at 1450 and
870 cm1 were assigned as the C–O stretching vibrations of
a small amount of carbonate that may be present.
b-TCP
[Ca3(PO4)2],
magnesium
stabilised
b-TCP
[(Ca,Mg)3(PO4)2] and Mg-whitlockite [Ca18Mg2H2(PO4)14] are
considered to be difficult to distinguish solely on PXRD data
alone; in the literature, their names are often mistakenly used
interchangeably.64 Our spectroscopic results suggest that the
whitlockite phase reported herein may exist with the generic
formula [Ca3yMgy(HPO4)z(PO4)22z/3] due to the presence of
FTIR peaks corresponding to HPO42.
A relatively large unit-cell group splitting is observed for n1 in
both IR and Raman spectra and for n3 in Raman spectra. The
spectroscopic data suggest that the mean wavenumbers of the P–
O stretches within each series are insensitive to the amount of
substituted magnesium ions. It is envisaged that it is due to two
factors, namely, (i) the weakening of intramolecular P–O bonds
due to the magnesium substitution (caused by the increased
magnesium interactions) and (ii) increasing repulsion potential of
the respective lattices in the same order.
This journal is ª The Royal Society of Chemistry 2008
6 | J. Mater. Chem., 2008, 18, 1–10
ART B807920J
20
25
30
35
40
45
50
55
1
1
5
5
10
10
15
15
20
20
25
30
35
40
45
50
55
Fig. 7 Rietveld plot showing the fit of the high resolution transmission PXRD data of hydroxyapatite for sample 0.5Mg-CaP using a single-phase
model and isotropic peak widths for all reflections. Dots indicate observed data, solid line calculated model, and vertical marks show the 2q positions of
the hkl reflections with the difference pattern shown below. Peak width anisotropy as a function of hkl is particularly marked for 00l reflections as shown
in the insert for the 002 peak at 25.9 2q.
Rietveld refinements using the high-resolution PXRD transmission data were used to obtain accurate unit-cell dimensions
for selected Mg-substituted samples and weight percent
proportions of HA to Mg-whitlockite present. A typical fit is
shown in Fig. 7. Rietveld refinement revealed anisotropic peak
broadening for HA with the width of all 00l reflections being
significantly narrower. This suggests preferential crystalline
growth along the c direction, which is consistent with the rod-like
forms as seen in Fig. 2(b).
The Rietveld fits could be improved by the use of separate
widths for the 00l reflections, with Rwp decreasing from 12.7% to
11.2% (Rexp ¼ 6.1%), though in practice this had a negligible
effect on the unit-cell dimensions derived from the data. Mixed
phase refinements with both HA and Mg-whitlockite present
produced similar quality fits as indicated by the Rwp.
Rietveld refinements of samples CaP, 0.5Mg-CaP, 1.5Mg-CaP
and 4Mg-CaP revealed 100 wt%, 100 wt%, 70 wt%, and 5 wt%
HA, respectively, and 0 wt%, 0 wt%, 30 wt%, and 95 wt% whitlockite, respectively. No other crystalline phases were observed in
any diffraction pattern with the sole exception of the 1.5Mg-CaP
PXRD pattern where a very small unidentified peak at 7.71 2q
was observed which may be due to a small contaminant.
Finally, and most importantly, refinements of the unit cell of the
HA samples show that a systematic volume decrease occurs with
increased levels of substituted magnesium in the samples (see ESI†
Fig.S3). Although there is a significant deviation from the line
calculated for the 4Mg-CaP cell volume, this is understandable as
refinement is based on the 5 wt% of HA present in this Mg-whitlockite sample but the EDS (attached to the SEM) Mg content
value is approximated from a large section of the whole sample.
Nevertheless, the overall decrease in unit cell volume of ca.
5 Å3 is clearly observed from sample CaP to 4Mg-CaP (assuming
this averaged Mg-content is used for the latter), and this reflects
the reduction in average cation size in exchanging Ca2+ with Mg2+
(cationic radii are approximately 1.06 and 0.72 Å, respectively,
for a similar coordination number).65 The small magnitude of the
value stems from the small number of Mg2+ ions being
substituted into the overall structure (which is less than one Mg
per unit cell). A similar effect is obtained from the refinement of
the whitlockite cell parameters, where the unit cell volume for
1.5Mg-CaP is approx. 3480 Å3. Increasing the % magnesium
from sample 1.5Mg-CaP to 4Mg-CaP results in a decrease in unit
cell volume of about 52 Å3. It should be noted that the Rietveld
refinement technique is not sensitive enough to determine Mg
concentration in each phase and so only unit cell parameters for
each phase present were refined with the structural parameters
being constrained to those determined by the single-crystal
models of unsubstituted hydroxyapatite and whitlockite.
A series of Mg-substituted calcium phosphates and biphasic
mixtures were synthesized using a CHFS system. The physical and
crystallographic properties of these materials can be closely
monitored.
Mg-substituted
hydroxyapatite
rods,
[Ca10xMgx(PO4)6(OH)2], were exclusively formed at low Mg
substitution levels (where x ¼ 0.2). However, at higher Mg
substitution levels, stabilization of phase-pure crystalline Mgwhitlockite [Ca3yMgy(HPO4)z(PO4)22z/3] was obtained (for y ¼
0.7). There do not appear to be any reports on direct and rapid
hydrothermal synthesis of crystalline Mg-whitlockite. A further
increase in magnesium substitutio, resulted in decreased crystallinity and generally increased surface area for the Mg-whitlockite
phase (up to y ¼ 1.6). The synthesis method employed herein also
J. Mater. Chem., 2008, 18, 1–10 | 7
ART B807920J
30
35
40
45
4. Conclusions
This journal is ª The Royal Society of Chemistry 2008
25
50
55
1
allows us to produce intimately mixed biphasic mixtures, which
could be extremely useful in applications such as controlled
resorbability bone grafts and related biomedical materials.
5
Acknowledgements
10
15
EPSRC is acknowledged for funding an EPSRC Advanced
Research Fellowship entitled ‘‘Next Generation Biomedical
Materials Using Supercritical Fluids’’ (JAD; grant GR/A11304).
The Higher Education Commission (HEC), Government of
Pakistan is thanked for a scholarship (AAC). Sun Chemicals is
thanked for an industrial case award (JG). P. Boldrin is thanked
for assisting in the preparation of this paper. Mick Willis and
Zofia Luklinska are thanked for their help with TEM. Nicky
Mordan (EDI, UCL) is thanked for her help regarding EDS
(SEM) of all the samples.
References
20
25
30
35
40
45
50
55
1 L. L. Hench, Journal of the American Ceramic Society, 1998, 81,
1705.
2 S. Kannan, I. A. F. Lemos, J. H. G. Rocha and J. M. F. Ferreira,
Journal of Solid State Chemistry, 2005, 178, 3190.
3 K. Yoshida, N. Kondo, H. Kita, M. Mitamura, K. Hashimoto
and Y. Toda, Journal of the American Ceramic Society, 2005, 88,
2315.
4 S. Kannan and J. M. F. Ferreira, Chemistry of Materials, 2006, 18,
198.
5 Y. Pan, J. L. Huang and C. Y. Shao, Journal of Materials Science,
2003, 38, 1049.
6 H. Yuan, C. A. van Blitterswijk, K. de Groot and J. D. de
Bruijn, Journal of Biomedical Materials Research Part A, 2006,
78A, 139.
7 I. Mandubalal, T. P. Sastry and R. V. S. Kumar, Journal of
Biomaterials Applications, 2005, 19, 341.
8 R. Z. LeGeros, Clinical Orthopaedics and Related Research, 2002, 81.
9 T. J. Webster, E. A. Massa-Schlueter, J. L. Smith and
E. B. Slamovich, Biomaterials, 2004, 25, 2111.
10 B. Wopenka and J. D. Pasteris, Materials Science & Engineering CBiomimetic and Supramolecular Systems, 2005, 25, 131.
11 R. A. Terpstra and F. C. M. Driessens, Calcified Tissue International,
1986, 39, 348.
12 I. R. Gibson and W. Bonfield, Journal of Materials Science-Materials
in Medicine, 2002, 13, 685.
13 I. Mayer, R. Schlam and J. D. B. Featherstone, Journal of Inorganic
Biochemistry, 1997, 66, 1.
14 S. R. Kim, J. H. Lee, Y. T. Kim, D. H. Riu, S. J. Jung, Y. J. Lee,
S. C. Chung and Y. H. Kim, Biomaterials, 2003, 24, 1389.
15 R. K. Tang, W. J. Wu, M. Haas and G. H. Nancollas, Langmuir,
2001, 17, 3480.
16 R. Lagier and C. A. Baud, Pathology Research and Practice, 2003,
199, 329.
17 E. Bertoni, A. Bigi, G. Cojazzi, M. Gandolfi, S. Panzavolta and
N. Roveri, Journal of Inorganic Biochemistry, 1998, 72, 29.
18 Z. Zyman, M. Tkachenko, M. Epple, M. Polyakov and M. Naboka,
Materialwissenschaft und Werkstofftechnik, 2006, 37, 474.
19 W. L. Suchanek, K. Byrappa, P. Shuk, R. E. Riman, V. F. Janas
and K. S. TenHuisen, Journal of Solid State Chemistry, 2004,
177, 793.
20 S. Kannan, J. H. G. Rocha and J. M. F. Ferreira, Journal of Materials
Chemistry, 2006, 16, 286.
21 A. Avaglioli, A. Krajewski, G. C. Celotti, A. Piancastelli, B. Bacchini,
L. Montanari and G. Zama, Biomaterials, 1996, 617.
22 A. Bigi, G. Falini, E. Foresti, M. Gazzano, A. Ripamonti and
N. Roveri, Journal of Inorganic Biochemistry, 1993, 49, 69.
23 R. N. Correia, M. C. F. Magalhaes, P. A. A. P. Marques and
A. M. R. Senos, Journal of Materials Science-Materials in Medicine,
1996, 7, 501.
24 N. C. Blumenthal, Clinical Orthopaedics and Related Research, 1989,
279.
25 C. Ergun, T. J. Webster, R. Bizios and R. H. Doremus, Journal of
1
Biomedical Materials Research, 2002, 59, 305.
26 R. Z. Legeros, D. Mijares, A. M. Gatti, and J. P. Legeros, Mgsubstituted tricalcium phosphates and biphasic calcium phosphates:
Formation and properties, Biomaterials 2004 Congress Managers,
3 5
Sydney, NSW 2001, Australia, Sydney, Australia, 2004, p. 1448.
27 Y. Yamasaki, Y. Yoshida, M. Okazaki, A. Shimazu, T. Uchida,
T. Kubo, Y. Akagawa, Y. Hamada, J. Takahashi and
N. Matsuura, Journal of Biomedical Materials Research, 2002, 62, 99.
28 W. L. Suchanek, K. Byrappa, P. Shuk, R. E. Riman, V. F. Janas and
K. S. TenHuisen, Biomaterials, 2004, 25, 4647.
29 C. M. Serre, M. Papillard, P. Chavassieux, J. C. Voegel and
10
G. Boivin, Journal of Biomedical Materials Research, 1998, 42,
626.
30 R. Z. LeGeros, A. M. Gatti, R. Kijkowska, D. Q. Mijares, and
J. P. LeGeros, Mg-substituted Tricalcium Phosphates: Formation and
Properties, Trans Tech Publications Ltd, Porto, Portugal, 2004, pp. 4
127–130.
15
31 I. V. Fadeev, L. I. Shvorneva, S. M. Barinov and V. P. Orlovskii,
Inorganic Materials, 2003, 39, 947.
32 J. M. Bouler, R. Z. LeGeros and G. Daculsi, Journal of Biomedical
Materials Research, 2000, 51, 680.
33 F. Gouin, J. Delecrin, N. Passuti, S. Touchais, P. Poirier and
J. V. Bainvel, Revue de Chirurgie Orthopedique et Reparatrice de l
Appareil Moteur, 1995, 81, 59.
20
34 G. Daculsi, M. B. Darc, P. Corlieu and M. Gersdorff, Annals of
Otology Rhinology and Laryngology, 1992, 101, 669.
35 J. Delecrin, S. Takahashi, F. Gouin and N. Passuti, Spine, 2000, 25,
563.
36 T. A. Fuierer, M. Lore, S. A. Puckett and G. H. Nancollas, Langmuir,
1994, 10, 4721.
25
37 R. Enderle, F. Gotz-Neunhoeffer, M. Gobbels, F. A. Muller and
P. Greil, Biomaterials, 2005, 26, 3379.
38 A. A. Chaudhry, S. Haque, S. Kellici, P. Boldrin, I. Rehman,
A. K. Fazal and J. A. Darr, Chemical Communications, 2006,
2286.
39 C. S. Chai, K. A. Gross and B. Ben-Nissan, Biomaterials, 1998, 19,
2291.
30
40 C. Kothapalli, M. Wei, A. Vasiliev and M. T. Shaw, Acta Materialia,
2004, 52, 5655.
41 C. R. Kothapalli, M. Wei, R. Z. LeGeros and M. T. Shaw, Journal of
Materials Science-Materials in Medicine, 2005, 16, 441.
42 G. Y. Hou, Z. H. Jin, X. H. Zheng and J. M. Qian, Rare Metal
Materials and Engineering, 2007, 36, 1649.
35
43 J. A. Darr and M. Poliakoff, Chemical Reviews, 1999, 99, 495.
44 A. Cabanas, J. A. Darr, E. Lester and M. Poliakoff, Chemical
Communications, 2000, 901.
45 A. Cabanas, J. A. Darr, E. Lester and M. Poliakoff, Journal of
Materials Chemistry, 2001, 11, 561.
46 P. Boldrin, A. K. Hebb, A. A. Chaudhry, L. Otley, B. Thiebaut,
40
P. Bishop and J. A. Darr, Industrial & Engineering Chemistry
Research, 2007, 46, 4830.
47 X. L. Weng, P. Boldrin, I. Abrahams, S. J. Skinner and J. A. Darr,
Chemistry of Materials, 2007, 19, 4382.
48 T. Adschiri, K. Kanazawa and K. Arai, Journal of the American
Ceramic Society, 1992, 75, 1019.
49 T. Adschiri, K. Kanazawa and K. Arai, Journal of the American
45
Ceramic Society, 1992, 75, 2615.
50 K. W. Sue, K. Kimura, M. Yamamoto and K. Arai, Materials
Letters, 2004, 58, 3350.
51 Y. Hakuta, H. Ura, H. Hayashi and K. Arai, Materials Chemistry and
Physics, 2005, 93, 466.
52 J. K. Cockcroft, 2005.
50
53 K. Sudarsanan and R. A. Young, Acta Crystallographica B, 1982, 24,
38.
54 L. W. Schroeder, B. Dickens and W. E. Brown, Journal of Solid State
Chemistry, 1977, 22, 253.
55 E. Guneri and M. Akkurt, G.U.Journal of Science, 2005, 18, 321.
56 H. Assaaoudi, Z. Fang, I. S. Butler, D. H. Ryan and J. A. Kozinski,
Solid State Sciences, 2007, 9, 385.
55
57 C. Calvo and R. Gopal, American Mineralogist, 1975, 60, 120.
58 I. Rehman, M. Karsh, L. L. Hench and W. Bonfield, Journal of
Biomedical Materials Research, 2000, 50, 97.
59 I. R. Gibson, I. Rehman, S. M. Best and W. Bonfield, Journal of
Materials Science-Materials in Medicine, 2000, 11, 533.
This journal is ª The Royal Society of Chemistry 2008
8 | J. Mater. Chem., 2008, 18, 1–10
ART B807920J
1
60 G. Penel, G. Leroy, C. Rey and E. Bres, Calcified Tissue International,
1998, 63, 475.
61 G. Gouadec and P. Colomban, Progress in Crystal Growth and
Characterization of Materials, 2007, 53, 1.
62 S. Koutsopoulos, Journal of Biomedical Materials Research, 2002, 62, 600.
63 I. Rehman and W. Bonfield, Journal of Materials Science-Materials in
Medicine, 1997, 8, 1.
64 M. Mathew and S. Takagi, Journal of Research of the National
Institute of Standards and Technology, 2001, 106, 1035.
65 R. D. Shannon, Acta Crystallographica., 1976, A32, 751.
1
5
5
10
10
15
15
20
20
25
25
30
30
35
35
40
40
45
45
50
50
55
55
This journal is ª The Royal Society of Chemistry 2008
J. Mater. Chem., 2008, 18, 1–10 | 9
ART B807920J
1
1
Authors Queries
Journal: JM
5
5
Paper: b807920j
Title: Synthesis and characterisation of magnesium substituted calcium phosphate bioceramic nanoparticles made via
continuous hydrothermal flow synthesis
10
Editor’s queries are marked like this... 1 , and for your convenience line numbers are inserted like this... 5
10
Query
Reference
Query
15
1
For your information: You can cite this article
before you receive notification of the page numbers
by using the following format: (authors), J. Mater.
Chem., 2008, DOI: 10.1039=b807920j.
20
2
Please supply a contents page entry: a figure (8 cm
wide by 4 cm high), plus text (<30 words)
emphasising the novel aspects of the work and
highlighting the materials applications or
properties. See the contents pages of any recent
issue of J. Mater. Chem. for examples.
15
20
25
Remarks
3
Ref. 26: Please provide the following details: lst of
editor names if any, publisher’s name, publisher’s
location.
4
Ref. 30: Please provide the following details: list of
editor names (if any), publisher’s name, publisher’s
location.
30
25
30
35
35
40
40
45
45
50
50
55
55
This journal is ª The Royal Society of Chemistry 2008
10 | J. Mater. Chem., 2008, 18, 1–10
ART B807920J