1 Introduction

A composite material consists of at least two chemically identified phases which are separated by interfaces. The properties of composites including mechanical properties are strongly influenced by a number of factors, e.g., filler shape, size and size distribution, properties and volume percentage of filler, matrix properties (e.g., molecular weight of polymer), dispersion of filler particles in the polymer matrix, and the state of the filler/matrix interface [1]. In the case of biocomposites, other factors such as biocompatibility of both the filler and matrix, the degradation rate of matrix and nontoxicity must be considered [2].

Nanofibrous materials containing different supplements are common examples of composites. Nanofibers are manufactured by various methods. There are drawing [3], template synthesis [4, 5], phase separation [6], self-assembly [7, 8], electrospinning [9], and melt-blown spinning [10]. Many of these methods have advantages for specific fiber formations, but are limited from an industrial processing standpoint.

Among them, electrospinning is the most promising method for mass production of one-by-one continuous nanofibers from various polymers due to its versatility, flexibility, and simplicity of fiber fabricating [9]. For medical application of nanofibrous materials is believed to be beneficial thanks to their unique characteristics, these being a high porosity, small pore size, and high specific surface area [9]. It is possible to change the physical properties of nanofibers based on polymers by embedding inorganic particles. Some approaches were usually used to improve mechanical properties of nanofibers materials. For instance, Wang et al. have applied titania to enhance the mechanical properties of fibers based on poly (lactic-co-glycolic acid). Their investigation indicated that the composite microsphere-sintered fibers are a promising scaffold for bone repair [11]. La Gatta et al. reported an injectable poly (caprolactone) (PCL)/ calcium sulfate (CS) system for bone regeneration. They mixed calcium sulfate hemihydrate (CHS) with a photo-cross linkable derivative of PCL, and the composite is proposed to be utilized in hard tissue repair particularly in the dental and orthopedic field [12]. Besides titania, tricalcium phosphate (TCP) and whisker (CS), hydroxyapatite (HA), and calcium carbonate (CaCO3) as a biodegradable inorganic materials are widely used in bone treatment as well. Hydroxyapatite (HA) ceramics have been widely applied as bone substitutes. Together with ß-tricalcium phosphate, they have been for nearly three decades the most extensively used substitution materials for artificial bone grafts [13]. Their chemical compositions are close to the mineral phase of bones and are an origin of their excellent biocompatibility to tissue bone. This meets the requirement of any materials designed for bone repair and augmentation [14]. Seema A et al. fabricated poly-L-lactide acid (PLLA)/hydroxyapatite (HA) hybrid membranes by electrospinning and studied the effect of HA filler on the Young’s modulus and tensile strength. They observed that HA nanoparticles were not only dispersed in PLLA but also reacted with the functional group of PLLA, resulting in strong surface bonding and high tensile strength of hybrid membrane. The cell adhesion and growth on the PLLA/HA hybrid membrane were far better than those on the pure PLLA membrane, which proves that the PLLA/HA hybrid membrane can be one of the promising biomaterials for bone tissue regeneration [15].

On the other hand, CaCO3 is one of the most common and inexpensive inorganic fillers that has been used in the nanocomposite preparation process. Fine grade nano-sized precipitated CaCO3 (NPCC) has the potential to be important functional filler in polymer systems such as polypropylene composites, poly (vinyl chloride) composites, and poly (ethylene terephthalate) composites that are being produced [16–19]. Extensive studies on the reinforcing effect of nano- and micro-sized CaCO3 particles have been carried out in different polymer matrices, such as high density polyethylene (HDPE), nylon, polypropylene (PP), polylactide, acrylonitrile butadiene styrene (ABS), and thermoplastic polyurethane (TPU). Using nano-sized CaCO3 particles significantly improved the mechanical properties as well as the surface smoothness of nanocomposites [20–23].

Polycaprolactone (PCL) is a type of synthetic, biodegradable polymer, applying in food packaging, tissue engineering, dressing for wound, and drug delivery [24–29]. It becomes one of the most promising biodegradable polymers currently available on the market. However, some shortcomings such as high costs, low melting temperature, and low mechanical properties restrict widespread industrial use of PCL. Electrospinning of PCL, PCL/HA, and CaCO3 has been widely investigated, the results show wide divergence from each other and some with clear contras. Patcharaporn W et al. [30] used electrospinning to fabricate novel bone-scaffolding materials from electro-spun mats of polycaprolactone filled with either calcium carbonate or hydroxyapatite nanoparticles. HA was successfully synthesized by the hydrolysis method, with the average particle size (based on visual observation) being about 230 nm. The observed increasing in the diameters of the as-spun fibers with the addition and increasing amounts of microparticles (0.5 and 1.0 % HA or CaCO3) fillers was probably responsible for the observed improvement in the tensile strength of the obtained fiber mats. These electro-spun fiber mats (i.e., PCL-FS for the neat electro-spun PCL fiber mat and PCL/HAp-FS for the electro-spun PCL fiber mat containing HAp nanoparticles) were used as scaffolding materials for the culture of mouse calvaria-derived, pre-osteoblastic cells, MC3T3-E1 [30]. Sanaz A et al. reported the effects of calcium carbonate (CaCO3) nanoparticles on the mechanical and thermal properties and surface morphology of polycaprolactone (PCL)/chitosan nanocomposites using a melt blending technique. Results revealed mechanical properties improved with the addition of up to 1 wt% of nano-sized CaCO3, and the thermal stability was best enhanced at 1 wt% of CaCO3 nanoparticles loading [31].

Although PCL polymer is widely used as implants, such as vascular prostheses, sutures, bone repairing etc. [32], PCL polymer has poor mechanical properties than bone. But the possibility to be mechanically strengthened and to be biodegradable makes it very promising as a candidate for bone replacement. The improvement of the mechanical properties of polymer can be achieved by either the modification of the structure of the polymer, or the strengthening of the polymer with fiber and/or filler as HA or CaCO3 particles. So, the purpose of using HA and CaCO3 particles in the PCL polymer matrix is to improve the mechanical properties such as the elastic modulus, to improve the bioactivity or bone-bonding properties and supply of Ca/P in the long run. Thus, the main goal of this research is to improve the mechanical of PCL composites by adding different mass fraction of CaCO3 and HA microparticles as a filler. An empirical equation to predict Young’s modulus of composite as a function of mass fraction is used to support the experimental data. The equation can be used for obtaining parameters of fibers at any filler concentrations. It can be used for obtaining material with defined mechanical parameters.

2 Materials and Methods

2.1 Materials

PCL (Mn 70,000–90,000) were supplied by Sigma-Aldrich. Also, the solvents dimethylformamide (DMF) (chemically pure) and dichloromethane (DCM) (chemically pure) were obtained from local suppliers. Calcium carbonate microparticles (average primary particle size ∼1.2 ± 0.1 μm) were synthesized using sodium carbonate (Na2CO3) and calcium chloride (CaCl2) that were purchased from Sigma-Aldrich-Fluka and used without any additional purification. Hydroxyapatite microparticles (average primary particle size ∼1.4 ± 0.5 μm) were prepared by ion exchange reaction using water solution of Na3PO4. Na3PO4 were purchased from Sigma-Aldrich-Fluka and used also without any additional purification. In all experiments, ultra pure water was used (Aquarias, Russia; Millipore Milli-Q, USA and Canada).

2.2 Preparation of Calcium Carbonate Particles

In order to produce micron and submicron HA particles, the calcium carbonate templates were first synthesized. CaCl2 and Na2CO3 were set at 1 M. Volumes of 9.8 ml of Na2CO3 and CaCl2 were dissolved in 20 ml water. The solution was stirred with ultrasonic treatment for 30 s. For sonication, an ultrasonic processor Bandelin SONOPULS HD 2070 was used, which operates at 20 kHz frequency and an acoustic power density of 1 W/cm2. The synthesized CaCO3 particles were carefully washed two times in water and one more time in ethanol in the glass filter and finally dried for 1 h at 80 °C. The ionic exchange reaction was used to prepare the hydroxyapatite particles; 2.5 mg of the prepared CaCO3 powder was dispersed in 1 ml of 1 M Na3PO4 for 24 h. After this step, the particles were washed in water for five times and dried for 1 h at 80 °C.

2.3 Preparation of PCL/HA, CaCO3 Composites

The polymer solutions for electrospinning were prepared by dissolving a certain amount of PCL pellets (9 wt%) in the mix of dimethylformamide (DMF)/dichloromethane (DCM) = 2/8 (v/v) at room temperature during 3 h. The HA and CaCO3 particles at different mass ratios, 2.5, 5, and 7.5 % w/w, were added to PCL solutions previously prepared. Prepared solutions were placed in a syringe whose needle inner diameter size was 0.21 mm. Feeding rate of solutions was fixed to 1.0 ml/h by syringe pump. The high voltage at 20 kV was used for electrospinning at room temperature and 30–40 % humidity condition. Distance between tip of the needle and ground collector plate was fixed to 130 mm. The rotating aluminum drum was used as the collector. The fabricated samples were dried one night at room temperature. Figure 1 shows a schematic of the fabrication and characterization process. The first step (1) is a preparation of polymer solution containing microparticles. The second (2) is a formation of materials using electrospinning method. The third (3) is a study of mechanical properties of fabricated nonwoven materials.

Fig. 1
figure 1

A schematic of the fabrication and characterization process, where (1) is a preparation of polymer solution containing microparticles, (2) is a formation of materials using electrospinning method, and (3) is study of mechanical properties

Full size image

2.4 Morphology

The morphology of hydroxyapatite, calcium carbonate microparticles, and PCL\wt% of HA or CaCO3 were characterized by scanning electron microscopy (SEM). SEM was performed with a MIRA II LMU instrument at an operating voltage of 20 kV and with a XL 30 ESEM FEG (Philips) at 15 kV. Micromorphology was investigated collecting images at a magnification ranging from 100 to 40,000 times.

2.5 Mechanical Testing

Mechanical integrity in terms of yield stress and tensile strength was investigated by using a DMA Q800 «TA Instruments» USA universal testing machine (gauge length is 24 mm), and the extension rate was kept at 0.0450 MPa/min. All measurements were carried out at room temperature. Mats of as-spun neat PCL and composite fibers were cut into a rectangular shape (10.23 × 6.45 × 0.12 mm).

2.6 Image Analysis

The average particle size of the calcium carbonate particles and HA particles and average fiber diameter were investigated by a set of SEM images in order to obtain a minimum of 100 measurements per sample. Image analysis and statistics were performed using Image J (NIH, http://rsb.info.nih.gov/ij/).

3 Results

3.1 Morphology of Nonwoven Composite Materials

SEM images of synthesized calcium carbonate and HA particles were shown on Fig. 2a, b. Good spherical shape calcium carbonate (vaterite) particles were obtained in ultrasonic treatment as shown in Fig. 2a. The average size of calcium carbonate particles produced during ultrasound treatment was 1.20 ± 0.1 μm. Figure 2b observed that the HA particles have good spherical shape, and these particles have a lot of thin needles on the surface and looks like a sea urchin. The average particle size was 1.40 ± 0.5 μm and the size of these needles is less than 100 nm. Figure 2c–i selected scanning electron micrographs of as-spun fibers from (c) as-spun PCL, (d) PCL/2.5 % HA, (e) PCL/5 % HA, (f) PCL/7.5 % HA, (g) PCL/2.5 % CaCO3, (h) PCL/5 % CaCO3, (i) PCL/7.5 % CaCO3, respectively.

Fig. 2
figure 2

a, b Scanning electron micrographs of synthesized of calcium carbonate and HA particles at micron size, respectively. Figure 2 (c–i) Selected scanning electron micrographs of as-spun fibers from c as-spun PCL, d PCL/2.5 % HA, e PCL/5 % HA, f PCL/7.5 % HA, g PCL/2.5 % CaCO3, h PCL/5 % CaCO3, and i PCL/7.5 % CaCO3, respectively

Full size image

Conditions and processing parameters of the PCL, hydroxyapatite, and calcium carbonate particles centered in the middle of the jet and stacked again along the fibers in short intervals. Therefore, the obtained morphology is the same as that of PCL; however, fiber diameter is increased dramatically [33].

Clearly, smooth fibers without the presence of beads were observed and randomly oriented as shown in Fig. 2c. Incorporation of both types of micron HA and CaCO3 particles in the spinning dopes resulted in the formation of rough fibers. Defects in the form beads and aggregates are observed in some of the fibers. Thus, for the polymer spun fibers (Fig. 2f, i) made from the composite containing higher concentrations (7.5 wt%) of HA and CaCO3, a considerable amount of defects, notches, and beads is observed. Each fiber (except those with beads) is found to be of fairly uniform diameter. The formation of beads was attributed to the precipitated HA and CaCO3 microparticles that were not distributed efficiently in the PCL matrix and therefore agglomerated in cluster. For as-spun neat PCL fibers, the average diameter was about ∼0.35 ± 0.10 μm, while, for as-spun composite fibers, it ranged between 0.66 ± 0.20 and 1.60 ± 0.39 μm with defects and 0.36 ± 0.15 and 0.59 ± 0.22 μm without defects as shown in Table 1. Apparently, the diameters of the as-spun composite fibers increased with the addition and increasing amounts of calcium carbonate and HA particles until 5 % mass fraction. The observed increase in the fiber diameters of the composite fibers in comparison with those of the neat fibers should be the result of the observed increase in the viscosity of solution for spinning due to the presence of the filler particles [34]. Also, the observed diameter of fiber reduced at 0.075 mass fractions of calcium carbonate and HA particles as shown in Table 1 when the average fiber diameter was calculated without defects. The observed reduction in the diameter at high mass fraction can be attributed to the effect of the increased viscosity of the latter dispersion during electrospinning.

Table 1 Average diameter size of composite fibers with and without defects which obtained from statistic analysis of SEM images at different mass fraction of filler
Full size table

3.2 Mechanical Properties of Nonwoven Composite Materials

Figure 3a, b shows the plot of tensile modulus (MPa) and tensile strength (MPa) as a function of the mass fraction (mf) of HA and CaCO3 microparticles in the composite. From Fig. 3a, b, compared to the PCL, the tensile stress (TS) of the PCL/ HA composite containing only 2.5 % of HA is increased by ∼523 % from 1.01 to 6.34 MPa, and the Young’s modulus (E) is increased by ∼936 % from 2.24 MPa to 23.21 MPa. Moreover, as the weight percent of HAs is increased from 2.5 to 5 %, the tensile stress increases by ∼140 % from 6.34 to 15.4 MPa and the Young’s modulus is increased by ∼58 % from 23.21 to 36.7 MPa. Also, the weight percent of HAs is increased from 5 to 7.5 %, the tensile stress increases by ∼86 % from 15.4 to 28.7 MPa, and the Young’s modulus is increased by ∼53 % from 36.7 to 56.3 MPa. Also, from Fig. 3, compared to the PCL, the tensile stress of the PCL/CaCO3 composite containing only 2.5 % of CaCO3 is increased by ∼236 % from 1.01 to 3.67 MPa, and the Young’s modulus is increased by ∼230 % from 2.24 to 7.4 MPa. Moreover, as the mass fraction of CaCO3 is increased from 2.5 to 5 %, the tensile stress increases by ∼94 % from 3.67 to 7.14 MPa and the Young’s modulus is increased by ∼54 % from 7.4 to 11.4 MPa. Also, the mass fraction of CaCO3 is increased from 5 to 7.5 %, the tensile stress increases by ∼45 % from 7.14 to 10.4 MPa, and the Young’s modulus is increased by ∼86 % from 11.4 to 21.3 MPa.

Fig. 3
figure 3

a The plot of Young’s modulus (MPa). b Tensile strength (MPa) as a function of the mass fraction of HA and CaCO3 in the composites

Full size image

3.3 Regression Analysis of Mechanical Properties

The method of least squares was used to determine the approximate solution of Young’s modulus of PCL + HA and PCL + CaCO3 composites [35]. This method is applied for estimating the relationships among variables. Variables in this method can be independent (mass fraction of added material) and dependent (Young’s modulus of composite). Young’s modulus is a function of mass fraction of added material. Experimental data for different mass fraction of hydroxyapatite and CaCO3 (Tables 2 and 3) is used for the required analysis.

Table 2 Young’s modulus of composite depends on mass fraction of hydroxyapatite
Full size table
Table 3 Young’s modulus of composite depends on mass fraction of CaCO3
Full size table

We set E = F(m f), e.g., Young’s modulus of composite is function of mass fraction of HA or CaCO3. Next, mass fraction is a function of volume fraction V f. So, finally, let E = F(V f).

There is rule of mixtures for solid composites:

$$ F\left({V}_{\mathrm{f}}\right)={V}_{\mathrm{f}}\cdot {E}_{\mathrm{f}}+\left(1\hbox{-} {V}_{\mathrm{f}}\right)\times {E}_{\mathrm{m}} $$
(1)

where V f —volume fraction of added material (HA or CaCO3), E f —Young’s modulus of added material (HA or CaCO3), and Em—Young’s modulus of basic material (PCL). The relation between volume fraction and mass fraction is given by Eq. (2), then the densities of blends were taken into consideration when volume fraction was calculated.

$$ {V}_{\mathrm{f}}=\frac{m_{\mathrm{f}}}{m_{\mathrm{f}}+\left(1\hbox{-} {m}_{\mathrm{f}}\right)\frac{\rho_{\mathrm{f}}}{\rho_{\mathrm{m}}}} $$
(2)

where ρ—density of blends. There are also mentioned rule of mixtures for porous composites [36, 37]:

$$ F\left({V}_{\mathrm{f}}\right)=K\cdot {V}_{\mathrm{f}}\cdot {E}_{\mathrm{f}}+\left(1-{V}_{\mathrm{f}}\right)\cdot {E}_{\mathrm{m}} $$
(3)

where K—coefficient that depends on chemical composition of matrix and filler.

Modify Eq. (3) slightly:

$$ F\left({V}_{\mathrm{f}}\right)={K}_1\times {V}_{\mathrm{f}}\times {E}_{\mathrm{f}}+{K}_2\times \left(1\hbox{-} {V}_{\mathrm{f}}\right)\times {E}_{\mathrm{m}} $$
(4)

where K 1, K 2—coefficients, that depend on specific blend. So, function F(V f) is a linear combination of two functions. The main goal of the method of least squares is to find such K 1, K 2 that minimize the sum

$$ {\displaystyle {\sum}_i\left[{E}_{\exp, i}-{F}_2\left({m}_{f,i}\right)\right]},\;\mathrm{where}\;i=1..4 $$

3.4 Numerical Solution

Our goal is to find coefficients K 1, K 2. For this purpose, the computer software Mathcad was used. Relying on the method of least squares, we can obtain coefficients K 1 and K 2. Obtained functions for HA and CaCO3 have been plotted using Gnuplot graphing utility. Plots are shown on Figs. 4 and 5: solid line—standard function (rule of mixtures), crosses—experimental data. On Figs. 4a, and 5a, plotted Young’s modulus of composite for mass fraction of filler vary from 0 to 100 %, and Figs. 4, 5) shows dependencies of mass fraction varying from 0 to 20 %.

Fig. 4
figure 4

a, b Young’s modulus (MPa) as a function of the mf in PCL + HA composite

Full size image
Fig. 5
figure 5

a, b Young’s modulus (MPa) as a function of the mf in PCL + CaCO3 composite

Full size image

We put together the experimental data from the current work and from [38]. Using the method described above, we calculate K 1, K 2 for a combined data (function E c,1(m f)) and plot a corresponding graph (Fig. 6) for PCL + HA blend.

Fig. 6
figure 6

Young’s modulus (MPa) as a function of the mf in PCL + HA composite using experimental data from [38]

Full size image

To predict the behavior of PCL + HA blend in the next set of experiments following approximate function of m f can be useful (Fig. 7); we used a polynomial (Eq. 5) to fit our experimental data.

Fig. 7
figure 7

Approximate value of Young’s modulus (MPa) using Eq. (5)

Full size image
$$ {E}_{\mathrm{c},2}\left({m}_{\mathrm{f}}\right)=2.792+762.5\times {m}_{\mathrm{f}}\hbox{-} 878.7\times {m_{\mathrm{f}}}^2+1344\times {m_{\mathrm{f}}}^3\hbox{-} 4631\times {m_{\mathrm{f}}}^4+5028\times {m_{\mathrm{f}}}^5 $$
(5)

Standard deviations, i.e., errors \( \sqrt{{\displaystyle \sum_i^n\left[{E_{\exp,}}_i-{E}_{c,k}\left({m}_{f,i}\right)\right]}\cdot \frac{1}{n-2}},i=1..4,k=1,2 \); calculated Young’s modulus are presented in Table 4:

Table 4 Standard deviations for calculated mechanical parameters
Full size table

4 Discussion

As illustrated in Fig. 3a, b, by increasing the amount of HA and CaCO3 particles, Young’s modulus and tensile strength are increasing. The percent elongation becomes much smaller in the composites with higher HA contents, as expected. This result can be explained by the load-transfer theory. Classical load-transfer theory suggests that the stress can transfer from polymer matrix to reinforcements through the interfacial layer. Reinforcements have higher modulus than the polymer matrix so that they can afford parts of stress to disperse the load in the matrix. Therefore, the strength of the reinforcement’s composites is improved by reinforcement’s addition [39].

There are three main mechanisms of load transfer from a polymer matrix to the filler. The first is micromechanical interlocking, which is likely to be difficult in the composites due to their atomically smooth surface. The second is chemical bonding between the filler and the fiber. The third mechanism is a weak van der Waals bonding between the filler and the fiber matrix [40]. The dispersion of the fillers in the matrix is also one of the important factors affecting the strength of the composites. It is well known that the agglomeration of the fillers can deteriorate the mechanical properties of the composites [41].

The HA and CaCO3 particles were covered with PCL fibers, which is an indicator of a good matrix\filler interface. The interfacial improvement leads to a better load transfer throughout the material, and result in better mechanical properties. The results seem to indicate that there is van der Waals bonding between the PCL chains and functional groups in both of HA and CaCO3 particles. Also, the suggested improvement is caused by hydrogen bond formation between surface of HA and CaCO3 particles and the PCL polymer [42, 43]. On the other hand, from the SEM image, the observed HA and CaCO3 particles were not homogeneously dispersed along the fiber and so the ultimate tensile strength values would not be as high as expected. Moreover, the increasing content of the fillers increases the weak van der Waals and hydrogen bonds interaction between the fillers and the fibers. However, van der Waals interactions are negligible at a lower mass fraction of filler, due to the directional anisotropy. That is explaining why the mechanical properties of composites increased with increasing mass fraction of fillers. Calcium carbonate particles reinforce the PCL fiber and increase its mechanical strength; however, the mechanical strength of PCL\(m f) CaCO3 less than the mechanical strength of PCL\(m f) HA. This can be due to the incomplete dispersion of CaCO3 particles along the fibers and may be the number of van der Waals and hydrogen bonds between the PCL chains and functional groups in CaCO3 particles which may negatively affect the mechanical properties.

5 Conclusions

In this study, PCL/HA, CaCO3 composites were fabricated by electrospinning technique. Different amounts of hydroxyapatite (HA) and calcium carbonate microparticles were loaded within the polymer fibers providing an opportunity to produce biomimetic materials for clinical applications. Both the Young’s modulus and the ultimate tensile stress increased with increases in mass fraction of HA and CaCO3 particles. The results seem to indicate that there is van der Waals bonding between the PCL chains and functional groups in both of the HA and CaCO3 particles. The uniform dispersion of fillers is important for enhancement of mechanical properties of the composite. The mechanical properties change significantly with an increasing of HA or CaCO3 mass fraction, for example, the Young’s modulus and the ultimate tensile strength of the composites containing 7.5 wt% HA were higher about 25 and 28 times compared to pure PCL fibers, respectively. Also, for composites containing 7.5 wt%, CaCO3 were higher to nine and ten times compared with the pure PCL fiber. Regression analysis of mechanical properties of binary composites (polymer matrix/filler) was done and as a result, the approximate equation for mechanical parameter prediction of nonwoven materials was obtained. The presented standard deviations of mechanical parameters indicate that results of analysis can be used to predict Young’s modulus of composite when mass fraction of fillers is between 0.00 and 0.1. Rule of mixtures is used mostly for solid composites in this study; we modified and used it to fit our experimental results. These materials can be used as scaffold combined with biosensors [44], for tissue engineering included bone repairing and augmentation, and also for smart textile with optimal properties related to water adsorption and air penetration.