1 Introduction

Since the first observation of conductive behavior in halogen-doped polyacetylene in 1977, there has been significant interests to explore conductive polymers for their unique physical and chemical properties [1]. The attributes that render these conducting polymers attractive include tunable electrical conductivity, structural flexibility, and processing adaptability. To date, a variety of conductive polymers including polyaniline, poly(3,4-ethylenedioxythiophene), polyacetylene, and polypyrrole (PPy) have been synthesized and studied with various success [2,3,4,5,6]. Among them, the PPy has attracted the most attention because of its impressive electrical, electrochemical, and optical characteristics, as well as excellent environmental stability. Among many forms of PPy, the hollow PPy microcapsules are of particular interest for promising potentials in rechargeable batteries, supercapacitors, and electrocatalyst supports [7,8,9,10].

Successful formation of hollow PPy microcapsules usually entails a template approach in which a sacrificial core is used to prepare a core@shell microsphere, followed by the selective removal of the core. For examples, core@PPy microspheres have been prepared for applications in bio-sensors, visual biomedical diagnostics, fuel cells, catalysis, and absorbers for electromagnetic waves [10,11,12,13,14,15,16,17,18,19]. In the literatures, the widely studied cores are silica and polystyrene (PS) microspheres [20,21,22,23,24,25,26,27]. Between them, the PS microspheres are recognized for a wider processing window, more freedoms in size availability and uniformity, as well as easier to be selectively removed by either chemical extraction or combustion.

During the synthesis of core@PPy microspheres, it is recognized that uniform and well-defined spherical core@PPy microspheres are difficult to obtain. So far, several groups have investigated relevant processing parameters responsible for the morphology formation during core@PPy synthesis. For instance, Yang et al. [20] reported that a pre-treatment of silica with different modified agents was critical for better morphology control in silica@PPy. In addition, Huang et al. [28] proposed an “inside-out” approach to prepare PS@PPy microspheres in which the pyrrole monomers were pre-absorbed inside the PS microspheres, followed by a drive-out step to render the polymerization of pyrrole on the PS surface. In their study, they observed that the morphology for PS@PPy microspheres was affected by varying the PS/pyrrole weight ratio or the addition rate of acid. Another study on the effect of hydrophilicity of the core during PPy formation by Cairns et al. [29] indicated that a better PPy coating was achieved on a hydrophobic poly(n-butyl methacrylate) (PBMA) surface instead of a hydrophilic poly(methyl methacrylate) (PMMA) one. To the best of our knowledge, however, the adjustment of chemical affinity of pyrrole monomers in an aqueous solution by the addition of organic solvent to control the resulting morphology of PS@PPy microspheres has not been investigated yet.

In this work, we employed a cosolvent approach to determine the optimized water/ethanol ratio for the formation of homogeneous PS@PPy microspheres. The effect of water/ethanol volume ratio was carefully studied with comprehensive physical and materials characterization. Subsequently, a chemical extraction process was adopted to produce uniform hollow PPy microcapsules.

2 Materials and method

2.1 Synthesis of PS and PS@PPy microspheres

PS microspheres with an average diameter of 490 nm were synthesized via an emulsifier-free emulsion polymerization process. The synthesis was conducted in a batch reactor equipped with a bladed pitched paddle impeller. The reactor contained 2 L DI water and was deoxygenated under N2 at 65 °C for 12 h. Then, styrene (monomer), C8H7NaO3S (co-monomer), and NaHCO3 (buffer) were mixed to form a homogeneous solution. After 15 min, K2S2O8 (initiator) was added, followed by stirring at 370 rpm for 16 h before cooling to 25 °C. Subsequently, the mixture was immersed in a water bath at 65 °C to evaporate residual solvent, resulting in PS microspheres in white powdery form.

To synthesize PS@PPy microspheres, 0.6 g PS microspheres and 0.1 g poly(N-vinylpyrrolidone) (PNVP; stabilizer) were mixed in 30 mL solution containing different volume ratios of DI water and ethanol (the ethanol volume percentage was varied from 1.5 to 60% whereas the PS/PPy weight ratio was maintained at 2.5). The mixture was subjected to ultrasonication for 4 h at 25 °C. Afterward, 1.6 g FeCl3·6H2O (oxidant) was dissolved for 1 h, followed by the addition of 249 μL pyrrole (98 wt%) to initiate the polymerization process. The polymerization reaction lasted for 24 h at 25 °C. Subsequently, the sample underwent three centrifugation/re-dispersion cycles during which the supernatants were replaced with DI water in order to remove residual inorganic by-products (FeCl2 and HCl). The resulting dispersion was dried at 60 °C for 12 h to obtain PS@PPy microspheres in black powders.

The PNVP, with a nominal molecular weight of 360,000, was purchased from Aldrich and used without further purification. Pyrrole (Alfa) and Styrene (Aldrich) were removed of inhibitor by passing through a column of activated basic alumina prior to use. FeCl3·6H2O (Aldrich), K2S2O8, C8H7NaO3S, and NaHCO3 were used as received.

2.2 Formation of hollow PPy microcapsules

Hollow PPy microspheres were fabricated by immersing 10 mg PS@PPy microspheres in 20 mL tetrahydrofuran (THF) under stirring at 25 °C for 24 h to dissolve the PS microspheres. The resulting hollow PPy microcapsules were isolated from THF by centrifugation at 6000 rpm for 15 min, followed by thorough ethanol washing for three times. Lastly, the sample was dried at 50 °C for 12 h before further material characterization. Figure 1 presents the schematic for the synthetic steps involved for the formation of hollow PPy microcapsules.

Fig. 1
figure 1

A schematic for the synthetic steps involved for the fabrication of PS@PPy microspheres and hollow PPy microcapsules

Full size image

2.3 Materials characterization

A scanning electron microscope (SEM; JEOL-LSM-6700) was used to observe morphologies of PS and PS@PPy microspheres, as well as hollow PPy microcapsules, respectively. To prepare SEM specimen, the samples were casted onto an ITO glass, followed by a Pt sputtering. Afterward, the specimens were maintained under 9.63 × 10−5 Pa, and the SEM observation entailed an acceleration voltage of 15 kV with an emission current of 10 μA. A transmission electron microscope (TEM; Philips/FEI TECNAI 20) was employed to observe the morphology of hollow PPy microcapsules. To prepare TEM specimen, the sample was dropped onto a copper grid, and residual solvent was evaporated at 60 °C for 12 h. Fourier transform infrared (FTIR) analysis was performed to validate the diffusion of pyrrole molecules to PS microspheres and the FTIR spectra were obtained using a FTIR spectrometer (PerkinElmer spectrum 100) in N2 atmosphere. Thermogravimetric analysis (TGA; TA Instruments Q500) was conducted to determine the exact amount of PPy in PS@PPy microspheres. The heating rate was 10 °C min−1 from 25 to 800 °C in N2 atmosphere. The zeta potential and polydispersity of PS and PS@PPy microspheres were determined via a particle analyzer (Beckman Coulter, Delsa Nano C). For zeta potential measurements, the pH value for the dispersion was adjusted by 0.1 M HCl and NaOH aqueous solution.

3 Results and discussion

Successful fabrication of uniform hollow PPy microcapsules requires monodispersive PS microspheres as the template with consistent coating of PPy. Therefore, the processing steps for the synthesis of homogeneous PS@PPy microspheres becomes rather critical. Figure 2a, b exhibit the SEM images of PS microspheres in low and high magnification. As shown, the PS microspheres appeared in spherical shape with a smooth surface morphology. In addition, their size was rather consistent and the average diameter was 490 nm with a standard deviation of 7.4 nm. It is to be noted that we did not observe any deformation as our synthetic process was rather robust and reproducible. However, the PS@PPy microspheres synthesized without the addition of ethanol revealed considerable deformation, as indicated in red circles in Fig. 2c. We believed that the formation of deformed PS@PPy microspheres was due to the diffusion of pyrrole molecules into the PS microspheres during the synthetic stage as both PS microspheres and pyrrole molecules are hydrophobic in nature. Therefore, in an aqueous medium, the pyrrole molecules preferred to stay with PS microspheres instead of dispensing in water. As a result, the pyrrole molecules acted as a plasticizer in softening the PS microsphere, causing the latter to deform moderately. Moreover, since those individual PS microspheres were not dispersed well in an aqueous solution, they tended to aggregate into large assembly. Thus, during the polymerization of pyrrole molecules, the PPy also covered and bonded those individual PS microspheres together, leading to irregularly-shaped agglomerates with sizes larger than 10 μm, as shown in Fig. 2d.

Fig. 2
figure 2

The SEM images of PS microspheres in a low and b high magnification, as well as PS@PPy microspheres showing c deformations (as indicated in red circles) and d sever aggregation. These PS@PPy microspheres are synthesized without the addition of ethanol

Full size image

In order to validate the diffusion of pyrrole molecules to PS microspheres during our synthetic step, we obtained FTIR spectra from PNVP-stabilized PS microspheres and pyrrole molecules, as well as sample with identical formulation of PS@PPy microspheres except without the addition of FeCl3·6H2O (labeled as PS-pyrrole in Fig. 3). This ensured that the polymerization of pyrrole molecules was not taking place so the individual pyrrole molecules were still entrapped in the PS microspheres. As shown in Fig. 3, there appeared two distinct absorption peaks associated with pyrrole molecules, and they were located at 736 and 3404 cm−1, respectively. These two signals corresponded to C=C bending and N–H stretching vibration. The PNVP-stabilized PS microspheres revealed obvious peaks at 699, 1453, 1494, 1602, 2929, and 3027 cm−1 from PS microspheres, as well as 1670 cm−1 from the pyrrolidone carbonyl group of PNVP. It is noted that our results agreed well with what was reported in earlier literature [22]. From the pyrrole-PS FTIR profile, we observed an identical pattern like PNVP-stabilized PS microspheres. In addition, a subdued absorption peak around 3440 cm−1 was recorded. This confirmed the presence of pyrrole molecules, providing a solid evidence about the diffusion of pyrrole molecules to the PS microspheres in an aqueous solution. A similar pattern was reported earlier by Huang et al. [28] in which a gradual disappearance of pyrrole molecular droplets after mixing pyrrole molecules and PS microspheres in an aqueous solution was observed under optical microscope observation. In addition, according to Zhang et al., the diffusion of pyrrole molecules to PS microspheres was anticipated because the PS microspheres were completely dissolved in pyrrole when the PS microspheres and pyrrole molecules were mixed together [30].

Fig. 3
figure 3

The FTIR spectra of pyrrole molecules, PNVP-stabilized PS microspheres, and PS microspheres mixed with pyrrole molecules

Full size image

To reduce undesirable deformation and aggregation of PS@PPy microspheres, ethanol was adopted as a cosolvent because the ethanol not only is miscible with water, but also exhibits greater affinity toward both pyrrole molecules and PS microspheres. With the presence of ethanol, we expected the pyrrole molecules to remain in the liquid phase rather than diffusing into the PS microspheres (organic phase). In addition, using ethanol as a cosolvent, the PS microspheres were likely to disperse more uniformly in the mixture. Both features suggested the feasibility of fabricating PS@PPy microspheres with significantly reduced deformation.

A series of PS@PPy microspheres were synthesized by varying the ethanol volume percentage from 1.5 to 60% while keeping the PS/pyrrole weight ratio at 2.5. The resulting SEM images are provided in Fig. 4. Apparently, a lower ethanol volume percentage of 1.5–10% engendered PS@PPy microspheres in spherical shape with subdued deformation and aggregation (shown in Fig. 4a–d). In contrast, with an ethanol volume percentage of 15% and above, there appeared noticeable irregularly-shaped PS@PPy microspheres (shown in Fig. 4e–h). However, the undesirable aggregation pattern was mostly inhibited as long as ethanol was added. In short, the optimized formula for fabricating homogeneous PS@PPy microspheres with significantly reduced deformation was 7 vol% of ethanol.

Fig. 4
figure 4

The SEM images of PS@PPy microspheres synthesized with different volume percentages of ethanol as a cosolvent: a 1.5, b 3.5, c 7, d 10, e 15, f 30, g 45 and h 60%

Full size image

To quantitatively determine the deformation ratio of PS@PPy microspheres synthesized from various volume percentages of ethanol, we selected 18 SEM images at 10,000 magnification from each sample, and calculated the number of deformed PS@PPy and un-deformed PS@PPy microspheres, respectively. To define a deformed PS@PPy microsphere, we first determined a (the longest diameter of a PS@PPy microsphere) and b (the shortest diameter of the same PS@PPy microsphere). Then, the x was obtained using the following equation;

$$ x = \left[ {\left( {a - b} \right)/\left( {\frac{a + b}{2}} \right)} \right] $$
(1)

where the x is a parameter indicative of deformation degree, and any PS@PPy microsphere with x larger than 7% is defined as a deformed PS@PPy microsphere. The deformation ratio, shown in Fig. 5 (left Y-axis), was then determined by dividing the number of deformed PS@PPy microspheres by the total number of PS@PPy microspheres in those 18 SEM images. Apparently, the deformation ratio was substantially reduced from 7 to 3.5 vol%.

Fig. 5
figure 5

a The deformation ratio (left Y-axis) and polydispersity (right Y-axis) of PS@PPy microspheres synthesized with different volume percentages of ethanol as a cosolvent. b The schematic illustrating the distribution of PS microspheres and PPy molecules in a cosolvent with different volume percentages of ethanol

Full size image

Chemically, the affinity of the pyrrole molecule is ethanol > PS microsphere > water. Therefore, we believed that a small amount of ethanol enables the pyrrole molecules to remain in the liquid medium instead of diffusing to the PS microspheres. At the same time, the PS microspheres became partially dispersible in the solution after ethanol was added. Consequently, severe aggregation of PS microspheres was avoided and the pyrrole molecules were able to polymerize on the PS microspheres forming the desirable PS@PPy microstructure. However, with the ethanol volume percentage reached 15 vol% and above, the PS microspheres and pyrrole molecules were mixed homogeneously in the mixture. As a result, the diffusion of pyrrole molecules to the PS microspheres took place again. This led to moderate increase in the number of deformed PS@PPy microspheres but at the same time, the extent of aggregation was still negligible. It is noted that the ethanol and water are totally miscible. Therefore, we were unable to observe any noticeable phase change associated with the re-distribution of PS microspheres and pyrrole molecules in the water/ethanol mixture.

The polydispersity of PS@PPy microspheres synthesized with different volume percentages of ethanol is also presented in Fig. 5 (right Y-axis). By definition, the polydispersity was obtained using the following equation;

$$ {\text{PDI}} = \left( {\frac{s}{{\bar{x}}}} \right)^{{\frac{1}{2}}} $$
(2)

where the s is the standard deviation of PS@PPy microspheres, and the \( \bar{x} \) is the average diameter of PS@PPy microspheres. For the as-synthesized PS microspheres, its polydispersity was 0.093, indicating its monodispersive nature. In contrast, the polydispersity of PS@PPy microspheres was above 0.1 and became larger with increasing ethanol amount. It was reasonable to expect a larger value of polydispersity once a PPy layer was deposited on the PS microspheres. Among these samples, the PS@PPy microspheres synthesized with 3.5 vol% of ethanol revealed the lowest polydispersity.

The average thickness of PPy shell for the PS@PPy microspheres was obtained by measuring the diameter of PS@PPy microspheres from SEM images, and subtrating that value by 490 nm (the average diameter of PS microsphere). Table 1 lists the average diameter of PS@PPy microspheres (from 50 PS@PPy microspheres) and their respective PPy shell thickness, from samples synthesized with different volume percentages of ethanol. Interestingly, there appeared similar PPy shell thickness among samples synthesized with ethanol volume percentage of 0–7%. For samples with ethanol above 15%, the PPy shell thickness was reduced to some degrees. We rationalized that a thinner PPy shell was attributed to a reduced chemical reactivity of FeCl3 because its solubility in the ethanol/water mixture was decreased with increasing ethanol amount. As a result, the complete polymerization of pyrrole molecules was not achieved leading to a moderate reduction in PPy shell thickness.

Table 1 Relevant properties of PS@PPy microspheres prepared with different volume percentages of ethanol
Full size table

We also attempted to estimate the PPy utilization rate (as listed in Table 1) during the synthesis of PS@PPy microspheres by first determining the number of PS microspheres \( (N_{PS} ) \) used in a batch and the weight of PPy shell for a single PS@PPy microsphere (\( W_{PPy} \)). The \( N_{PS} \) was obtained by dividing the total weight of PS microspheres (0.6 g) by the weight of a single PS microsphere (\( W_{PS} \)). The \( N_{PS} \), \( W_{PS} \), and \( W_{PPy} \) for a single PS@PPy microsphere were calculated by following equations;

$$ W_{PS} = \frac{4}{3} \times \pi \times r_{ps}^{3} \times D_{PS} $$
(3)
$$ W_{PPy} = \frac{4}{3}\pi \left[ {\left( {r_{PS - PPy}^{3} } \right) - \left( {r_{PS}^{3} } \right)} \right] \times D_{PPy} $$
(4)
$$ N_{PS} = \frac{0.6}{{W_{PS} }} $$
(5)

where the \( r_{PS} \) and the \( r_{PS - PPy} \) are the average radius of PS and PS@PPy microsphere, and the \( D_{PS} \) and the \( D_{PPy} \) are the density of polystyrene (1.05 g cm−3) and polypyrrole (1.46 g cm−3), respectively. Afterwards, the PPy utilization rate could be estimated by dividing the total weight of PPy shell by the amount of pyrrole (0.24 g) used in a batch as following equation;

$$ PPy\,\,utilization \left( \% \right) = \frac{{N_{PS} \times W_{PPy} }}{0.24} $$
(6)

After further simplification, the PPy utilization rate could be obtained in following equation;

$$ PPy\,utilization \left( \% \right) = \left[ {\left( {\frac{{r_{PS - PPy} }}{{r_{PS} }}} \right)^{3} - 1} \right] \times \frac{{D_{PPy} }}{{D_{PS} }} \times 2.5 $$
(7)

The TGA profiles for PS microspheres, as well as PS@PPy microspheres synthesized with ethanol volume percentage of 7 and 15% are provided in Fig. 6. Apparently, these samples revealed a notable weight loss between 300 and 400 °C, which was due to the decomposition of PS microspheres. In addition, the weight loss at 800 °C was 92 and 83% for PS@PPy microspheres synthesized with ethanol volume percentage of 15 (curve b) and 7% ethanol (curve c), respectively. Because the TGA data were recorded in N2 atmosphere, according to earlier literature [31], the PPy was unable to be decomposed entirely but converted to a carbon-rich substance at elevated temperature. Our TGA results were consistent with what was reported by Huang et al. in which a ~ 33% weight loss of PPy at 700 °C was observed [28]. In addition, the residual weight of PS@PPy microspheres after the decomposition of PS microspheres represented the exact amount of PPy loading in PS@PPy microspheres. From TGA profiles, the difference in the PPy loading between cure b and c was in reasonable agreement with the thickness difference of PPy shell from PS@PPy microspheres in Table 1.

Fig. 6
figure 6

The TGA curves of (a) PS microspheres, as well as PS@PPy microspheres synthesized with ethanol volume percentage of (b) 15 and (c) 7%, respectively

Full size image

Figure 7 shows the zeta potential as a funciton of pH for PS microspheres as well as PS@PPy microspheres synthesized with 3.5, 7, and 60 vol% of ethanol. Apparently, the PS microsphere exhibited negative surface charges from pH 2 to pH 11. Its negative charge was attributed to the sulfonate groups from the synthetic stage. In contrast, the PS@PPy microsphere revealed a transition from positive surface charges at low pH value to negative surface charges at high pH value. The isoelectric point for the PS@PPy microsphere synthesized with 3.5, 7, and 60 vol% of ethanol was 5.53, 6.44, and 3.73, respectively. According to literature [32], the positive surface charge of PS@PPy microsphere was due to the protonation of nitrogen atoms of polypyrrole and simultaneous dissociation of doped Cl− ions when the pH value was below its isoelectric point. On the other hand, when the pH value was increased above its isoelectric point, the PS@PPy microsphere demonstrated a negative surface charge attributed to the deprotonation of the protonated nitrogen atoms on polypyrrole and selective adsorption of OH− ions on the surface of polypyrrole. Notably, the PS@PPy microsphere synthesized with 60 vol% of ethanol exhibited a moderate shift of isoelectric point to a lower pH value as compared to those synthesized with 3.5 and 7 vol% of ethanol. This moderate shift was likely influenced by the thickness of PPy layer as a thicker PPy shell rendered a larger isoelectric point because of less influence from the negative charges on the PS microsphere.

Fig. 7
figure 7

The zeta potential as a function of pH value for PS microspheres and PS@PPy microspheres synthesized with ethanol volume percentage of 3.5, 7, and 60%

Full size image

Figure 8a exhibits a SEM image of hollow PPy microcapsules whose average diameter was around 527 nm. Some fractured PPy microspheres were also observed, which confirmed that the THF was able to dissolve the PS microsphere selectively. From the SEM image, upon the removal of PS microspheres, some of the remaining PPy microspheres suffered from structural damage due to lack of mechanical strength from the PPy shell. Inset in Fig. 8a revealed a hollow PPy microcapsule with a surface hole. This image indicated that the PPy structure was relatively robust and the dissolution of PS microsphere was indeed feasible. Figure 8b demonstrates a TEM image showing a high resolution picture of individual hollow PPy microcapsules. The color contrast among these hollow PPy microcapsules was attributed to residual PS after THF extraction; a darker image suggested the presence of residual PS whereas a transparent image indicated the complete removal of PS. Noticeably, the PPy shell layer was mostly intact with a thickness of 20 nm (insert image in Fig. 8b).

Fig. 8
figure 8

The a SEM and b TEM images of PPy hollow microspheres obtained after immersing the PS@PPy microspheres in THF. These PS@PPy microspheres are synthesized with 7 vol% of ethanol as a cosolvent. Inset in a shows a hollow PPy microcapsule with a hole on its PPy shell. Inset in b shows a representative hollow PPy microcapsule showing the complete removal of PS core

Full size image

The formation of microcapsules via the template approach is quite interesting. Currently, it has not been demonstrated yet to insert additional materials once the microcapsules are formed. According to the literature, the only feasible route to date is to design/synthesize the desirable materials as the shell layer, and after removing the core constituent, the shell layer, i.e., the microcapsules, becomes a composite [33].

In principle, the size of the PPy microcapsules could be readily adjusted by selecting PS microspheres in different sizes. In our laboratory, we are in the process of utilizing these PPy microcapsules as the substrate to support nanoparticulate IrO2 nanoparticles for the application of bio-stimulating electrode to promote cell growth. We anticipate to obtain tangible data within a year and publish our results afterwards.

4 Conclusions

We developed a templated approach to fabricate uniform hollow PPy microcapsules in diameter of 527 nm with shell thickness of 20 nm. These hollow PPy microcapsules were prepared by chemical extraction of PS from the PS@PPy microspheres. The homogeneous PS@PPy microspheres were synthesized by optimizing a water/ethanol ratio during the synthetic stage. Without the addition of ethanol, FTIR and TGA confirmed the diffusion of pyrrole molecules to the PS microspheres, engendering noticeable deformation and aggregation of PS@PPy microspheres. By optimizing the volume percentage of ethanol in an aqueous solution, proper dispersion of PS microspheres and pyrrole molecules were obtained. As a result, a desirable PS@PPy structure with significantly reduced deformation and aggregation was achieved. The PS microspheres carried negative surface charges, and after the deposition of PPy, the PS@PPy microspheres revealed a transition from positive to negative surface charges with increasing pH values.