Available online at www.sciencedirect.com

Acta Biomaterialia 5 (2009) 1241–1252

www.elsevier.com/locate/actabiomat

Amino acid synergetic effect on structure, morphologyand surface properties of biomimetic apatite nanocrystals

Barbara Palazzo a,b, Dominic Walsh c, Michele Iafisco a,b, Elisabetta Foresti a,Luca Bertinetti d, Gianmario Martra d, Claudia Letizia Bianchi e

Giuseppe Cappelletti e, Norberto Roveri a,*

a Department of Chemistry ‘‘G. Ciamician”, Alma Mater Studiorum, University of Bologna, Via Selmi 2, 40126 Bologna, Italyb Department of Medical Science, University of Eastern Piedmont, Via Solaroli 4, 28100 Novara, Italy

c Centre for Organized Matter Chemistry, School of Chemistry, University of Bristol, Bristol BS8 1TS, UKd Department of Chemistry IFM and NIS Center of Excellence, University of Turin, Via P. Giuria 7, 10125 Torino, Italy

e Department of Physical Chemistry and Electrochemistry, Universita di Milan, Via Golgi 19, 20133 Milano, Italy

Received 17 July 2008; received in revised form 17 October 2008; accepted 30 October 2008Available online 21 November 2008

Abstract

Alanine, arginine and aspartic acid have been used as co-reagents during the synthesis of a biomimetic calcium-deficient hydroxyap-atite (CDHA) for the synergistic coupling of synthesis and functionalization. The surface and bulk characterizations are consistent with abinding model in which the amino acid is preferentially bound to the CDHA nanocrystal surface by its lateral chain group. The f-poten-tial measurements show that the amino acid-functionalized CDHA surface charge shifts towards neutral compared to CDHA synthe-sized in the absence of amino acids. Amino acids bound to the crystal induce crystal growth inhibition predominantly at the Ca-richsurfaces during the initial stages of crystallization. Moreover, high-resolution transmission electron microscopy measurements suggesta model for needle-shaped CDHA nanocrystals formation in the presence of either arginine or aspartic acid based on the oriented aggre-gation of primary crystallite domains specifically along the c-axis direction and the self-assembly of preformed nanoparticles. The resultshave significant importance for the control of the shape, morphology and aggregation of the CDHA nanocrystals, while the observedsurface modifications are of marked importance for the nature, stability and reactivity of the functionalized surfaces produced.� 2009 Published by Elsevier Ltd. on behalf of Acta Materialia Inc.

Keywords: Hydroxyapatite; Amino acids; Surface functionality; Biomimetism; Nanocrystals

1. Introduction

To understand how crystals with classical shapes can beso remarkably different when formed by organisms, severalstudies on the effects of additives on crystal morphologyhave been undertaken. By comparing shape changes tothe molecular structures of the modified crystal surfaces,particularly those of newly expressed faces, the investiga-tions can postulate the nature of the dominant interactions

1742-7061/$ - see front matter � 2009 Published by Elsevier Ltd. on behalf of

doi:10.1016/j.actbio.2008.10.024

* Corresponding author. Tel.: +39 0512099486; fax: +39 0512099593.E-mail address: norberto.roveri@unibo.it (N. Roveri).

and stereochemical relationships responsible for the alteredcrystal shapes [1].

Hydroxyapatite (HA) [Ca5(PO4)3OH], the thermody-namically most stable calcium phosphate phase under phys-iological conditions, represents the model compound of theinorganic constituent of bone, teeth and many pathologicalcalcifications [2–5]. HA has been the object of extensiveresearch in numerous interdisciplinary areas to determineboth a better understanding of the formation mechanismsin natural mineralization processes and its applicability asa biomedical or industrial material [6,7]. There exist numer-ous synthetic strategies for producing HA and substitutedHA, including wet methods, hydrothermal and ultrasonic

Acta Materialia Inc.

1242 B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252

nebulization methods, sol–gel and solid-state synthesis [8,9].Size, morphology and surface properties represent the phys-ico-chemical features that can be tailored in synthetic HAcrystals to optimize their suitability for specific biomedicalapplications [10–12]. The effect of biological molecules onHA crystal growth, and consequently the above properties,have been widely studied and directly related to physiologi-cal or pathological calcification processes [13,14].

Amino acids are biological molecules of particular inter-est, due to their great importance for living organisms. Theeffects of hydroxyproline, tyrosine, serine, glycine, cysteine,cystine, glutamine and lysine on HA crystallization havebeen studied by the ‘‘Chemo-stat method” [15–17]. Usingthis technique, the chemical potentials of the species inthe working solution are kept constant during the crystal-lization process by the stoichiometric addition of reactants,and therefore the crystal growth reaction is performedunder pseudo-steady-state conditions. Diverse inhibitingactivities were observed as a function of the different aminoacid side groups.

All these studies have utilized biomolecules as simplegrowth inhibitors of HA crystallization, rather than con-sidering their use as a strategy to fine-tune the bioactivityof the apatite crystals. In fact, considering that the presenceof proteins in biological materials is intrinsic to the bioac-tivity of HA, amino acids can be considered as agents thatcan increase the bioactivity of synthetic HA. This idea wasfirst developed by Gonzales et al. [18], who obtained, byhydrothermal crystallization, stable aqueous colloids ofpositively charged amino acid-functionalized HA nanorodsless than 80 nm in length and ca. 5 nm in width. Greatlyelongated HA nanorods, of ca. 150 � 3.5 nm, were pre-pared with aspartic acid. The above work inspired furtherinvestigations into the nature and stability of the interac-tion of a series of amino acids with HA [19–22].

Together with the ‘‘Chemo-stat method”, these studiesalso suggested that crystallization kinetics is affectedthrough the inhibition by amino acids of the active growthsites on the surface of HA crystals. The active growth siteon the apatite surface could be the calcium, phosphate orhydroxyl group, and the interference of one or more ofthese occur independently of the amino acid lateral groupused. These experiments have been conducted either undersupersaturation conditions [15–17] or at high temperature[19–22].

The complex behaviour observed by these studies sug-gests that there may be a complex multi-site adsorptionprocess, perhaps associated with different faces of the HAcrystal. Changing pH may effectively ‘‘switch” certain sitesfor adsorption on or off, and it is further likely that the nat-ure of the interaction between the amino acid and the CHA(e.g. a surface complex between COO� and Ca2+) surfacemay be different at different sites.

The aim of this work was to further investigate both therole of surface Ca2+ on amino acid linking and the role ofamino acid charge in determining the resulting amino acidfunctionalized HA structure, morphology and surface

properties. For this purpose, calcium-deficient hydroxyap-atite (CDHA) nanocrystals have been synthesized undermild conditions, and with differently charged aminoacids—alanine, arginine and aspartic acid—as co-reagents,with comparison made to bone-like CDHA.

2. Materials and methods

2.1. Materials

All the common high-purity chemical reagents were sup-plied by Sigma-Aldrich. L-Arginine (98% pure), DL-alanine(99% pure) and DL-aspartic acid (98% pure) were also pur-chased from Sigma–Aldrich. Ultrapure water (0.22 lS,25 �C) was used.

2.2. Synthesis

CDHA nanocrystals were synthesized by a modificationof the method of Liou et al. [23]. The nanocrystals synthe-sized in the absence of amino acids (control) were precipi-tated from an aqueous solution of (CH3COO)2Ca (75 mM)by the slow addition (1 drop s�1) of an aqueous solution ofH3PO4 (50 mM), keeping the pH at a constant value of 10by the addition of (NH4)OH solution. The amino acid-functionalized CDHA nanocrystals were synthesized bydissolving an appropriate amount of the amino acid inthe calcium solution. The amino acid concentration was150 mM in order to have a fixed Ca2+/PO4

3�/amino acidmolar ratio of 3/2/6. The resulting suspension was stirredat room temperature for 24 h, then repeatedly washed withwater by centrifugation (PK121 multispeed centrifuge fromThermo Electron Corporation) at 5000 rpm (6527g) for10 min and freeze dried at �60 �C under vacuum (3 mbar)for 12 h.

2.3. Electron microscopy investigation

Transmission electron microscopy (TEM) investigationswere carried out by using a JEOL 1200 EX microscope fit-ted with link elemental dispersive X-ray analysis (EDXA)detectors and a JEOL 3010 UHR operating at 300 kV.General view images were typically obtained at �80,000,while high-resolution images of single particles were typi-cally taken at �400,000. The powdered samples were ultra-sonically dispersed in ultrapure water and a few droplets ofthe slurry were then deposited on perforated carbon foilssupported on conventional copper microgrids. Mean parti-cle dimensions were determined from TEM images of 50randomly selected individual nanoparticles.

Scanning electron microscopy (SEM) observations werecarried out with a JEOL 840 A microscope.

The specimens were mounted on aluminium stubs usingcarbon tape and they were covered with a coating of Au–Pd, approximately 10 nm thick, using a coating unit(Polaron Sputter Coater E5100, Polaron Equipment, Wat-ford, Hertfordshire, UK).

B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252 1243

2.4. X-ray diffraction analysis

X-ray diffraction (XRD) powder patterns were col-lected using an Analytical X’Pert Pro equipped withX’Celerator detector powder diffractometer using CuKa radiation generated at 40 kV and 40 mA. The instru-ment was configured with 0.25� divergence and receivingslits. A quartz sample holder was used. The 2h range wasfrom 5 to 60�, with a step size of 0.05� and a countingtime of 3 s.

The degree of CDHA crystallinity was calculatedaccording to the formula [24]:

Crystallinity ¼ 100� C=ðAþ CÞ

where C is the area of the peaks in the diffraction pattern(‘‘the crystalline area”) and A is the area between the peaksand the background (‘‘the amorphous area”).

Crystal domain sizes along the CDHA axis directionswere calculated applying the Scherrer equation [25]:

LðhklÞ ¼0:94k

cos hffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiDr2 � D2

0

q� �� �

where h is the diffraction angle for plane (hkl), Dr and D0

are the widths (in radians) of reflection (hkl) at half heightfor the synthesized and the reference HA materials, respec-tively, and k = 1.5405 A.

2.5. Specific surface area determination

Measurements were undertaken using a Carlo ErbaSorpty 1750 instrument by measuring N2 absorption at77 K and adopting the well-known BET procedure [26].

2.6. X-ray photoelectron spectroscopy analysis

X-ray photoelectron spectroscopy (XPS) measurementswere performed in an M-Probe Instrument (Surface Sci-ence Instruments) equipped with a monochromatic Al Ka

source (1486.6 eV) with a spot size of 200 � 750 lm andpass energy of 25 eV, providing a resolution of 0.74 eV.With a monochromatic source, an electron flood gun wasused to compensate the build-up of positive charge onthe insulator samples during the analyses: 10 eV wasselected to perform measurements on these samples. Theequipment has a fixed degree of surface sensitivity due tothe photoelectron collections at a fixed take-off angle (angleof X-ray incidence a = 71� (relative to sample normal),electron emission angle b = 0� (relative to sample normal),angle between the X-ray axis and the electron analyzer axisf = 71� (fixed, non-variable)). Surface analysis with a sam-pling volume that extends from the surface to a depth ofmaximum 40–50 A can be performed.

The accuracy of the reported binding energies (BEs) wasestimated to be ±0.2 eV. The quantitative data were alsoaccurately checked and reproduced several times (at least

10 times for each sample) and the percentage error was esti-mated to be ±1%.

2.7. Fourier transform and attenuated total reflectance

infrared spectroscopy

Infrared spectra were recorded on a Thermo Nicolet 380Fourier transform infrared (FT-IR) spectrometer equippedwith a commercial attenuated total reflectance (ATR)accessory.

The FT-IR spectra were recorded from 4000 to 400cm�1 at 2 cm�1 resolution using a Bruker IFS 66v/S spec-

trometer using KBr pellets.The ATR-IR spectra were recorded with the cell empty

to be used as a blank for subsequent experiments. Sampleswere made by placing the powder sample onto the Ge ATRcrystal. Spectra were collected by averaging 256 scans at4 cm�1 resolution.

2.8. Thermogravimetric and differential scanning

calorimetric analysis

Thermogravimetric analysis (TGA) investigations werecarried out on dried samples using a Thermal AnalysisSDT Q 600 (TA Instruments, New Castle, DE, USA).Heating was performed in a nitrogen flow (100 ml min�1)using an alumina sample holder at a rate of 5 �C min�1

up to 1200 �C. The weight of the samples was approxi-mately 10 mg.

2.9. Zeta potential analysis

Electrophoretic determinations were performed by aCoulter DELSA apparatus. A COULTER DELSA 440instrument measured the electrophoretic velocities of sus-pended particles by measuring the Doppler shift of scat-tered laser light simultaneously at four different scatteringangles: 7.5, 15.0, 22.5 and 30.0�. The suspensions of modi-fied HA were prepared as follows: 0.05 g l�1 of functional-ized HA in 10�2 M KNO3 (constant ionic strength), atspontaneous constant pH.

2.10. Chemical analysis

Calcium and phosphorous content were determined byinductively coupled plasma-optical emission spectrometry(ICP-OES; Liberty 200, Varian, Clayton South, Australia):HA and functionalized HA nanocrystals were dissolved in1 wt.% ultrapure nitric acid. The following analytical wave-lengths were chosen: Ca = 422 and P = 213 nm.

2.11. Statistics analysis

Syntheses were performed in triplicate and results arepresented as mean values ± standard deviation (SD). Dif-ferences were considered statistically significant at a signif-icance level of 90%.

1244 B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252

3. Results and discussions

3.1. Structural and morphological characterization

CDHA nanocrystals were synthesized in the presence ofdifferently charged amino acids (alanine, arginine andaspartate) (Scheme 1), according to a sol–gel synthesismaintaining a Ca2+/PO4

3�/amino acid molar ratio of3/2/6. It was noted that alanine and aspartic acid(pI = 6.01 and 2.7, respectively) had a negative charge atthe pH (10) synthesis, while arginine (pI = 11.15) was pos-itively charged. In addition bone-like CDHA nanocrystalswithout additive were produced as a control. XRD pat-terns of all the crystallized inorganic phases (Fig. 1a–d)showed characteristic diffraction maxima of an apatite sin-gle phase (JCPDS 9-432).

They are consistent with a poorly crystalline HA resem-bling the deprotienated bone apatite phase very closely(Fig. 1e). The CDHA synthesized in the presence of aminoacids (Fig. 1b–d) showed a very slight decrease in thedegree of crystallinity compared to the CDHA control.This effect was constant and reproducible, and was quanti-fied according to a previously reported formula [25]. Thedegree of crystallinity of the control CDHA nanocrystalswas 65 ± 5%, which decreased slightly to about 60 ± 5%for CDHA nanocrystals synthesized in the presence ofthe amino acids, independently of the amino acid charge.This is probably due to the presence of the amorphousamino acid phase.

Furthermore, the CDHA synthesized in the presence ofthe amino acids (Fig. 1b–d) show more broadened XRDpeaks compared with the control CDHA (Fig. 1a), whichindicates a relative decrease in the primary domain size.The determination of the primary domain size along thec-axis and the diagonal between the a and b directions, cal-culated by Scherrer’s formula using measurements of peakwidths at half height for the (002) and the (310) reflectionsof each XRD pattern, are shown in Table 1.

The introduction of the amino acid during the CDHAnanocrystals synthesis induced an appreciable modificationin primary crystallite domains along the c direction or

Scheme 1. Scheme of amino acids.

along its transversal direction with respect to CDHA nano-crystals synthesized in the absence of amino acid.

TEM images of CDHA synthesized in the absence ofamino acids are reported in Fig. 2. In panel A, a generalview at low magnifications is shown, while panels B andC show details at high resolution. It can be observed thatparticles exhibit a plate-like morphology (length and widthof 25 ± 5 and 15 ± 3 nm, respectively), with the maindimension elongated towards the c-axis, as derived by theorientation of fringes, 0.34 nm apart, related to (002)planes (Fig. 2, panel B). Edge on plates orientated alongthe c-axis were observed to have a thickness of 8 ± 2 nm(Fig. 2, panel C).

Interestingly, the diffraction fringes along all the crystal-lographic directions extended up to the surfaces of the par-ticles, whilst the borders appeared quite irregular,indicating that no specific crystalline planes are actuallyexposed on the surface of the particles.

The introduction of alanine, an apolar amino acid, duringthe CDHA synthesis does not induce a significant modifica-tion in nanocrystal morphology or in their dimensions(Fig. 3a, panel A). However, as observed in images at highmagnification, lattice fringes running along the particlesvanished near the surface, replaced by a region having athickness in the range 0.5–1.0 nm, exhibiting a contrasttypical of an amorphous phase (Fig. 3a, panel B). Such afeature indicates that the presence of the amino acid duringthe formation of particles affected their surface structuralorder, even when it did not induce a significant modificationin the morphology of the nanocrystals, or in their dimension.

In the case of arginine-functionalized CDHA, particlesexhibited a needle-like morphology (Fig. 3b, panel A), withlength up to 70 nm, width of 6 ± 2 nm (Fig. 3b, panel B)and thickness, measured for orientated particles, of4 ± 1 nm (Fig. 3b, panel C). In addition, most particlesappeared stacked one upon another along their length,mainly in a head-to-tail arrangement, forming 100- to200-nm-long secondary particles (Fig. 3b, panel D). Also,these particles exhibited a surface amorphous layer as inthe previous case (data not shown).

For CDHA particles grown in the presence of asparticacid, morphological features similar to those ones exhibitedby arginine–CDHA were observed (Fig. 3c, panels A andB). Nevertheless, high-magnification images revealed thata number of them were coated in an amorphous layer upto ca. 3.0 nm thick (Fig. 3c, panel C). The analysis of thecontrast of such a layer by means of the calculation ofthe autocorrelation function [27] allowed the presence ofa crystal mineral to be excluded. It can thus be proposedthat this amorphous phase could have resulted from theaspartate on the surface of CDHA.

A comparison between particle dimensions and domainsize (determined by TEM and XRD, respectively) isreported in Table 1.

The obtained data suggest that needle-shaped CDHAnanocrystals form in the presence of either arginine oraspartic acid, based on the oriented aggregation of the

Fig. 1. XRD patterns of CDHA nanocrystals synthesized in the absence of amino acid (a) and in the presence of alanine (b), arginine (c) and aspartate (d).The XRD pattern of a biological apatite obtained from deproteinated bone (e) is shown for comparison.

B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252 1245

primary crystallite domains specifically along the c-axisrather than conventional solution-based mechanisms.

The size of the crystal domains is governed by interac-tions with the amino acid additives that restrict the growthof the primary crystallites by surface capping. This isresponsible for the domain size decrease in all the aminoacid-functionalized CDHA nanocrystals compared to theunfunctionalized CDHA. While the apolar alanine induces

the inhibition of crystal domain growth without affectingthe particles’ dimensions, both the charged amino acidsdecrease the crystals domain dimensions parallel to a andb (XRD data) and also make the crystals’ width close toa single domain dimension (high-resolution transmissionelectron microscopy (HRTEM) analyses). Probable‘‘amino acid residue”-induced electrostatic attractionsbetween crystal domains in turn accelerate the rate of

Table 1Comparison between particle dimensions and domain size (determined byTEM and XRD, respectively) for control CDHA nanocrystals and forCDHA nanocrystals synthesized in the presence of different amino acids.

D002 (nm) D310 (nm) Length (nm) Width (nm)

CDHA 23 ± 5 9 ± 3 25 ± 5 15 ± 3Alanine–CDHA 18 ± 2 6 ± 2 25 ± 5 15 ± 3Arginine–CDHA 19 ± 2 6 ± 2 70 ± 5 6 ± 2Aspartate–CDHA 15 ± 3 5 ± 2 70 ± 5 6 ± 2

1246 B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252

aggregation and promote a one-dimensional aggregation toproduce needle-shaped nanocrystals that are highly elon-gated along the c-axis. The concurrent crystal domain sizegrowth inhibition (XRD data) and particle size enhance-ment along the c direction (HRTEM analyses) can beexplained by CDHA primary particle self-assembly.

SEM images of CDHA synthesized in the absence ofamino acids and in the presence of alanine, arginineand aspartate are reported in Fig. 4a–d, respectively,which show the different morphologies of the assemblednanocrystal clusters. The clusters resulting from theassembly of plate-shaped nanocrystals (non-functional-ized CDHA and alanine-functionalized CDHA) showsmooth surfaces. Conversely, the surfaces of clustersassembled from needle-like arginine- and aspartate-func-tionalized CDHA nanocrystals are characteristicallyrough.

3.2. Spectroscopic and thermal characterization

The ATR-IR spectrum of the CDHA synthesized in theabsence of amino acids reveals the presence of adsorptionbands at 1466 and 1422 cm�1 and at 1545 cm�1, whichare consistent with carbonate type-A (hydroxyl site) andtype-B (phosphate site) substituted HA (Fig. 5) [28]. Thisfinding was not observed in the case of almost stoichiome-

Fig. 2. TEM images. (A) General view of CDHA nanoparticles synthesized inhigh-resolution images of particles with the c-axis parallel or perpendicular to

tric HA nanocrystals synthesized in the same conditions,which show a prevalent type-B substitution [29].

The ATR-IR spectrum of alanine-functionalized CDHAnanocrystals resembles that of non-functionalized CDHA,except for the presence of –CH stretching bands (at 2850and 2917 cm�1; data not shown).

The arginine-functionalized CDHA nanocrystals ATR-IR spectrum (Fig. 6b) shows bands corresponding toC@O stretches (1660 cm�1), COO� asymmetric stretches(1570, 1555, 1450 and 1410 cm�1). In the ATR-IR spec-trum of arginine-functionalized CDHA there is a markedshift of the band corresponding to the CN stretch from1330 cm�1 in the pure amino acid sample to 1314 cm�1.Moreover, a shift in the band corresponding to the C@Ostretch from 1670 cm�1 in the pure amino acid sample to1660 cm�1 can be seen.

The ATR-IR spectrum of aspartic acid-functionalizedCDHA nanocrystals (Fig. 7b) shows bands correspondingto a COO� asymmetric stretch (1590 cm�1) and NH2-H+

symmetric stretches (1485 and 1420 cm�1). The COO-

stretch shifts from 1600 cm�1 in the pure aspartic acid to1590 cm�1 in the functionalized CDHA nanocrystals. Theshift of the characteristic amino acids band compositestowards a lower wave number is consistent with theincrease in the CN and COO� bond lengths. This findingis due to the effective binding of amino acid to CDHAnanocrystals by its lateral residue. The coexistence andconsequent convolution of different bands, both due tothe amino acid carboxylate and amino groups, in the1550–1420 cm�1 region prevent determination of the extentand type of apatite carbonation, the characteristic bands ofwhich appear in the same spectral region.

Thermal analyses carried out on the amino acid-func-tionalized CDHA nanocrystals confirm a polar amino acidpresence even after extensive washing of the functionalizednanocrystals. In fact, the weight loss in the temperaturerange 200–600 �C can be attributed to biomolecular degra-

the absence of amino acids (original magnification: �100,000); (B and C)the image plane, respectively.

Fig. 3. TEM images of CDHA nanoparticles synthesized in the presence of amino acids. (a) Alanine. Panel A: general view; panel B: detail of the surfacetermination of a particle at high resolution. (b) Arginine. Panels A and B: general view; panel C: view of particles stacked on each other; panel D: high-resolution image of a particle oriented with the c-axis perpendicular to the image plane. (c) Aspartate. Panels A and B: general view; panel C: detail of thesurface termination of a particle at high resolution.

B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252 1247

dation. Weight loss increasing in the nanocomposites withrespect to the non-functionalized CDHA can be clearlyseen in Fig. 8. Moreover, the TGA weight loss (Dm) inthe temperature range 200–600 �C is consistent with an

amino acid content order in the functionalized nanocrys-tals corresponding to: aspartic acid–CDHA � arginine–CDHA > alanine–CDHA (Table 2). Furthermore, a con-sistent residue percentage decrease in the functionalized

Fig. 4. SEM images of nanocrystals clusters of CDHA synthesized in the absence of amino acids (a) and in the presence of alanine (b), arginine (c) andaspartate (d).

Fig. 5. ATR-IR spectrum of the CDHA synthesized in the absence ofamino acids revealing the presence of adsorption bands at 1466 and1422 cm�1 and at 1545 cm�1, which are consistent with a carbonate type-A- and type-B-substituted HA.

1248 B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252

CDHA compared to the CDHA synthesized in the absenceof amino acids can be observed. This decrease has beenobserved to follow the order: aspartic acid–CDHA � argi-nine–CDHA > alanine–CDHA � CDHA, which suggeststhat the polar amino acids are not only strictly linked tothe resulting apatite crystal surface, but also surround theprimary particles, which are self-assembling to form theelongated nanocrystal.

In Table 3, XPS and ICP-OES analysis results of controlCDHA and amino acid-functionalized CDHA nanocrys-tals are reported. The Ca/P molar ratio of 1.53 determined

in bulk by ICP analyses for CDHA nanocrystals corre-sponds to the stoichiometric value expected for a cal-cium-deficient apatite and confirms that the formedapatite is substituted at both the phosphate site and thehydroxyl site, which does not affect the Ca/P value. Fur-thermore, it can be seen that the bulk Ca/P molar ratioshows only a slight change upon the addition of aminoacids. The bulk Ca/P molar ratio reduces to a value of1.2 when determined on the crystal’s surface by XPS anal-ysis. This finding indicates an enhancement of surface cal-cium deficiency compared to the bulk calcium deficiency.The surface Ca/P molar ratios of amino acid-functional-ized CDHA nanocrystals are slightly higher than 1.2. How-ever, the amino acid binding onto the functionalizedCDHA nanocrystal’s surface is shown by the modificationof the surface relative Ca/N ratio (Table 3).

XPS analyses of spectral features of the C 1s region werecarried out for amino acid-functionalized CDHA nano-crystals in order to determine the BE and, therefore, com-position of the surface C components.

The C region of the spectrum showed (Fig. 9a–c) a def-inite C 1s shape fitted by four components at different BEs,of about 284.6, 286.4, 288.1 and 289.3 eV, attributed tobound C–C and C–H, C–N, C@O and apatite carbonategroups, respectively. The lowering of the intensity of theC 1s peak at 289.3 eV in the aspartate-functionalizedCDHA could correspond to a decrease in CO3 on thenanocrystal’s surface due to its substitution with the aminoacid’s carboxylate functions.

Zeta potential measurements (Table 4) show that theCDHA has an enhanced negative charge compared to

Fig. 6. ATR-IR spectra of the CDHA synthesized in the absence of amino acids (a), arginine-functionalized CDHA (b) and arginine-functionalizedCDHA (c).

B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252 1249

that reported in the literature (�10 mV) [20], which is inagreement with its surface calcium deficiency. Aminoacid-functionalized CDHA nanocrystals have a decreasednegative charge compared to CDHA synthesized in theabsence of amino acids. At the synthesis pH value (10),alanine and aspartic acid (pI = 6.01 and 2.7, respectively)had a negative charge, whereas arginine (pI = 11.15) waspositively charged. This fact, together with the f potentialdata, is consistent with a binding model in which alanineweakly bonds to the apatite surface calcium sites throughits carboxylate group, aspartate links to the same sitethrough a cooperative linkage involving both the lateralchain group and the amino acidic carboxylate. On theother hand, it is reasonable that arginine is preferentiallybound to the surface phosphate group through its aminolateral group, even if also the amino acidic carboxylateparticipates in the surface link, bounding itself to the cal-cium site.

In this way, alanine, aspartate and arginine expose theiramino groups at the nanocrystal/solvent interface. At the fpotential working pH, the alanine, arginine and aspartateamino groups are positively charged.

4. Conclusions

The use of differently charged amino acids with the aimof both functionalizing the CDHA surface and altering thecrystal’s morphology and assembly has been investigated.Hydroxyapatite nanocrystals with amino acid surface func-tionalities and different morphologies, dependent on theamino acid used as the co-reagent during the synthesis,have been obtained. It was found that, while the introduc-tion of the amino acids during the CDHA nanocrystal syn-thesis induces a modification in the primary crystallitedomains along the c direction and along its transversedirections, only polar amino acids induce a morphological

Fig. 7. ATR-IR spectra of the CDHA synthesized in the absence of amino acids (a), aspartate-functionalized CDHA (b) and aspartate-functionalizedCDHA (c).

Fig. 8. TGA curves obtained for CDHA nanocrystals synthesized in theabsence of amino acids (a) and in the presence of alanine (b), arginine (c)and aspartate (d).

Table 2Thermal stability of CDHA and CDHA nanocrystals synthesized in thepresence of different amino acids.

CDHA CDHA–alanine

CDHA–arginine

CDHA–aspartate

Dm (%)200–600 �C 4 5 11 11Residue (%) at

1200 �C85 82 67 67

Table 3Surface and bulk Ca/P molar ratios and surface Ca/N molar ratio ofcontrol CDHA and amino acid-functionalized CDHA nanocrystals.

Ca/P (XPS) Ca/N (XPS) Ca/P (ICP-OES)

Control CDHA 1.2 — 1.53Alanine–CDHA 1.3 25.9 1.62Arginine–CDHA 1.4 1.7 1.50Aspartic acid–CDHA 1.3 4.4 1.53

1250 B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252

Fig. 9. C 1s spectral region obtained by XPS analyses for (a) alanine-functionalized CDHA, (b) aspartic acid-functionalized CDHA and (c) arginine-functionalized CDHA. Characteristic peaks: (A) C–C and C–H bonds; (B) C–N bond; (C) C@O bond; and (D) apatite carbonate group.

Table 4f potential measurements for CDHA synthesized in the absence of aminoacids and amino acid-functionalized CDHA nanocrystals.

pH of suspension f potential (mV)

CDHA 7.5 �21.2 ± 1.4Alanine–CDHA 7.8 �5.8 ± 1.5Aspartic acid–CDHA 7.6 �6.9 ± 1.2Arginine–CDHA 7.6 �4.1 ± 1.0

B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252 1251

and dimensional variation in CDHA nanocrystals. Consid-ering that the crystal domains sizes appear slightlydecreased by the presence of amino acid while the nano-crystal grows unidirectionally, a self-assembly mechanismcan be supposed. The f potential measurements show thatthe amino acid-functionalized CDHA surface charge isinverted compared to the unfunctionalized CDHA shifting

towards neutrality. The role of the lateral residue in thebinding mechanisms, involving either the carboxylategroup (in the case of alanine and aspartic acid) or theamino lateral group in the case of arginine, has been under-lined. This chemical surface modification should dramati-cally affect the HA biological properties, and offers thepotential for the nanocrystals to be used for optimizingthe preparation of apatite as a biomaterial. Moreover,the possibility of tailoring the morphology of crystals anddriving their assembly can also help to understand howthe classical shapes of crystals can be remarkably modifiedby biomolecules when formed by organisms.

Acknowledgements

We acknowledge financial support from the Universityof Bologna, the Italian Ministero dell’Istruzione, Universi-

1252 B. Palazzo et al. / Acta Biomaterialia 5 (2009) 1241–1252

ta’ e Ricerca (MIUR), PRIN project number 2006032335.L.B. and G.M. are grateful to Regione Piemonte (ProgettoNANOMAT, Docup 2000–2006, Linea 2.4a).

References

[1] Banfield JF, Welch SA, Zhang H, Ebert TT, Penn RL. Aggregation-based crystal growth and microstructure development in natural ironoxyhydroxide biomineralization products. Science 2000;289:751–4.

[2] Glimcher MJ, Bonar LC, Grynpas MD, Landis WJ, Roufosse AH.Recent studies of bone mineral: is the amorphous calcium phosphatetheory valid? J Cryst Growth 1981;53:100–19.

[3] Worms D, Weiner S. Mollusk shell organic matrix: Fourier transforminfrared study of the acidic macromolecules. J Exp Zool1985;237:11–20.

[4] Mann S. Oxford chemistry masters, 5. Biomineralization: principlesand concepts in bioinorganic materials chemistry. New York: OxfordUniversity Press; 2001.

[5] Lowenstam HA, Weiner S. On biomineralization. New York: OxfordUniversity Press; 1989.

[6] Roveri N, Palazzo B. Hydroxyapatite nanocrystals as bone tissuesubstitute. In: Kumar CSSR, editor. Nanotechnologies for the LifeSciences. Tissue, Cell and Organ Engineering, vol. 9. Wein-heim: Wiley-VCH; 2007. p. 283–306.

[7] Tanaka H, Miyajima K, Nakagaki M, Shimabayashi S. Interactionsof aspartic acid, alanine and lysine with hydroxyapatite. Chem PharmBull 1989;37:2897–901.

[8] Walsh D, Arcelli L, Swinerd V, Fletcher J, Mann S, Palazzo B.Aerosol-mediated fabrication of porous thin films using ultrasonicnebulization. Chem Mater 2007;19:503–8.

[9] Villacampa AI, Garcıa-Ruiz JM. Synthesis of a new hydroxyapatite–silica composite material. Cryst Growth 2000;211:111–5.

[10] Tampieri A, Celotti G, Landi E. From biomimetic apatites tobiologically inspired composites. Anal Bioanal Chem2005;381:568–76.

[11] Palazzo B, Sidoti MC, Roveri N, Tampieri A, Sandri M, Bertolazzi L,et al. Controlled drug delivery from porous hydroxyapatite grafts: anexperimental and theoretical approach. Mater Sci Eng C2005;25:207–13.

[12] Palazzo B, Iafisco M, Laforgia M, Margiotta N, Natile G, BianchiCL, et al. Biomimetic hydroxyapatite–drug nanocrystals as potentialbone substitutes with antitumor drug delivery properties. Adv FunctMater 2007;17:2180–8.

[13] Mann S, Webb J, Williams RJP. Biomineralization: chemical andbiochemical perspective. Weinheim: VCH; 1989.

[14] Boskey AL. Phospholipids and calcification: an ovevwiew. In: Ali SY,editor. Cell mediated calcification and matrix vesicles. Amster-dam: Elsevier; 1986. p. 175–9.

[15] Koutsopoulos S, Dalas E. Hydroxyapatite crystallization in thepresence of serine, tyrosine and hydroxyproline amino acids withpolar side groups. J Cryst Growth 2000;216:443–9.

[16] Koutsopoulos S, Dalas E. Hydroxyapatite crystallization in thepresence of amino acids with uncharged polar side groups: glycine,cysteine, cystine and glutamine. Langmuir 2001;17:1074–9.

[17] Koutsopoulos S, Dalas E. The crystallization of hydroxyapatite in thepresence of lysine. J Colloid Interface Sci 2000;231:207–12.

[18] Gonzalez-McQuire R, Chane-Ching JY, Vignaud E, Lebugle A,Mann S. Synthesis and characterization of amino acid-functionalizedhydroxyapatite nanorods. J Mater Chem 2004;14:2277–81.

[19] Boanini E, Torricelli P, Gazzano M, Giardino R, Bigi A. Nanocom-posites of hydroxyapatite with aspartic acid and glutamic acid and theirinteraction with osteoblast-like cells. Biomaterials 2006;27:4428–33.

[20] Jack KS, Vizcarra TG, Trau M. Characterization and surfaceproperties of amino acid-modified, carbonate-containing hydroxyap-atite particles. Langmuir 2007;23:12233–42.

[21] Rosseeva EV, Golovanova OA, Frank-Kamenetskaya OV. Theinfluence of amino acids on the formation of nanocrystallinehydroxyapatite. Glass Phys Chem 2007;33:283–6.

[22] Huang S, Zhou K, Li Z. Inhibition mechanism of aspartic acid oncrystal growth of hydroxyapatite. T Nonferr Metal Soc2007;17:612–6.

[23] Liou SC, Chen SY, Lee HY, Bo JS. Structural characterization ofnanosized calcium deficient apatite powders. Biomaterials2004;25:189–96.

[24] Sherman BC. 2004. Magnesium Omeprazole. US Patent 6,713,495.[25] Klug HP, Alexander LE. X-ray diffraction procedures for polycrys-

talline and amorphous materials. New York: Wiley; 1954.[26] Brunauer S, Emmet PH, Teller E. Adsorption of gases in multimo-

lecular layers. J Amer Chem Soc 1938;60:309–19.[27] Celotti G, Tampieri A, Sprio S, Landi E, Bertinetti L, Martra G, et al.

Crystallinity in apatites: how can a truly disordered fraction bedistinguished from nanosize crystalline domains? J Mater Sci MaterMed 2006;17:1079–87.

[28] Sønju Clasen AB, Ruyter IE. Quantitative determination of type Aand type B carbonate in human deciduous and permanent enamel bymeans of Fourier transform infrared spectrometry. Adv Dent Res1997;11:523–7.

[29] Guaber SPA, Gazzaniga G, Roveri N, Rimondini L, Palazzo B,Iafisco M, Gualandi P. Biologically active nanoparticles of acarbonate-substituted hydroxyapatite, process for their preparationand composition incorporating the same. EU Patent 005146; 2006.