Synthesis and characterization of nanofibrous hollow microspheres with tunable size and morphology via thermally induced phase separation technique

Abstract

In this paper, we present a facile thermally induced phase separation (TIPS) technique for the synthesis of size and morphology controllable poly(L-lactic acid) (PLLA) nanofibrous hollow microspheres (NHMs). The effects of propanetriol (PT) volume, polymer concentration, dropping rate, stirring speed and addition sequence on the physical properties of NHMs including particle size and surface morphology were systematically investigated. The results revealed that the obtained NHMs became more round in shape and more uniform in size with the increase of the PT volume. When the concentration of polymer solution was increased to 4%, the obtained NHMs showed a homogeneous nanofibrous structure, while their size became bigger and more uniform than those of prepared in lower polymer concentrations. The increased dropping rate of PT resulted in the packed nanofibrous structure and decreased microsphere size, whereas the increased stirring speed led to a reduced microsphere size. The uniformity and morphology of microspheres are also strongly influenced by the addition sequence. Furthermore, the in vitro cytotoxicity assay demonstrated that NHMs possess good biocompatibility. As a result, the synthesized NHMs would have significant potential in a variety of biomedical applications such as drug delivery and tissue regeneration.

---

Introduction

Microspheres composed of biodegradable and non-biodegradable polymers have aroused considerable interest during the past decades, due to their potential utility in diverse fields such as drug delivery and tissue engineering.¹ Since the use of nonbiodegradable polymers are greatly limited by safety concerns,¹ a variety of biocompatible and biodegradable polymers were designed and fabricated into microspheres.^2,3 Among these, poly(lactic acid) (PLA) and its copolymers with glycolic acid have been increasingly utilized as microsphere materials.^4,5 PLA, as one of the most promising polymers has been approved by the United States Food and Drug Administration (FDA), is widely used in biomedical and pharmaceutical applications due to its excellent mechanical properties, tunable degradation rate, low toxicity and immunological response.⁶

The most common fabrication techniques of polymeric microspheres include phase separation,⁷ emulsion solvent extraction/evaporation,⁸ spray drying,⁹ cold precipitation,¹⁰ electrospray¹¹ and so on. Although all of these techniques have shown some success, each has certain disadvantages. For example, sphere size and size distributions are often poorly controllable for the conventional solvent evaporation and spraying approaches.^12,13 The particle size (diameter and size distribution) and surface characteristic (charge and hydrophobicity) of microspheres are two crucial parameters for their biomedical applications, which can influence the adsorption and phagocytosis of cells as well as the bioavailability and stability of microspheres.¹⁴ With the advent of nanotechnology, some conventional techniques have been used to fabricate polymeric microspheres with special characteristics such as porosity, high specific surface area, low density, and low coefficient of thermal expansion.¹⁵ Thermally induced phase separation (TIPS) technique, which was previously used to fabricate nanofibrous scaffolds for various tissue engineering applications,¹⁶ has been recently considered as an alternative method for the production of porous microspheres from some biopolymers.¹⁷ More recently, Ma's group has reported a novel injectable polymeric hollow microspheres by phase separation of functional poly(L-lactic acid) (PLLA). They demonstrated that the as-prepared microspheres have opening hole on the nanofibrous shells, and therefore are suitable for cartilage and bone regeneration.^18,19 However, there is lack of research on the control of particle size and morphology of nanofibrous microspheres, which are affected by various factors during the fabrication process.

In the present study, biodegradable PLLA was used to prepare nanofibrous hollow microspheres (NHMs) via a facile TIPS technique. The properties of NHMs in respect to the particle size and surface morphology can be altered by varying the fabrication parameters, such as propanetriol (PT) volume, polymer concentration, dropping rate, stirring speed and addition sequence. The effects of processing parameters on hollow and nanofibrous structure were emphatically investigated. In addition, the biocompatibility of NHMs was also verified, which makes NHMs more practical for application in biomedical fields.

Experimental

Materials

PLLA with an inherent viscosity of 1.93 dL g⁻¹ and a weight-average molecular weight (M_w) of 247 000 g mol⁻¹ was purchased from Daigang Biomaterials Inc. (Jinan, China). PLLA was purified by dissolving in chloroform, recrystallized in anhydrous ethyl alcohol and dried for use. Tetrahydrofuran (THF), PT, Triton® X-100 and bovine serum albumin (BSA) were obtained from Sigma-Aldrich (Shanghai) Trading Co., Ltd (Shanghai, China). Fetal bovine serum (FBS), penicillin–streptomycin, trypsin, Dulbecco's modified Eagle's medium (DMEM) high glucose media were all obtained from Gibco Life Technologies Co. (Grand Island, USA). Paraformaldehyde was purchased from Beijing Dingguo Changsheng Biotechnology Co., Ltd (Beijing, China). 4′,6-Diamidino-2-phenylindole (DAPI) was purchased from Bestbio (Shanghai, China). 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) was purchased from Shanghai Yuanxiang medical equipment Co., Ltd. Deionized (DI) water (18.2 MΩ cm⁻² at 25 °C) was used for all experiments. All other chemicals were obtained from Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). All other reagents were analytical grade and used without further purification.

Preparation of NHMs

NHMs were prepared according to a modified TIPS technique.¹⁹ For NHMs fabrication, five synthesis variables were systematically examined in this study. The detail recipe for preparation conditions is listed in Table 1. Each batch was performed in triplicate in beakers with a volume of 50 mL. In a typical fabrication procedure, PT was dropwise introduced into a solution of 10 mL PLLA in THF at 50 °C. The mixture was vigorous stirring for a determined duration of 10 min, and then the as-synthesized mixture solution was quickly poured into the liquid nitrogen for phase separated for about 30 min and immersed into DI water for solvent exchange at 4 °C for 24 h. The as-synthesized samples were washed three times with excessive amount of DI water to remove PT residue. After rinsing with DI water, the resultant microspheres were then vacuum freeze-dried for 2 days. The final product was NHMs.

Table 1 Brief summary of the experimental parameters, and the corresponding average size of microspheres

No.	PT (mL)	PLLA/THF (w/v%)	Dropping rate (mL h^−1)	Stirring speed (rpm)	Sequence	Particle size (μm)
PT-1	5	2	90	800	PLLA-PT	—
PT-2	10	2	90	800	PLLA-PT	—
PT-3	20	2	90	800	PLLA-PT	103.7 ± 17.6
PT-4	30	2	90	800	PLLA-PT	69.2 ± 12.8
P-1	30	1	90	800	PLLA-PT	85.2 ± 20.5
P-2	30	2	90	800	PLLA-PT	69.2 ± 12.8
P-3	30	4	90	800	PLLA-PT	94.7
DR-1	30	2	40	800	PLLA-PT	128.0 ± 27.2
DR-2	30	2	90	800	PLLA-PT	69.2 ± 12.8
DR-3	30	2	140	800	PLLA-PT	76.5 ± 14.1
SR-1	30	2	90	400	PLLA-PT	127.5 ± 27.2
SR-2	30	2	90	800	PLLA-PT	69.2 ± 12.8
SR-3	30	2	90	1200	PLLA-PT	63.8 ± 13.2
S-1	30	2	90	800	PT-PLLA	69.2 ± 12.8
S-2	30	2	90	800	PLLA-PT	62.5 ± 13.0

---

Morphological characterization

Morphological characterization was conducted by using Hitachi S-4800 (Hitachi Ltd, Japan) field emission scanning electron microscope (FESEM). NHMs were sprinkled over conductive adhesive on a cupreous stub and then were sparked with gold. The morphology was observed by FESEM using an accelerating voltage of 10 kV.

Microsphere size analysis

The average diameter of NHMs was obtained from at least 50 measurements on a typical FESEM image by using Image J 1.40 G software (NIH, USA).

Cell culture experiments

Mouse fibroblast L929 cell lines were purchased from Shanghai Institute of Cell Biology, Chinese Academy of Sciences (Shanghai, China). L929 cells were cultured in DMEM containing 10% FBS, 100 units per mL penicillin and 100 μg mL⁻¹ streptomycin. The cells were maintained at 37 °C in a humidified atmosphere containing 5% CO₂. The media were changed every day, and the cells were detached with 0.25% trypsin growth medium before confluence.

In vitro cytotoxicity assay

The in vitro cytotoxicity was estimated by the standard cell viability MTT assay.²⁰ Briefly, the cells were pre-seeded in 96-well plates at a density of 10⁴ cells per well and incubated for 24 h at 37 °C to allow cell attachment. Afterwards, the cells were washed with pre-warmed PBS, and the medium was replaced with fresh medium containing the NHMs at different concentrations (200, 400 and 800 μg mL⁻¹). At the end of incubation time (24 or 48 h), each well of the plates was washed with PBS and incubated in 400 μL DMEM with supplement 40 μL MTT solution (5 mg mL⁻¹ in PBS) for another 4 h at 37 °C in the incubator. Then the medium was removed and the dark blue formazan crystals were solubilized in 200 μL dimethyl sulfoxide (DMSO). After vibrating for 30 min to dissolve the formazan, an aliquot of 100 μL was taken from each well and transferred to a fresh 96-well plate. The absorbance of each well was measured at 495 nm wavelength on a microplate reader (Multiskan MK3, Thermo LabSystems). All the results were expressed as mean values of five replicates. Cells incubated in the absence of NHMs were used as the blank control and the viability was set as 100%. The cell viability was calculated as the following equation:
where A_Test and A_Control are the absorbance of cell treated with and without NHMs, respectively.

For confocal laser scanning microscopy (CLSM) observation, L929 cells (2 × 10⁴ cells per dish) were seeded on 20 mm glass bottom culture dishes (Nest Biotechnology Co., Ltd, China) and incubated for 24 h at 37 °C. The cells were washed with PBS, and subsequently incubated with or without NHMs (800 μg mL⁻¹) in the culture medium at 37 °C for 24 or 48 h. After removing the medium, the cells were washed twice with pre-warmed PBS and fixed with 4% paraformaldehyde solution for 20 min at 4 °C. Thereafter, the fixed cells were rinsed three times with PBS and stained with 100 nM DAPI solution for 5 min.²¹ Then the samples were extensively washed with PBS to remove unbound dyes and observed by using a Carl Zeiss LSM 700 confocal microscope (He–Ne and Ar lasers).

Statistical analysis

All experiments were conducted at least three times and all data were expressed as the mean ± standard deviation (S.D). One-way analysis of variance (one-way ANOVA) and Scheffe's post hoc test were used for statistical analyses. The statistical significance for all tests were set at *P < 0.05 and **P < 0.01.

Results and discussion

Effect of PT volume

As the same condition of PLLA concentration, drop rate, stirring speed and sequence, the effects of PT volume on the formation of microspheres were first investigated. As shown in Fig. 1, remarkable changes were observed in the morphology of the products as the volume of PT increased. For products synthesized with 5 mL PT, a lamellar morphology with a rough surface was observed, but no sphere was formed. (Fig. 1A–C). When 10 mL PT was used, a lamellar structure was still maintained for obtained products, and their surface changed into interconnected nanofibrous structure, as shown in Fig. 1D–F. When the volume of PT was increased to 20 mL, the lamellar morphology disappeared and microspheres started to form (Fig. 1G–I). The surface of the microspheres was compact with little shrink, which can be confirmed by FESEM image in Fig. 1I. The average diameter of the obtained microspheres was 103.7 μm with a wide particle size distribution. Further increasing the PT volume to 30 mL, the surface of the microspheres changed into a homogeneous interconnected nanofibrous structure, as illustrated in Fig. 1J–L. The prepared NHMs are monodisperse with no the sign of bulk aggregation and the average particle size was 69.2 μm with a relatively narrow size distribution. In order to observe the internal structure, an aqueous dispersion of NHMs was exposed to liquid nitrogen under ultrasound. It can be seen that the internal structure is hollow (Fig. 2).

Fig. 1 FESEM micrographs of the samples synthesized at different propanetriol volume: (A–C) PT-1; 5 mL, (D–F) PT-2; 10 mL, (G–I) PT-3; 20 mL, and (J–L) PT-4; 30 mL.

Fig. 2 FESEM micrograph of the cross-section of NHM synthesized at 30 mL PT.

As shown in Scheme 1, a hypothetical scenario of NHMs formation according to the reported literature is as follows,¹⁹ during the initial addition of a small amount of PT (less than 10 mL) to the PLLA/THF solution under mechanical stirring, the PT was dispersed as “water phase” droplets in the “oil phase” PLLA/THF solution which led to the formation “water-in-oil (W/O)” type emulsion (“droplet-in-drop” structure, PT-in-PLLA/THF). The emulsion stability is known to be associated with the equilibrium phase behavior in water-oil systems.²² As more glycerol was added, there was a “phase inversion” phenomenon in which the dispersed “water phase” (PT) became a continuous phase and “oil phase” (PLLA/THF) fraction became a dispersed phase which led to the formation of “water-in-oil-water” (W/O/W) type double emulsion (PT-in-PLLA/THF-in-PT) to form microspheres.¹⁸ The initial formation of small and viscous PT drops remained stocked inside the newly formed PLLA microspheres. Upon quenching in liquid nitrogen and the subsequent extraction of the solvent THF and PT, the double emulsions became hollow microspheres. The formation of nanofibrous structure was mainly due to the phase separation of the polymer solution.¹⁶ In brief, when the solution was placed in liquid nitrogen, the homogeneous PLLA/THF system would be thermodynamically unstable under the stage of rapid cooling. As a result, the PLLA solution was separated into polymer-rich phase and a polymer-lean phase. After that, the “solvent” (THF) was directly exchanged with a “different solvent” (DI water) and then sublimated by freeze-drying. After the polymer-rich phase solidifying, the desired and unique nanofibrous matrix structure was created.

Scheme 1 Schematic depicting the thermally induced phase separation technique to fabricate PLLA NHMs.

Effect of PLLA concentration

In order to determine the most appropriate PLLA concentration, we first prepared NHMs by adding 1% PLLA/THF solution. The SEM images in Fig. 3A–C showed that the surface of the obtained NHMs was shrinkage with collapse, and the nanofibrous structure could not be observed. Only the pore morphologies could still be maintained. When 2% PLLA/THF solution was used, the nanofibers started to form on the surface of the NHMs and NHMs became more uniform in size. As the concentration of PLLA/THF was increased to 4%, the surface morphology of the NHMs changed into homogeneous interconnected nanofibers, and the size of NHMs became bigger and particle size distribution became more uniform. Average particle size of NHMs increased from 69.2 to 94.7 μm when the PLLA concentration increased from 1% to 4%. Since PLLA is a polymer with a high molecular weight, the presence of PLLA in the external PT phase may increase the viscosity of the double emulsion, which results in an increased difficulty in breaking up the emulsion into small droplets, and thus facilitate to yield bigger microspheres.

Fig. 3 FESEM micrographs of the samples synthesized at different PLLA concentration: (A–C) P-1; 1%, (D–F) P-2; 2%, and (G–I) P-3; 4%.

Effect of PT dropping rate

As shown in Fig. 4A–C, when 40 mL h⁻¹ of the PT dropping rate was used for the preparation of the NHMs, crude nanofibrous structure was observed on the surface of the obtained products. However, the more fine nanofibrous structure could be observed onto the microspheres with the dropping rate of PT increasing from 90 to 140 mL h⁻¹. At the dropping rate of 90 mL h⁻¹, porous structure of NHMs is not destroyed (Fig. 4D–F) and the cross-connected and microporous nanofibers are clearly shown. A unique structure composed of nanofibers could be observed when the PT dropping rate increased to 140 mL h⁻¹ (Fig. 4G–I). When the PT dropping rate was increased from 40 to 140 mL h⁻¹, the average diameter of microspheres decreased from 128.0 to 76.5 μm. During the addition of PT to the PLLA solution under stirring, the PT was dispersed as liquid drops in the PLLA/THF solution. When PT was added at slow speed, longer time interval and smaller PT droplets quantity led to the decreased inner “water phase” number and correspondingly the increased size of NHMs.

Fig. 4 FESEM micrographs of the samples synthesized at different propanetriol dropping rate: (A–C) DR-1; 40 mL h⁻¹, (D–F) DR-2; 90 mL h⁻¹, and (G–I) DR-3; 140 mL h⁻¹.

Effect of stirring speed

At same concentration of PLLA solution, dropping rate, PT volume and dropping sequence, the effects of the stirring speed on the microsphere size were then studied. Stirring speed is the dominating factor for the microsphere size. As seen from Fig. 5, the morphology of NHMs prepared at 400 rpm was similar to that prepared at 800 or 1200 rpm. The NHMs were all highly dispersed with interconnected nanofibrous structures. When the stirring speed increased from 400 to 1200 rpm, the average particle size of NHMs decreased from 127.5 to 63.8 μm. While when the stirring speed was controlled at 1200 rpm, the particles size distribution was wide. By increasing the stirring speed, therefore, more energy was provided to disperse the “O/W” type emulsion (PT dispersed in the PLLA/THF solution). Higher shear forces were created, acted on the emulsion globules and then broke up the droplets, thereby forming smaller microspheres.²³

Fig. 5 FESEM micrographs of the samples synthesized at different stirring speed: (A–C) SR-1; 400 rpm, (D–F) SR-2; 800 rpm, and (G–I) SR-3; 1200 rpm.

Effect of addition sequence

The addition sequence of the components also affected the nanostructure of obtained NHMs. It was evident that the addition sequence strongly affected the uniformity and morphology of microspheres, as shown in Fig. 6. The result demonstrated that the NHMs prepared by the method of S-1 were polydispersed, while those prepared by S-2 were uniform. Furthermore, in the former case, the surface of NHMs was obvious shrinkage and not spherical, and the smaller particles adhered on the larger spheres to form coagulates due to the high specific surface energy of small particles. The median particle diameter reached to 69.23 μm. The morphology of the NHMs was only tiny pores without apparent nanofibers. The NHMs obtained in the method of S-2 were relatively uniform-sized distribution and the pores on the surface were well interconnected. During the addition of PLLA solution to PT under stirring, the polymer solution was dispersed as oil drops in PT. A low oil volume yielded a concentrated polymer solution and PT as a continuous phase was viscous, so that it was more difficult for the PLLA solution to be broken up into uniform and smaller oil droplets during the stirring process. Moreover, the polymer solution coagulated fast and was lack of sufficient stabilization, so yielded a tighter structure and irregular shapes.

Fig. 6 FESEM micrographs of the samples synthesized at different addition sequences: (A–C) S-1; PLLA-PT, and (D–F) S-2; PT-PLLA.

In vitro cytotoxicity effect on L929 cells

We further evaluated the cytotoxicity of the synthesized NHMs, which is very important for their biomedical applications. The cytotoxicity of NHMs was firstly determined by measuring the viability of L929 cells after incubation with NHMs. From Fig. 7A, no visible cytotoxicity and no significant difference were observed with and without the presence of NHMs in L929 cells even at high NHMs concentration of 800 μg mL⁻¹. Furthermore, it was found that similar results were obtained in CLSM observations. As shown in Fig. 7B, the cells were typically well-spread, and had no apparent change in cell morphology and cell density. Therefore, all these results indicated that the prepared NHMs possess good biocompatibility.

Fig. 7 (A) The cytotoxicity of NHMs against L929 cells incubated at concentration of 200, 400 and 800 μg mL⁻¹ for 24 and 48 h. (B) Confocal laser scanning microscopy images of L929 cells incubated with NHMs at concentration of 800 μg mL⁻¹ for 24 and 48 h. The scale bars represent 10 μm.

Conclusions

In summary, monodispersed NHMs with controllable size and morphology were successfully developed by utilizing a phase separation technique. The size and surface nanofibrous morphology of the obtained microspheres can be well controlled by adjusting the processing conditions, such as PT volume, PLLA concentration, dropping rate, stirring speed and addition sequence. The NHMs possess well-interconnected nanofibrous structures and good biocompatibility, and should therefore find wide applications in biomedical fields such as drug delivery and tissue engineering, especially as injectable cell or drug carriers for tissue regeneration.

Acknowledgements

This study was financially supported by the National Natural Science Foundation of China (31271028 and 21444002), Innovation Program of Shanghai Municipal Education Commission (13ZZ051), Open Foundation of State Key Laboratory for Modification of Chemical Fibers and Polymer Materials (LK1416) and Chinese Universities Scientific Fund (BCZD201506).

References

1. V. R. Sinha and A. Trehan, J. Controlled Release, 2003, 90, 261–280.
2. H. K. Kim, H. J. Chung and T. G. Park, J. Controlled Release, 2006, 112, 167–174.
3. T. K. Mandal, L. A. Bostanian, R. A. Graves, S. R. Chapman and T. U. Idodo, Eur. J. Pharm. Biopharm., 2001, 52, 91–96.
4. G. Crotts and T. G. Park, J. Controlled Release, 1995, 35, 91–105.
5. N. Rescignano, Y. Gonzalez-Alfaro, E. Fantechi, M. Mannini, C. Innocenti, E. Ruiz-Hitzky, J. M. Kenny and I. Armentano, Eur. Polym. J., 2015, 62, 145–154.
6. A. J. Lasprilla, G. A. Martinez, B. H. Lunelli, A. L. Jardini and R. M. Filho, Biotechnol. Adv., 2012, 30, 321–328.
7. J. J. Blaker, J. C. Knowles and R. M. Day, Acta Biomater., 2008, 4, 264–272.
8. C. J. Ke, T. Y. Su, H. L. Chen, H. L. Liu, W. L. Chiang, P. C. Chu, Y. Xia and H. W. Sung, Angew. Chem., Int. Ed. Engl., 2011, 50, 8086–8089.
9. R. de Smet, S. Verschuere, L. Allais, G. Leclercq, M. Dierendonck, B. G. de Geest, I. van Driessche, T. Demoor and C. A. Cuvelier, Biomacromolecules, 2014, 15, 2301–2309.
10. H. S. Yoo, Colloids Surf., B, 2006, 52, 47–51.
11. H. Valo, L. Peltonen, S. Vehvilainen, M. Karjalainen, R. Kostiainen, T. Laaksonen and J. Hirvonen, Small, 2009, 5, 1791–1798.
12. M. Enayati, Z. Ahmad, E. Stride and M. Edirisinghe, J. R. Soc., Interface, 2010, 7, 667–675.
13. C. Berkland, K. Kim and D. W. Pack, J. Controlled Release, 2001, 73, 59–74.
14. N. Bock, M. A. Woodruff, D. W. Hutmacher and T. R. Dargaville, Polymers, 2011, 3, 131–149.
15. H. Ghanbar, C. J. Luo, P. Bakhshi, R. Day and M. Edirisinghe, Mater. Sci. Eng., C, 2013, 33, 2488–2498.
16. C. L. He, W. Nie and W. Feng, J. Mater. Chem. B, 2014, 2, 7828–7848.
17. D. P. Go, D. J. E. Harvie, N. Tirtaatmadja, S. L. Gras and A. J. O'Connor, Part. Part. Syst. Charact., 2014, 31, 685–698.
18. Z. Zhang, M. J. Gupte, X. Jin and P. X. Ma, Adv. Funct. Mater., 2015, 25, 350–360.
19. X. H. Liu, X. B. Jin and P. X. Ma, Nat. Mater., 2011, 10, 398–406.
20. K. X. Qiu, C. L. He, W. Feng, W. Z. Wang, X. J. Zhou, Z. Q. Yin, L. Chen, H. S. Wang and X. M. Mo, J. Mater. Chem. B, 2013, 1, 4601–4611.
21. W. Feng, X. J. Zhou, C. L. He, K. X. Qiu, W. Nie, L. Chen, H. S. Wang, X. M. Mo and Y. Z. Zhang, J. Mater. Chem. B, 2013, 1, 5886–5898.
22. F. Bouchama, G. A. van Aken, A. J. E. Autin and G. Koper, Colloids Surf., A, 2003, 231, 11–17.
23. Y. Y. Yang, T. S. Chung and N. P. Ng, Biomaterials, 2001, 22, 231–241.

---

Footnote

† Electronic supplementary information (ESI) available: Details about the particle size distribution of NHMs synthesized at different condition. See DOI: 10.1039/c5ra11525f

---