- Email: [email protected]
Janus emulsion mediated porous scaffold bio-fabrication
Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Janus emulsion mediated porous scaffold bio-fabrication Ildiko Kovach a , Jens Rumschöttel a , Stig E. Friberg b , Joachim Koetz a,∗ a b
Institut für Chemie, Universität Potsdam, Karl-Liebknecht-Straße 24-25, D-14476 Potsdam, Germany Ugelstad Laboratory, NTNU, Trondheim, Norway
a r t i c l e
i n f o
Article history: Received 3 March 2016 Received in revised form 22 April 2016 Accepted 5 May 2016 Available online 10 May 2016 Keywords: Janus emulsions Calcium phosphates Gelatin-chitosan scaffolds,
a b s t r a c t A three dimensional biopolymer network structure with incorporated nano-porous calcium phosphate (CaP) balls was fabricated by using gelatin-chitosan (GC) polymer blend and GC stabilized olive/silicone oil Janus emulsions, respectively. The emulsions were freeze-dried, and the oil droplets were washed out in order to prepare porous scaffolds with larger surface area. The morphology, pore size, chemical composition, thermal and swelling behavior was studied by Scanning Electron Microscopy (SEM), Fourier Transform Infrared Spectroscopy (FTIR) and micro-Differential Scanning Calorimetry (micro-DSC). Microscopic analysis confirmed that the pore size of the GC based sponges after freeze-drying may be drastically reduced by using Janus emulsions. Besides, the incorporation of nanoporous calcium phosphate balls is also lowering the pore size and enhancing thermal stability. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The purpose of tissue engineering is to fabricate biocompatible and biodegradable scaffolds with similar mechanical properties of the targeted tissue for repair and replacement. Natural polymers possess biocompatibility, high porosity, biomechanical compatibility, and biodegradability. Consequently, they are considered as biodegradable biomaterials which are used clinically and are still attracting much attention for their application as material for tissue engineering [1]. Gelatin is one of the most popular amphiphilic biopolymer scaffold materials for tissue regeneration. Because gelatin is derived from natural collagen, which forms the extracellular matrix of the bone, it can provide a native like environment for the cells. The main disadvantage of gelatin as a tissue substitute is that gelatin is unstable in aqueous medium. To overcome this weakness gelatin can be associated with other naturally obtained polymers, in particular with chitosan [2]. Structure formation can be realized through electrostatic interactions (polyelectrolyte complex formation), physical binding (blends), or covalent cross-linking (network formation) [3–5]. Studies provide that the dissolution degree of the gelatinchitosan blends is reduced by increasing the chitosan concentration [5]. As well as gelatin, chitosan has qualified itself for tissue engineering, on account of its outstanding properties, namely high
∗ Corresponding author. E-mail address: [email protected] (J. Koetz). http://dx.doi.org/10.1016/j.colsurfb.2016.05.018 0927-7765/© 2016 Elsevier B.V. All rights reserved.
biodegradability, biocompatibility, non-toxicity and anti-microbial properties, osteoconductivity, hemostaticity and last but not least of its low costs [6,7]. Chitosan based scaffolds and gels have been studied extensively for wound healing, tissue engineering and tissue repair of bones, cartilage, liver, or nerve tissues [8,9]. The attention towards biopolymer hard tissue engineering has focused on composites. Since these materials are close to mimic the construction of natural bones, calcium orthophosphate crystals within a collagen fiber have been favored. In composite materials the high mechanical stability of ceramics is advantageously coupled with good elasticity of the polymers [10–12]. Calcium phosphate (CaP) chitosan composites are claimed to have outstanding osteoblast adhesion, migration, differentiation, and proliferation. Furthermore, the bioactivity and cell growth factor are outstanding [13]. Apart from scaffold design chitosan and gelatin are widely applied to influence the mineralization process of calcium orthophosphates, e.g., by constructing diverse hybrid structures [14]. Wang et al. precipitated flower like hybrid chitosan/hydroxyapatite composites under stirring [15]. Li et al. produced calcium phosphate nanoparticles in a matrix mediated synthesis and claimed that the chitosan/pectin network modulates nucleation and growth of the hydroxyapatite crystals [16]. In addition, there are various methods of preparing porous scaffolds using biopolymers, which include fiber bonding, melt molding solvent casting/particulate leaching, gas foaming and phase separation and freeze-drying, among others [17,18]. It is a challenge for tissue engineering to reconstruct certain tissues with appropriate pore size, interconnectivity and porosity. The
348
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
effect of the pore size on tissue regeneration was demonstrated by several studies, suggesting that the ideal pore size is: -
5 m for neovascularization, 5–15 m for fibroblast ingrowth, 20 m for hepatocytes ingrowth, 200–350 m for osteoconduction, 20–125 m for mammalian skin regeneration [2].
With a reference to our latest result, a supramolecular CaP cardhouse structure (of about 5–15 m in size) was synthesized in presence of gelatin/chitosan polymer blend. In other words, nanoporous calcium phosphate balls are formed, built up by individual flat CaP platelets [19]. In the present contribution, the freeze-drying method was applied to fabricate porous biopolymer ceramic scaffolds with gelatin-chitosan (GC) networks modified with hybrid calcium phosphate balls [19]. Our approach was to introduce the so called Janus emulsions as a template phase for controlling the surface area and pore size of the three-dimensional structure formed, followed by the freeze-dry technique to fabricate a 3D scaffold. Janus emulsions are dual emulsions of commonly known multiple emulsions, where two non-mixable oil components, e.g., silicone and vegetable oil, are combined in one drop. Introduced by Nisisako et al. [20], Janus emulsions immediately attracted extensive attention, because of the compelling correlation between interfacial tension and the drop topology [21–23]. Droplet topology can be controlled, e.g., in a one-step high energy mixing procedure [24]. Immiscible liquids control the morphology of patchy emulsions, but for the present contribution these emulsions were prepared by traditional vibration emulsification, according to Hasinovic et al. [25,26]. One criterion for the spontaneous Janus droplet formation, numerically evaluated [27] and experimentally determined [28], is that the interfacial tension of the more hydrophobic oil towards the aqueous phase is less than the sum of the two remaining interfacial tensions [27]. However, most relevant for the present publication are a new type of Janus emulsions, introduced by Kovach et al. [29], where the oil droplets are stabilized by means of biopolymers, namely gelatin (G) and chitosan (C). The authors claimed that in a first step the GC complexes were strongly adsorbed at the olive oil/water interface, before a rigid skin-like polymer composite layer is formed [29]. GC polymer blends can build up three dimensional networks and stabilize completely engulfed Janus emulsions, and can control the formation of supramolecular ordered CaP nanoporous card-house structures [29,19]. Taking this into account, the aim of the present research was to utilize CaP balls for
constructing porous composite scaffolds and applying Janus emulsions as a template phase to affect the pore size of the GC scaffold. 2. Experimental 2.1. Materials The low molecular weight chitosan with a degree of deacetylation of 81.2%, and a moisture content ≤12 wt% was obtained from Sigma-Aldrich® , and used without further treatment. Gelatin powder (Type A, isoelectric point ≈7, Bloom number 140) with a moisture content ≤11 wt% was purchased from Carl Roth® . The polymers were used without any further treatment. Silicone oil (SiO) (viscosity: 10–mPa s), olive oil (OO) and ethanol ≥99.5% were obtained from Sigma-Aldrich® . Recently, we were able to show that well defined supramolecular structured, spherical CaP card-house structures are formed in the presence of gelatin-chitosan blends in a wet chemical procedure at 90 ◦ C [19]. The resulting CaP powder (without calcification) was used here as given. All reagents were dissolved in Millipore Milli-Q deionized water. 2.2. Preparation of the gelatin-chitosan (GC) blend Chitosan powder (2.5 wt%) was dissolved in 0.1 mol acetic acid, and the solution was homogenized by stirring overnight. Gelatin (2.5 wt%) was dissolved in 0.1 mol acetic acid, under continuous stirring and heating up to 40 ◦ C for 5 min. The 2.5 wt% GC blend was prepared by mixing the 2.5 wt% chitosan solution with the 2.5 wt% gelatin solution, in the ratio of 1:1, under constant stirring at room temperature for 72 h. 2.3. Preparation of GC-CaP suspensions The nano-porous, ball-like calcium phosphate particles (with diameter between 5 and 15 m) synthesized according to Ref. [19], were added to the 2.5 wt% GC blend by stirring for 2 min, in order to obtain the GC-CaP suspension. The resulting GC-CaP suspension contains 0.5 wt% calcium phosphate. 2.4. Preparation of Janus GC and Janus GC-CaP emulsions The GC blend solution and GC-CaP suspension were used as the aqueous phase for preparing Janus emulsions. Each mixture consists of 0.15 g silicone oil, 0.15 g olive oil and 0.7 g of the GC solution or GC-CaP suspension. The emulsification with the oil components was made in a 2 ml Eppendorf tube by mixing with Minishaker IKA (Roth® ) at 2500 rpm in order to produce completely engulfed
Fig. 1. Micrograph of the GC blend Janus emulsion (A) and the corresponding GC-CaP Janus emulsion (B).
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Janus emulsions according to Ref. [28], that means by mixing the aqueous phase with the two oil phases in a one-step procedure. Fig. 1 shows the GC blend Janus emulsion (A) and the corresponding GC-CaP Janus emulsion (B), respectively. Emulsions are freshly prepared direct before the freeze-drying process. The emulsions were solidified by immersing into liquid nitrogen.
349
2.5. Scaffold fabrication by freeze-drying and characterization The aqueous GC blend and the Janus emulsions were submerged into liquid nitrogen, the end of the Eppendorf tubes were cut, and freeze-dried at 0 ◦ C for 12–48 h. The freeze-drying time depends on the nature of the samples. The morphology of the freeze-dried scaffolds was determined by means of scanning electron microscopy
Fig. 2. Photograph of GC (A), GC-CaP composite (B), Janus GC (C), Janus GC-CaP composite (D) after freeze-drying.
Fig. 3. SEM micrographs (cross section) of the freeze-dried scaffold materials: GC (A), GC-CaP composite (B), Janus GC (C), Janus GC-CaP composite (D).
350
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Fig. 4. SEM micrographs of GC-CaP composite at different magnification.
(S-4800 from Hitachi® ). The scaffold samples were sited on an aluminum sample holder and coated with platinum layer to make the sample conductive. Janus GC and Janus GC-CaP composite scaffolds were investigated in cryo-mode, because of the presence of oil rests. Firstly, the samples were cooled by plunging into liquid nitrogen, then freeze fractured at −180 ◦ C, etched for 45 s at −98 ◦ C, and sputtered with platinum in the GATAN® Alto 2500 cryo preparation chamber. Afterwards, the samples were placed into the microscope chamber. Pore sizes were measured by using the iTEM software from Olympus Soft Imaging Solutions (OSIS). Statistical data were carried out from 200 to 300 data points, using more than one micrograph. Due to the deformation of the pores the pore perimeter was measured instead of the diameter. The chemical composition of the scaffolds was investigated using attenuated total reflection (ATR) with NEXUS FTIR spectrometer (Thermo-Nicolet® , Diamond) with a scanning range of 400–4000 cm−1 and a resolution of 2 cm−1 at a scan rate of 32 scans/sample. ATR correction was done via Omnic 8.1.11 (Thermo Fisher Scientific Inc® ). DSC experiments were performed with a Setaram® Micro-DSC VII with a fixed heating and cooling rate of 1 ◦ C min−1 . About 4 mg of a freeze-dried sample (2.5 wt% gelatin and 2.5 wt% chitosan solution, GC blend, and GC-CaP composite) was cut to a form of the aluminum pan and placed inside the pan without compression and scanned between +25 ◦ C and +100 ◦ C. In order to get a better resolution, the heating curves are plotted between +50 ◦ C and +100 ◦ C (compare Fig. 5). Scaffold samples were measured against empty pan. Measurements were recorded in triplicate. Enthalpy is considered as the endothermic peak area, determined by Calisto Processing Thermal Analysis Software. 2.6. Swelling tests Cylinder shaped freeze-dried scaffold samples were weighed (Wdry ) and immersed in glacial ethanol. The samples were ultrasound treated for 10 min. After the samples were removed from the ethanol, the ethanol wet weight (Wethanol ) of the scaffolds was measured. Then the ethanol was replaced with ethanol water mixture series (90%, 80%, 70%, 60%, 50%) each for 30 min and finally with distillated water, near gentle shaking. After 24 h immersion in water at room temperature the wet water weight (Wwater ) was measured. All experiments were performed in triplicate and the results were expressed as an average value.
Fig. 5. Fourier Transform Infrared Spectra of GC, CaP, GC-CaP composite (A); Janus GC and Janus GC-CaP composite (B).
3. Results and discussion 3.1. Morphological characterization All scaffolds are soft sponge-like (Fig. 2), except the GC-CaP composite which under dry condition is brittle and less flexible than the other scaffolds. This means that the small addition of CaP significantly changes the mechanical properties of the polymer. Janus emulsion supported scaffolds are pale yellow colored (Fig. 2C, D), indicating that the scaffolds contain olive oil residues. The GC polymer blend adsorbs rapidly on the olive oil surface in contrast to the case of silicone oil [29]. Emulsions can be formed by using the oils, separately. In both cases the size of the drops decreases. Olive oil emulsions are stable for a longer time, but not yet silicon oil emulsions. In presence of GC blends completely engulfed Janus drops exist, which means that the silicon drops are trapped in the olive oil, i.e., an olive layer covers the silicon droplet. When this layer breaks the Janus drop turns to be partially engulfed. Silicon oil GC emulsions are destabilizing faster than olive oil and/or Janus emulsions, as to be seen experimentally (not shown here). However, scaffold fabrication failed by using GC stabilized emulsions with only one oil component, i.e., olive oil or silicone oil. In presence of olive oil, dried substance remains in clumps with a deep
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
yellow color, without stable frame. Because the droplet of the olive oil is much smaller than the Janus droplet and during the drying process it impregnates the polymer surface. Attempts with silicon oil alone were not successful, silicon oil flew out in the very beginning of the drying process. By this movement bigger pores and a room inside of the scaffold is formed. The advantage of applying Janus droplets is that the two oils hinder each other during the preparation process and porous scaffolds can be prepared. Fig. 3A shows homogenous elongated pore structures with a heterogeneous pore size. The average pore size (perimeter) of the GC scaffold is 720 ± 540 m, but the pores are isolated from each other. Nwe et al. [30] show similar elongated pore structures, which are typical for shrimps´ı shell chitosan freeze-dried scaffolds. SEM micrographs of the freeze-dried 2.5 wt% polymer solutions alone (compare Fig. S1) display a typical lamellar morphology, along the freezing direction in both cases. GC-CaP composite scaffolds, presented in Fig. 3B, show a drastically changed morphology. The size of the average pore perimeter had decreased to 14 ± 11 m, and the pore interconnectivity was highly increased. It has to be mentioned here that the interconnected geometry of the scaffolds is very important for the cell growth, to insure that the cells are supplied with oxygen and nutrients. The structure of individual calcium phosphate particles inside the scaffold will be discussed in more detail below. The remaining oil residues in the Janus supported scaffolds are disturbing the vacuum of the microscopy chamber, therefore, the micrographs were scanned in cryo-mode. The surface of Janus GC scaffolds is visibly dissimilar from that of the GC scaffolds. Janus supported GC scaffolds present a lamellar network structure with interconnected pores (Fig. 3C). The average pore size is 77 ± 33 m. However, the SEM micrograph of Janus GC-CaP composite scaffolds (Fig. 3D) still contain oil residues, making the CaP observation difficult. Some of the CaP particles were found, but are covered with oil. The average pore size is 238 ± 103 m. The higher pore size range can be explained with the presence of CaP balls preventing the smaller Janus droplet formation during the emulsification process. For a more comprehensive characterization of the incorporated CaP in the GC network additional SEM micrographs at higher magnification were performed. In Fig. 4 one can see supramolecular arranged calcium phosphate balls inside of the GC network, the CaP nano-porous aggregates of about 10 m in size are build up by individual thin platelets. The nano-porous CaP balls are only weakly embedded into the polymer network as to be seen in Fig. 4. Based on the fact that the calcium phosphate aggregates have the same electrical charge as the network (Zeta potential: +15 ± 2 mV [19]), one may assume that electrostatic repulsion forces hinder a stronger binding between the network and the CaP nano-balls. The supramolecular structure of the crystal balls is arranged as a card house, built up by individual dicalcium phosphate dihydrate (DCPD) platelets in a metastable calcium orthophosphate form. Dorozhkin recently discussed that DCPD efficiently resorbs under physiological conditions, and thereby can support bone remineralization [31].
351
to CO stretching vibration of amid I. at 1660 cm−1 and amid II. at 1559 cm−1 . In addition, absorption peak at 1650 cm−1 is an N H bending group of chitosan amid I. The bands at 1410, 1311 and 1260, 1140 cm−1 , are assigned to CH2 bending; 1140 and 1074 cm−1 to C O C stretching were confirmed [32,33,13]. GC-CaP composites (Fig. 5A) show the characteristic groups of phosphate and polymers. However, the absorption peaks attributed to the polymers are much weaker than in the GC scaffold specimens, due to less polymer volume. Typical P O stretching mode of the phosphate group appears at 1063 cm−1 , the bending modes are found at 602, and 561 cm−1 ; bending vibrations of P O H appear at 897 cm−1 [19]. The FTIR spectrum of the GC-CaP composite is similar to the GC spectrum. The bands at 1072 cm−1 correspond to P O stretching. This result reveals that the CaP is incorporated into the polymer sponge [34]. FTIR spectrum of the Janus supported GC in absence and presence of CaP particles are provided in Fig. 5B. The bending and stretching groups of the polymers can be identified in the spectra. Differences, sharper peaks around 2900 cm−1 can be attributed to −CH3 stretching vibration of the SiO; an extra band appears at 1750 cm−1 , this is characteristic of the carbonyl group (C O) of esters, attributed to the vegetable oil [35]. In addition, absorption bands at 1250 cm−1 , and at 790 cm−1 are explained as stretching group of Si CH3 . This result indicates oil rests in the samples [36]. 3.3. Thermal analysis Thermal stability of the GC blend scaffold and the GC-CaP composite scaffold was analyzed and compared to that of the freeze-dried pure polymers. Thermal analysis of biopolymer scaffolds commonly is carried out by means of DSC in combination with TGA, however, preparation of most pharmaceutical products with biopolymers, e.g., chitosan, usually do not involve heating above 100 ◦ C. Therefore, we present a detailed investigation by using a high sensitive differential scanning micro-calorimeter. Fig. 6 shows the DSC heat curves of freeze-dried pure gelatin and chitosan (A), and their freeze-dried blend and the blend composite (B). The specimens, except the pure chitosan exhibited a broad endothermic peak at different positions ranging from about 60 ◦ C up to 95 ◦ C. The endothermic peak of the gelatin observed at 63 ± 0.2 ◦ C (compare Fig. 6) can be related to the triple helix random coil transition, according to Fernandes et al. [37]. Chitosan shows a straight line which starts to rise at 95 ◦ C, this can be attributed to bound water evaporation. Unfortunately, we could not follow further changings, due to the heating limitation of the instrument.
3.2. FTIR characterisation Fourier Transform Infrared Spectroscopy (FTIR) is a valuable technique for characterizing organic and certain inorganic materials. It can provide fast information about the molecular structure, bond behavior and phase purity and was used to gain information about the chemical composition of final samples. FTIR spectrum of the pure freeze-dried GC blend is included in Fig. 5A. The broad peak between 3600 and 3100 cm−1 is caused by the stretching vibration O H and/or N H; the absorption peak occurs at 2934 cm−1 is a stretching vibration of CH2 originated from pyranose ring or C H stretching in CH3 . Strong bands attributed
Fig. 6. DSC thermograms of freeze-dried gelatin and chitosan solutions (A), GC blend and GC-CaP composite scaffold (B).
352
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Table 1 Weight of the different freeze-dried scaffolds (including standard deviation) in comparison to the scaffolds swollen by ethanol and water. Wdry (g) GC GC-CaP JE GC JE GC-CaP
0.018 0.022 0.164 0.141
± ± ± ±
Wethanol (g) 0.001 0.003 0.007 0.035
0.273 0.224 0.310 0.298
± ± ± ±
Wwater (g)
0.062 0.033 0.044 0.043
0.426 0.417 0.381 0.368
± ± ± ±
0.039 0.096 0.007 0.035
112 ◦ C,
Silva et al. found a water evaporation peak at and the transition temperature exhibit at 243 ◦ C [38]. In the gelatin-chitosan blend an endothermic peak appears, which is shifted to 69.8 ± 0.3 ◦ C. A similar effect was observed in a gelatin-cellulose-soy protein film by Li et al., where the endothermic peak of the complex film is shifted to higher temperature in comparison to the single components [39]. These results indicate the existence of H-bonding forces between chitosan and gelatin. The thermal stability of the GC was further improved by adding CaP particles, the endothermic peak of the GC-CaP composite shifts now to 84.5 ± 2.1 ◦ C and became broader. The associated enthalpy value of the GC is 3.53 ± 0.11 Jg−1 , and of the GC-CaP composite is 6.47 ± 0.25 Jg−1 . Unfortunately, our Janus supported samples show no detectable changes in the DSC curves because of the oil residues. 3.4. Swelling behavior All scaffolds used for the swelling test have a similar cylinder shape (r ≈ 0.4 h ≈ 1 cm). The dissimilarity of the dry weights is perceptible between the scaffolds in absence and presence of Janus droplets. Therefore, the ethanol and water absorption given in grams and the swelling ratio and the porosity were not calculated. The GC scaffold has the highest ethanol and water absorption capacity (compare Table 1), and the size of the scaffold was expanded greatly, as to be seen in Fig. S2. GC-CaP composite scaffolds swell less because of their compact structure. Janus templated scaffolds still contain oil at the beginning. After the purification process the water absorption capacity was approximated to the GC scaffolds, but the size of the scaffolds was not drastically expanded, which can be explained with the higher surface area. 4. Conclusions Gelatin-chitosan composite scaffolds were successfully fabricated by a drying technique, and nano-porous calcium phosphate balls of about 10 m in size were incorporated, resulting in GC scaffolds with smaller pores and higher temperature tolerance. The approach to use GC stabilized Janus emulsions, was also successful. In contrast to single emulsions with one oil component, i.e., olive oil or silicone oil, which are not stable during the freezedrying scaffold preparation process, engulfed olive oil/silicone oil Janus emulsions can be used as a template phase to build up a GC composite scaffold with significant smaller pores. Although, the scaffolds need to be purified after the freeze-drying process, because the scaffolds are not completely oil free, further work must focus on this problem. Authors contribution I.K. and J.K. designed the experiments. I.K. performed the experimental work. J.R. was responsible for the DSC measurement. I.K., J.K., and S.E.F. drafted the manuscript. Each author read and approved the final manuscript.
Conflict of interest The authors declare that they have no competing interest. The author retains the right to include the journal article, in full or in part, in a thesis or dissertation. Acknowledgements The authors are thankful to Dr. Brigitte Tiersch and Sibylle Rüstig for the scanning electron microscopy images. SEF thanks Ugelstad Laboratory for support. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb.2016.05. 018. References [1] L.S. Nair, C.T. Laurencin, Prog. Polym. Sci 32 (2007) 762. [2] B. Dhandayuthapani, Y. Yoshida, T. Maekawa, D.S. Kumar, Int. J. Polym. Sci. (2011), http://dx.doi.org/10.1155/2011/290602. [3] Y. Yin, Z. Li, Y. Sun, K. Yao, J. Mat. Sci. Lett. 40 (17) (2005) 4649. [4] E. Chiono, G. Pulieri, G. Vozzi, A. Ciardelli, P. Ahluwalia, Giusti, J. Mater. Sci. Med. 19 (2) (2008) 889. [5] E. Pulieri, V. Chiono, G. Ciardelli, G. Vozzi, A. Ahluwalia, C. Domenici, F. Vozzi, P. Giusti, J. Biomed. Mater. Res. A 86 (2) (2008) 311. [6] R. Jayakumar, M. Prabaharan, S.V. Nair, H. Tamura, Biotechnol. Adv. 28 (1) (2010) 142. [7] J. Zhang, W. Xia, P. Liu, Q. Cheng, T. Tahirou, W. Gu, B. Li, Mar. Drugs 8 (7) (2010) 1962. [8] M. Dash, F. Chiellini, R.M. Ottenbrite, E. Chiellini, Prog. Polym. Sci. 36 (2011) 981. [9] F. Croisier, C. Jérôme, Eur. Polym. J. 49 (4) (2013) 780. [10] S.V. Dorozhkin, Biomatter 1 (1) (2011) 3. [11] S.V. Dorozhkin, Materials 6 (2013) 3840. [12] Y.M. Lee, Y.-J. Park, S.-J. Lee, Y. Ku, S.-B. Han, S.-M. Choi, P.R. Klokkevold, C.P. Chung, J. Periodontol. 71 (2000) 410. [13] C.E. Tanase, M.I. Popa, L. Verestiuc, Mater. Lett. 65 (2011) 1681. [14] N. Roveri, M. Iafisco, Nanotechnol. Sci. Appl. 3 (2010) 107. [15] J. Wang, Z. Shi, S.- Li, H. Zhang, Z. Wu, C. Jiang, C. Yang, Tian, ACS Appl. Mater. Interfaces 6 (16) (2014) 14522. [16] J. Li, D. Zhu, J. Yin, Y. Liu, F. Yao, K. Yao, Mater. Sci. Eng. C 30 (2010) 795. [17] N. Zhu, X. Chen, in: Rosario Pignatello (Ed.), Advances in Biomaterials Science and Biomedical Applications, InTech, 2013, http://dx.doi.org/10.5772/53461, ISBN: 9535110514 9789535110514 (Chapter 12). [18] S.K.L. Levengood, M. Zhang, J. Mater. Chem. B 2 (2014) 3161. [19] S. Kovach, C. Kosmella, C. Prietzel, J. Bagdahn, Koetz, Colloids Surf. B 132 (2015) 246. [20] T. Nisisako, S. Okushima, T. Torii, Soft Matter 1 (2005) 23. [21] N. Pannacci, H. Bruus, D. Bartolo, I. Etchart, T. Lockhart, Y. Hennequin, H. Willaime, P.N. Tabeling, Phys. Rev. Lett. 101 (2008) 164502. [22] J. Guzowski, P.M. Korczyk, S. Jakiela, P. Garstecki, Soft Matter 8 (2012) 7269. [23] M.J. Neeson, D.Y.C. Chan, R.F. Tabor, F. Grieser, R.R. Dagastine, Soft Matter 8 (2012) 11042. [24] L. Ge, W. Shao, S. Lu, R. Guo, Soft Matter 10 (2014) 4498. [25] H. Hasinovic, S.E. Friberg, Langmuir 27 (2011) 6584. [26] H. Hasinovic, S.E. Friberg, I. Kovach, J. Koetz, Colloid Polym. Sci. 292 (2014) 2319–2324. [27] S.E. Friberg, I. Kovach, J. Koetz, ChemPhysChem 14 (2013) 3772. [28] J. Kovach, S.E. Koetz, Friberg, Colloids Surf. A 441 (2014) 66–71. [29] J. Kovach, S.E. Won, J. Friberg, Koetz, Colloid Polym. Sci. 294 (2016) 705. [30] N. Nwe, T. Furuike, H. Tamura, Materials 2 (2009) 374. [31] S.V. Dorozhkin, J. Mater. Sci. M 24 (2013) 1335. [32] N. Sagar, V.P. Soni, J.R. Bellare, J. Biomed. Mater. Res. B 10 (3) (2012) 624. [33] C. Isikli, V. Hasirci, N. Hasirci, J. Tissue Eng. Regener. Med. 6 (2) (2012) 135. [34] B. Li, Y. Wang, D. Jia, Y. Zhou, J. Biomater. Sci. 22 (2011) 505. [35] C.H. Kim, C.-K. Joo, H.J. Chun, B.R. Yoo, D. Il Noh, Y.B. Shim, Appl. Surf. Sci. 262 (2012) 146. [36] M.M. Mossoba, V. Milosevic, M. Milosevic, J.K. Kramer, H. Azizian, Anal. Bioanal. Chem. 389 (2007) 87. [37] F.M. Fernandes, I. Manjubala, E. Ruiz-Hitzky, Phys. Chem. Chem. Phys. 13 (11) (2011) 4901. [38] C.L. Silva, J.C. Pereira, A. Ramalho, A.A.C.C. Pais, J.J.S. Sousa, J. Membr. Sci. 320 (2008) 268. [39] C. Li, J. Luo, Z. Qin, H. Chen, Q. Gau, J. Li, RSC Adv. 5 (2015) 56518.
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Janus emulsion mediated porous scaffold bio-fabrication Ildiko Kovach a , Jens Rumschöttel a , Stig E. Friberg b , Joachim Koetz a,∗ a b
Institut für Chemie, Universität Potsdam, Karl-Liebknecht-Straße 24-25, D-14476 Potsdam, Germany Ugelstad Laboratory, NTNU, Trondheim, Norway
a r t i c l e
i n f o
Article history: Received 3 March 2016 Received in revised form 22 April 2016 Accepted 5 May 2016 Available online 10 May 2016 Keywords: Janus emulsions Calcium phosphates Gelatin-chitosan scaffolds,
a b s t r a c t A three dimensional biopolymer network structure with incorporated nano-porous calcium phosphate (CaP) balls was fabricated by using gelatin-chitosan (GC) polymer blend and GC stabilized olive/silicone oil Janus emulsions, respectively. The emulsions were freeze-dried, and the oil droplets were washed out in order to prepare porous scaffolds with larger surface area. The morphology, pore size, chemical composition, thermal and swelling behavior was studied by Scanning Electron Microscopy (SEM), Fourier Transform Infrared Spectroscopy (FTIR) and micro-Differential Scanning Calorimetry (micro-DSC). Microscopic analysis confirmed that the pore size of the GC based sponges after freeze-drying may be drastically reduced by using Janus emulsions. Besides, the incorporation of nanoporous calcium phosphate balls is also lowering the pore size and enhancing thermal stability. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The purpose of tissue engineering is to fabricate biocompatible and biodegradable scaffolds with similar mechanical properties of the targeted tissue for repair and replacement. Natural polymers possess biocompatibility, high porosity, biomechanical compatibility, and biodegradability. Consequently, they are considered as biodegradable biomaterials which are used clinically and are still attracting much attention for their application as material for tissue engineering [1]. Gelatin is one of the most popular amphiphilic biopolymer scaffold materials for tissue regeneration. Because gelatin is derived from natural collagen, which forms the extracellular matrix of the bone, it can provide a native like environment for the cells. The main disadvantage of gelatin as a tissue substitute is that gelatin is unstable in aqueous medium. To overcome this weakness gelatin can be associated with other naturally obtained polymers, in particular with chitosan [2]. Structure formation can be realized through electrostatic interactions (polyelectrolyte complex formation), physical binding (blends), or covalent cross-linking (network formation) [3–5]. Studies provide that the dissolution degree of the gelatinchitosan blends is reduced by increasing the chitosan concentration [5]. As well as gelatin, chitosan has qualified itself for tissue engineering, on account of its outstanding properties, namely high
∗ Corresponding author. E-mail address: [email protected] (J. Koetz). http://dx.doi.org/10.1016/j.colsurfb.2016.05.018 0927-7765/© 2016 Elsevier B.V. All rights reserved.
biodegradability, biocompatibility, non-toxicity and anti-microbial properties, osteoconductivity, hemostaticity and last but not least of its low costs [6,7]. Chitosan based scaffolds and gels have been studied extensively for wound healing, tissue engineering and tissue repair of bones, cartilage, liver, or nerve tissues [8,9]. The attention towards biopolymer hard tissue engineering has focused on composites. Since these materials are close to mimic the construction of natural bones, calcium orthophosphate crystals within a collagen fiber have been favored. In composite materials the high mechanical stability of ceramics is advantageously coupled with good elasticity of the polymers [10–12]. Calcium phosphate (CaP) chitosan composites are claimed to have outstanding osteoblast adhesion, migration, differentiation, and proliferation. Furthermore, the bioactivity and cell growth factor are outstanding [13]. Apart from scaffold design chitosan and gelatin are widely applied to influence the mineralization process of calcium orthophosphates, e.g., by constructing diverse hybrid structures [14]. Wang et al. precipitated flower like hybrid chitosan/hydroxyapatite composites under stirring [15]. Li et al. produced calcium phosphate nanoparticles in a matrix mediated synthesis and claimed that the chitosan/pectin network modulates nucleation and growth of the hydroxyapatite crystals [16]. In addition, there are various methods of preparing porous scaffolds using biopolymers, which include fiber bonding, melt molding solvent casting/particulate leaching, gas foaming and phase separation and freeze-drying, among others [17,18]. It is a challenge for tissue engineering to reconstruct certain tissues with appropriate pore size, interconnectivity and porosity. The
348
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
effect of the pore size on tissue regeneration was demonstrated by several studies, suggesting that the ideal pore size is: -
5 m for neovascularization, 5–15 m for fibroblast ingrowth, 20 m for hepatocytes ingrowth, 200–350 m for osteoconduction, 20–125 m for mammalian skin regeneration [2].
With a reference to our latest result, a supramolecular CaP cardhouse structure (of about 5–15 m in size) was synthesized in presence of gelatin/chitosan polymer blend. In other words, nanoporous calcium phosphate balls are formed, built up by individual flat CaP platelets [19]. In the present contribution, the freeze-drying method was applied to fabricate porous biopolymer ceramic scaffolds with gelatin-chitosan (GC) networks modified with hybrid calcium phosphate balls [19]. Our approach was to introduce the so called Janus emulsions as a template phase for controlling the surface area and pore size of the three-dimensional structure formed, followed by the freeze-dry technique to fabricate a 3D scaffold. Janus emulsions are dual emulsions of commonly known multiple emulsions, where two non-mixable oil components, e.g., silicone and vegetable oil, are combined in one drop. Introduced by Nisisako et al. [20], Janus emulsions immediately attracted extensive attention, because of the compelling correlation between interfacial tension and the drop topology [21–23]. Droplet topology can be controlled, e.g., in a one-step high energy mixing procedure [24]. Immiscible liquids control the morphology of patchy emulsions, but for the present contribution these emulsions were prepared by traditional vibration emulsification, according to Hasinovic et al. [25,26]. One criterion for the spontaneous Janus droplet formation, numerically evaluated [27] and experimentally determined [28], is that the interfacial tension of the more hydrophobic oil towards the aqueous phase is less than the sum of the two remaining interfacial tensions [27]. However, most relevant for the present publication are a new type of Janus emulsions, introduced by Kovach et al. [29], where the oil droplets are stabilized by means of biopolymers, namely gelatin (G) and chitosan (C). The authors claimed that in a first step the GC complexes were strongly adsorbed at the olive oil/water interface, before a rigid skin-like polymer composite layer is formed [29]. GC polymer blends can build up three dimensional networks and stabilize completely engulfed Janus emulsions, and can control the formation of supramolecular ordered CaP nanoporous card-house structures [29,19]. Taking this into account, the aim of the present research was to utilize CaP balls for
constructing porous composite scaffolds and applying Janus emulsions as a template phase to affect the pore size of the GC scaffold. 2. Experimental 2.1. Materials The low molecular weight chitosan with a degree of deacetylation of 81.2%, and a moisture content ≤12 wt% was obtained from Sigma-Aldrich® , and used without further treatment. Gelatin powder (Type A, isoelectric point ≈7, Bloom number 140) with a moisture content ≤11 wt% was purchased from Carl Roth® . The polymers were used without any further treatment. Silicone oil (SiO) (viscosity: 10–mPa s), olive oil (OO) and ethanol ≥99.5% were obtained from Sigma-Aldrich® . Recently, we were able to show that well defined supramolecular structured, spherical CaP card-house structures are formed in the presence of gelatin-chitosan blends in a wet chemical procedure at 90 ◦ C [19]. The resulting CaP powder (without calcification) was used here as given. All reagents were dissolved in Millipore Milli-Q deionized water. 2.2. Preparation of the gelatin-chitosan (GC) blend Chitosan powder (2.5 wt%) was dissolved in 0.1 mol acetic acid, and the solution was homogenized by stirring overnight. Gelatin (2.5 wt%) was dissolved in 0.1 mol acetic acid, under continuous stirring and heating up to 40 ◦ C for 5 min. The 2.5 wt% GC blend was prepared by mixing the 2.5 wt% chitosan solution with the 2.5 wt% gelatin solution, in the ratio of 1:1, under constant stirring at room temperature for 72 h. 2.3. Preparation of GC-CaP suspensions The nano-porous, ball-like calcium phosphate particles (with diameter between 5 and 15 m) synthesized according to Ref. [19], were added to the 2.5 wt% GC blend by stirring for 2 min, in order to obtain the GC-CaP suspension. The resulting GC-CaP suspension contains 0.5 wt% calcium phosphate. 2.4. Preparation of Janus GC and Janus GC-CaP emulsions The GC blend solution and GC-CaP suspension were used as the aqueous phase for preparing Janus emulsions. Each mixture consists of 0.15 g silicone oil, 0.15 g olive oil and 0.7 g of the GC solution or GC-CaP suspension. The emulsification with the oil components was made in a 2 ml Eppendorf tube by mixing with Minishaker IKA (Roth® ) at 2500 rpm in order to produce completely engulfed
Fig. 1. Micrograph of the GC blend Janus emulsion (A) and the corresponding GC-CaP Janus emulsion (B).
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Janus emulsions according to Ref. [28], that means by mixing the aqueous phase with the two oil phases in a one-step procedure. Fig. 1 shows the GC blend Janus emulsion (A) and the corresponding GC-CaP Janus emulsion (B), respectively. Emulsions are freshly prepared direct before the freeze-drying process. The emulsions were solidified by immersing into liquid nitrogen.
349
2.5. Scaffold fabrication by freeze-drying and characterization The aqueous GC blend and the Janus emulsions were submerged into liquid nitrogen, the end of the Eppendorf tubes were cut, and freeze-dried at 0 ◦ C for 12–48 h. The freeze-drying time depends on the nature of the samples. The morphology of the freeze-dried scaffolds was determined by means of scanning electron microscopy
Fig. 2. Photograph of GC (A), GC-CaP composite (B), Janus GC (C), Janus GC-CaP composite (D) after freeze-drying.
Fig. 3. SEM micrographs (cross section) of the freeze-dried scaffold materials: GC (A), GC-CaP composite (B), Janus GC (C), Janus GC-CaP composite (D).
350
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Fig. 4. SEM micrographs of GC-CaP composite at different magnification.
(S-4800 from Hitachi® ). The scaffold samples were sited on an aluminum sample holder and coated with platinum layer to make the sample conductive. Janus GC and Janus GC-CaP composite scaffolds were investigated in cryo-mode, because of the presence of oil rests. Firstly, the samples were cooled by plunging into liquid nitrogen, then freeze fractured at −180 ◦ C, etched for 45 s at −98 ◦ C, and sputtered with platinum in the GATAN® Alto 2500 cryo preparation chamber. Afterwards, the samples were placed into the microscope chamber. Pore sizes were measured by using the iTEM software from Olympus Soft Imaging Solutions (OSIS). Statistical data were carried out from 200 to 300 data points, using more than one micrograph. Due to the deformation of the pores the pore perimeter was measured instead of the diameter. The chemical composition of the scaffolds was investigated using attenuated total reflection (ATR) with NEXUS FTIR spectrometer (Thermo-Nicolet® , Diamond) with a scanning range of 400–4000 cm−1 and a resolution of 2 cm−1 at a scan rate of 32 scans/sample. ATR correction was done via Omnic 8.1.11 (Thermo Fisher Scientific Inc® ). DSC experiments were performed with a Setaram® Micro-DSC VII with a fixed heating and cooling rate of 1 ◦ C min−1 . About 4 mg of a freeze-dried sample (2.5 wt% gelatin and 2.5 wt% chitosan solution, GC blend, and GC-CaP composite) was cut to a form of the aluminum pan and placed inside the pan without compression and scanned between +25 ◦ C and +100 ◦ C. In order to get a better resolution, the heating curves are plotted between +50 ◦ C and +100 ◦ C (compare Fig. 5). Scaffold samples were measured against empty pan. Measurements were recorded in triplicate. Enthalpy is considered as the endothermic peak area, determined by Calisto Processing Thermal Analysis Software. 2.6. Swelling tests Cylinder shaped freeze-dried scaffold samples were weighed (Wdry ) and immersed in glacial ethanol. The samples were ultrasound treated for 10 min. After the samples were removed from the ethanol, the ethanol wet weight (Wethanol ) of the scaffolds was measured. Then the ethanol was replaced with ethanol water mixture series (90%, 80%, 70%, 60%, 50%) each for 30 min and finally with distillated water, near gentle shaking. After 24 h immersion in water at room temperature the wet water weight (Wwater ) was measured. All experiments were performed in triplicate and the results were expressed as an average value.
Fig. 5. Fourier Transform Infrared Spectra of GC, CaP, GC-CaP composite (A); Janus GC and Janus GC-CaP composite (B).
3. Results and discussion 3.1. Morphological characterization All scaffolds are soft sponge-like (Fig. 2), except the GC-CaP composite which under dry condition is brittle and less flexible than the other scaffolds. This means that the small addition of CaP significantly changes the mechanical properties of the polymer. Janus emulsion supported scaffolds are pale yellow colored (Fig. 2C, D), indicating that the scaffolds contain olive oil residues. The GC polymer blend adsorbs rapidly on the olive oil surface in contrast to the case of silicone oil [29]. Emulsions can be formed by using the oils, separately. In both cases the size of the drops decreases. Olive oil emulsions are stable for a longer time, but not yet silicon oil emulsions. In presence of GC blends completely engulfed Janus drops exist, which means that the silicon drops are trapped in the olive oil, i.e., an olive layer covers the silicon droplet. When this layer breaks the Janus drop turns to be partially engulfed. Silicon oil GC emulsions are destabilizing faster than olive oil and/or Janus emulsions, as to be seen experimentally (not shown here). However, scaffold fabrication failed by using GC stabilized emulsions with only one oil component, i.e., olive oil or silicone oil. In presence of olive oil, dried substance remains in clumps with a deep
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
yellow color, without stable frame. Because the droplet of the olive oil is much smaller than the Janus droplet and during the drying process it impregnates the polymer surface. Attempts with silicon oil alone were not successful, silicon oil flew out in the very beginning of the drying process. By this movement bigger pores and a room inside of the scaffold is formed. The advantage of applying Janus droplets is that the two oils hinder each other during the preparation process and porous scaffolds can be prepared. Fig. 3A shows homogenous elongated pore structures with a heterogeneous pore size. The average pore size (perimeter) of the GC scaffold is 720 ± 540 m, but the pores are isolated from each other. Nwe et al. [30] show similar elongated pore structures, which are typical for shrimps´ı shell chitosan freeze-dried scaffolds. SEM micrographs of the freeze-dried 2.5 wt% polymer solutions alone (compare Fig. S1) display a typical lamellar morphology, along the freezing direction in both cases. GC-CaP composite scaffolds, presented in Fig. 3B, show a drastically changed morphology. The size of the average pore perimeter had decreased to 14 ± 11 m, and the pore interconnectivity was highly increased. It has to be mentioned here that the interconnected geometry of the scaffolds is very important for the cell growth, to insure that the cells are supplied with oxygen and nutrients. The structure of individual calcium phosphate particles inside the scaffold will be discussed in more detail below. The remaining oil residues in the Janus supported scaffolds are disturbing the vacuum of the microscopy chamber, therefore, the micrographs were scanned in cryo-mode. The surface of Janus GC scaffolds is visibly dissimilar from that of the GC scaffolds. Janus supported GC scaffolds present a lamellar network structure with interconnected pores (Fig. 3C). The average pore size is 77 ± 33 m. However, the SEM micrograph of Janus GC-CaP composite scaffolds (Fig. 3D) still contain oil residues, making the CaP observation difficult. Some of the CaP particles were found, but are covered with oil. The average pore size is 238 ± 103 m. The higher pore size range can be explained with the presence of CaP balls preventing the smaller Janus droplet formation during the emulsification process. For a more comprehensive characterization of the incorporated CaP in the GC network additional SEM micrographs at higher magnification were performed. In Fig. 4 one can see supramolecular arranged calcium phosphate balls inside of the GC network, the CaP nano-porous aggregates of about 10 m in size are build up by individual thin platelets. The nano-porous CaP balls are only weakly embedded into the polymer network as to be seen in Fig. 4. Based on the fact that the calcium phosphate aggregates have the same electrical charge as the network (Zeta potential: +15 ± 2 mV [19]), one may assume that electrostatic repulsion forces hinder a stronger binding between the network and the CaP nano-balls. The supramolecular structure of the crystal balls is arranged as a card house, built up by individual dicalcium phosphate dihydrate (DCPD) platelets in a metastable calcium orthophosphate form. Dorozhkin recently discussed that DCPD efficiently resorbs under physiological conditions, and thereby can support bone remineralization [31].
351
to CO stretching vibration of amid I. at 1660 cm−1 and amid II. at 1559 cm−1 . In addition, absorption peak at 1650 cm−1 is an N H bending group of chitosan amid I. The bands at 1410, 1311 and 1260, 1140 cm−1 , are assigned to CH2 bending; 1140 and 1074 cm−1 to C O C stretching were confirmed [32,33,13]. GC-CaP composites (Fig. 5A) show the characteristic groups of phosphate and polymers. However, the absorption peaks attributed to the polymers are much weaker than in the GC scaffold specimens, due to less polymer volume. Typical P O stretching mode of the phosphate group appears at 1063 cm−1 , the bending modes are found at 602, and 561 cm−1 ; bending vibrations of P O H appear at 897 cm−1 [19]. The FTIR spectrum of the GC-CaP composite is similar to the GC spectrum. The bands at 1072 cm−1 correspond to P O stretching. This result reveals that the CaP is incorporated into the polymer sponge [34]. FTIR spectrum of the Janus supported GC in absence and presence of CaP particles are provided in Fig. 5B. The bending and stretching groups of the polymers can be identified in the spectra. Differences, sharper peaks around 2900 cm−1 can be attributed to −CH3 stretching vibration of the SiO; an extra band appears at 1750 cm−1 , this is characteristic of the carbonyl group (C O) of esters, attributed to the vegetable oil [35]. In addition, absorption bands at 1250 cm−1 , and at 790 cm−1 are explained as stretching group of Si CH3 . This result indicates oil rests in the samples [36]. 3.3. Thermal analysis Thermal stability of the GC blend scaffold and the GC-CaP composite scaffold was analyzed and compared to that of the freeze-dried pure polymers. Thermal analysis of biopolymer scaffolds commonly is carried out by means of DSC in combination with TGA, however, preparation of most pharmaceutical products with biopolymers, e.g., chitosan, usually do not involve heating above 100 ◦ C. Therefore, we present a detailed investigation by using a high sensitive differential scanning micro-calorimeter. Fig. 6 shows the DSC heat curves of freeze-dried pure gelatin and chitosan (A), and their freeze-dried blend and the blend composite (B). The specimens, except the pure chitosan exhibited a broad endothermic peak at different positions ranging from about 60 ◦ C up to 95 ◦ C. The endothermic peak of the gelatin observed at 63 ± 0.2 ◦ C (compare Fig. 6) can be related to the triple helix random coil transition, according to Fernandes et al. [37]. Chitosan shows a straight line which starts to rise at 95 ◦ C, this can be attributed to bound water evaporation. Unfortunately, we could not follow further changings, due to the heating limitation of the instrument.
3.2. FTIR characterisation Fourier Transform Infrared Spectroscopy (FTIR) is a valuable technique for characterizing organic and certain inorganic materials. It can provide fast information about the molecular structure, bond behavior and phase purity and was used to gain information about the chemical composition of final samples. FTIR spectrum of the pure freeze-dried GC blend is included in Fig. 5A. The broad peak between 3600 and 3100 cm−1 is caused by the stretching vibration O H and/or N H; the absorption peak occurs at 2934 cm−1 is a stretching vibration of CH2 originated from pyranose ring or C H stretching in CH3 . Strong bands attributed
Fig. 6. DSC thermograms of freeze-dried gelatin and chitosan solutions (A), GC blend and GC-CaP composite scaffold (B).
352
I. Kovach et al. / Colloids and Surfaces B: Biointerfaces 145 (2016) 347–352
Table 1 Weight of the different freeze-dried scaffolds (including standard deviation) in comparison to the scaffolds swollen by ethanol and water. Wdry (g) GC GC-CaP JE GC JE GC-CaP
0.018 0.022 0.164 0.141
± ± ± ±
Wethanol (g) 0.001 0.003 0.007 0.035
0.273 0.224 0.310 0.298
± ± ± ±
Wwater (g)
0.062 0.033 0.044 0.043
0.426 0.417 0.381 0.368
± ± ± ±
0.039 0.096 0.007 0.035
112 ◦ C,
Silva et al. found a water evaporation peak at and the transition temperature exhibit at 243 ◦ C [38]. In the gelatin-chitosan blend an endothermic peak appears, which is shifted to 69.8 ± 0.3 ◦ C. A similar effect was observed in a gelatin-cellulose-soy protein film by Li et al., where the endothermic peak of the complex film is shifted to higher temperature in comparison to the single components [39]. These results indicate the existence of H-bonding forces between chitosan and gelatin. The thermal stability of the GC was further improved by adding CaP particles, the endothermic peak of the GC-CaP composite shifts now to 84.5 ± 2.1 ◦ C and became broader. The associated enthalpy value of the GC is 3.53 ± 0.11 Jg−1 , and of the GC-CaP composite is 6.47 ± 0.25 Jg−1 . Unfortunately, our Janus supported samples show no detectable changes in the DSC curves because of the oil residues. 3.4. Swelling behavior All scaffolds used for the swelling test have a similar cylinder shape (r ≈ 0.4 h ≈ 1 cm). The dissimilarity of the dry weights is perceptible between the scaffolds in absence and presence of Janus droplets. Therefore, the ethanol and water absorption given in grams and the swelling ratio and the porosity were not calculated. The GC scaffold has the highest ethanol and water absorption capacity (compare Table 1), and the size of the scaffold was expanded greatly, as to be seen in Fig. S2. GC-CaP composite scaffolds swell less because of their compact structure. Janus templated scaffolds still contain oil at the beginning. After the purification process the water absorption capacity was approximated to the GC scaffolds, but the size of the scaffolds was not drastically expanded, which can be explained with the higher surface area. 4. Conclusions Gelatin-chitosan composite scaffolds were successfully fabricated by a drying technique, and nano-porous calcium phosphate balls of about 10 m in size were incorporated, resulting in GC scaffolds with smaller pores and higher temperature tolerance. The approach to use GC stabilized Janus emulsions, was also successful. In contrast to single emulsions with one oil component, i.e., olive oil or silicone oil, which are not stable during the freezedrying scaffold preparation process, engulfed olive oil/silicone oil Janus emulsions can be used as a template phase to build up a GC composite scaffold with significant smaller pores. Although, the scaffolds need to be purified after the freeze-drying process, because the scaffolds are not completely oil free, further work must focus on this problem. Authors contribution I.K. and J.K. designed the experiments. I.K. performed the experimental work. J.R. was responsible for the DSC measurement. I.K., J.K., and S.E.F. drafted the manuscript. Each author read and approved the final manuscript.
Conflict of interest The authors declare that they have no competing interest. The author retains the right to include the journal article, in full or in part, in a thesis or dissertation. Acknowledgements The authors are thankful to Dr. Brigitte Tiersch and Sibylle Rüstig for the scanning electron microscopy images. SEF thanks Ugelstad Laboratory for support. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb.2016.05. 018. References [1] L.S. Nair, C.T. Laurencin, Prog. Polym. Sci 32 (2007) 762. [2] B. Dhandayuthapani, Y. Yoshida, T. Maekawa, D.S. Kumar, Int. J. Polym. Sci. (2011), http://dx.doi.org/10.1155/2011/290602. [3] Y. Yin, Z. Li, Y. Sun, K. Yao, J. Mat. Sci. Lett. 40 (17) (2005) 4649. [4] E. Chiono, G. Pulieri, G. Vozzi, A. Ciardelli, P. Ahluwalia, Giusti, J. Mater. Sci. Med. 19 (2) (2008) 889. [5] E. Pulieri, V. Chiono, G. Ciardelli, G. Vozzi, A. Ahluwalia, C. Domenici, F. Vozzi, P. Giusti, J. Biomed. Mater. Res. A 86 (2) (2008) 311. [6] R. Jayakumar, M. Prabaharan, S.V. Nair, H. Tamura, Biotechnol. Adv. 28 (1) (2010) 142. [7] J. Zhang, W. Xia, P. Liu, Q. Cheng, T. Tahirou, W. Gu, B. Li, Mar. Drugs 8 (7) (2010) 1962. [8] M. Dash, F. Chiellini, R.M. Ottenbrite, E. Chiellini, Prog. Polym. Sci. 36 (2011) 981. [9] F. Croisier, C. Jérôme, Eur. Polym. J. 49 (4) (2013) 780. [10] S.V. Dorozhkin, Biomatter 1 (1) (2011) 3. [11] S.V. Dorozhkin, Materials 6 (2013) 3840. [12] Y.M. Lee, Y.-J. Park, S.-J. Lee, Y. Ku, S.-B. Han, S.-M. Choi, P.R. Klokkevold, C.P. Chung, J. Periodontol. 71 (2000) 410. [13] C.E. Tanase, M.I. Popa, L. Verestiuc, Mater. Lett. 65 (2011) 1681. [14] N. Roveri, M. Iafisco, Nanotechnol. Sci. Appl. 3 (2010) 107. [15] J. Wang, Z. Shi, S.- Li, H. Zhang, Z. Wu, C. Jiang, C. Yang, Tian, ACS Appl. Mater. Interfaces 6 (16) (2014) 14522. [16] J. Li, D. Zhu, J. Yin, Y. Liu, F. Yao, K. Yao, Mater. Sci. Eng. C 30 (2010) 795. [17] N. Zhu, X. Chen, in: Rosario Pignatello (Ed.), Advances in Biomaterials Science and Biomedical Applications, InTech, 2013, http://dx.doi.org/10.5772/53461, ISBN: 9535110514 9789535110514 (Chapter 12). [18] S.K.L. Levengood, M. Zhang, J. Mater. Chem. B 2 (2014) 3161. [19] S. Kovach, C. Kosmella, C. Prietzel, J. Bagdahn, Koetz, Colloids Surf. B 132 (2015) 246. [20] T. Nisisako, S. Okushima, T. Torii, Soft Matter 1 (2005) 23. [21] N. Pannacci, H. Bruus, D. Bartolo, I. Etchart, T. Lockhart, Y. Hennequin, H. Willaime, P.N. Tabeling, Phys. Rev. Lett. 101 (2008) 164502. [22] J. Guzowski, P.M. Korczyk, S. Jakiela, P. Garstecki, Soft Matter 8 (2012) 7269. [23] M.J. Neeson, D.Y.C. Chan, R.F. Tabor, F. Grieser, R.R. Dagastine, Soft Matter 8 (2012) 11042. [24] L. Ge, W. Shao, S. Lu, R. Guo, Soft Matter 10 (2014) 4498. [25] H. Hasinovic, S.E. Friberg, Langmuir 27 (2011) 6584. [26] H. Hasinovic, S.E. Friberg, I. Kovach, J. Koetz, Colloid Polym. Sci. 292 (2014) 2319–2324. [27] S.E. Friberg, I. Kovach, J. Koetz, ChemPhysChem 14 (2013) 3772. [28] J. Kovach, S.E. Koetz, Friberg, Colloids Surf. A 441 (2014) 66–71. [29] J. Kovach, S.E. Won, J. Friberg, Koetz, Colloid Polym. Sci. 294 (2016) 705. [30] N. Nwe, T. Furuike, H. Tamura, Materials 2 (2009) 374. [31] S.V. Dorozhkin, J. Mater. Sci. M 24 (2013) 1335. [32] N. Sagar, V.P. Soni, J.R. Bellare, J. Biomed. Mater. Res. B 10 (3) (2012) 624. [33] C. Isikli, V. Hasirci, N. Hasirci, J. Tissue Eng. Regener. Med. 6 (2) (2012) 135. [34] B. Li, Y. Wang, D. Jia, Y. Zhou, J. Biomater. Sci. 22 (2011) 505. [35] C.H. Kim, C.-K. Joo, H.J. Chun, B.R. Yoo, D. Il Noh, Y.B. Shim, Appl. Surf. Sci. 262 (2012) 146. [36] M.M. Mossoba, V. Milosevic, M. Milosevic, J.K. Kramer, H. Azizian, Anal. Bioanal. Chem. 389 (2007) 87. [37] F.M. Fernandes, I. Manjubala, E. Ruiz-Hitzky, Phys. Chem. Chem. Phys. 13 (11) (2011) 4901. [38] C.L. Silva, J.C. Pereira, A. Ramalho, A.A.C.C. Pais, J.J.S. Sousa, J. Membr. Sci. 320 (2008) 268. [39] C. Li, J. Luo, Z. Qin, H. Chen, Q. Gau, J. Li, RSC Adv. 5 (2015) 56518.