Preparation and characterization of biodegradable nanoparticles based on amphiphilic poly(3-hydroxybutyrate)–poly(ethylene glycol)–poly(3-hydroxybutyrate) triblock copolymer

EUROPEAN POLYMER JOURNAL

European Polymer Journal 42 (2006) 2211–2220

www.elsevier.com/locate/europolj

Macromolecular Nanotechnology

Cheng Chen

a,b,*

, Chung Him Yu a, Yin Chung Cheng a, Peter H.F. Yu a, Man Ken Cheung a

a Open Laboratory of Chirotechnology of the Institute of Molecular Technology for Drug Discovery and Synthesis, Department of Applied Biology and Chemical Technology, Hong Kong Polytechnic University, Hong Kong, PR China b Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, PR China

Received 19 March 2006; received in revised form 3 July 2006; accepted 3 July 2006 Available online 30 August 2006

Abstract Amphiphilic triblock copolymers of poly(3-hydroxybutyrate)–poly(ethylene glycol)–poly(3-hydroxybutyrate) (PHB– PEG–PHB) were directly synthesized by the ring-opening copolymerization of b-butyrolactone monomer using PEG as macroinitiator. Their structure, thermal properties and crystallization were investigated by 1H NMR, differential scanning calorimetry (DSC) and X-ray diffraction. It was found that both PHB and PEG blocks were miscible. With the increase in the PHB block length, the triblock copolymers became amorphous because amorphous PHB block remarkably depressed the crystallization of the PEG block. Biodegradable nanoparticles with core-shell structure were prepared in aqueous solution from the amphiphilic triblock copolymers, and characterized by 1H NMR, SEM and fluorescence. The hydrophobic PHB segments formed the central solid-like core, and stabilized by the hydrophilic PEG block. The nanoparticle size was close related to the initial concentrations of the nanoparticle dispersions and the compositions of the triblock copolymers. Moreover, the PHB–PEG–PHB nanoparticles also showed good drug loading properties, which suggested that they were very suitable as delivery vehicles for hydrophobic drugs.  2006 Elsevier Ltd. All rights reserved. Keywords: Poly(3-hydroxybutyrate); Triblock copolymer; Amphiphilic; Nanoparticle

1. Introduction

*

Corresponding author. Tel.: +852 3400 8731; fax: +852 2364 9932. E-mail addresses: [email protected], bcchen92@ polyu.edu.hk (C. Chen).

Microbial poly(3-hydroxybutyrate) (PHB) is produced by many strains of bacteria as an intracellular storage material of carbon and energy. Because it is produced from renewable resources and biodegraded to carbon dioxide and water, PHB is often

0014-3057/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2006.07.001

MACROMOLECULAR NANOTECHNOLOGY

Preparation and characterization of biodegradable nanoparticles based on amphiphilic poly(3-hydroxybutyrate)–poly(ethylene glycol)– poly(3-hydroxybutyrate) triblock copolymer

MACROMOLECULAR NANOTECHNOLOGY

2212

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

described as environmentally friendly plastics [1]. So far PHB has attracted industrial attention as environmentally degradable materials for a wide range of agricultural, marine, and medical applications. As produced in bacteria, PHB has a relatively high molecular weight and excellent crystallization resulting from completely chirality in molecular chains, which is unsuitable for molecular design of specialty polymers such as amphiphilic block copolymers, e.g., with poly(ethylene glycol) (PEG) or poly(ethylene oxide) (PEO). Diblock amphiphilic copolymers of PHB and monomethoxy poly(ethylene glycol) (mPEG) were synthesized in an one-step process by catalyzed transesterification in the melt [2]. The resulting diblock copolymers were self-assembled into sterically stabilized colloidal suspensions of PHB crystalline lamellae. Although microbial PHB in the diblock copolymer showed more crystalline and water insoluble, the architecture of the copolymers was not still suitable as drug carries because the crystalline PHB segments might cause the dense core and was difficult to encapsulate drugs. ABAtype triblock copolymers consisting of PEO and PHB were investigated by coupling two chains of PEO with a low-molecular-weight isotactic PHB chain in the middle (PEO–PHB–PEO) [3]. The crystallinity of the PHB block in the copolymers clearly increased compared with that of the pure PHB precursor, and their critical micelle concentrations (CMCs) were in the range of 13–1100 mg/L [4]. In fact, the CMC values of the PEO–PHB–PEO copolymers were too high, especially comparing with that of poly(e-carprolactone) (PCL)–PEG and poly(L-lactide)(PLLA)–PEG (2.5–35 mg/L) [5,6]. The results implied that the PEO–PHB–PEO nanoparticles were very easy to dissociate upon dilution in the blood stream after intravenous injection, which was supported by the fact that all PEO–PHB–PEO copolymers were soluble in water [4]. It is worthy to be noted that intracellular native PHB is amorphous. During the recovery process, PHB becomes crystalline, and is hardly degraded by intracellular depolymerases [7]. In addition to the biosynthesis by bacterial fermentation, PHB can be synthesized by anionic polymerization of b-butyrolactone (BL) monomer. The chemical syntheses of low-molecular-weight PHB was performed using very sophisticated step-by-step polycondensation methods [8]. Because BL monomer consisted of R and S conformations, the chemical synthe-

sized PHB was completely amorphous and atactic without film-forming properties even at very high molecular weights. Moreover, [R,S]-3-hydroxybutyric oligomers were non-toxic and biocompatible, and can potentially serve to modify pharmacological properties in the form of chemical conjugates for drug delivery [9]. Therefore, it was proposed that atactic PHB were used to prepare the amphiphilic copolymer instead of microbial PHB, and new copolymer should possess some novel properties because amorphous PHB was close to its original state in the cells. According to our previous studies [10], the amphiphilic triblock copolymer PHB–PEG–PHB exhibited much smaller CMCs, especially compared with those of PEO–PHB– PEO copolymers. Moreover, they also showed shorter biodegradation period relatively to microbial PHB in vitro in the presence of enzyme. Pyrene was used as imitative drug and encapsulated into the PHB–PEG–PHB nanoparticles. It was found that the release of pyrene from the PHB– PEG–PHB nanoparticles exhibited the second order exponential decay behavior. These features suggested that the PHB–PEG–PHB nanoparticles were potential drug carriers for hydrophobic drugs. As one of our series of studies on the PHB–PEG– PHB nanoparticles as drug carriers, this article presented the ring-opening copolymerization of bbutyrolactone to produce triblock copolymers containing the PEG block in the middle and the PHB block as the end blocks (PHB–PEG–PHB). The chemical structure and properties of the triblock copolymers were also investigated. The nanoparticles were prepared with the amphiphilic copolymers in aqueous solution, and its formation was studies by 1H NMR, SEM and fluorescence by choosing pyrene as an imitative drug. 2. Materials and methods 2.1. Materials All chemicals, except otherwise stated, were purchased from Aldrich Chemical Co. b-Butyrolactone (BL) monomer and dichloromethane (CH2Cl2) were dried with calcium hydride and distilled under nitrogen atmosphere and lower pressure prior to use, respectively. PEG [number-average molecular weight (Mn) = 4000] was used after purification by precipitation method. Other organic solvents were analytical grade and used as received.

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

2.2.1. Polymerization of the triblock copolymers The purified BL monomer and PEG were placed in a dried polymerization flask with appropriate ratio, and stannous octoate [Sn(Oct)2] in desired concentrations was added as a solution in dried CH2Cl2. The reactants were mixed and then dried under reduced pressure at 40 C for 1 h to completely remove CH2Cl2. The reactions were carried out under nitrogen atmosphere at desired temperatures. After specified reaction time, the reacted products were dissolved in chloroform, and then precipitated in hexane several times to remove the unreacted BL monomer. The isolated products were dried in vacuum at 40 C for 48 h. 2.2.2. Nanoparticle preparation Aqueous dispersions of the nanoparticles were prepared by a precipitation/solvent evaporation technique without any surfactants for the investigation of their native ability of forming nanoparticles. A triblock copolymer solution in acetone was added dropwise to distilled and deionized water (DD H2O) under ultrasonic situation. Acetone was removed under lower pressure and ultrasonic situation. All dispersions were filtered using disposable 0.45 lm Millipore filters, without significant effect on the particle yield or size distribution. For NMR experiments, deuterated water (D2O) was used instead of DD H2O. 2.2.3. Drug loading content and drug loading efficiency (DLC and DLE) Pyrene was chosen as an imitative drug to determine DLC and DLE of the nanoparticles because of its unique fluorescence and hydrophobic characters. The pyrene solutions in DD H2O were prepared with different concentrations of 0.1–104 mg/L. Its absorbance at 333.1 nm in the fluorescence excitation spectra was measured to generate a calibration curve for quantitative analysis. The nanoparticle dispersions were prepared in the presence of pyrene (6 · 107 M), and their initial concentrations were 1 mg/mL. Excess pyrene was removed by dialysis in DD H2O for 2 days. Finally, the nanoparticle dispersions containing pyrene were centrifuged to separate the nanoparticles loaded pyrene. By comparing the change in the intensity at 333.1 nm, the amount of loaded pyrene was known. DLC and DLE were calculated as following:

DLC = weight of loaded pyrene in nanoparticles/weight of nanoparticles · 100%. DLE = weight of loaded pyrene in nanoparticles/ weight of pyrene added · 100%.

2.3. Measurements The NMR analysis of the specimens was carried out on a Varian Inova 500-MHz NMR spectrometer. Chemical shifts were given in ppm using tetramethylsilane (TMS) in CDCl3 as internal reference and sodium 3-trimethylsilylpropionate-d4 (TSP) in D2O as external reference. Differential scanning calorimetry (DSC) was carried out on a Mettler DSC 30 differential scanning calorimeter under nitrogen atmosphere at a heating rate of 10 C/min. Wideangle X-ray diffraction (WAXD) experiments were performed with a Philips PW 1700 X-ray diffractometer using Cu Ka X-rays with a voltage of 30 kV and a current of 20 mA in the 2h range 5– 40. The morphology of the nanoparticles was observed using a field-emission scanning electron microscopy (FE-SEM, JEOL JSM-6335F). All samples were sputter coated with a thin layer of golden in a vacuum. Fluorescence measurements were carried out with a Perkin–Elmer LS50B Luminescence spectrometer. The excitation wavelength was 339 nm, and the emission wavelength was 394 nm. Both excitation and emission bandwidths were 2.5 nm. The particle size was measured using a Malvern Zetasizer 3000HSA equipped with a 10 mW He–Ne laser (633 nm) and operating at 20 C and an angle of 90. The scanning number was 10. 3. Results and discussion 3.1. Synthesis and characterization of the triblock copolymers Triblock PHB–PEG–PHB copolymers were directly synthesized by ring-opening polymerization of BL monomer using PEG as macroinitiators. The method was easier and more convenient than those previously reported [2,3,8] because it simplified the reaction steps. After purification, 13C NMR was used to characterize the reacted products. Fig. 1 shows the 13C NMR spectrum of sample 1 (seen in Table 1). A strong peak was found at 70.5 ppm, which was attributed to the methylene carbon atoms of PEG block. Four relative weak resonances were also detected at 169.3, 67.6, 40.8 and 19.8 ppm.

MACROMOLECULAR NANOTECHNOLOGY

2.2. Experimental procedures

2213

MACROMOLECULAR NANOTECHNOLOGY

2214

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

Fig. 1. 13C NMR spectrum of the reacted products for sample 1, besides that of pure PEG.

They were assigned to carboxylic carbon, methine, methylene and methyl carbon atoms in 3-hydroxybutyrate units because their chemical shifts were in agreement with that of microbial PHB [11], which demonstrated that PHB block were synthesized by the ring-opening polymerization of BL monomer. The end groups of PEG methylene carbon was found at 72.7 ppm. After reaction, they almost disappeared, and a new peak appeared at 69.0 ppm. The results indicated that the hydroxyl end groups of PEG chain were consumed in the reaction, and then formed new ester bonds, resulting in upfield shifting of the PEG methylene carbon end groups from 72.7 to 69.0 ppm. Moreover, the peak ratio at 69.0–70.5 ppm was almost the same as that at 72.7–70.5 ppm, implying that no free PEG chains remained in the reacted product. Hence, it is con-

cluded that the reacted products were triblock copolymers consisted of a PEG central block and two PHB lateral blocks, and an ester linkage existed between PEG and PHB blocks. The scheme of synthesis of the triblock copolymers was shown in Scheme 1. Table 1 presents the characteristics of various PHB–PEG–PHB triblock copolymers synthesized at different reaction conditions. Triblock copolymers were named using the acronym HBx–EGy– HBx. In these acronyms, HB and EG represented the PHB and the PEG blocks, respectively; x and y represented the number-average degree of polymerization of the PHB and PEG blocks. Considering that PEG was almost monodispersed (Mw/ Mn = 1.08), the molecular weight of the PHB block in the triblock copolymers can also be determined from the ratio between integral of peaks for the PEG and PHB segments. Hence, the conversion of the BL monomer and Mn of the triblock copolymers were calculated according to the following equations: Conversion of BL monomerðBL Con:Þ ¼ I PHB =ðI PHB þ I BL Þ  100% M n ¼ DPPEG  44 þ DPPHB  86 where IPHB and IBL expressed the integration intensity of the PHB block at 1.24 ppm and unreacted BL monomer at 1.54 ppm; DPPEG = y = MnPEG/ 44, DPPHB = 2x = 4DPPEG · (HB/EG); 44 and 86 were the molar masses of EG and HB repeat units, respectively. As seen in Table 1, increasing reaction temperature and time favored to increase the conversion of BL monomer and Mn of the triblock copolymers. When reaction temperature was 170 C and reaction

Table 1 Characteristics of the PHB–PEG–PHB triblock copolymers synthesized in different conditions Samples

Structure (HBx–EGy–HBx)

Temperature (C)

Time (h)

Catalyst (%)

BL/PEG in feed

HB/EGa in products

BL Con.b (%)

Mnb (copolymer)

1 2 3 4 5 6 7 8 9

7–91–7 14–91–14 24–91–24 35–91–35 55–91–55 90–91–90 101–91–101 187–91–187 368–91–368

130 150 170 170 170 170 170 170 170

10 10 6 6 6 8 10 10 12

0.1 0.1 0.1 0.3 0.3 0.3 0.3 0.3 0.3

0.1 0.1 0.1 0.3 0.5 0.5 0.5 1.0 2.0

0.04 0.08 0.13 0.19 0.26 0.49 0.55 1.03 2.02

87 93 92 93 98 98 99 98 100

5274 6416 8169 10,700 12,173 19,547 21,373 36,243 67,391

a b

Calculated for the integration of NMR resonances belonging to PEG block at 3.64 ppm and to PHB block at 5.26 ppm. Calculated by 1H NMR.

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

2215

O Sn(Oct)2 +

H

O

CH 2

CH2

O

OH

T, t

m

PEG

CH3

BL monomer CH 3 O

CH

O CH2

O

CH 3

C

O

CH

CH2

C

O

CH2

CH2

O

CH2

CH2

n

PHB m-1

PHB-PEG-PHB Scheme 1. Synthesis of the amphiphilic PHB–PEG–PHB triblock copolymers.

3.2. Thermal properties and crystallization of the PHB–PEG–PHB triblock copolymers Fig. 2 shows that the DSC curves of PEG and the PHB–PEG–PHB triblock copolymers with different PHB block lengths at a heating rate of 10 C/min from 60 C to 100 C (run I). PEG was partially crystalline polymer with a melting point about 59.9 C. As the PHB block length increased, the melting peaks of the triblock copolymers shifted to lower temperatures. Considering that atactic PHB was amorphous, its contribution to the melting peak

Samples

3 5 6 8 9

-40

-20

0

20

40

60

80

100

Temperature (°C)

Fig. 2. DSC curves of PEG and the PHB–PEG–PHB triblock copolymers with different PHB block lengths at a heating rate of 10 C/min from 60 to 100 C.

might be negligible. Hence, the melting peaks of the PHB–PEG–PHB triblock copolymers should mainly result from the melting of the PEG block. Table 2 lists the glass transition temperatures (Tg), melting temperatures (Tm) and melting enthalpies (DHm) of PEG and the PHB–PEG–PHB triblock copolymers. When the PHB block length increased from 7 to 55, the Tms of the triblock copolymers reduced from 50.7 to 36.3 C. Meanwhile, their melting enthalpies decreased from 102.9 to 33.9 J/g. This implied that the crystallization of the PHB– PEG–PHB triblock copolymers became worse. Considering that the melting peaks mainly resulted from the melting of the PEG block, the crystallinity (Xc) of the triblock copolymers was estimated from DHm assuming the heat of fusion DH m of 100% wt% crystalline PEG was 208 J/g [12]. The Xc results were also listed in Table 2. For samples 1, 3 and 5, their Xc was 49.4%, 25.4% and 16.2%, respectively. The results mean that the introduction of

MACROMOLECULAR NANOTECHNOLOGY

PEG 1

Endo

time was 10 h, the conversion of BL monomer reached to 99%. This was an expectable behavior because the ring-opening copolymerization may be fully completed. It is also seen in Table 1 that the BL/PEG ratio in feed had an important influence on Mn of the triblock copolymers. As it was raised from 0.1 to 2.0, the Mn of the triblock copolymers was increased from several thousands to about 60,000. For the triblock copolymers with lower Mn, such as sample 5, the HB/EG ratio in the product (0.26) was clearly lower than the BL/PEG ratio in feed (0.5). Considering its lower conversion of BL monomer, the result could be assigned to the fact that BL monomer did not fully react. For the triblock copolymers with higher Mn, they showed the same HB/EG ratio as that of BL/PEG in feed. Moreover, they also showed the high conversion of BL monomer. Therefore, simply by changing the Mn of PEG and the ratio of monomer to PEG, the compositions and Mn of the triblock copolymers can be easily adjusted and controlled. Compared with traditionally biosynthesis, the ring-opening copolymerization of BL monomer was easier to obtain copolymers with various chemical structures and compositions.

2216

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

Table 2 DSC data of PEG and the PHB–PEG–PHB triblock copolymers with different PHB block lengths Samples

Run I Tm (C)

DHm (J/g)

Xc (%)

Tg (C)

Tcc (C)

DHcc (J/g)

Tm (C)

DHm (J/g)

Xc (%)

– – – – 35.4 33.3 29.9

59.9 50.7 42.2 36.3 – – –

185.7 102.9 52.8 33.9 – – –

89.2 49.4 25.4 16.2 – – –

– – 43.9 41.0 31.8 29.7 24.8

– – 14.5 0.65 – – –

– – 10.9 32.1 – – –

55.4 60.2 44.3 38.9 34.9 – –

170.1 86.8 48.0 30.0 – – –

81.7 41.7 23.1 14.4 – – –

the PHB block remarkably depressed the crystallization of the PEG block. In the case of samples 6, 8 and 9, no melting peak was detected, implying that all of them were amorphous. However, a single Tg was observed, and it increased with the PHB block length (Fig. 2 and Table 2). It was known that the Tgs of PEG and atactic PHB was about 65 and 0.6 C, respectively [13,14]. Tgs of the triblock copolymers dependent on the compositions were found in the middle of PEG and atactic PHB Tg. Therefore, both PEG and PHB blocks in the copolymers were miscible in the amorphous state. Fig. 3 shows DSC melting curves of PEG and the PHB–PEG–PHB triblock copolymers with different PHB block lengths non-isothermally crystallized from the melt at a cooling rate of 10 C/min (run II). No Tg was detected for PEG nor sample 1 because they crystallized too rapidly. For other triblock copolymers, single Tgs were obviously shown in the DSC melting curves, and increased with the

Samples PEG 1 3 5

Endo

MACROMOLECULAR NANOTECHNOLOGY

PEG 1 3 5 6 8 9

Run II

Tg (C)

6 8 9

-60

-40

-20

0

20

40

60

80

Temperature (°C)

Fig. 3. DSC melting curves of PEG and the PHB–PEG–PHB triblock copolymers with different PHB block lengths nonisothermally crystallized from the melt at a cooling rate of 10 C/min.

PHB block length. The results indicated that both PEG and the PHB blocks were miscible in the melt, similar to the result of atactic PHB and PEG blends [14]. The increase in Tg also suggested that the chain mobility of the triblock copolymers became weak because the segments had to be mobile at higher temperature. During the heating process, the DSC melting curve of PEG showed multiple melting peaks which resulted from the recrystallization, while those of samples 1, 3 and 5 only showed a single melting peak. Their melting peaks gradually shifted to lower temperatures as the PHB block length increased, and the peak width clearly increased. The results indicated that the crystal perfection of the triblock copolymers decreased. The imperfect crystals might be melted at lower temperature, and the melting range of the triblock copolymers was expanded. Moreover, the Xc of the triblock copolymers also remarkably reduced with the PHB block length. When the PHB block length reached to 55, its Xc reduced to 14% (Table 2). The longer the PHB block length, the lower the Xc of the copolymers. The results further confirmed that the presence of the PHB block hindered the crystallization of the PEG block. When the PHB block was long enough, the melting peak of the triblock copolymers disappeared, as shown for samples 6, 8 and 9. Hence, together with the DSC results in run I, samples 6, 8 and 9 should be fully amorphous. For samples 3 and 5, their cold crystallization peak (Tcc) was observed and exhibited the similar trend as Tg. Cold crystallization mainly resulted from the inability of all crystallizable chains to crystallize fully during the cooling cycles. The remaining chains could recrystallize again when the segments regained mobility upon heating above Tg. Consequently, the increase in Tcc can also be explained by considering the chain mobility. The X-ray diffraction results of pure PEG and samples 3, 5 and 6 are shown in Fig. 4. Two crystal-

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

2217

Samples

Sample 6

6

5 3

PEG CH2 PEG

20

30

40

5

4

3

2

1

2θ

Chemical shift (ppm)

Fig. 4. X-ray diffraction profiles of pure PEG and the PHB– PEG–PHB triblock copolymers for samples 3, 5 and 6.

Fig. 5. 1H NMR spectrum of the PHB–PEG–PHB nanoparticles for sample 6 in D2O.

line characteristic peaks of PEG were observed for three triblock copolymers, while the diffraction assigned to the PHB block was not clearly found. The diffraction intensities of the triblock copolymers remarkably weakened relatively to those of pure PEG, implying that the Xc of the PEG block obviously decreased and their crystal perfection also became worse. Compared the diffractions of both PEG and the triblock copolymers, it was found that the d-spacing values of the PEG block were constant for all crystallographic planes, which suggested that the PEG unit cells were not changed in the copolymers. The PHB segments were excluded from the crystal lattice of PEG during the crystallization process. For other triblock copolymers with longer PHB block, no diffraction peak was found (no shown), indicating that they hardly crystallize.

seen as a symmetric singlet. Hence, it should be fully solvated to form the hydrophilic outer shell and then stabilize the nanoparticles. Only very small fraction of PHB groups was detected in the spectrum, which resulted from the different environments. This indicates that the majority of the PHB protons was in a solid environment and consequently was not detected in the NMR experiment because the spectrophotometer was configured for normal liquid-state NMR. The hydrophobic PHB segments would entrap in the central solid-like core to minimize the interaction with water. Fig. 6 shows the SEM micrograph of the PHB–PEG–PHB nanoparticles for sample 9. The image showed that the PHB–PEG–PHB copolymers

3.3. Characterization and properties of the PHB–PEG–PHB nanoparticles 3.3.1. Characterization of the PHB–PEG–PHB nanoparticles According to the preparation procedure of nanoparticle dispersions, the microphase separation of both blocks was induced by the diffusion of acetone into water. 1H NMR was employed to study the structure of the PHB–PEG–PHB nanoparticles, since the nanoparticles were in the same conformational state when they were dispersed in water and in D2O. Fig. 5 shows the 1H NMR spectrum of the PHB–PEG–PHB nanoparticles for sample 6 in D2O. The methylene groups of PEG block was clearly shown at 3.67 ppm, and its resonance was

Fig. 6. SEM micrograph of the PHB–PEG–PHB nanoparticles for sample 9.

MACROMOLECULAR NANOTECHNOLOGY

10

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220 Sample 6 Concentration (g/L) 1. 1 2. 0.1 3. 0.01 4. 0.001 5. 0.0001 6. 0.00001 7. pyrene in H2O

I1 I3

Intensity

formed spherical, discrete particles in aqueous medium. Because the PHB–PEG–PHB copolymers were amorphous and its Tg was lower than room temperature, the nanoparticles were very soft and easy to deform. Hence, they showed the irregular shape in solid state. Fluorescence technique was also used to confirm the formation of the PHB–PEG–PHB nanoparticles using pyrene as hydrophobic probe because it was preferentially dissolved into the hydrophobic core. When pyrene environment changed from a polar to a non-polar one, there was a remarkable change in its fluorescence spectra, such as an increase in the quantum yield and a decrease in its intensity ratio (I1/I3) of the first and third highest emission peaks [15]. The ratio can change from 1.8 in water to 0.6 in organic solvents [15] or 1.0 in the presence of anionic surfactant micelles [16]. Therefore, the ratio was an indicator of where pyrene was located. Fig. 7(a) showed fluorescence emission spectra of the PHB–PEG–PHB copolymers for sample 6 at different concentrations in the presence of pyrene (6 · 107 M) and that in pure DD H2O. The pyrene concentration was lower than its saturated concentration in water (7 · 107 M) in order to prevent the formation of microcrystals. Pyrene in DD H2O had lower fluorescence intensity because of its short lifetime (200 ns) and low quantum yield. For the nanoparticle dispersions of sample 6, with the increase in copolymer concentrations, the emission intensity of pyrene generally increased, suggesting that pyrene was shifted from polar water environment to less polar one because pyrene had a longer lifetime and a higher quantum yield in a non-polar environment. Moreover, the I1/I3 ratio of pyrene in water was 1.85, the same as the literature value. The I1/I3 ratio for sample 6 was about 1.1–1.4, higher than that the expected values in a hydrophobic environment. The higher ratio can be attributed to the low copolymer concentration (1 · 105 g/L) and the polar surface of PEG [17]. Fig. 7(b) shows the corresponding excitation spectra of sample 6. When the copolymer concentrations increased from 1 · 105 to 1 g/L, the characteristic (0, 0) band of pyrene shifted from 333.1 to 336.6 nm, which were the characteristic values of pyrene in water and in the hydrophobic domain. Therefore, it can be concluded that the PHB–PEG–PHB nanoparticle were formed in aqueous solution once the amphiphilic triblock copolymers dispersed in water, and pyrene was located into the hydrophobic PHB core.

1 2 3 4 5 6 7

360

380

400

420

440

460

480

500

Wavelength (nm)

Intensity

MACROMOLECULAR NANOTECHNOLOGY

2218

Sample 6 Concentration (g/L) 1. 1 2. 0.1 3. 0.01 4. 0.001 5. 0.0001 6. 0.00001

1

2

43 5 6

270

280

290

300

310

320

330

340

350

360

Wavelength (nm) Fig. 7. Fluorescence emission (a) and excitation (b) spectra of the PHB–PEG–PHB nanoparticles for sample 6 with different concentrations in the presence of pyrene (6 · 107 M).

3.3.2. Size of the PHB–PEG–PHB nanoparticles It was known that a triblock copolymer chain can undergo either an intrachain association to form a loop or an interchain bridge between two different clusters. The interchain association can be suppressed by diluting the initial organic solution and increasing the speed of the microphase inversion [18]. Fig. 8 showed the effect of initial copolymer concentrations on particle size. The results showed that particle size increased with increasing initial copolymer concentrations. This can be ascribed to a higher probability of intrachain assembly at higher polymer concentration. When the PEG block was identical, particle size was also dependent on the PHB block length because a longer PHB block can enhance its assembly and lead to a larger core. In addition, the PHB–PEG–PHB nanoparticles also showed highly monodispersity, especially

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220 140

7

1 mg/mL 10 mg/mL

120

2219

DLC DLE

60

6

DLC (mg/g)

80 60

40 4 30

3

40

2

20

1

20

10 20

Sample 3

Sample 6

Sample 8

for the nanoparticles with the shorter PHB block length [polydispersity (Dv/Dn) = 1.2–1.8, where Dv and Dn were the volume and number-average particle diameters, respectively]. 3.3.3. DLC and DLE of the PHB–PEG–PHB nanoparticles In order to study the capacity of the PHB–PEG– PHB nanoparticles to load hydrophobic drugs, pyrene was used as an imitative drug. As shown in Fig. 9, after loaded pyrene, the size of the PHB– PEG–PHB nanoparticles clearly increased. With the increase in the PHB block length, the size of the PHB–PEG–PHB nanoparticles loaded pyrene increased obviously because of higher DLC. Moreover, the loading of pyrene did not affect the size

140 blank nanoparticles nanoparticles loaded pyrene

100 80

60

80

100

120

140

160

180

200

PHB block length

Sample 9

Fig. 8. The effect of initial concentrations and the PHB block length on particle size at copolymer concentrations of 1 and 10 mg/mL. Data present the mean and standard deviation of 10 independent experiments.

120

40

Fig. 10. DLC and DLE of the PHB–PEG–PHB nanoparticles using pyrene as imitative drug as a function of the PHB block length.

distribution of the PHB–PEG–PHB nanoparticles, and their polydispersity was among 1.1–1.6. Fig. 10 shows DLC and DLE of the PHB–PEG– PHB nanoparticles with different PHB block lengths. It was seen that the DLC and DLE linearly increased with the PHB block length. When the PHB block length increased from 24 to 187, DLC was raised from 1.4 to 6.7 mg/g. Meanwhile, DLE was raised from 13.9% to 60.9%. Physical encapsulation of hydrophobic drugs into polymeric nanoparticles was mainly driven by the hydrophobic interactions between drugs and hydrophobic segments of polymers. Generally, drug hydrophobicity was of very important to drug encapsulation. Considering that pyrene was inherently hydrophobic, it was easy to be entrapped by the longer PHB block because of its stronger hydrophobic interaction. In addition, the crystallinity of the amphiphilic copolymers also affected their DLC and DLE. Higher crystallinity in the nanoparticle core would decrease DLC and DLE because only the amorphous phase was likely to accommodate drug molecules. For the PHB–PEG–PHB nanoparticles, the hydrophobic cores were constructed with the amorphous PHB segments, whose architecture favored to encapsulate hydrophobic drugs.

60

4. Conclusions

40 20 0 Sample 3

Sample 6

Sample 9

Fig. 9. Particle size of the PHB–PEG–PHB nanoparticles before and after loading pyrene. Data present the mean and standard deviation of 10 independent experiments.

By the ring-opening copolymerization of BL monomer, amphiphilic triblock copolymers of PHB–PEG–PHB with different PHB block lengths were synthesized for potential drug delivery. Relatively to previous methods, the present method was easier and more convenient. Single Tg depended on

MACROMOLECULAR NANOTECHNOLOGY

0

Particle size (nm)

5

DLE (%)

Particle size (nm)

50

100

MACROMOLECULAR NANOTECHNOLOGY

2220

C. Chen et al. / European Polymer Journal 42 (2006) 2211–2220

the compositions was found, indicating that the PHB and PEG blocks were miscible. The PHB block can seriously hinder the crystallization of PEG block. The triblock copolymers with longer PHB block were amorphous. The results of 1H NMR, SEM and fluorescence suggested that the amphiphilic triblock copolymers can assemble to nanoparticles with core-shell structure in aqueous solution, and the hydrophobic PHB core was stabilized by hydrophilic PEG shell. When the PEG block length was identical, the particle sizes were affected by initial concentrations and their PHB block lengths. By using pyrene as imitative drug, it was also found that DLC and DLE of the PHB–PEG–PHB nanoparticles linearly increased with the PHB block length. More detail works are in progress. Acknowledgements Financial support from the University Grant Council of Hong Kong for Grant # PolyU 5299/ 01P, PolyU 5257/02M, and PolyU 5403/03M, and from the University Grants Committee Area of Excellence Scheme (Hong Kong) AoE/P-10/01 is greatly appreciated. References [1] Doi Y. Microbial polyester. New York: VCH Publisher; 1990.

[2] Ravenelle F, Marchessault RH. Biomacromolecules 2002;3:1057–64. [3] Li J, Li X, Ni XP, Leong KW. Macromolecules 2003;36: 2661–7. [4] Li J, Li X, Ni XP, Tan NK. Langmuir 2005;21:8681–5. [5] Yasugi K, Nagasaki Y, Kato M, Kataoka K. J Controlled Release 1999;62:89–100. [6] Hagan SA, Coombes AGA, Garnett MC. Langmuir 1996;12:2153–61. [7] Lee SY. Biotech Bioeng 1996;49:1–14. [8] Kemnitzer JE, Gross RA, McCarthy SP. J Environ Polym Degrad 1995;3:37–47. [9] Piddubnyak V, Kurcok P, Matuszowicz A, Glowala M, Kierzkowska AF, Jedlinski Z, et al. Biomaterials 2004;25:5271–9. [10] Chen C, Yu CH, Cheng YC, Yu Peter HF, Cheung MK. Biomaterials 2006;27:4804–14. [11] Chen C, Fei B, Peng SW, Wu H, Zhuang YG, Chen XS, et al. J Polym Sci: Part B: Polym Phys 2002;40:1893–903. [12] Cooke J, Viras K, Yu GE, Sun T, Yonemitsu T, Ryan AJ, et al. Macromolecules 1998;31:3030–9. [13] He Y, Zhu B, Kai WH, Inoue Y. Macromolecules 2004;37:3337–45. [14] Shuai XT, He Yong, Na YH, Inoue Y. J Appl Polym Sci 2001;80:2600–8. [15] Wilhelm M, Zhao CL, Wang YC, Xu RL, Winnik MA. Macromolecules 1991;24:1033–40. [16] Kalyanasundaram K, Thomas JK. J Am Chem Soc 1977;99:2039–45. [17] Gan ZH, Jim TF, Li M, Zhao Y, Wang SG, Wu C. Macromolecules 1999;32:590–4. [18] Zhao Y, Liang HJ, Wang SG, Wu C. J Phys Chem 2001;105:848–51.