Combining Breath Figures and Supercritical Fluids To Obtain Porous Polymer Scaffolds.

Supercritical fluids technology is a clean methodology to foam polymeric materials. However, this technique provides only the formation of inner porosity, whereas the so-called skin layer is commonly observed at the polymer surface. This article describes a new method for the preparation of outer and inner porous poly(ε-caprolactone) (PCL) scaffolds by combination of supercritical CO2 (SCCO2) foaming and the breath figures technique. In the first step, experiments with a SCCO2 reactor were performed at 35-45 °C, 100-250 bar, and 1-20 min depressurization time. The effect of these parameters in the formation of inner porosity was investigated for an adequate optimization. In a late stage, to provide also surface porosity to the polymeric samples and remove the skin layer, the breath figures technique was employed. The evaluation of porosity was determined by scanning electronic microscopy, mercury porosimetry, and micro X-ray computerized tomography scanning processing the images obtained with the ImageJ software. The results of this study using these two complementary techniques showed the existence of interconnectivity between inner and outer porosity of the samples. Furthermore, thermal transitions and crystallinity of the PCL samples have been analyzed by differential scanning calorimetry. Finally, a preliminary biological evaluation of the resulting scaffolds with mouse endothelial cells (C166-GFP) was performed to assess their biocompatibility and cellular viability.

Introduction

Porous biodegradable polymer scaffolds are being applied in tissue regeneration processes[1] as the number of people who receive life-saving organs and tissue transplantation is limited by a shortage of donor tissue. A great variety of biomaterials as synthetic polymers, ceramics, and naturally derived proteins are being used for the preparation of these scaffolds.[2] Different techniques have been developed to fabricate these polymers into porous scaffolds for tissue engineering applications.[3−5] In this sense, supercritical CO2 (SCCO2) is a well-known technology to obtain porous structures of polymeric materials. Its advantages with respect to other techniques rely on its green character (CO2 is recyclable and elude the undesirable use of organic solvents), low economical costs, easy achievable critical points, nontoxicity, nonflammability, and easy control of the porosity and pore size by the depressurization process when using the appropriate temperature and pressure conditions.[6,7] However, the main disadvantage that this technique presents when foaming polymer samples is the appearance of the so-called “skin layer” which avoid the formation of external porosity. Regarding the skin layer, the rapid diffusion of the embedded fluid out of the sample edges results in the formation of this dense nonporous skin layer that can be decreased, to a certain extent, by an increase in pressure.[8]

Different approaches have been carried out to prevent the formation of a skin layer and form an entirely porous structure. In that sense, Mooney et al.[9] incorporated NaCl particles to the polymer solution of poly(d,l-lactic-co-glycolic acid) before gas foaming. The leaching of this porogen after fabrication of the polymer foam created an interconnected open pore network. Barry et al.[10] in 2004 simply removed the skin layer from the scaffolds before cell culture when they used the gas-foaming method to create poly(ethyl methacrylate)/tetrahydrofurfuryl methacrylate foams. Hori et al.[11] prepared skinless polymer foams by treating poly(methyl methacrylate) (PMMA) with SCCO2, that is, by immersing the PMMA in SCCO2 with ethanol to nucleate, and then incubated in a hot water bath to induce nuclear growth. The obtained polymer foams had a low density, and cells were connected to each other to make channels through the skin layer.

Our group described recently the preparation of poly(ε-caprolactone) (PCL) biodegradable porous membranes with outer and inner porosity just by SCCO2 treatment, displaying surface hierarchical macroporosity which could be tailored by careful control of the pressure, in the range of 150–250 bar, and depressurization processes in several steps, showing also pore interconnectivity between both membrane faces. In this case, the thickness of the employed membranes played an important role to obtain outer and inner porosity, as those with thickness higher than 90 μm showed the appearance of the so-called skin layer.[12]

In the present work are combined the SCCO2 technology, which provides the internal porosity, with the breath figures[13] (BF) technique which affords the external porosity in PCL samples independently of the sample thickness. This technique is based on the short immersion of the porous samples in an organic solvent in a moist atmosphere. During the evaporation of the thin solvent layer deposited at the polymer surface, the interfacial temperature decreases thus leading to water vapor condensation.[14] The condensation process finally produces water droplets at the sample surface which, after evaporation of both solvent and condensed water, provide the external porosity. SCCO2 pressure, temperature, and depressurization are crucial parameters that were evaluated to obtain the internal porosity, as well as different immersion times for the BF technique were employed to obtain connection between the external and internal porosity. The thermal properties of the foamed PCL are also described in terms of changes in glass transition temperature, crystallinity, and melting point. Finally, cell culture studies were carried out on the prepared porous PCL samples using mouse endothelial cells analyzing the metabolic activity with a view to possible applications in tissue engineering.

Results and Discussion

Control of the Internal Porosity of the PCL Scaffolds

Although all polymeric samples treated exclusively with SCCO2 showed a nonporous outer skin layer with variable thickness, the pore distribution as well as the average pore size depends on the experimental conditions. Therefore, the optimization of the inner porosity was carried out by varying three parameters: temperature, pressure, and depressurization time.

The first parameter explored was the temperature. Figure shows the scanning electronic microscopy (SEM) micrographs for samples prepared at different temperatures, between 35 and 45 °C, keeping the pressure constant at 200 bar and the depressurization time at 5 min. On the one hand, at 35 and 37 °C, no porosity can be observed in the center of the cylindrical samples. At 35 °C, the thickness of the porous layer is about 400 μm, and at 37 °C, the thickness of the porous layer is about 800 μm. The absence of porosity in the center of the polymer is associated with the low temperature used. At this temperature, the viscosity of the polymeric matrix (largely below the melting point of PCL) is too high and limits the diffusion of CO2 within the saturation time employed.[15] This fact can also be due to the noncompleted solubility of CO2 in the polymer sample.

Figure 1

SEM micrographs for samples treated with SCCO2 at 200 bar, 5 min depressurization time, and different saturation temperatures: (a) 35, (b) 37, (c) 40, and (d) 45 °C.

On the other hand, as depicted in Figure , at temperatures between 40 and 45 °C, the polymer samples showed inner porosity in the whole fractured area. Table shows the average pore size of the samples for the different conditions employed. At 40 °C a more homogeneous distribution with an average pore size of 77 ± 17 μm can be seen, while at 45 °C this value is 106 ± 29 μm and the size distribution is more heterogeneous. This observation can be related to a partial collapse of the pores due to a process temperature close to the melting temperature of the polymer after the absorption of CO2. The larger pore sizes observed at 45 °C are due to the fact that the viscosity of the polymer is lower and the gas diffusivity increases allowing the pores to grow.[15,16]

Table 1

SCCO2 Conditions (Temperature, Pressure, and Depressurization Time) and Average Pore Size for the Different Experiments Obtained Using the ImageJ Software

Sample	Temperature [°C]	Pressure [bar]	Depr. time [min]	Average pore size [μm]
E1-T35	35	200	5	44 ± 15
E1-T37	37	200	5	56 ± 15
E1-T40	40	200	5	77 ± 17
E1-T45	45	200	5	106 ± 29
E4-P100	40	100	5	375 ± 265
E4-P150	40	150	5	100 ± 30
E4-P250-5	40	250	5	65 ± 16
E4-P250-10	40	250	10	70 ± 20
E2-DP60	40	200	1	25 ± 10
E2-DP600	40	200	10	110 ± 25
E2-DP1200	40	200	20	195 ± 80

As a result of the experiments depicted above, the temperature employed for the rest of experiments was 40 °C, as at this value pores are obtained in the whole sample while remaining below the melting temperature. Interestingly, this temperature has also been selected and used by other authors[17,18] that employed PCL as polymeric material.

The second parameter explored to control the internal structure of the PCL scaffolds was the pressure employed. In Figure , the SEM images of the porous samples obtained varying the pressure employed are represented. At 100 bar (Figure a) (the lowest pressure tested), a mixture of areas with rather large pores (up to 500 μm in diameter) and solid areas can be distinguished. Using these conditions, the SCCO2 has not been dissolved enough to generate the appropriate number of nuclei that can grow afterward. A rather homogeneous distribution can be observed in the fractured area for the other three experimental pressures (Figure b–d). Moreover, the average pore size decreases when the pressure increases (Table ). This effect is associated with an increase in dissolved CO2 in the polymer matrix, which generates more available nuclei for the formation and growth of the pores. As a consequence, more pores showing smaller size are generated.[15]

Figure 2

SEM micrographs for samples treated with SCCO2 at 40 °C, 5 min depressurization time, and different saturation pressures: (a) 100, (b) 150, (c) 200, and (d) 250 bar.

Finally, the third experimental condition analyzed was the depressurization time. In Figure , the different SEM images obtained by varying the depressurization time are depicted. When the depressurization time increases, that is, depressurization takes place at slower rate, the average pore size also increases (see also Table ). The reason for this observation is that the nucleation process competes with the growth of the pores, as a result of the gas diffusion through them. On the one hand, when the depressurization rate is high, nucleation takes place faster and a great amount of pores are generated. The development of each pore will occur fast, so that the effects of diffusion will be negligible, and the resulting structure will have an uniform pore size distribution. On the other hand, when the nucleation process is slow, the resulting pores will be bigger than the rest because of a higher diffusion of the gas, and the final structure will have a larger pore size distribution.[19]

Figure 3

SEM micrographs for samples treated with SCCO2 at 40 °C, 200 bar and different depressurization times: (a) 1, (b) 5, (c) 10, and (d) 20 min.

Analysis of the Pore Interconnection

A major requirement for the application of porous materials for cell growth is the formation of interconnected pores that enable the cells to migrate and simultaneously permit the flow of nutrients inside the scaffold. As depicted in Figure , both closed pores as well as a majority of interconnected pores have been observed in the cross section SEM micrographs of the samples. Besides, the high porosity values obtained by mercury porosimetry (Table ) revealed the existence of connection between the pores. It is important to note that non-interconnected pores would not have permit mercury to fill all the free space in the sample. Interestingly, it is worth mentioning that porosity increases when pressure increases at a given depressurization time and can be, therefore, modulated.

Figure 4

SEM micrographs of the cross section of samples treated with SCCO2 at (a) 200 bar and 10 min depressurization time and (b) 250 bar and 5 min depressurization time.

Table 2

Porosity Measured by Mercury Porosimetry for Samples Treated with SCCO2 at Different Saturation Pressures

sample	pressure [bar]	depressurization time [min]	porosity [%]
E4-P100	100	5	71.9
E4-P150	150	5	76.8
E4-P200	200	5	79.1

Although, by micro-computerized tomography, it has also been possible to show the good interconnection between de inner pores, we will discuss these analyses in the next section.

Effect of SCCO2 on the Thermal Transitions and Crystallinity of the Polymer

Differential scanning calorimetry (DSC) analyses were made to determine the influence of the SCCO2 on the thermal transitions of the processed porous PCL. For comparative purposes, as a control sample, a nontreated PCL was employed and the measured Tg was −60 °C and the Tm was 73 °C, while the crystallinity was 50.6%. Table shows the values obtained for a selection of samples, and it can be observed that the Tg in all cases is very similar to the one of the control sample, with variations between −62 and −59 °C. Nevertheless, the changes observed in the Tm are greater, going from 6 to 14 °C less than the untreated PCL independently of the pressures employed or the depressurization times used. The difference between the thermal transitions after the BF technique in comparison to the ones just treated with SCCO2 under same conditions is also negligible. Regarding the crystallinity, it was observed that, in general, these were higher in comparison to the control PCL. This behavior, also reported by other authors,[17,20,21] can be explained due to the fact that the experimental temperature is close to the Tm, resulting in a higher mobility of the chains, thus enabling the polymer chain reorganization and improving the crystal formation. It is also note worth that the Tm decrease can be attributed to the formation of smaller crystals, as they do not have enough time to crystallize as bigger ones.

Table 3

DSC Results for Untreated PCL, PCL Treated under Different Pressure, and Depressurization Time Conditions with the SCCO2 Foaming Technique, and PCL Samples after the BF Technique

sample	pressure [bar]	depressurization time [min]	immersion in CHCl3 [s]	Tg [°C]	Tm [°C]	crystallinity [%]
PCL control				–60	73	50.6
E4-P100B	100	5		–59	65	56.0
E4-P150C	150	5		–60	67	54.3
E1-T40	200	5		–62	59	45.3
E1-T40B	200	5	5	–61	67	57.2
E4-P250I	250	5		–61	64	51.4
E4-P250C	250	5	5	–61	65	55.3
E4-P250E	250	10		–59	67	51.2
E4-P250F	250	10	5	–59	67	67.7
E4-P250G	250	20		–62	65	52.3
E4-P250H	250	1		–61	65	51.1

Surface Porosity and Interconnection with Inner Porosity

After the optimization of the parameters of the SCCO2 technique, tests were conducted to study the formation process of the pores on the surface of the scaffolds by the BF technique. This study was focused on the samples obtained at 250 bar saturation pressure and 40 °C temperature which were the most homogeneous in terms of pore size. The interconnection of the pores was studied by SEM (Figures and 6) and by microcomputerized X-ray tomography (Figure ).

Figure 5

SEM micrographs for (a) nonporous surface of a PCL sample after the SCCO2 technique; and size of the superficial pores for samples treated with SCCO2 at 250 bar and 10 min depressurization time, immersed in CHCl3 for (b) 3, (c) 5, and (d) 7 s.

Figure 6

SEM micrographs of inner and outer porosity of scaffolds treated with SCCO2 at 40 °C, 250 bar, 5 min depressurization time and (a) 5 s in CHCl3, (b) 7 s in CHCl3, (c) 10 s in CHCl3 and (d) 30 s in CHCl3.

Figure 7

Micro-CT images showing the interconnection between inner and outer pores for a sample treated at 40 °C, 250 bar, 5 min depressurization time and 8 s immersion in CHCl3. (a) Cross section and (b) vertical section of the sample.

From the SEM analysis of the samples shown in Figure , it can be observed that for immersion times between 5 and 7 s in CHCl3, in areas where the dense skin layer thickness is in the range of 10–30 μm, the pores formed at the surface are clearly interconnected with those inside the sample (Figures and 7). This means, for samples treated with SCCO2 at high pressures (250 bar) and long immersion times for BF, connection between inner porosity (SCCO2) and outer porosity (BF) is obtained. For 1 and 3 s immersion time, that is, shorter immersion times, the solvent did not penetrate the skin layer and the interconnection did not occur. Additional experiments were carried out at longer immersion times (from 8 s until 30 s) to study whether the interconnection can be still improved (see Figure ). For samples treated with SCCO2 under the same conditions (250 bar, 40 °C and 5 min depressurization time), the connection between the inner and the outer pores is appreciable with no significant differences. From 10 s onward, the effect is not further improved because a wide (ca. 300 μm) nonporous outer layer is formed. This effect of the immersion time may be due to two different reasons. On the one hand, the outer skin of the polymer has been, at least to some extent, dissolved in CHCl3, reducing the thickness of the dense skin. On the other hand, the swelling of the PCL surface by the CHCl3 employed and its evaporation leads to deeper pores.

Another interesting aspect to be observed is the differences in terms of pore size obtained by the SCCO2 technique and BF. Although BF generally form pores with sizes ranging from 5 to 20 μm, foaming with SCCO2 produced larger pores (50–150 μm for the conditions used in this work). This aspect can be advantageous in tissue engineering due to the fact that small pore sizes improve cell response on the surface, while inner bigger pores allow that, after cell seeding, the tissue penetrates in the scaffold and improves the integration of the material until its degradation.[22]

Evaluation of the Cell Adhesion and Biocompatibility of the PCL Porous Scaffolds

To test suitability of the PCL scaffolds for tissue engineering purposes, a cytocompatibility assay has been performed. Endothelial C166-GFP cell line has been used, because of their autofluorescence and its optimal behavior in terms of cell growing over biomaterials.

After 7 days of culture, all samples allowed endothelial cell adhesion and proliferation. As it can be observed in Figure , nontreated PCL scaffolds allowed only small cell clumps proliferation, supporting a reduced number of cells with a poorly attached morphology over their surface. Nevertheless, this situation changed for the three treated PCL samples. It was possible to detect bigger cell colonies adhered in the PCL surface and inside the pores, at several depth levels (Figure b–d). Especially, cell proliferation was increased over E4-P250-5 and E4-P250-10, where endothelial cells located and grew into interconnected scaffold cavities.

Figure 8

Cell spreading on scaffolds after 7 days of culture for (a) nontreated PCL, (b) PCL after SCCO2 at 40 °C, 150 bar, 5 min depressurization time, and 7 s in CHCl3, (c) PCL after SCCO2 at 40 °C, 250 bar, 5 min depressurization time, and 7 s in CHCl3, and (d) PCL after SCCO2 at 40 °C, 250 bar, 10 min depressurization time, and 7 s in CHCl3. (e) alamarBlue metabolic assay after 7 days of culture.

In addition, a metabolic activity study of these samples was also performed to quantify cell proliferation. As a result, all scaffolds showed similar values of cell activity (Figure e), demonstrating that treatment did not decrease biocompatibility. On the contrary, a slight upward trend was detected over specific scaffolds, such as E4-P250-5.

Regarding these preliminary biological results, double SCCO2 and BF techniques over PCL scaffolds improved cell behavior in comparison with nontreated PCL. This finding suggests an interesting application in bone and cartilage tissue engineering purposes. On the one hand, PCL characteristics include high mechanical strength, bioresorbability, and a proper degradation rate compatible with bone tissue proliferation.[23] On the other hand, the introduction of inner and surface porosity in bone implants has shown to increase cell culture activity.[24] Moreover, after bone matrix synthesis, sample–tissue interface is enhanced, improving implant fixation and promoting final success. Although endothelial cell results are interesting by themselves, mainly because vascular growth is also needed in these bone regenerative applications, a deeper study including osteoblasts and/or osteoclasts and specific differentiation markers would ensure this clinical use.

Conclusions

The combination of BF and supercritical fluids (SCCO2) techniques allows the preparation of PCL porous scaffolds with interconnected outer and inner porosity. Their inner porosity was obtained by SCCO2 treatments by keeping the temperature constant at 40 °C, decreasing the pore size when increasing the pressure, and increasing when slowing the depressurization rate, showing all samples an outer skin layer. SCCO2 porous scaffolds were found to be interconnected. The SCCO2 treatment effect on the PCL thermal transitions showed constant Tg and Tm and crystallinity slightly increased. By BF, samples previously treated by SCCO2 at 40 °C and 250 bar were submitted to different CHCl3 immersion times (up to 10 s), providing surface porosity that was found by SEM and micro X-ray computerized tomography scanning analysis, to be connected to the inner one, showing pores from 5 to 20 μm by BF and from 50 to 150 μm by SCCO2. Finally, preliminary cell adhesion and biocompatibility of the porous scaffolds exhibited good cell adhesion and proliferation into the interconnected scaffolds, maintaining the biocompatibility and improving cell behavior with respect to nontreated PCL.

Experimental Section

Materials

PCL (50 000 g mol–1) was supplied by Perstorp. Liquid carbon dioxide was purchased from Carburos Metálicos with a 99.99% purity. Chloroform (CHCl3) was supplied by Sigma-Aldrich.

Preparation of the Samples

The samples were prepared by extrusion of PCL pellets. The process conditions were 60 °C and velocity, with a nozzle diameter of 3 mm. In each experiment, three samples were tested at the same time and placed in 4 mm diameter tubes to keep the cylindrical shape.

Preparation of the Porous Structures

Inner Porosity

PCL cylinders were placed in a SCCO2 reactor (Thar R100W, 104 mL) which operated under different conditions: 35–45 °C temperature, 100–250 bar pressure, and 1–20 min depressurization time. The SCCO2 system consists of a CO2 tank, which delivers CO2 to a high pressure pump precooled using a cryostat at 4 °C, reaching the reactor at the desired pressure and temperature. The system remained at these conditions for a processing time of 90 min using a CO2 flow of 5 g/min for all the cases and then depressurization takes place until atmospheric pressure.

Superficial Porosity

Samples with different inner pore sizes (between 50 and 150 μm) were chosen to obtain superficial porosity by the BF technique. For this purpose, PCL cylinders pretreated with SCCO2 were immersed during different times (1–30 s) in chloroform under saturated relative humidity (over 90%) conditions at room temperature in a closed chamber. The samples were left to dry inside the chamber for 24 h.

Characterization

Determination of Pore Size

The porous scaffolds were frozen in liquid nitrogen and fractured. Their cross sections were analyzed by SEM (XL30ESEM Philips) at an accelerating voltage of 25 kV. The morphology of the pores, the thickness of the skin layer (in the case of samples treated only by SCCO2), and the average pore size were determined by this technique. SEM micrographs were analyzed with the ImageJ software. Average pore sizes were determined by measuring the diameter of 30 pores of the image and obtaining their average and standard deviation.

Mercury Porosimetry

The determination of the percentage of porosity was performed using the unit PoreMaster-33, Quantachrome Instruments.

Computerized Tomography

The interconnectivity of the pores was studied by computerized tomography with the unit CT-SCAN-XT-H-160 (Nikon). This technique allows to take two-dimensional images while the sample rotates 360° and to obtain a volumetric three-dimensional reconstruction with the proper software.

Differential Scanning Calorimetry

Thermal transitions and crystallinity of the samples were analyzed by DSC (Mettler Toledo DSC 822e). The polymer was heated from −90 to 100 °C at a rate of 10 °C/min. The melting temperature, Tm, was determined using the value of the endothermic transition peak, and the glass transition temperature, Tg, was taken as the middle point of the transition. The percentage of crystallinity was calculated through the ratio between ΔH and ΔH0.

Biological Studies

Cell Culture and Cell Seeding

The cell studies were carried out using C166-GFP, a mouse endothelial cell line (ATCC CRL-2583). For culturing cells on the scaffolds, the endothelial cells were seeded singly over the samples in supplemented Dulbecco’s modified Eagle’s medium, and the polymers were placed in a 24-well plate in maintenance medium, incubated at 37 °C with 5% CO2 in a humidified incubator. For experiments on PCL scaffolds, cells were seeded at a density of 5000 cells/scaffold, and three scaffolds of each sample type were used for each experiment. Samples were observed and fluorescent images were captured at 168 h using an Olympus BX51 microscope.

Metabolic Activity Study

Metabolic activity of cells was measured by alamarBlue assay at 72 and 168 h. AlamarBlue dye (10% of the culture volume) was added to each well, containing living cells seeded over samples, and incubated for 90 min. The fluorescence (λex/λem 535/590 nm) of each well was measured using a plate-reader (Synergy HT, BioTek).