Polyelectrolyte Multilayer-Calcium Phosphate Composite Coatings for

Article pubs.acs.org/Biomac

Polyelectrolyte Multilayer-Calcium Phosphate Composite Coatings for Metal Implants Alon Elyada, Nissim Garti, and Helga Füredi-Milhofer* Casali Center for Applied Chemistry, The Institute of Chemistry, The Hebrew University of Jerusalem, Edmond J. Safra Campus, Givat Ram, Jerusalem 9190401, Israel ABSTRACT: The preparation of organic−inorganic composite coatings with the purpose to increase the bioactivity of bioinert metal implants was investigated. As substrates, glass plates and rough titanium surfaces (Ti-SLA) were employed. The method comprises the deposition of polyelectrolyte multilayers (PEMLs) followed by immersion of the coated substrate into a calcifying solution of low supersaturation (MCS). Single or mixed PEMLs were constructed from poly-L-lysine (PLL) alternating with poly-L-glutamate, (PGA), poly-Laspartate (PAA), and/or chondroitin sulfate (CS). ATR-FTIR spectra reveal that (PLL/PGA)10 multilayers and mixed multilayers with a (PLL/PGA)5 base contain intermolecular β-sheet structures, which are absent in pure (PLL/PAA)10 and (PLL/CS)10 assemblies. All PEML coatings had a grainy topography with aggregate sizes and size distributions increasing in the order: (PLL/PGA)n < (PLL/PAA)n < (PLL/CS)n. In mixed multilayers with a (PLL/PGA)n base and a (PLL/PAA)n or (PLL/CS)n top, the aggregate sizes were greatly reduced. The PEMLs promoted calcium phosphate nucleation and early crystal growth, the intensity of the effect depending on the composition of the terminal layer(s) of the polymer. In contrast, crystal morphology and structure depended on the supersaturation, pH, and ionic strength of the MCS, rather than on the composition of the organic matrix. Crystals grown on both uncoated and coated substrates were mostly platelets of calcium deficient carbonate apatite, with the Ca/P ratio depending on the precipitation conditions.

1. INTRODUCTION Metals and metal alloys are frequently used as implant materials for load bearing applications, such as hip replacement, because of their biocompatibility and favorable mechanical properties.1−3 However, most are bioinert and cannot actively interact with the surrounding tissue.2 In order to facilitate bone ingrowth, different surface treatments were suggested. The rationale is to retain the key physical properties while modifying only the outermost surfaces to influence biointeraction. One approach involves the deposition of calcium phosphate coatings, which are described to induce an increased bone-toimplant contact,4 to improve the implant fixation,5 and to facilitate the bridging of small gaps between implant and surrounding bone.6 Biomimetic methods have been developed to deposit such coatings at room or body temperature from low concentration calcifying solutions onto pretreated titanium surfaces.7−9 It has been shown that it is possible to coprecipitate organic macromolecules, such as proteins8 and antibiotics9 with the deposited calcium phosphate coatings. A more recent approach is coating of the implant surfaces with an organic matrix in the form of polyelectrolyte multilayers (PEMLs),10 into which relatively large concentrations of bioactive molecules, including growth hormones, can be incorporated.11,12 In this context, calcium phosphate deposition on different PEMLs has also been investigated13−15 In a previous work16−18 some of us proposed a biomimetic method for the preparation of a new type of organic/inorganic composite coatings, consisting of (PLL/PGA)n multilayers (where PLL is poly-L-lysine and PGA is poly-L-glutamate) © XXXX American Chemical Society

interspersed with octacalcium phosphate (OCP) crystals. Such coatings were prepared from PEMLs alternating with layers of amorphous calcium phosphate (ACP). The final construct was immersed into a metastable calcifying solution (MCS), thus initiating amorphous−crystalline phase transformation19−21 within and upon the organic matrix. Under the experimental conditions employed, spherulitic aggregates of relatively large (micron-sized) OCP crystals were grown, anchoring the composite coating to the underlying substrate.16 In the present work we used a similar approach but without the addition of ACP to the organic matrix. The matrix consisted of PEMLs with the cationic polyelectrolyte PLL alternating with one or two of the anionic polyelectrolytes, PGA, poly-Laspartate (PAA), and/or chondroitin sulfate (CS). Calcium phosphate crystallization was initiated from MCS by the functional groups of the respective anionic polyelectrolyte(s) in the PEML. Thus, the addition of inorganic seed material was avoided and a more intimate relationship within the organic/ inorganic composite was created. As substrates, smooth glass surfaces and extremely rough titanium surfaces were used, and the relationship between the PEMLs and the underlying substrate was investigated. In addition, the influence of the PEML’s composition, secondary molecular structure, and aggregation state on calcium phosphate nucleation and the crystals composition, structure and morphology were studied. Received: April 28, 2014 Revised: August 3, 2014

A

dx.doi.org/10.1021/bm5006245 | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

Figure 1. Three-dimensional images from AFM topography measurements of PEML coatings on glass plates, 10 μm × 10 μm. (a−d) (PLL/PGA)n with n = 3 (a), n = 5 (b), n = 7 (c), and n = 10 (d). (e) (PLL/PAA)7. 2.2. Methods. 2.2.1. Preparation of the Coatings. Generally, coatings were prepared by the following steps: • Layer-by-layer deposition of polyelectrolytes (PE) with alternating charges to obtain PEMLs; • In situ calcium phosphate crystal growth by immersion of PEML coated glass or titanium plates into MCS. 2.2.2. Polyelectrolyte Multilayer Buildup. All PEMLs were prepared by alternated adsorption of positively and/or negatively charged PEs. Glass or titanium plates were immersed into solutions of the PEs for 10 min each. After each adsorption step, the plates were washed in a buffer solution two times, to remove excess PE solution. Two types of PEML films were thus prepared and investigated: (a) PEMLs composed of 3−10 PLL/X bilayers, where X is one of the polyanions, PGA, PAA, or CS. (b) Mixed PEMLs consisting of five bilayers of (PLL/PGA) followed by five bilayers of either (PLL/PAA) or (PLL/CS). In most samples, the top layer was negatively charged, except when the coating was imaged with CLSM; in this case, fluorescein isothiocyanate (FITC)-labeled PLL (PLLFITC, Mw 30−70 kDa, Sigma) was deposited as the top layer on the multilayered substrates. The exact compositions of the respective PEMLs are given in the results section. At the end of the adsorption process, the plates were washed with TDW and kept refrigerated for at least 12 h until further use in order to allow the multilayer to stabilize. 2.2.3. In Situ Crystal Growth. MCS was prepared by mixing equal volumes of CaCl2 and Na2HPO4 solutions (Ca/P atomic ratio 1.6) in HEPES buffer, pH 8, and was filtered before use. The initial supersaturation for each experiment is given in the results section. MCS used for the kinetic tests was stable (if undisturbed) for at least 4 days. For kinetic experiments, coated glass or titanium plates were immersed in vials, containing 20 mL of MCS and incubated at 25 °C. Half milliliter samples of the supernatant were taken at specified time

2. EXPERIMENTAL SECTION 2.1. Materials. 2.1.1. Chemicals. Analytical grade chemicals and tridistilled water (TDW) were used for all experiments. HEPES buffer solution was prepared by dissolving 25 mmol/L HEPES (N-[2hydroxyethyl] piperazine-N′-2-ethanesulfonic acid, Merck) and 50 mmol/L NaCl (J.T. Baker) in TDW, and adjusting the pH to 8 with 1 M HCl. The following polymers were purchased from Sigma and/or Alamanda Polymers (Huntsville AL, USA): poly-L-lysine (PLL, Sigma, Cat. No. P2636, Mw 30−70 Kda, or Alamanda Cat. No. PLKB250, Mw 52 kDa), poly-L-glutamic acid (PGA, Sigma, Cat. No. P4886, Mw 50−100 Kda, or Alamanda Cat. No. PLE400, Mw 60 kDa), poly-Laspartic acid (PAA, Sigma, Cat. No. P5387, Mw 5−15 Kda, or Alamanda Cat. No. PLD100, Mw 14 kDa), and chondroitin sulfate (CS, from bovine cartilage, Sigma, Cat. No. C6737, Mw 15−40 kDa). All polymers were used as received. Solutions of PE at 1 mg/mL in HEPES buffer were freshly prepared not longer than 1 day before use. Calcium chloride and sodium phosphate stock solutions were prepared from CaCl2·2H2O (Merck) and Na2HPO4 (BDH), which were dried overnight before weighting and dissolving in TDW. HCl was added to the sodium phosphate solution to readjust the pH to 8. Each of the stock solutions contained, in addition, 0.05% sodium azide to prevent bacterial contamination. 2.1.2. Substrates. Sandblasted-acid etched titanium plates (TiSLA22) were received courtesy of Morphoplant GmbH, Bochum, Germany. The plates were washed with TDW and used as received. Glass plates (Borofloat 33) were purchased from Schott, Jena, Germany. Before coating, glass plates were sonicated in detergent, acetone, ethanol, and TDW. Each cycle lasted 15 min. The plates were then immersed for a minimum of 2 h in ammonium persulfate (Nochromix, Godax laboratories) solution in sulfuric acid, then washed with 50% ethanol solution and finally TDW. Hydroxyapatite nanopowder (HAP) was purchased from Sigma. B


Biomacromolecules

Article

intervals within immersion periods of 4 days and analyzed for calcium and phosphate content (see below). Incubation of some of the samples was extended up to 7 days to enable the crystals to grow for morphological and physicochemical characterization. At the end of this period, samples were washed thoroughly with TDW to remove all accidental deposition and then dried, observed by high-resolution scanning electron microscopy (HR-SEM) and energy-dispersive X-ray spectroscopy (EDX) and/or characterized by attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) and X-ray spectroscopy. 2.2.4. Adhesive Tape Test. In order to test the adhering of the crystals to the substrate, we applied the adhesive tape test according to ASTM protocol.23 An adhesive tape suitable for the test (Elcometer) was placed over the substrate and then pulled off. Before and after the test the surface was imaged by HR-SEM. 2.2.5. Instrumental Analysis. (i) Imaging of PEMLs and crystals: Substrates coated with PEMLs were imaged and characterized by atomic force microscopy in the tapping mode (AFM; Scanning Probe Microscope, Dimension 3100 Nanoscope V from Veeco/Bruker, Santa Barbara CA, USA) and confocal laser scanning microscopy (CLSM; Leica TCS SP5, Confocal microscope from Leica Microsystems, Mannheim, Germany). For AFM analysis the plates were used as is, while for CLSM analysis the plates were wetted with TDW before observation. Imaging of CaPO4 crystals was performed by HR-SEM (Sirion and/or XHR-SEM Magellan 400L, both from FEI company, Hillsboro Oregon, USA) with EDX attachments for elemental analysis. Before analysis, glass samples were coated with a gold−palladium layer, whereas Ti-SLA samples were directly imaged. (ii) Physicochemical characterization: PEML and CaPO4 crystals were characterized by ATR-FTIR (Alpha FT-IR spectrometer, Bruker Optik, Ettlingen, Germany). The spectra were recorded with 50 scans. The ATR diamond was wetted with D2O before each sample was scanned. X-ray diffraction (XRD) measurements were performed on a D8 Advance Diffractometer (Bruker AXS, Karlsruhe, Germany). Crystals grown on glass plates were scratched and transferred onto a sample holder before measurement. (iii) Kinetics of crystallization: Calcium and phosphate concentrations were determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES, Optima 3000, PerkinElmer, MA, USA). Samples of ∼0.5 mL were taken from MCS vials at specified time intervals, filtered, and diluted 10 times to achieve the instrument measuring range. 2.3. Treatment of Data. 2.3.1. Calculation of the Supersaturation. The relative supersaturation (σ) is defined as24

coatings on glass surfaces revealed a grainy topography as assessed by AFM and CLSM imaging (Figures 1−3 and Table 1). The aggregate sizes and size distributions were dependent Table 1. Aggregate Sizes of Different PEML Compositions multilayer composition

mean size (μm2)

(PLL/PGA)10 (PLL/PAA)10 (PLL/CS)7 (PLL/PGA)5−(PLL/PAA)5 (PLL/PGA)5−(PLL/CS)5

0.503 ± 0.426 5.828 ± 6.379 10.396 ± 6.824 0.101 ± 0.210 0.363 ± 0.513

on the type of the anionic layer as well as on the number of bilayers, n. The grainy structure of the (PLL/PGA)n seems to evolve throughout the PEM buildup, the density and height of the aggregates increasing with increasing n (Figure 1a−d). The height of a (PLL/PGA)10 assembly reaches approximately 100 nm (Figure 1d), which is lower than previously observed in a fluidic environment (approx 280 nm).27 The difference may be due to the drying process employed in our experiments. Compared to the relatively regular aggregates of (PLL/PGA)n multilayers, the multilayers of (PLL/PAA)n (n = 7 and 10) and (PLL/CS)7 created relatively large, irregular aggregates with wide size distributions (Figures 1e and 2b,c). The (PLL/CS)7− PLLFITC aggregates were the largest, even though fewer bilayers have been built up (Figure 2c and Table 1). When multilayers of (PLL/PAA)5−PLL or (PLL/CS)5−PLL were built on a (PLL/PGA)5 base, the resulting aggregates were much smaller with narrower size distributions (Figure 3 and Table 1). 3.1.2. Ti-SLA Surfaces. On highly porous Ti-SLA plates the polymer coating adopts the substrate contours as represented in Figures 4 (CSLM) and 5 (SEM). Figure 4a,b shows that the distribution of the PEML on an x−y plane is changing along the z-axis, as the two focal planes seem to create complementary imaging of the polymer adsorption. This indicates that the polymer is not located at one elevation line, but inside and around the pores. Even in the presence of much larger amounts of polymer the Ti pores are clearly expressed (Figure 5). 3.2. Secondary Molecular Structure of Polyelectrolyte Multilayers. Glass slides coated with PEMLs were scanned with ATR-FTIR spectrometer. In this section we are interested primarily in the transmittance “window” region of 1500−1700 cm−1 where there is no overlap with peaks from glass or D2O and where the amide I band, which contains information on the secondary structure of polypeptides,28−33 appears. (PLL/ PGA)10 multilayers show three significant peaks in this range: 1556/1564, 1617, and 1642 cm−1 (Figure 6a); the first is assigned to ionized COO− side chains of PGA,29 the second peak is in the range of 1609−1620 cm−1 in which peaks were assigned to antiparallel intermolecular β-sheet structure of PLL−PGA complexes in the multilayer,28−32 while the broad band at 1642 cm−1 has been assigned to α helices and or random structures.28−30,33 The IR spectra of (PLL/PAA)10 multilayers show only the 1642 cm−1 peak (Figure 6b), but when (PLL/PAA)5 multilayers were built on a (PLL/PGA)5 base (Figure 6c), all three above-mentioned peaks (1566, 1618, and 1643 cm−1) appeared. These results are also in accordance with the literature data.29,30 In the case of (PLL/CS)10 (Figure 6d), an amide I peak at 1638 cm−1, stemming from the PLL molecule, and a smaller

σ = (IP1/ n − K sp1/ n)/K sp1/ n where IP is the ionic activity product, Ksp is the solubility product, and n is the number of ions in the formula unit for the particular calcium phosphate phase. For convenience, in most examples we expressed the supersaturation with regard to OCP (σOCP). The activity of free ions was calculated from the Davies equation using dissociation constants25 for the following soluble species: H3PO4, H2PO4−, HPO42−, PO43−, CaH2PO4+, CaHPO4, CaPO4−, CaOH+, CaHCO3+, and CaHCO3, and assuming constant ionic strength and pH (0.093 M and 8.0, respectively). 2.3.2. Treatment of CLSM Images. Size distributions of the PEML aggregates were obtained from CLSM images by using a computer program.26

3. RESULTS 3.1. Aggregation of Polyelectrolyte Multilayers at Interfaces. 3.1.1. Glass Surfaces. All investigated PEML C


Biomacromolecules

Article

Figure 4. CLSM micrographs of Ti-SLA plate coated with (PLL/ PGA)10−PLLFITC: (a) 30th focal plane and (b) 60th focal plane. Arrows are pointing toward some of the complementary areas between the two images.

Figure 5. SEM micrographs of (a) uncoated Ti SLA plate and (b) (PLL/PGA)20−(PLL/PAA)5 on Ti-SLA.

The peak at 1609 cm−1 becomes more pronounced when five bilayers of PLL/CS are adsorbed upon five bilayers of PLL/ PGA (Figure 6e) because of additional absorption at this frequency related to the intermolecular β-sheet structure of the PLL/PGA double layers. 3.3. Crystallization Induced by PEMLs. 3.3.1. Crystallization Kinetics. In order to assess the capacity of coatings to induce de novo crystallization of calcium phosphates, glass slides, uncoated or coated with PEMLs, were immersed into a MCS of σOCP = 2.76, which was prepared as described in section 2.2. Samples of the supernatant were taken at specific time intervals, and the calcium and phosphorus concentrations were measured by ICP-AES. Figure 7a shows changes of solution calcium concentration as a function of time for glass plates coated with (PLL/PGA)10 (curve 1) and (PLL/PAA)10 (curve 2) compared to an uncoated glass plate (curve 3). The kinetic curve obtained for a mixed multilayer (PLL/PGA)5− (PLL/PAA)5 coating is similar to curve 2 in Figure 7a (data not shown). Figure 7b shows precipitation kinetics induced by a (PLL/CS)10 and a (PLL/PGA)5−(PLL/CS)5 coating, while Figure 7c is the positive control, where MCS was seeded with HAP. The corresponding phosphate concentration curves (not shown) were approximately parallel. Each of the curves in Figure 7a,b is the result of two parallel experiments in which samples were taken at different successive time intervals. When MCS was seeded with HAP (Figure 7c), the induction time was shorter than 1 h, and after approximately 8 h, the Ca and phosphate concentration approached a plateau at ∼0.5 mmol/L of Ca. The results clearly show a significant influence of the composition of the multilayer on the length of the induction time, τ, which is inversely proportional to the rate of calcium phosphate nucleation.

Figure 2. CLSM micrographs (average intensity projection along zaxis) and the corresponding particle size distribution (below) of different polymer layers on glass: (a) (PLL/PGA)10−PLLFITC; (b) (PLL/PAA)10−PLLFITC; and (c) (PLL/CS)7−PLLFITC.

Figure 3. CLSM micrographs (average intensity projection along zaxis) and the corresponding particle size distribution (below) of different polymer layers on glass: (a) (PLL/PGA)5−(PLL/PAA)5− PLLFITC and (b) (PLL/PGA)5−(PLL/CS)5−PLLFITC.

peak at 1609 cm−1, corresponding to the asymmetrical stretching mode of ionized (−COO−) groups of CS, were observed (assigments according to refs 34 and 35). D


Biomacromolecules

Article

Figure 6. FTIR spectra of PEMLs on glass: (a) (PLL/PGA)10; (b) (PLL/PAA)10; (c) (PLL/PGA)5−(PLL/PAA)5; (d) (PLL/CS)10; and (e) (PLL/ PGA)5−(PLL/CS)5.

phosphate ions in a nonstoichiometric apatitic environment, while the latter may be due to the presence of CO32− ions in nonstoichiometric apatites.36,37 In the ATR-FTIR spectrum of crystals grown on glass plates (Figure 10b), the 1115 cm−1 peak is discernible, while the 1020 cm−1 peak is obscured by the overlapping spectrum of the glass. XRD spectra (data not shown) exhibit the main diffraction peaks characteristic of hydroxyapatite but no peaks characteristic of octacalcium phosphate. We therefore conclude that the crystals are calcium-deficient carbonate apatite. 3.3.3. Mechanical Properties. SEM images of glass plates with different coatings before and after applying the ASTM adhesive tape test23 are shown in Figure 11. Figure 11a,b shows that the (PGA/PLL)10 multilayer strongly adheres to the underlying glass plate. (Note that in this experiment we are interested primarily in the strength of adhesion of the coating to the substrate. For the morphology of the aggregates, depicted in Figure 11a, see also the corresponding AFM and CLSM images in Figures 1d and 2a.) Unlike the PEML, calcium phosphate crystals readily detach, when grown directly on the glass plates (Figure 11c,d). However, when crystals were grown on the (PLL/PGA)10 multilayer, the primary crystal layer could not be removed by the adhesion tape, whereas the ball-like crystals were readily removed (Figure 11e,f). Thus, it appears that strong bonds existed between the primary crystals and the PEML, while the ball-like crystals grew upon the primary crystal layer.

It appears that the order of efficiency of the different substrates in inducing calcium phosphate crystallization is as follows: HAP ≫ (PLL/PGA)10 > [(PLL/PAA)10 ≃ (PLL/PGA)5 −(PLL/PAA)5 ] > uncoated glass ≃ [(PLL/PGA)5 −(PLL/CS)5 ] > (PLL/CS)10

3.3.2. Crystal Composition, Morphology, and Structure. After incubation of 1 week in MCS, all substrates were examined by SEM and EDX spectroscopy. Typical SEM images of both uncoated and PEML coated glass slides revealed platelet-like calcium phosphate crystals, which were evenly spread over large areas of the substrate, and, in addition, spherulitic aggregates of platelet-like crystals (Figure 8a,b). The Ca/P atomic ratio of the crystals varied between 1.35 to 1.45 (Figure 8c), independently on the composition of the underlying PEML. Crystals that were grown on Ti-SLA plates coated with (PLL/PGA)20−PLL/PAA)5 from MCS with a higher supersaturation (σOCP = 3.25) are shown in Figure 9. Just like the corresponding PEML coating (Figure 4) these crystals adopt the Ti contour, thus retaining the porousness of the surface. ATR-FTIR spectra of crystals grown on titanium (Figure 10a) reveal one main vibration peak at 1020 cm−1 and a smaller one at 1115 cm−1. The former vibration is mainly due to E


Biomacromolecules

Article

Figure 7. Changes of calcium concentration with time during calcification experiments. Calcification initiators: (a) glass substrate, △ (PLL/PGA)10; ○ (PLL/PAA)10; □ uncoated glass; (b) glass substrate, ◊ (PLL/PGA)5−(PLL/CS)5; ▽ (PLL/CS)10; (c) substrate, 5 mg of HAP powder. Initial conditions: supersaturation of MCS σOCP = 2.76, Ca/P atomic ratio = 1.6, pH = 8, and ionic strength = 0.093 M.

4. DISCUSSION The above results may facilitate our understanding of the role of biological macromolecules in the deposition of mineral (calcium carbonate and calcium phosphate) during biomineralization.38,39 Such knowledge is of great importance as the basis for the development of biomimetic, osteo-inductive, and osteo-conductive organic−inorganic composite coatings for bioinert artificial implant materials, such as metals, metal alloys, certain polymers, etc. In order to keep our system relevant to biomineralization we first deposited an organic matrix on a smooth (glass) or extremely rough (Ti-SLA) substrate surface. In a second step the coated substrate was incubated into a metastable calcifying solution, MCS, in order to evaluate its influence on de novo nucleation and/or growth of calcium phosphate crystals. 4.1. Polyelectrolyte Multilayers: Structure and Surface Topography. A convenient way to construct the organic matrix is the layer-by layer deposition approach,10 which involves alternate deposition of positively and negatively charged polyelectrolytes. In all experiments, deposition started with layers of positively charged PLL, which alternated with layers of negatively charged polyelectrolytes. PGA and PAA were chosen as models for acidic proteins, which are known to play a decisive role in biomineralization.39 The glycosaminoglycan chondroitin sulfate (CS) was chosen because it is part of the native extracellular matrix, has an anti-inflammatory effect, and appears to enhance bone remodeling and new bone

Figure 8. (a,b) Typical SEM micrographs of calcium phosphate crystals grown on uncoated glass plates and glass coated with PEML. In the above example, the PEML composition was (PLL/PGA)5− (PLL/PAA)5; panel b is an enlargement of the marked area in panel a. (c) EDX spectrum showing an average Ca/P = 1.39. Initial conditions of MCS: Ca/P = 1.6, σOCP = 2.76, pH = 8.0, and ionic strength = 0.093 M.

formation.35 CS has also been observed to possess a unique potential to adsorb large amounts of the growth hormone BMP-2, which is essential in inducing mineralization in humans.12 Assuming that the structure of the PEMLs at the solid/liquid interface will have a decisive influence on calcium phosphate crystal nucleation, we first paid attention to the aggregation (Figures 1−5) and molecular structure (Figure 6) of freshly deposited PEMLs at glass and Ti-SLA surfaces. The results obtained on secondary molecular structure of the PEMLs are consistent with previous reports.28−30,32−34,40 Our FTIR spectra show that (PLL/PGA)n multilayers form a large F


Biomacromolecules

Article

Figure 11. SEM images before (left column) and after (right column) adhesive tape test. Glass plates coated with (a,b) (PLL/PGA)10; (c,d) calcium phosphate crystals grown on uncoated glass; and (e,f) crystals grown on (PLL/PGA)10. Initial conditions: σOCP = 2.76, pH = 8, and ionic strength = 0.093 M.

Figure 9. (a,b) SEM micrographs of crystals grown on Ti-SLA plate coated with (PLL/PGA)20−(PLL/PAA)5; panel b is an enlargement of the marked area in panel a. (c) EDX spectrum; Ca/P atomic ratio = 1.52. Initial conditions of MCS: σOCP = 3.25, Ca/P = 1.6, pH = 8, and ionic strength = 0.093 M.

(PLL/PGA)5 basis, a peak at 1610 cm−1, characteristic of intermolecular β-sheet structures, appeared (Figure 6c) or, in the case of (PLL/CS), was enhanced (Figure 6e). At this stage we shall assume that the additional (or enhanced) signals in the mixed multilayer spectra stem from the (PLL/PGA)5 base, i.e., that combinations of signals from the different types of

amount of intermolecular β-sheet structure (Figure 6a), which is due to PGA−PLL complexing,30,32 but no such structure is evident from (PLL/PAA)n spectra (Figure 6b). However, when (PLL/PAA)5 or (PLL/CS)5 double-layers were coadsorbed to a

Figure 10. FTIR spectra of crystals grown on (a) Ti-SLA plate coated with (PLL/PGA)20−(PLL/PAA)5 and (b) glass coated with (PLL/PGA)10− (PLL/PAA)5 (dashed line, spectrum of glass without coating). G


Biomacromolecules

Article

propensity of crystal nucleation, we envisage a mechanism similar to the one suggested by Addadi et al. for calcite crystal nucleation by assemblies of organic macromolecules.45 We propose that the crystallization process begins with attracting Ca2+ ions to the negatively charged surface, creating a surface concentration necessary for nucleation, followed by binding of the ions to specific functional groups of the PEML. The binding of calcium ions is followed by phosphate ions to produce initial crystal nuclei. Considering this mechanism, the results of the kinetic experiments shown in Figure 7 may be interpreted as follows: With hydroxyapatite powder as positive and uncoated glass plates as negative controls, we see that both polypeptide assemblies (PLL/PGA)10 and (PLL/PAA)10 promote calcium phosphate nucleation, with (PLL/PGA)10 being the most effective (Figure 7a). The difference can be explained by the fact that the (PLL/PGA)10 assembly adopts the β-sheet conformation to a large extent, while the (PLL/PAA)10 assembly does not (Figure 6 and refs 14, 29, and 30). It can be shown that in β-sheet assemblies the repeat distances between carboxylate groups that point in the same direction (6.5−7 Å) are commensurate with the distances of calcium ions exposed in the c direction at the (100) crystal plane of OCP (6.87 Å).46,47 The (PLL/PGA)10 assembly therefore facilitates nucleation of platelet-like crystals of OCP by orienting the adsorbed calcium ions. The fact that mixed multilayers of the composition (PLL/PGA)5−(PLL/PAA)5, regardless of their partial β-sheet content, show roughly the same lag time for CaP nucleation as the (PLL/PAA)10 assembly implies that the (PLL/PGA)5 part of the mixed film is oriented toward the substrate, while the (PLL/PAA)5 part is facing the solid/ solution interface. Experiments published by Pilbat et al.30 show that adding PAA to a (PLL/PGA)5 multilayer is indeed a relatively smooth process, which does not cause major distortions of the PLL/PGA secondary structure. We may thus assume a regular buildup of the (PLL/PAA)5 layers upon the (PLL/PGA)5 base without much intermixing of the different domains. Consequently the adsorbed calcium ions would see the terminal (PLL/PAA)5 assembly rather than the underlying (PLL/PGA)5 base. For the (PLL/CS)10 and (PLL/PGA)5 (PLL/CS)5 multilayers the lag time could not be precisely determined with the current experimental setup. The corresponding kinetic curve for the mixed multilayer (Figure 7b) is very similar to the curve obtained with uncoated glass as a substrate (Figure 7a), whereas the lag time obtained with the (PLL/CS)10 assembly probably exceeds the measured time of metastability of the MCS. However, it may be concluded that the (PLL/CS)10 assembly is a very weak promoter, if not an inhibitor of CaP nucleation. This observation is at first glance surprising, considering that CS is very negative, consisting of repeated disaccharide groups, which display acidic carboxylic and sulfate groups. It has indeed been shown48 that in a calcifying solution CS can promote the formation of highly ordered HAP crystallite assemblies. However, in (PLL/CS)n films the ratio of lysine monomers over disaccharide units was found to be approximately 2, i.e., it was shown that the film contains two lysine per one disaccharide monomer unit.35 This renders the (PLL/CS)n films very close to neutrality and thus upsets the effect of the strongly negative CS molecules on templating CaP nucleation. In addition, the softness and gel-like nature of the (PLL/CS)n assembly (Figure 2c and refs 35 and 40) may also play a role.

multilayers are observed. This conclusion is supported by the results of previous authors. Thus, Pilbat et al.30 found that when PAA was added to a (PGA/PLL)5−PGA film, PGA/PLL and PAA/PLL complexes coexisted with their unaltered secondary structures in the mixed film. Likewise Grohmann et al.40 showed the content of random coil conformation increasing with the successive adsorption of (PLL/CS) bilayers onto a (PLL/PGA)5 base. The authors proposed a bizonal model, where the zone in proximity to the substrate surface was predominantly composed of chains with intermolecular β-sheet structure, while toward the film−buffer interface chains in random coiled conformation were dominant. The observations of a grainy surface topography of the PEML films are in general agreement with those of Grohmann et al.27 whose experiments were carried out in a fluidic environment. Aggregate sizes and the respective size distributions drastically increase in the order: (PLL/PGA)10 ≪ (PLL/ PAA)10 ≪ (PLL/CS)7 (Figure 2), i.e., the film with the highest local order imposed by the formation of β-sheet structures has the smallest and most regular aggregates (See also Figure 1). The large irregular aggregates obtained in (PLL/CS)n films (Figure 2c) are apparently due to the high water binding potential of CS (for eight bilayer constructs, a hydration of 63% has been measured35), which results in a soft, hydrogel-like structure of the polysaccharide. When (PLL/PAA) 5 or (PLL/CS) 5 multilayers were deposited on a (PLL/PGA)5 base (Figure 3), the resulting surface aggregates were significantly smaller and size distributions were narrower (see also Table 1). Thus, it appears that the underlying base directs the formation of aggregates in subsequent multilayers. 4.2. Nucleation and Growth of Calcium Phosphate Crystals Templated by PEMLs. The composition and structure of calcium phosphate crystals precipitating from aqueous solutions depends on the initial conditions of precipitation, i.e., the supersaturation, ionic strength, pH, and temperature of the calcifying solution. In neutral and slightly basic solutions of moderate or high supersaturation, spherulitic aggregates of amorphous calcium phosphate (ACP) are usually formed first41 and induce, after an induction period, secondary precipitation of a crystalline phase, i.e., OCP and/or calcium deficient hydroxyapatite.19,20,42 Formation of hydroxyapatite without an amorphous precursor phase has been reported at low supersaturations.43,44 In the present investigation of calcium phosphate nucleation templated with PEMLs we tried to avoid the formation and phase transformation of ACP, which would ultimately result in the formation of spherulitic crystalline aggregates.16−18 Therefore, the supersaturation was kept as low as possible within the metastable range, i.e., σOCP = 2.76−3.25. Given the results of the kinetic experiments (Figure 7) we assume that the organic matrix is involved in the induction of heterogeneous nucleation of the inorganic phase. The question is then: Do changes in the surface topography and/or the molecular structure have a significant influence on its propensity to induce nucleation? Considering that relatively large changes in the surface topography between (PLL/PAA)10 (Figure 2b) and (PLL/PGA)5−(PLL/PAA)5 (Figure 3a) produce just minor changes in the lag time preceding rapid crystal growth (note that curve 2 in Figure 7a is representative of both systems), we may surmise that the surface topography alone has very little influence. Then, in order to understand, whether the molecular structure of the PEMLs influences their H


Biomacromolecules

Article

The morphology and structure of the crystals ultimately obtained were independent of the substrate composition and could be identified as calcium-deficient carbonate apatite. On smooth substrates (coated or uncoated glass plates), even layers of platelet-like crystals formed in close proximity to the substrate with additional spherulitic aggregates (average Ca/P = 1.39−1.42) on top of the layers (Figure 8), while on rough porous surfaces (Ti-SLA plates) the coating (both polymer and crystals, Ca/P ≈ 1.52) followed the substrate contours (Figures 5 and 9). The plate- or ribbon-like crystal habit is typical for both precipitated and biological apatite49,50 and has been taken as evidence that these crystals were formed via an OCP intermediate.51 Evenly spread platelets have also been obtained previously on uncoated silica surfaces13 as well as on PLL, CS, and/or phosvitin terminated films.15 The formation of spherulitic aggregates of radially oriented crystals (see Figure 8a and refs 7, 14, 16, and 52) is fairly typical for templated calcium phosphate crystallization and may be caused by the following mechanisms: (a) Crystallization initiated by spherulitic ACP particles. This mechanism has been demonstrated by embedding ACP particles between PEMLs and immersing the construct into a calcifying solution.16 It has also been shown19,20,41 that when calcium phosphate is precipitated from aqueous solutions of pH ≥ 7.5 and σOCP ≥ 4.5, the first precipitate consists of spherulitic aggregates of amorphous particles (ACP). Such particles may adsorb on a substrate coated with PEMLs and, by way of phase transformation, yield spherulitic crystalline aggregates. (b) Secondary crystallization. If, at low supersaturations, heterogeneous nucleation and subsequent crystal growth are initiated by the organic matrix, at prolonged immersion the first crystal layer may initiate secondary crystallization. In that case, the secondary crystalline precipitate may self-assemble as spherulitic aggregates. We assume that the spherulitic crystal aggregates shown in Figure 8a may be due to such mechanism. This assumption is corroborated by the fact that these crystal aggregates are relatively easily removed by the adhesive tape test, while the evenly spread platelet-like crystals are tightly bound to the underlying polymer (Figure 11e,f). Kinetic studies to confirm this assumption are planned. From the standpoint of implantology, coating of ultrahydrophilic, highly porous substrates, such as Ti-SLA plates by the above-described organic−inorganic composite should have several advantages: 1. The preparation is simple and cost-effective, and the organic matrix of the composite coating can be used as a carrier of bioactive molecules, such as growth hormones11,12 and/or drugs. The inorganic phase should enhance the propensity of the coating for biomineralization [note that hydroxyapatite powder induces calcium phosphate deposition much faster than any of the polyelectrolytes studied in this investigation (Figure 7c)]. 2. During the coating procedure, the porous architecture of the titanium substrate is preserved, with the calcified polymer localized within and around the pores (Figure 9). Thus, the coated surface area is enlarged, facilitating bone ingrowth and incorporation of the artificial implant into the surrounding tissue. 3. The organic−inorganic composite tightly adheres to the substrate (Figure 11e,f), which may be especially

important in dental implantology, where the implant is screwed into the surrounding bone tissue.

5. CONCLUSIONS The formation and characteristics of organic−inorganic composite coatings consisting of polyelectrolyte multilayers (PEMLs) and calcium phosphate crystals on smooth (glass) and rough (Ti-SLA) surfaces have been studied. It was found that the PEMLs influence calcium phosphate crystal nucleation and early crystal growth kinetics, as measured by the lag time before the onset of fast precipitation. The ability of the PEML to promote calcium phosphate nucleation depended on the composition and molecular structure of the terminal layer(s). The structure and morphology of the ensuing crystals depended on the calcifying solution (i.e., its supersaturation, pH, ionic strength, and temperature), the time of aging, etc., rather than on the properties of the PEML. The coatings tightly adhere to the underlying substrate, and the pores of rough substrates are preserved; thus, integration of a coated substrate into the surrounding tissue should be facilitated. The results point to a simple and cost-effective method for the enhancement of the bioactivity of artificial implants, which are used for bone and/or tooth replacement and regeneration.

■

AUTHOR INFORMATION

Corresponding Author

*(H.F.-M.) E-mail: [email protected]. Phone: 972-26585885. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS It is a pleasure to thank Prof. Dr. H. P. Jennissen, Institute of Physiological Chemistry, University of Duisburg-Essen, Essen, and Dr. M. Laub, CEO Morphoplant GmbH, Bochum, Germany for making Ti-SLA plates available. Thanks are also due for imaging and/or analytical work to Mrs. Evgenia Blayvas, Unit for Nanoscopic Characterization, Dr. Naomi MelamedBook, Institute for Life Sciences and Mr. Ofir Tirosh, Institute of Earth Sciences, all from the Faculty of Science, the Hebrew University of Jerusalem, Jerusalem, Israel. The financial support granted by the German−Israeli Foundation (GIF, grant I-90739, 3/2006) is gratefully acknowledged.

■

REFERENCES

(1) Breme, J.; Steinhäuser, E.; Paulus, G. Commercially pure titanium Steinhäuser plate-screw system for maxillofacial surgery. Biomaterials 1988, 9 (4), 310−313. (2) Liu, X.; Chu, P. K.; Ding, C. Surface modification of titanium, titanium alloys, and related materials for biomedical applications. Mater. Sci. Eng., R 2004, 47 (3−4), 49−121. (3) Balazic, M.; Kopac, J.; Jackson, M. J.; Ahmed, W. Review: titanium and titanium alloy applications in medicine. Int. J. Nano Biomater. 2007, 1 (1), 3−34. (4) Barrère, F.; van der Valk, C. M.; Meijer, G.; Dalmeijer, R. A. J.; de Groot, K.; Layrolle, P. Osteointegration of biomimetic apatite coating applied onto dense and porous metal implants in femurs of goats. J. Biomed. Mater. Res., Part B 2003, 67B (1), 655−665. (5) Soballe, K.; Hansen, E. S.; Brockstedt-Rasmussen, H.; Bunger, C. Hydroxyapatite coating converts fibrous tissue to bone around loaded implants. J. Bone Jt. Surg., Br. Vol. 1993, 75-B (2), 270−278.

I


Biomacromolecules

Article

(6) Stephenson, P. K.; Freeman, M. A. R.; Revell, P. A.; Germain, J.; Tuke, M.; Pirie, C. J. The effect of hydroxyapatite coating on ingrowth of bone into cavities in an implant. J. Arthroplasty 1991, 6 (1), 51−58. (7) Wen, H. B.; de Wijn, J. R.; Cui, F. Z.; de Groot, K. Preparation of calcium phosphate coatings on titanium implant materials by simple chemistry. J. Biomed. Mater. Res. 1998, 41 (2), 227−236. (8) Wen, H. B.; de Wijn, J. R.; van Blitterswijk, C. A.; de Groot, K. Incorporation of bovine serum albumin in calcium phosphate coating on titanium. J. Biomed. Mater. Res. 1999, 46 (2), 245−252. (9) Stigter, M.; de Groot, K.; Layrolle, P. Incorporation of tobramycin into biomimetic hydroxyapatite coating on titanium. Biomaterials 2002, 23 (20), 4143−4153. (10) Decher, G. Fuzzy nanoassemblies: Toward layered polymeric multicomposites. Science 1997, 277 (5330), 1232−1237. (11) Crouzier, T.; Ren, K.; Nicolas, C.; Roy, C.; Picart, C. Layer-bylayer films as a biomimetic reservoir for rhBMP-2 delivery: Controlled differentiation of myoblasts to osteoblasts. Small 2009, 5 (5), 598− 608. (12) Jennissen, H. P.; Madenci, S.; Lüers, S.; Laub, M. Monitoring polyelectrolyte multilayer assembly and stability on non-transparent rough metal surfaces. Biomed. Eng. 2013, 58, Suppl. 1. (13) Ngankam, P. A.; Lavalle, P.; Szyk, L.; Decher, G.; Schaaf, P.; Cuisinier, F. J. G. Influence of polyelectrolyte multilayer films on calcium phosphate nucleation. J. Am. Chem. Soc. 2000, 122 (37), 8998−9005. (14) Ball, V.; Michel, M.; Boulmedais, F.; Hemmerle, J.; Haikel, Y.; Schaaf, P.; Voegel, J. C. Nucleation kinetics of calcium phosphates on polyelectrolyte multilayers displaying internal secondary structure. Cryst. Growth Des. 2005, 6 (1), 327−334. (15) Abdelkebir, K.; Morin-Grognet, S.; Gaudière, F.; Coquerel, G.; Labat, B.; Atmani, H.; Ladam, G. Biomimetic layer-by-layer templates for calcium phosphate biomineralization. Acta Biomater. 2012, 8 (9), 3419−3428. (16) Dutour-Sikirić, M.; Gergely, C.; Elkaim, R.; Wachtel, E.; Cuisinier, F. J. G.; Füredi-Milhofer, H. Biomimetic organic−inorganic nanocomposite coatings for titanium implants. J. Biomed. Mater. Res., Part A 2009, 89A (3), 759−771. (17) Schade, R.; Dutour-Sikirić, M.; Lamolle, S.; Ronold, H. J.; Lyngstadass, S. P.; Liefeith, K.; Cuisinier, F. J. G.; Füredi-Milhofer, H. Biomimetic organic−inorganic nanocomposite coatings for titanium implants. In vitro and in vivo biological testing. J. Biomed. Mater. Res., Part A 2010, 95A (3), 691−700. (18) Füredi-Milhofer, H.; Bar Yosef-Ofir, P.; Sikirić, M.; Gergely, C.; Cuisinier, F. Organic-inorganic nanocomposite coatings for implant materials and methods of preparation thereof. Patent WO2004047880A1, 2004. (19) Brečević, L.; Füredi-Milhofer, H. Precipitation of calcium phosphates from electrolyte solutions. II. Formation and transformation of the precipitates. Calcif. Tissue Res. 1972, 10 (1), 82−90. (20) Bar-Yosef Ofir, P.; Govrin-Lippman, R.; Garti, N.; FürediMilhofer, H. The influence of polyelectrolytes on the formation and phase transformation of amorphous calcium phosphate. Cryst. Growth Des. 2004, 4 (1), 177−183. (21) Rabadjieva, D.; Gergulova, R.; Titorenkova, R. Biomimetic transformations of amorphous calcium phosphate: kinetic and thermodynamic studies. J. Mater. Sci.: Mater. Med. 2010, 21 (9), 2501−2509. (22) Jennissen, H. P.; Zumbrink, T.; Chatzinikolaidou, M.; Steppuhn, J. Biocoating of Implants with Mediator Molecules: Surface Enhancement of Metals by Treatment with Chromosulfuric Acid. Materialwiss. Werkstofftech. 1999, 30 (12), 838−845. (23) Standard Test Methods for Measuring Adhesion by Tape Test. In Paint Tests for Chemical, Physical, and Optical Properties; Appearance; ASTM: West Conshohocken, PA, 2009; p D3359-09e2. (24) Nancollas, G. H. In Vitro Studies of Calcium Phosphate Crystallization. In Biomineralization. Chemical and Biochemical Perspectives; Mann, S., Webb, J., Williams, R. J. P., Eds.; VCH Verlagsgesellschaft: Weinheim, Germany, 1989; pp 157−187.

(25) Lu, X.; Leng, Y. Theoretical analysis of calcium phosphate precipitation in simulated body fluid. Biomaterials 2005, 26 (10), 1097−1108. (26) NIH. ImageJ. http://rsbweb.nih.gov/ij. (27) Grohmann, S.; Rothe, H.; Frant, M.; Liefeith, K. Colloidal force spectroscopy and cell biological investigations on biomimetic polyelectrolyte multilayer coatings composed of chondroitin sulfate and heparin. Biomacromolecules 2011, 12 (6), 1987−1997. (28) Boulmedais, F.; Schwinté, P.; Gergely, C.; Voegel, J. C.; Schaaf, P. Secondary structure of polypeptide multilayer films: An example of locally ordered polyelectrolyte multilayers. Langmuir 2002, 18 (11), 4523−4525. (29) Debreczeny, M.; Ball, V.; Boulmedais, F.; Szalontai, B.; Voegel, J. C.; Schaaf, P. Multilayers built from two component polyanions and single component polycation solutions: A way to engineer films with desired secondary structure. J. Phys. Chem. B 2003, 107 (46), 12734− 12739. (30) Pilbat, A. M.; Ball, V.; Schaaf, P.; Voegel, J. C.; Szalontai, B. Partial poly(glutamic acid) ↔ poly(aspartic acid) exchange in layer-bylayer polyelectrolyte films. Structural alterations in the threecomponent architectures. Langmuir 2006, 22 (13), 5753−5759. (31) van Stokkum, I. H. M.; Linsdell, H.; Hadden, J. M.; Haris, P. I.; Chapman, D.; Bloemendal, M. Temperature-Induced Changes in Protein Structures Studied by Fourier Transform Infrared Spectroscopy and Global Analysis. Biochemistry 1995, 34 (33), 10508−10518. (32) Itoh, K.; Tokumi, S.; Kimura, T.; Nagase, A. Reinvestigation on the buildup mechanism of alternate multilayers consisting of poly(Lglutamic acid) and poly(L-, D-, and DL-lysines). Langmuir 2008, 24 (23), 13426−13433. (33) Susi, H.; Timasheff, S. N.; Stevens, L. Infrared spectra and protein conformations in aqueous solutions: I. The amide I band in H2O and D2O solutions. J. Biol. Chem. 1967, 242 (23), 5460−5466. (34) Rhee, S. H.; Tanaka, J. Self-assembly phenomenon of hydroxyapatite nanocrystals on chondroitin sulfate. J. Mater. Sci.: Mater. Med. 2002, 13 (6), 597−600. (35) Crouzier, T.; Picart, C. Ion pairing and hydration in polyelectrolyte multilayer films containing polysaccharides. Biomacromolecules 2009, 10 (2), 433−442. (36) Rey, C.; Shimizu, M.; Collins, B.; Glimcher, M. Resolutionenhanced fourier transform infrared spectroscopy study of the environment of phosphate ion in the early deposits of a solid phase of calcium phosphate in bone and enamel and their evolution with age: 2. Investigations in the v3 PO4 domain. Calcif. Tissue Int. 1991, 49 (6), 383−388. (37) Koutsopoulos, S. Synthesis and characterization of hydroxyapatite crystals: A review study on the analytical methods. J. Biomed. Mater. Res. 2002, 62 (4), 600−612. (38) Lowenstam, H. A.; Weiner, S. Biomineralization Processes. In On Biomineralization; Oxford University Press: New York, 1989; pp 25−49. (39) Weiner, S.; Addadi, L. Acidic macromolecules of mineralized tissues: The controllers of crystal formation. Trends Biochem. Sci. 1991, 16 (0), 252−256. (40) Grohmann, S.; Rothe, H.; Liefeith, K. Investigations on the secondary structure of polypeptide chains in polyelectrolyte multilayers and their effect on the adhesion and spreading of osteoblasts. Biointerphases 2012, 7, 1−4. (41) Eanes, E. D.; Gillessen, I. H.; Posner, A. S. Intermediate states in the precipitation of hydroxyapatite. Nature 1965, 208 (5008), 365− 367. (42) Despotović, R.; Filipović-Vinceković, N.; Füredi-Milhofer, H. Precipitation of calcium phosphates from electrolyte solutions. Calcif. Tissue Res. 1975, 18 (1), 13−26. (43) Boskey, A. L.; Posner, A. S. Formation of hydroxyapatite at low supersaturation. J. Phys. Chem. 1976, 80 (1), 40−45. (44) Nancollas, G. H.; Tomazic, B. Growth of calcium phosphate on hydroxyapatite crystals. Effect of supersaturation and ionic medium. J. Phys. Chem. 1974, 78 (22), 2218−2225. J


Biomacromolecules

Article

(45) Addadi, L.; Moradian, J.; Shay, E.; Maroudas, N. G.; Weiner, S. A chemical model for the cooperation of sulfates and carboxylates in calcite crystal nucleation: Relevance to biomineralization. Proc. Natl. Acad. Sci. U.S.A. 1987, 84 (9), 2732−2736. (46) Brown, W. E. Octacalcium phosphate and hydroxyapatite: Crystal structure of octacalcium phosphate. Nature 1962, 196 (4859), 1048−1050. (47) Füredi-Milhofer, H.; Moradian-Oldak, J.; Weiner, S.; Veis, A.; Mintz, K. P.; Addadi, L. Interactions of matrix proteins from mineralized tissues with octacalcium phosphate. Connect. Tissue Res. 1994, 30 (4), 251−264. (48) Jiang, H.; Liu, X.-Y.; Zhang, G.; Li, Y. Kinetics and template nucleation of self-assembled hydroxyapatite nanocrystallites by chondroitin sulfate. J. Biol. Chem. 2005, 280 (51), 42061−42066. (49) Moradian-Oldak, J.; Weiner, S.; Addadi, L.; Landis, W. J.; Traub, W. Electron imaging and diffraction study of individual crystals of bone, mineralized tendon and synthetic carbonate apatite. Connect. Tissue Res. 1991, 25 (3−4), 219−228. (50) Bocciarelli, D. S. Morphology of crystallites in bone. Calcif. Tissue Res. 1970, 5 (1), 261−269. (51) Elliott, J. C. Structure and Chemistry of the Apatites and Other Calcium Orthophosphates; Elsevier: New York, 1994; Vol. 18, p 155. (52) Balas, F.; Kawashita, M.; Nakamura, T.; Kokubo, T. Formation of bone-like apatite on organic polymers treated with a silane-coupling agent and a titania solution. Biomaterials 2006, 27 (9), 1704−1710.

K


Polyelectrolyte Multilayer-Calcium Phosphate Composite Coatings for

Recommend Documents