**Effect of CaCO<sup>3</sup> Nanoparticles on the Mechanical and Photo-Degradation Properties of LDPE**

**Paula A. Zapata 1,\*, Humberto Palza <sup>2</sup> , Boris Díaz <sup>1</sup> , Andrea Armijo <sup>1</sup> , Francesca Sepúlveda <sup>1</sup> , J. Andrés Ortiz <sup>1</sup> , Maria Paz Ramírez <sup>1</sup> and Claudio Oyarzún 1**


Received: 14 November 2018; Accepted: 21 December 2018; Published: 31 December 2018

**Abstract:** CaCO<sup>3</sup> nanoparticles of around 60 nm were obtained by a co-precipitation method and used as filler to prepare low-density polyethylene (LDPE) composites by melt blending. The nanoparticles were also organically modified with oleic acid (O-CaCO3) in order to improve their interaction with the LDPE matrix. By adding 3 and 5 wt% of nanofillers, the mechanical properties under tensile conditions of the polymer matrix improved around 29%. The pure LDPE sample and the nanocomposites with 5 wt% CaCO<sup>3</sup> were photoaged by ultraviolet (UV) irradiation during 35 days and the carbonyl index (CI), degree of crystallinity (χc), and Young's modulus were measured at different times. After photoaging, the LDPE/CaCO<sup>3</sup> nanocomposites increased the percent crystallinity (χc), the CI, and Young's modulus as compared to the pure polymer. Moreover, the viscosity of the photoaged nanocomposite was lower than that of photoaged pure LDPE, while scanning electron microscopy (SEM) analysis showed that after photoaging the nanocomposites presented cavities around the nanoparticles. These difference showed that the presence of CaCO<sup>3</sup> nanoparticles accelerate the photo-degradation of the polymer matrix. Our results show that the addition of CaCO<sup>3</sup> nanoparticles into an LDPE polymer matrix allows future developments of more sustainable polyethylene materials that could be applied as films in agriculture. These LDPE-CaCO<sup>3</sup> nanocomposites open the opportunity to improve the low degradation of the LDPE without sacrificing the polymer's behavior, allowing future development of novel eco-friendly polymers.

**Keywords:** CaCO<sup>3</sup> nanoparticles; polyethylene nanocomposites; photoaged polyethylene

#### **1. Introduction**

Inorganic fillers are incorporated into a polyolefin to form composites with enhanced mechanical, thermal, and barrier properties compared to the polymer matrix [1]. Nanocomposites are a class of filled polymers in which nanometric inorganic fillers are incorporated into the polymer matrix with property enhancements at much lower concentrations than those of microfillers. Calcium carbonate is one of the most commonly used inorganic fillers in thermoplastic polymers, such as poly(vinyl chloride) and polypropylene, to improve their mechanical properties. CaCO<sup>3</sup> nanoparticles, in particular, have been incorporated into a polyethylene (PE) matrix by the melting process, increasing Young's modulus with the filler concentration and decreasing both the upper yield point and elongation at break compared to pure PE [2–5]. Although CaCO<sup>3</sup> is a well-known filler in polymer composites for mechanical

reinforcement, at the nanometric scale it can add other functionalities to the polymer matrix, such as barrier properties and antimicrobial behavior [6,7].

Recently, the effect of the incorporation of nano-particulate calcium carbonate hollow spheres (3, 10 and 25 wt%) in high-density polyethylene (HDPE) by extrusion was studied. They found a crystallinity decrease with increasing filler content. There was found a typical increase of Young's modulus (E) (ca. 17%) with increasing concentration of hollow spheres of CaCO<sup>3</sup> filler due to the rigidity of the filler particles and the strong interaction of the filler with the polymer matrix, and it was companied by the corresponding decrease of the upper yield point and elongation at break [2].

One of the major drawbacks of polymer nanocomposites is the high agglomeration of the fillers, which can be reduced by surface modifications. For instance, Lazzeri et al. [8] studied the influence of the organic surface modification in CaCO<sup>3</sup> nanoparticles (70 nm) by stearic acid (SA) treatment on the mechanical properties of HDPE composites. Incorporation of 10 vol% of CaCO<sup>3</sup> to HDPE increased a rise in yield stress in all composites, but the yield stress decreases with increasing SA content. The author explained this behavior by stating that the addition of SA to the surface of the particles should reduce the stress transfer ability of the interface and even its thickness, leading to a softer interface. The influence of nano-CaCO<sup>3</sup> and its surface modification have also been studied in polypropylene (PP) composites prepared by the melting process. In particular, nano-CaCO<sup>3</sup> (diameter ca. 44 nm) was modified with stearic acid [1,9,10] and palmitic acid [11], with the addition of modified CaCO<sup>3</sup> increasing tensile strength, Young's modulus, and melting point. Another route to improve the dispersion of nano-CaCO<sup>3</sup> in PP matrices is by the addition of a small amount of a non-ionic modifier during melt extrusion. In this case Young's modulus increased slightly with amount of CaCO<sup>3</sup> load, while the yield strength of PP decreased [12].

On the other hand, to improve the physicochemical properties of the polymer some researchers have treated the surface polymer using thin-layer technology, including oxygen and nitrogen plasma discharge, deposition of functional coatings (i.e., diamond-like carbon (DLC)) among others. For example, low-density polyethylene (LDPE) increased its surface hardness 7 times after layer deposition by DLC coating. Those techniques allowed giving desirable surface properties to the polymer [13].

Despite the relevance for society of the degradation properties of plastic materials, the environmental stability of polymer/calcium carbonate nanocomposites has been barely reported. The effect of nano-CaCO<sup>3</sup> on the natural photo-aging degradation of PP was studied outdoors during 88 days [14]. The degradation polymer was studied by Fourier transform infrared (FTIR) spectroscopy and pyrolysis gas chromatography-mass spectroscopy (PGC-eMS). The PP/CaCO<sup>3</sup> nanocomposites showed higher photo-degradation than neat PP. The authors explained this behavior as due to the functional groups on the surface of nanoparticles catalyzing the photo-oxidation reaction of PP. There are adsorbed hydroxyl groups on the surface, which is active in photo-chemical reactions. Morreale et al. [15] studied the accelerated weathering behavior of PP/CaCO<sup>3</sup> micro- and nanocomposites, showing that the nanosized filler may lead to a faster photo-oxidation rate than that of pure polypropylene. In particular, nanosized calcium carbonate caused faster photodegradation rates than microsized calcium carbonate.

Achieving the bidodegradation of commercial commodity plastics is an enormous environmental challenge due to the increased social demand for higher sustainability processes. The addition of additive/filler to accelerate the photodegradation of these polymers can be associated with an early decrease in the mechanical property even during use. Therefore, nanoparticles able to accelerate the photodegradation together with improving the mechanical behavior can compensate for the latter issue.

Considering what was mentioned above, the present work studies the effect of adding pure and organic-modified CaCO<sup>3</sup> nanoparticles into a non-polar LDPE matrix. The effect of different amounts of pure CaCO<sup>3</sup> nanoparticles and oleic acid-modified-CaCO<sup>3</sup> on the thermal and mechanical properties of polyethylene were studied. The effect of CaCO<sup>3</sup> nanoparticles on the photoaging process of LDPE was further investigated.

#### **2. Experiment**

#### *2.1. Materials*

Polyethylene was purchased from Aldrich, density: 0.925 g cm−<sup>3</sup> , melt index: 25 g/10 min (190 ◦C/2.16 kg); impact strength: 45.4 J/m (Izod, ASTM D 256, −50 ◦C). CaCO<sup>3</sup> nanoparticles were synthesized by a precipitation method [16]. Briefly, the reagents used were sodium carbonate, Na2CO<sup>3</sup> (Merck, Darmstadt, Germany, 99.9%), calcium nitrate, Ca(NO3)<sup>2</sup> (Aldrich, Darmstadt, Germany, 99%), sodium nitrate NaNO<sup>3</sup> (Aldrich, 99%), sodium hydroxide, NaOH pellets (Mallinckrodt Chemicals., Dublin, Ireland, ≥98%), and distilled water. Oleic acid (Aldrich, reagent grade, 98%) was used for the modification of the CaCO<sup>3</sup> nanoparticles.

#### *2.2. CaCO<sup>3</sup> Nanoparticle Synthesis*

The CaCO<sup>3</sup> nanoparticles were obtained by a method reported by Babou-Kammoe et al. [16]. First, sodium carbonate (Na2CO3) (0.042 g) was dissolved in deionized water (80 mL) with sodium hydroxide (NaOH) (1.25 g) and sodium nitrate NaNO<sup>3</sup> (0.612 g). In a second step, calcium nitrate (Ca(NO3)2) (0.944 g) was dissolved in deionized water (80 mL) and the resultant mixture formed a precipitate. The calcium nitrate solution was added dropwise to the sodium carbonate solution with continuous stirring during 4 h at 25 ◦C. The resultant mixture formed a precipitate which was separated from the water by filtering off. The nanoparticles were dried at 60 ◦C during 24 h and were characterized.

#### *2.3. Organic Modification of CaCO<sup>3</sup> Nanoparticles (O-CaCO3)*

The nanoparticles were modified with oleic acid [17]. 1-Hexane (100 mL) and oleic acid (200 µL) were mixed with stirring. Then 1 g of CaCO<sup>3</sup> nanoparticles was added to the solution at 60 ◦C with vigorous stirring during 5 h. The nanoparticles were then filtered, washed with ethanol, and vacuum-dried at 100 ◦C during 24 h [18].

#### *2.4. Low-Density Polyethylene (LDPE)/CaCO<sup>3</sup> and LDPE/O-CaCO<sup>3</sup> Nanocomposite*

The nanocomposites were prepared using a Brabender Plasti-Corder (Duisburg, Germany) internal mixer at 150 ◦C and a speed of 110 rpm, during 10 min. The nanocomposites with 3, 5, and 8 wt% of CaCO<sup>3</sup> nanoparticles were obtained by mixing predetermined amounts of the CaCO<sup>3</sup> as filler and neat LDPE under a nitrogen atmosphere. The samples were press-molded at 190 ◦C at a pressure of 50 bar during 3 min and cooled under pressure by flushing the press with cold water.

#### *2.5. Nanoparticles and Composite Characterization*

The morphology of the CaCO<sup>3</sup> was analyzed by transmission electron microscopy (TEM) (JEOL ARM 200 F, Boston, MA, USA) operating at 20 kV. Samples for TEM measurements were prepared by placing a drop of CaCO<sup>3</sup> nanoparticles on a carbon-coated standard copper grid (400 mesh).

The X-ray diffraction (XRD) patterns of the CaCO<sup>3</sup> nanoparticles were studied on a Siemens D5000 diffractometer (Berlin, Germany), using Ni-filtered Cu Kα radiation (λ = 0.154 nm). The diffraction patterns were recorded in the 2θ = 5–80◦ range.

FTIR measurements of CaCO<sup>3</sup> and modified nanoparticles (O-CaCO3) were performed in a Bruker Vector 22 FTIR spectrometer (Karlsruhe, Germany). The infrared (IR) spectra were collected in the 4000 to 500 cm−<sup>1</sup> range, with a resolution of 4 cm−<sup>1</sup> at room temperature.

The tensile properties of the neat polyethylene (neat LDPE) and composites (LDPE/CaCO<sup>3</sup> and LDPE/O-CaCO3) were determined on an HP model D-500 dynamometer (Buenos Aires, Argentina). The materials were molded for 3 min in a hydraulic press, HP Industrial Instruments, at a pressure

of 50 bar and a temperature of 170 ◦C, and then cooled under pressure with water circulation. Films around 0.05 mm thickness were obtained. Dumbbell-shaped samples with an effective length of 30 mm and a width of 5 mm were cut from the compression-molded sheets. The samples were tested at a rate of 50 mm/min at 20 ◦C. Each set of measurements was repeated at least four times.

#### 2.5.1. Photo-Exposure

#### Photoaging

Polymer films of 0.02 mm of thickness and the dimensions and 4 cm were irradiated using a Microscal Light exposure unit and Suntest/Atlas XLS 2200 W (Linsengericht, Germany) using a solar standard filter (borosilicate), which provides 550 W m−<sup>2</sup> (Irradiance acc. ISO 4892/DIN 53387) in the 300–800 nm wavelength region. The temperature was kept constant at 45 ◦C during the testing. Exposed samples of 1 cm × 1 cm were periodically taken out and characterized. The irradiation side of the sample was alternated every 3 days. At different aging times the oxidation rates were determined on an FTIR spectrometer using the standard carbonyl index method. FTIR spectra were obtained on a Perkin Elmer BX-FTIR (Waltham, MA, USA). The polymer degradation was determined using the carbonyl index (CI) as the ratio of the optical density of the ketone carbonyl absorptions bands at 1715 cm−<sup>1</sup> and the optical density corresponding to CH<sup>2</sup> scissoring peak at 1465 cm−<sup>1</sup> [19].

Differential scanning calorimetry (DSC) was studied on a METTLER DSC823 (Columbus, OH, USA) The melting temperature and enthalpy of fusion of the neat and nanocomposite samples were determined before and after photoaging. The measurements were made at a heating rate of 10 ◦C·min−<sup>1</sup> in an inert atmosphere. The samples were heated from 25 ◦C to 180 ◦C and then cooled to 25 ◦C at the same rate. Percent crystallinity (χc) was determined using Equation (1):

$$\chi\text{c} = \frac{\Delta H\_l}{(1 - \Phi)\Delta H\_0} \times 100\tag{1}$$

where ∆*H<sup>l</sup>* is the melting enthalpy (J g−<sup>1</sup> ) of the polymer nanocomposite, ∆H<sup>0</sup> is the enthalpy corresponding to the melting of a 100% crystalline sample (289 J g−<sup>1</sup> ) [20], and Φ is the weight fraction of the filler in the nanocomposit. The standard deviation of the T<sup>c</sup> and T<sup>m</sup> measurements was ca. ±2 ◦C.

Thermogravimetric analysis (TGA) experiments were performed on a Netzsch TG 209 F1 Libra Instrument (Selb, Germany). The films were heated from 25 ◦C to 600 ◦C at a rate of 10 ◦C·min−<sup>1</sup> and the nitrogen flow was kept constant at 60 mL·min−<sup>1</sup> . The TGA analysis also verified the content of CaCO<sup>3</sup> in the LDPE/CaCO<sup>3</sup> nanocomposites. The LDPE/CaCO<sup>3</sup> nanocomposites with 5 wt% showed a 4.65 wt% of the CaCO<sup>3</sup> nanoparticle content after the melting process.

Viscosimetric analysis before and after irradiations was carried out in o-dichlorobenzene at 135 ◦C in a Viscosimatic-Sofica viscometer (Santiago, Chile)

The surface morphology of the polymers before and after photoaging was characterized by scanning electron microscopy (SEM) using a Philips XL30 model instrument (Billerica, MA, USA).

#### **3. Results and Discussion**

#### *3.1. Nanoparticle Characterization*

The morphology of the nanoparticles was studied by TEM as shown in Figure 1. The nanoparticles synthesized by the coprecipitation method have an average diameter of ca. 60 nm and irregular morphology. The yield of this method was ca. 75%. The crystalline phases of the CaCO<sup>3</sup> nanoparticles were studied by XRD (Figure 2). Nanoparticles have two characteristic phases as concluded by analyzing the Bragg reflections: calcite, associated with peaks at 29◦ and 32◦ from the (104) and (006) crystal planes, respectively; and aragonite, with peaks at 27◦ , 30◦ , and 45◦ from the (111), (021), and (221) planes, respectively [21].

O bond and calcite vibrations.

*Molecules* **2018**, *23*, x FOR PEER REVIEW 5 of 13

**Figure 1.** Transmission electron microscopy (TEM) image of the CaCO3 nanoparticles. **Figure 1.** Transmission electron microscopy (TEM) image of the CaCO<sup>3</sup> nanoparticles. **Figure 1.** Transmission electron microscopy (TEM) image of the CaCO3 nanoparticles.

**Figure 2.** X-ray diffraction (XRD) spectra of CaCO3 nanoparticles. **Figure 2.** X-ray diffraction (XRD) spectra of CaCO3 nanoparticles. **Figure 2.** X-ray diffraction (XRD) spectra of CaCO<sup>3</sup> nanoparticles.

The FTIR spectrum of calcined samples (CaCO3) (Figure 3b), shows the presence of calcium carbonate (CaCO3) bands at 715, 880, 1490, 1804, 2530, 2900, and 2998 cm−1. The band at 1490 cm−<sup>1</sup> correspond to essentially asymmetric and symmetric lengthening of the O–C–O bond. Absorption bands centered at 715, 880 and 1490 cm−1 are characteristics of the calcite phase of CaCO3 [22]. The method used for the organic modification of nanoparticles was based on that reported by Li and Zhu [17] and it was verified by FTIR spectra, where the peaks corresponding to the alkyl chain (CH2) of oleic acid appear at 2920 cm−1 and 2855 cm−1 (Figure 3). Moreover, a small spectral line at 1710 cm−<sup>1</sup> corresponding to the stretching of the carbonyl group of oleic acid indicates that the carboxylic acid group of oleic acid, −COOH, reacted with surface hydroxyl groups from the starch nanoparticles [23]. Other peaks at 1590 cm−1 due to carboxylate groups, and the peaks at 1550 cm−1 and 1430 cm−1 that indicate the presence of COO−, are overlapped with the characteristic band at 1490 cm−1 of the O–C– The FTIR spectrum of calcined samples (CaCO3) (Figure 3b), shows the presence of calcium carbonate (CaCO3) bands at 715, 880, 1490, 1804, 2530, 2900, and 2998 cm−1. The band at 1490 cm−<sup>1</sup> correspond to essentially asymmetric and symmetric lengthening of the O–C–O bond. Absorption bands centered at 715, 880 and 1490 cm−1 are characteristics of the calcite phase of CaCO3 [22]. The method used for the organic modification of nanoparticles was based on that reported by Li and Zhu [17] and it was verified by FTIR spectra, where the peaks corresponding to the alkyl chain (CH2) of oleic acid appear at 2920 cm−1 and 2855 cm−1 (Figure 3). Moreover, a small spectral line at 1710 cm−<sup>1</sup> corresponding to the stretching of the carbonyl group of oleic acid indicates that the carboxylic acid group of oleic acid, −COOH, reacted with surface hydroxyl groups from the starch nanoparticles [23]. Other peaks at 1590 cm−1 due to carboxylate groups, and the peaks at 1550 cm−1 and 1430 cm−1 that indicate the presence of COO−, are overlapped with the characteristic band at 1490 cm−1 of the O–C– O bond and calcite vibrations. The FTIR spectrum of calcined samples (CaCO3) (Figure 3b), shows the presence of calcium carbonate (CaCO3) bands at 715, 880, 1490, 1804, 2530, 2900, and 2998 cm−<sup>1</sup> . The band at 1490 cm−<sup>1</sup> correspond to essentially asymmetric and symmetric lengthening of the O–C–O bond. Absorption bands centered at 715, 880 and 1490 cm−<sup>1</sup> are characteristics of the calcite phase of CaCO<sup>3</sup> [22]. The method used for the organic modification of nanoparticles was based on that reported by Li and Zhu [17] and it was verified by FTIR spectra, where the peaks corresponding to the alkyl chain (CH2) of oleic acid appear at 2920 cm−<sup>1</sup> and 2855 cm−<sup>1</sup> (Figure 3). Moreover, a small spectral line at 1710 cm−<sup>1</sup> corresponding to the stretching of the carbonyl group of oleic acid indicates that the carboxylic acid group of oleic acid, −COOH, reacted with surface hydroxyl groups from the starch nanoparticles [23]. Other peaks at 1590 cm−<sup>1</sup> due to carboxylate groups, and the peaks at 1550 cm−<sup>1</sup> and 1430 cm−<sup>1</sup> that indicate the presence of COO−, are overlapped with the characteristic band at 1490 cm−<sup>1</sup> of the O–C–O bond and calcite vibrations.

*Molecules* **2018**, *23*, x FOR PEER REVIEW 6 of 13

**Figure 3.** Fourier transform infrared (FTIR) spectra of (**a**) Oleic acid, (**b**) CaCO3 nanoparticles, and (**c**) **Figure 3.** Fourier transform infrared (FTIR) spectra of (**a**) Oleic acid, (**b**) CaCO<sup>3</sup> nanoparticles, and (**c**) nanoparticles modified with oleic acid (O-CaCO<sup>3</sup> ).

#### *3.2. Composite Characterization*

*3.2. Composite Characterization* 

nanoparticles modified with oleic acid (O-CaCO3).

#### 3.2.1. Thermal Properties

3.2.1. Thermal Properties The crystallization temperature (Tc), melting temperature (Tm), and degree of cristallinity (χc) were analyzed by DSC and the thermal stability obtained by TGA of the neat LDPE and LDPE/CaCO3 nanocomposites are shown in Table 1. The crystallization temperature, melting temperature, and degree of crystallinity (χc) did not change with the incorporation of the nanoparticles, meaning that the presence of these nanoparticles did not affect the crystallization process of the polymer matrix. Similar results have been reported by other authors when different nanoparticles like ZnO, clay, silica, silver, and TiO2 were incorporated into LDPE. This behavior may be correlated to the minimal volume fraction of the nanoparticles incorporated into the composite [24–26]. Also, the similar thermal properties of the matrix and composites would suggest analogous processing conditions as The crystallization temperature (Tc), melting temperature (Tm), and degree of cristallinity (χc) were analyzed by DSC and the thermal stability obtained by TGA of the neat LDPE and LDPE/CaCO<sup>3</sup> nanocomposites are shown in Table 1. The crystallization temperature, melting temperature, and degree of crystallinity (χc) did not change with the incorporation of the nanoparticles, meaning that the presence of these nanoparticles did not affect the crystallization process of the polymer matrix. Similar results have been reported by other authors when different nanoparticles like ZnO, clay, silica, silver, and TiO<sup>2</sup> were incorporated into LDPE. This behavior may be correlated to the minimal volume fraction of the nanoparticles incorporated into the composite [24–26]. Also, the similar thermal properties of the matrix and composites would suggest analogous processing conditions as that of LDPE at a hypothetical industrial-scale production of these nanocomposites.

that of LDPE at a hypothetical industrial-scale production of these nanocomposites. In the initial degradation step of the decomposition temperature, at 2% weight loss (T2), the nanocomposites (LDPE/CaCO3) were slightly more stable than LDPE nanocomposites at ca. 5%. For 10% weight loss (T10), for 50% weight loss (T50), and the temperature for the maximum rate of weight loss (Tmax) did not change with the nanoparticle incorporation compared to the pure neat LDPE under inert conditions. It is well known that the incorporation of different kinds of nanofillers into a polymer can act as a superior insulator and mass transport barrier for the volatile products generated during decomposition, increasing the thermal degradation temperatures. However, these processes are relevant for high aspect ratio nanoparticles such as layered clays. Spherical-like particles with low aspect ratio should not trigger these mechanisms and only an adsorption process can explain changes in the degradation, as reported by our group on spherical silica nanoparticles [27]. In our case, the spherical-like CaCO3 nanoparticles were not able to disrupt the diffusion nor adsorb volatile In the initial degradation step of the decomposition temperature, at 2% weight loss (T2), the nanocomposites (LDPE/CaCO3) were slightly more stable than LDPE nanocomposites at ca. 5%. For 10% weight loss (T10), for 50% weight loss (T50), and the temperature for the maximum rate of weight loss (Tmax) did not change with the nanoparticle incorporation compared to the pure neat LDPE under inert conditions. It is well known that the incorporation of different kinds of nanofillers into a polymer can act as a superior insulator and mass transport barrier for the volatile products generated during decomposition, increasing the thermal degradation temperatures. However, these processes are relevant for high aspect ratio nanoparticles such as layered clays. Spherical-like particles with low aspect ratio should not trigger these mechanisms and only an adsorption process can explain changes in the degradation, as reported by our group on spherical silica nanoparticles [27]. In our case, the spherical-like CaCO<sup>3</sup> nanoparticles were not able to disrupt the diffusion nor adsorb volatile compounds and, therefore, no changes were observed in TGA analysis.

compounds and, therefore, no changes were observed in TGA analysis.


**Table 1.** Thermal properties of polyethylene (PE)/CaCO<sup>3</sup> nanocomposites before photoaging.

Tc: Crystallization temperature, η: Viscosity, T<sup>m</sup> = melting temperature; χ<sup>c</sup> = percent crystallinity, T<sup>2</sup> = decomposition temperature at 2% weight loss; T<sup>10</sup> = decomposition temperature at 10% weight loss; T<sup>50</sup> = decomposition temperature at 50% weight loss, Tmax = temperature for the maximum rate of weight loss (Tmax); O-CaCO<sup>3</sup> = modified nanoparticles. The standard deviation of the viscosity measurements is <sup>±</sup>0.03 dLg−<sup>1</sup> . The standard deviation of the T<sup>m</sup> and T<sup>c</sup> measurements are ca. ±2 ◦C. The thermogravimetric analysis (TGA) has a standard deviation of ca. ±2 ◦C.

#### 3.2.2. Mechanical Properties

The mechanical properties of the neat LDPE and the LDPE/CaCO<sup>3</sup> nanocomposites are displayed in Table 2 and Figure 4. An increase of Young's modulus results from adding the CaCO<sup>3</sup> nanoparticles in comparison with the neat LDPE. This performance was more pronounced with 5 wt% of nanoparticles for the LDPE/O-CaCO<sup>3</sup> nanocomposites, as Young's modulus increased ca. 29% compared to neat LDPE. Morreale et al. [3] found that 10 wt% of CaCO<sup>3</sup> fillers (50–100 nm) improved Young's modulus just ca. 20% compared to neat LDPE, due to the presence of the nanoparticle agglomeration. The increase in the modulus in our case must be caused by the strong interaction between the polymer and the nanoparticles, improving the dispersion of the particles [1]. Similar results were found by Lapcík et al. [2], who stated that due to the rigidity of the filler particles and the interaction of the filler with the polymer matrix, a reinforcement improvement can be obtained. The yield stress remained unaffected with the addition of CaCO<sup>3</sup> to neat polyethylene. A similar behavior was found for nanocomposites based on high-density polyethylene with CaCO<sup>3</sup> nanoparticles (ca. 60 nm) [9].

On the other hand, the deformation at break decreased when 5 wt% of the CaCO<sup>3</sup> nanoparticles were incorporated, probably due to many defects in the polymer matrix leading to ductility reduction [10]. The decrease of the deformation at break of LDPE/O-CaCO<sup>3</sup> (5 wt%) was slightly lower, and this may be due to the modifier improving the interaction between nanoparticles and LDPE [1].


**Table 2.** Mechanical properties of LDPE and LDPE/CaCO<sup>3</sup> nanocomposites.

E = Young's modulus; σy = yield stress; ε Break = deformation at break.

*Molecules* **2018**, *23*, x FOR PEER REVIEW 8 of 13

**Figure 4.** Stress-strain curves for neat LDPE, LDPE/CaCO3 with 3 and 5 wt%, and LDPE/O-CaCO3 with 5 wt% nanocomposites. **Figure 4.** Stress-strain curves for neat LDPE, LDPE/CaCO<sup>3</sup> with 3 and 5 wt%, and LDPE/O-CaCO<sup>3</sup> with 5 wt% nanocomposites.

#### *3.3. Photoaging Analysis 3.3. Photoaging Analysis*

#### 3.3.1. Thermal and Mechanical Properties 3.3.1. Thermal and Mechanical Properties

Crystallization temperature (Tc), melting temperature (Tm), degree of cristallinity (χc), thermal stability analysis, and viscosity of the neat LPE and LDPE/CaCO3 nanocomposites after photoaging are displayed in Table 3. After photoaging, the χc for nanocomposites increased slightly compared to pure LDPE and LLDPE/CaCO3 before irradiation. This behavior has been attributed to recrystallization due to LDPE scission of end chains producing mobile small chain fragments able to undergo reorganization and recrystallization [14,28]. This scission is confirmed also finding that after photoaging the viscosity decreased due to the formation of low molecular weight compounds during aging (Table 3). It should be noted that this behavior is slightly greater for nanocomposites than for neat LDPE. These results show that the incorporation of nanoparticles into the polymer accelerates its degradation. After photoaging, both decomposition temperatures, (T10) and Tmax, did not change. In previous work using Ca and Fe stereates as PE degradant, the authors found a slight decrease in T10, and they explained this behavior as due to the prooxidative nature of stereate during the Crystallization temperature (Tc), melting temperature (Tm), degree of cristallinity (χc), thermal stability analysis, and viscosity of the neat LPE and LDPE/CaCO<sup>3</sup> nanocomposites after photoaging are displayed in Table 3. After photoaging, the χ<sup>c</sup> for nanocomposites increased slightly compared to pure LDPE and LLDPE/CaCO<sup>3</sup> before irradiation. This behavior has been attributed to recrystallization due to LDPE scission of end chains producing mobile small chain fragments able to undergo reorganization and recrystallization [14,28]. This scission is confirmed also finding that after photoaging the viscosity decreased due to the formation of low molecular weight compounds during aging (Table 3). It should be noted that this behavior is slightly greater for nanocomposites than for neat LDPE. These results show that the incorporation of nanoparticles into the polymer accelerates its degradation. After photoaging, both decomposition temperatures, (T10) and Tmax, did not change. In previous work using Ca and Fe stereates as PE degradant, the authors found a slight decrease in T10, and they explained this behavior as due to the prooxidative nature of stereate during the photoaging process [29].

photoaging process [29]. **Table 3.** Thermal properties of PE/CaCO<sup>3</sup> nanocomposites after photoaging.


Neat LDPE N/A 0.18 105 107 40 423 467 LDPE/CaCO3 5 0.13 106 108 44 426 460 Tc: crystallization temperature, Tm = melting temperature; χc = percent crystallinity. T10 = Tc: crystallization temperature, T<sup>m</sup> = melting temperature; χ<sup>c</sup> = percent crystallinity. T<sup>10</sup> = decomposition temperature at 10% weight loss; Tmax = temperature for the maximum rate of weight loss (Tmax). Photoaging during 10 days.

decomposition temperature at 10% weight loss; Tmax = temperature for the maximum rate of weight loss (Tmax). Photoaging during 10 days. The mechanical properties were evaluated after 10 days of photoaging as displayed in Table 4. The polymers were difficult to break into pieces by hand after 10 days of irradiation, confirming the strong degradation. Young's modulus for the LDPE/CaCO3 increased after photoaging compared to photoaged neat PE. Young's modulus increased after photo-oxidation mainly due to a significant The mechanical properties were evaluated after 10 days of photoaging as displayed in Table 4. The polymers were difficult to break into pieces by hand after 10 days of irradiation, confirming the strong degradation. Young's modulus for the LDPE/CaCO<sup>3</sup> increased after photoaging compared to photoaged neat PE. Young's modulus increased after photo-oxidation mainly due to a significant embrittlement of the material and recrystallization phenomena caused by scission reactions [15]. These results further confirmed that nanoparticles accelerated the degradation of the polymer.

embrittlement of the material and recrystallization phenomena caused by scission reactions [15]. These results further confirmed that nanoparticles accelerated the degradation of the polymer.


**Table 4.** Mechanical properties of LDPE and LDPE/CaCO<sup>3</sup> nanocomposites after irradiation during 10 days. **Table 4.** Mechanical properties of LDPE and LDPE/CaCO3 nanocomposites after irradiation during

*Molecules* **2018**, *23*, x FOR PEER REVIEW 9 of 13

E = Young's modulus; σy = yield stress; ε Break = deformation at break. The photoaging during 10 days. E = Young's modulus; σy = yield stress; ε Break = deformation at break. The photoaging during 10 days.

#### 3.3.2. Carbonyl Index 3.3.2. Carbonyl Index

The degradation, calculated by FTIR, and the carbonyl index (CI) after 35 days of irradiation are displayed in Figure 5. The carbonyl index was measured for neat LDPE and the LDPE/CaCO<sup>3</sup> sample with 5 wt% of nanoparticles. The LDPE/CaCO<sup>3</sup> nanocomposite with 5 wt% had a higher carbonyl index than LDPE, showing an influence in the degradation of the polymers with nanoparticle incorporation. The IR spectra of the photoaged polymer (LDPE/CaCO3) has a strong peak at 1720 cm−<sup>1</sup> , which is related to the C=O stretching vibration of the carbonyl group (Figure 6) [30]. The second band around 3400 cm−<sup>1</sup> is related to the hydroxyl group, which indicates the generation of hydroperoxides and hydroxyl species. Furthermore, the intensity of the carbonyl and hydroxyl bands grew with increasing exposure time [14]. The intensity peaks at 2930, 2850, 1470, and 720 cm−<sup>1</sup> , corresponding to the alkyl chain, decreased slightly. Carboxylic acid salts can be formed by the reaction between the carboxylic acids coming from photoaged PE and the basic fillers. Li et al. [14] explained that PP/CaCO<sup>3</sup> shows a higher degradation rate than neat PP, due to functional groups on the surface of the nanoparticles catalyzing the photooxidant ion reaction of PP. There are absorbed hydroxyl groups on the surface of the nanofillers, which are active in photo-chemical reactions. Therefore, the hydrophilic surface of the nanoparticles is responsible for the increased polymer degradation. Further degradation in the abiotic environment is through the Norrish type I and II mechanism, giving rise to esters and ketones [31]. The degradation, calculated by FTIR, and the carbonyl index (CI) after 35 days of irradiation are displayed in Figure 5. The carbonyl index was measured for neat LDPE and the LDPE/CaCO3 sample with 5 wt% of nanoparticles. The LDPE/CaCO3 nanocomposite with 5 wt% had a higher carbonyl index than LDPE, showing an influence in the degradation of the polymers with nanoparticle incorporation. The IR spectra of the photoaged polymer (LDPE/CaCO3) has a strong peak at 1720 cm−1, which is related to the C=O stretching vibration of the carbonyl group (Figure 6) [30]. The second band around 3400 cm−1 is related to the hydroxyl group, which indicates the generation of hydroperoxides and hydroxyl species. Furthermore, the intensity of the carbonyl and hydroxyl bands grew with increasing exposure time [14]. The intensity peaks at 2930, 2850, 1470, and 720 cm−1, corresponding to the alkyl chain, decreased slightly. Carboxylic acid salts can be formed by the reaction between the carboxylic acids coming from photoaged PE and the basic fillers. Li et al. [14] explained that PP/CaCO3 shows a higher degradation rate than neat PP, due to functional groups on the surface of the nanoparticles catalyzing the photooxidant ion reaction of PP. There are absorbed hydroxyl groups on the surface of the nanofillers, which are active in photo-chemical reactions. Therefore, the hydrophilic surface of the nanoparticles is responsible for the increased polymer degradation. Further degradation in the abiotic environment is through the Norrish type I and II mechanism, giving rise to esters and ketones [31].


photoaging for 35 days.

*Molecules* **2018**, *23*, x FOR PEER REVIEW 10 of 13

*Molecules* **2018**, *23*, x FOR PEER REVIEW 10 of 13

**Figure 6.** Infrared (IR) spectra of initial LDPE/CaCO3 nanocomposites and LDPE/CaCO3 after photoaging for 35 days. **Figure 6.** Infrared (IR) spectra of initial LDPE/CaCO<sup>3</sup> nanocomposites and LDPE/CaCO<sup>3</sup> after photoaging for 35 days. **Figure 6.** Infrared (IR) spectra of initial LDPE/CaCO3 nanocomposites and LDPE/CaCO3 after

SEM images of LDPE and LDPE/CaCO3 with 5 wt% of CaCO3 before and after photoaging are shown in Figure 7. Before irradiation, LDPE and LDPE/CaCO3 images exhibit a smooth and homogeneous surface morphology (Figure 7a,c). After irradiation, the morphology is changed, the LDPE and LDPE/CaCO3 nanocomposites presented some cavities, with nanoparticles producing larger ones (Figure 7b,d). After irradiation, the nanocomposites undergo greater acceleration of photodegradation than neat LDPE, confirming the results shown above by CI, mechanical properties, and viscosity. SEM images of LDPE and LDPE/CaCO<sup>3</sup> with 5 wt% of CaCO<sup>3</sup> before and after photoaging are shown in Figure 7. Before irradiation, LDPE and LDPE/CaCO<sup>3</sup> images exhibit a smooth and homogeneous surface morphology (Figure 7a,c). After irradiation, the morphology is changed, the LDPE and LDPE/CaCO<sup>3</sup> nanocomposites presented some cavities, with nanoparticles producing larger ones (Figure 7b,d). After irradiation, the nanocomposites undergo greater acceleration of photodegradation than neat LDPE, confirming the results shown above by CI, mechanical properties, and viscosity. SEM images of LDPE and LDPE/CaCO3 with 5 wt% of CaCO3 before and after photoaging are shown in Figure 7. Before irradiation, LDPE and LDPE/CaCO3 images exhibit a smooth and homogeneous surface morphology (Figure 7a,c). After irradiation, the morphology is changed, the LDPE and LDPE/CaCO3 nanocomposites presented some cavities, with nanoparticles producing larger ones (Figure 7b,d). After irradiation, the nanocomposites undergo greater acceleration of photodegradation than neat LDPE, confirming the results shown above by CI, mechanical properties, and viscosity.

**Figure 7.** Scanning electron microscopy (SEM) images of initial and photoaged PE and PE/SNp during 35 days of photoaging: (**a**) PE initial; (**b**) PE aged; (**c**) PE/CaCO3 initial; and (**d**) PE/CaCO3 aged. **Figure 7.** Scanning electron microscopy (SEM) images of initial and photoaged PE and PE/SNp during 35 days of photoaging: (**a**) PE initial; (**b**) PE aged; (**c**) PE/CaCO3 initial; and (**d**) PE/CaCO3 aged. **Figure 7.** Scanning electron microscopy (SEM) images of initial and photoaged PE and PE/SNp during 35 days of photoaging: (**a**) PE initial; (**b**) PE aged; (**c**) PE/CaCO<sup>3</sup> initial; and (**d**) PE/CaCO<sup>3</sup> aged.

#### **4. Conclusions**  The co-precipitation method was used to produce CaCO3 (60 nm), which were then modified **4. Conclusions 4. Conclusions**

organically with oleic acid (O-CaCO3). Young's modulus increased ca. 29% for LDPE/O-CaCO3 compared to the neat LDPE Regarding polymer photoaging, the degree of crystallinity (χc) increased with photoaging, and The co-precipitation method was used to produce CaCO3 (60 nm), which were then modified organically with oleic acid (O-CaCO3). Young's modulus increased ca. 29% for LDPE/O-CaCO3 compared to the neat LDPE The co-precipitation method was used to produce CaCO<sup>3</sup> (60 nm), which were then modified organically with oleic acid (O-CaCO3). Young's modulus increased ca. 29% for LDPE/O-CaCO<sup>3</sup> compared to the neat LDPE.

this effect was higher for LDPE/CaCO3 (ca. 19%) nanocomposites than for neat LDPE (ca. 8%), attributed to recrystallization of the polymer. The viscosity of LDPE decreased by ca. 59% after photoaging and around 72% for LDPE/CaCO3, as indicated by the decreased molecular weight of the polymer due to chain scissions, and the pronounced effect of the nanoparticles in the polymer Regarding polymer photoaging, the degree of crystallinity (χc) increased with photoaging, and this effect was higher for LDPE/CaCO3 (ca. 19%) nanocomposites than for neat LDPE (ca. 8%), attributed to recrystallization of the polymer. The viscosity of LDPE decreased by ca. 59% after photoaging and around 72% for LDPE/CaCO3, as indicated by the decreased molecular weight of the polymer due to chain scissions, and the pronounced effect of the nanoparticles in the polymer Regarding polymer photoaging, the degree of crystallinity (χc) increased with photoaging, and this effect was higher for LDPE/CaCO<sup>3</sup> (ca. 19%) nanocomposites than for neat LDPE (ca. 8%), attributed to recrystallization of the polymer. The viscosity of LDPE decreased by ca. 59% after photoaging and around 72% for LDPE/CaCO3, as indicated by the decreased molecular weight of the polymer due to chain scissions, and the pronounced effect of the nanoparticles in the polymer degradation. Young's modulus increased ca. 16% for LDPE/O-CaCO<sup>3</sup> after photoaging because the nanoparticles accelerate the polymer's degradation. The degradation of the films obtained was

confirmed by the carbonyl index, where carbonyl bands appear more intense. LDPE/CaCO<sup>3</sup> with 5 wt% had a high carbonyl index, showing an influence in the degradation of the polymers with the incorporation of nanoparticles.

**Author Contributions:** Conceptualization, B.D. and A.A.; Methodology, M.P.R.; Validation, J.A.O.; Formal Analysis, F.S. and C.O.; Investigation, P.A.Z. and H.P.; Resources, P.A.Z.; Data Curation, F.S. and C.O.; Writing-Original Draft Preparation, P.A.Z.; Writing-Review & Editing, P.A.Z., F.S., H.P.; Visualization, F.S.; Supervision, P.A.Z. and H.P.; Project Administration, P.A.Z.; Funding Acquisition, P.A.Z.

**Funding:** This research was funded by [FONDECYT Regular Project] grant number [1170226] and FIA, [Fundación para la Innovación Agraria] under FIA project "FIA-PYT-2013-0018" (http://www.fia.cl/); and "Gobierno Regional Metropolitano de Santiago" (GORE-RM). P.A.Z. acknowledges the financial support of Project [DICYT] grant number [051641ZR\_DAS], Vicerrectoria de Investigación, Desarrollo e Innovación, Universidad de Santiago de Chile. H.P. acknowledges the financial support of the [FONDECYT Project] grant number [1150130].

**Acknowledgments:** P.A.Z. acknowledges the financial support under FONDECYT Regular Project 1170226; FIA, "Fundación para la Innovación Agraria" under FIA project "FIA-PYT-2013-0018" (http://www.fia.cl/); and "Gobierno Regional Metropolitano de Santiago" (GORE-RM). P.A.Z. acknowledges the financial support of Project DICYT, 051641ZR\_DAS, Vicerrectoria de Investigación, Desarrollo e Innovación, Universidad de Santiago de Chile. H.P. acknowledges the financial support of the FONDECYT Project 1150130.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Sample Availability:** Samples of the compounds are not available from the authors.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Fabrication of Spherical Titania Inverse Opal Structures Using Electro-Hydrodynamic Atomization**

#### **Jong-Min Lim \* and Sehee Jeong**

Department of Chemical Engineering, Soonchunhyang University, 22 Soonchunhyang-ro, Shinchang-myeon, Asan-si, Chungcheongnam-do 31538, Korea; jsh951004@naver.com

**\*** Correspondence: jmlim@sch.ac.kr; Tel.: +82-41-530-4961

Academic Editors: Marinella Striccoli, Roberto Comparelli and Annamaria Panniello Received: 29 September 2019; Accepted: 26 October 2019; Published: 30 October 2019

**Abstract:** Spherical PS/HEMA opal structure and spherical titania inverse opal structure were fabricated by self-assembly of colloidal nanoparticles in uniform aerosol droplets generated with electro-hydrodynamic atomization method. When a solution of PS/HEMA nanoparticles with uniform size distribution was used, PS/HEMA nanoparticles self-assembled into a face-centered cubic (FCC) structure by capillary force with the evaporation of the solvent in aerosol droplet, resulting in a spherical opal structure. When PS/HEMA nanoparticles and anatase titania nanoparticles were dispersed simultaneously into the solution, titania nanoparticles with relatively smaller size were assembled at the interstitial site of PS/HEMA nanoparticles packed in the FCC structure, resulting in a spherical opal composite structure. Spherical titania inverse opal structure was fabricated after removing PS/HEMA nanoparticles from the spherical opal composite structure by calcination.

**Keywords:** colloidal crystal; inverse opal; electro-hydrodynamic atomization; photonic ball; titania

#### **1. Introduction**

Photonic bandgap is a phenomenon that originated from periodically arranged materials having different refractive indices. It can be controlled by changing the structure of periodicity and the refractive index of constituent material. Yablonovitch et al. have demonstrated the fabrication of 3D photonic crystal having a photonic bandgap in the microwave region [1]. Since then, there have been various studies on the fabrication of 3D photonic crystal structures [2–7]. With a bottom-up approach, colloidal self-assembled structures have been widely adopted to make 3D photonic crystal structures due to their simplicity and cost-effectiveness. As the solvent of colloidal dispersion evaporates slowly, nanoparticles with uniform size distribution are self-assembled into a hexagonal crystal lattice (i.e., face-centered cubic (FCC) structure) to form an opal structure. In addition, an inverse opal structure with more robust photonic bandgap can be prepared by filling interstitial sites of the opal structure using materials with high refractive index and removing colloids having a hexagonal crystal lattice [3,5,6,8].

One of the main issues in the fabrication of 3D colloidal photonic crystals (e.g., opal and inverse opal structures) is the control of their shape and size in a reproducible manner. In order to control size and shape, various studies have prepared spherical 3D colloidal crystals using uniform-sized droplets as confined geometries [6–9]. Electro-hydrodynamic atomization method, also known as electrospray, has been developed for large-scale production of spherical 3D colloidal crystals. Moon et al. have demonstrated electro-hydrodynamic atomization for large-scale production of spherical polystyrene opal structures and spherical silica inverse opal structures [10]. Since they used water as a solvent, an additional process was required for the evaporation of the solvent. In addition, spherical titania inverse opal structures could not be fabricated due to poor dispersion stability of titania nanoparticles in aqueous solution. Hong et al. have prepared spherical silica opal structures without an additional solvent evaporation process using ethanol as a solvent in electro-hydrodynamic atomization [11]. They

also used crosslinked polystyrene nanoparticles in toluene for electro-hydrodynamic atomization. However, spherical structures with irregularly packed colloidal crystal shells and hollow cores were fabricated due to the extremely fast evaporation rate of toluene. shells and hollow cores were fabricated due to the extremely fast evaporation rate of toluene. In this study, we used poly [styrene-co-(2-hydroxyethyl methacrylate)] nanoparticles (PS/HEMA nanoparticles) with enhanced stability in ethanol to fabricate spherical PS/HEMA opal

hydrodynamic atomization. However, spherical structures with irregularly packed colloidal crystal

*Molecules* **2019**, *24*, x FOR PEER REVIEW 2 of 9

nanoparticles in aqueous solution. Hong et al. have prepared spherical silica opal structures without an additional solvent evaporation process using ethanol as a solvent in electro-hydrodynamic

In this study, we used poly [styrene-co-(2-hydroxyethyl methacrylate)] nanoparticles (PS/HEMA nanoparticles) with enhanced stability in ethanol to fabricate spherical PS/HEMA opal structures and spherical titania inverse opal structures by electro-hydrodynamic atomization. Without the additional evaporation process, we could prepare compact spherical PS/HEMA opal structures regularly packed in an FCC structure. When a mixture of PS/HEMA nanoparticles in ethanol and anatase titania nanoparticles in methanol is used for electro-hydrodynamic atomization, titania nanoparticles of relatively small size were assembled at the interstitial site of PS/HEMA nanoparticles packed in the FCC structure, resulting in a spherical opal composite structure. After calcination to remove PS/HEMA nanoparticles, spherical titania inverse opal could be fabricated. Since the electro-hydrodynamic atomization method can rapidly prepare a large amount of spherical PS/HEMA opal structures and spherical titania inverse opal structures, it can expedite the commercial application of spherical opal and spherical inverse opal structures in various areas, including reflective mode display, photo catalysis, solar cell electrode materials, and analytical systems [12–17]. structures and spherical titania inverse opal structures by electro-hydrodynamic atomization. Without the additional evaporation process, we could prepare compact spherical PS/HEMA opal structures regularly packed in an FCC structure. When a mixture of PS/HEMA nanoparticles in ethanol and anatase titania nanoparticles in methanol is used for electro-hydrodynamic atomization, titania nanoparticles of relatively small size were assembled at the interstitial site of PS/HEMA nanoparticles packed in the FCC structure, resulting in a spherical opal composite structure. After calcination to remove PS/HEMA nanoparticles, spherical titania inverse opal could be fabricated. Since the electro-hydrodynamic atomization method can rapidly prepare a large amount of spherical PS/HEMA opal structures and spherical titania inverse opal structures, it can expedite the commercial application of spherical opal and spherical inverse opal structures in various areas, including reflective mode display, photo catalysis, solar cell electrode materials, and analytical systems [12–17].

#### **2. Results 2. Results**

#### *2.1. Characterization of Monodisperse PS*/*HEMA Nanoparticles and Titania Nanoparticles 2.1. Characterization of Monodisperse PS/HEMA Nanoparticles and Titania Nanoparticles*

Figure 1a shows an SEM image of monodisperse PS/HEMA nanoparticles. PS/HEMA nanoparticles were hexagonally packed (i.e., FCC structure), confirming uniform size distribution of PS/HEMA nanoparticles. Figure 1b shows a TEM image of titania nanoparticles. These titania nanoparticles had an anatase phase based on power x-ray diffraction as in Figure 1c. Figure 1a shows an SEM image of monodisperse PS/HEMA nanoparticles. PS/HEMA nanoparticles were hexagonally packed (i.e., FCC structure), confirming uniform size distribution of PS/HEMA nanoparticles. Figure 1b shows a TEM image of titania nanoparticles. These titania nanoparticles had an anatase phase based on power x-ray diffraction as in Figure 1c.

**Figure 1.** (**a**) SEM image of PS/HEMA latex nanoparticles. (**b**) TEM image and (**c**) x-ray diffraction data of titania colloidal nanoparticles. **Figure 1.** (**a**) SEM image of PS/HEMA latex nanoparticles. (**b**) TEM image and (**c**) x-ray diffraction data of titania colloidal nanoparticles.

*2.2. Preparation of Uniform Aerosol Droplets Using Electro-hydrodynamic Atomization* 

#### *2.2. Preparation of Uniform Aerosol Droplets Using Electro-hydrodynamic Atomization Molecules* **2019**, *24*, x FOR PEER REVIEW 3 of 9

AC electric field in the range of 1.2–1.8 kV/mm intensity with frequency of 1–5 kHz was used to maintain stable Taylor cone jet mode. The frequency of the AC electric field did not significantly affect the size of aerosol droplets because the frequency range (i.e., 1–5 kHz) was high enough [11]. AC electric field in the range of 1.2–1.8 kV/mm intensity with frequency of 1–5 kHz was used to maintain stable Taylor cone jet mode. The frequency of the AC electric field did not significantly affect the size of aerosol droplets because the frequency range (i.e., 1–5 kHz) was high enough [11].

#### *2.3. Fabrication of Spherical PS*/*HEMA Opal Structures Structures 2.3. Fabrication of Spherical PS/HEMA Opal Structures Structures*

Figures 2a–c and 2d–f show SEM and digital camera images of the spherical opal structure consisting of 280 and 210-nm-sized monodisperse PS/HEMA nanoparticles, respectively. As shown in Figure 2a,d, spherical opal structure could be fabricated as the solvent evaporated. As shown in magnified SEM images (Figure 2b,e), monodisperse PS/HEMA nanoparticles self-assembled into a hexagonal crystal lattice, showing the (111) plane of the FCC structure. Figure 2a–c and Figure 2d–f show SEM and digital camera images of the spherical opal structure consisting of 280 and 210-nm-sized monodisperse PS/HEMA nanoparticles, respectively. As shown in Figure 2a,d, spherical opal structure could be fabricated as the solvent evaporated. As shown in magnified SEM images (Figure 2b,e), monodisperse PS/HEMA nanoparticles self-assembled into a hexagonal crystal lattice, showing the (111) plane of the FCC structure.

**Figure 2.** (**a**,**b**) SEM images and (**c**) digital camera image of spherical opal structures made of 280 nmsized PS/HEMA nanoparticles. (**d**, **e**) SEM images and (**f**) digital camera image of spherical opal structures from 210-nm-sized PS/HEMA nanoparticles. **Figure 2.** (**a**,**b**) SEM images and (**c**) digital camera image of spherical opal structures made of 280 nm-sized PS/HEMA nanoparticles. (**d**,**e**) SEM images and (**f**) digital camera image of spherical opal structures from 210-nm-sized PS/HEMA nanoparticles.

#### *2.4. Fabrication of Spherical Titania Inverse Opal Structures 2.4. Fabrication of Spherical Titania Inverse Opal Structures*

Since both PS/HEMA nanoparticles and anatase titania nanoparticles could make stable colloidal dispersion in the mixture of ethanol and methanol at a 3:1 volumetric ratio, we could use mixed colloidal dispersion for electro-hydrodynamic atomization. Figure 3a,b show SEM images of spherical opal composite structures consisting of PS/HEMA nanoparticles and anatase titania nanoparticles. As shown in Figure 3c,d, spherical titania inverse opal structures could be obtained after removing PS/HEMA nanoparticles by 500 °C calcination. The inset of Figure 3b,d shows the fast Fourier transform (FTT) of the SEM images. Since both PS/HEMA nanoparticles and anatase titania nanoparticles could make stable colloidal dispersion in the mixture of ethanol and methanol at a 3:1 volumetric ratio, we could use mixed colloidal dispersion for electro-hydrodynamic atomization. Figure 3a,b show SEM images of spherical opal composite structures consisting of PS/HEMA nanoparticles and anatase titania nanoparticles. As shown in Figure 3c,d, spherical titania inverse opal structures could be obtained after removing PS/HEMA nanoparticles by 500 ◦C calcination. The inset of Figure 3b,d shows the fast Fourier transform (FTT) of the SEM images.

*Molecules* **2019**, *24*, x FOR PEER REVIEW 4 of 9

**Figure 3.** (**a**,**b**) SEM images of spherical opal composite structures consisting of PS/HEMA nanoparticles and titania nanoparticles. (**c**, **d**) SEM images of spherical titania inverse opal structure. Inset shows the fast Fourier transform (FTT) of the SEM images. **Figure 3.** (**a**,**b**) SEM images of spherical opal composite structures consisting of PS/HEMA nanoparticles and titania nanoparticles. (**c**,**d**) SEM images of spherical titania inverse opal structure. Inset shows the fast Fourier transform (FTT) of the SEM images.

#### **3. Discussion**

and the periodicity of the structure.

**3. Discussion**  The size of the spherical opal structure was determined by the size of the aerosol droplet produced by electro-hydrodynamic atomization and amounts of nanoparticles in an aerosol droplet. Since the concentration of nanoparticles in the solution was kept constant, uniform aerosol droplets The size of the spherical opal structure was determined by the size of the aerosol droplet produced by electro-hydrodynamic atomization and amounts of nanoparticles in an aerosol droplet. Since the concentration of nanoparticles in the solution was kept constant, uniform aerosol droplets should be stably generated to produce spherical opal structures with uniform size distribution.

should be stably generated to produce spherical opal structures with uniform size distribution. Uniform aerosol droplets of several tens to several hundreds of micrometers in size could be generated in a controlled manner using electro-hydrodynamic atomization method. When colloidal dispersion was injected into the capillary needle at a constant flow rate using a syringe pump, droplets were formed at the end of the capillary. When gravity became greater than restoring surface tension, the droplet detached from the capillary and dripped through the ring electrode. In the electro-hydrodynamic atomization method, AC electric field was applied to a stainless capillary needle and a ring electrode. Because AC electric field deforming electrical tangential stress was applied to the meniscus, when the electric field increased, the size of the dripping droplet became smaller due to electrical tangential stress known as dripping mode. When the applied electric field exceeded the threshold value, the meniscus became a cone shape known as Taylor cone jet mode [18]. In the Taylor cone jet mode, droplets with uniform size distribution could be generated stably. As Uniform aerosol droplets of several tens to several hundreds of micrometers in size could be generated in a controlled manner using electro-hydrodynamic atomization method. When colloidal dispersion was injected into the capillary needle at a constant flow rate using a syringe pump, droplets were formed at the end of the capillary. When gravity became greater than restoring surface tension, the droplet detached from the capillary and dripped through the ring electrode. In the electro-hydrodynamic atomization method, AC electric field was applied to a stainless capillary needle and a ring electrode. Because AC electric field deforming electrical tangential stress was applied to the meniscus, when the electric field increased, the size of the dripping droplet became smaller due to electrical tangential stress known as dripping mode. When the applied electric field exceeded the threshold value, the meniscus became a cone shape known as Taylor cone jet mode [18]. In the Taylor cone jet mode, droplets with uniform size distribution could be generated stably. As electric field strength further increased, multiple unstable jets were formed at the end of the stainless steel capillary known as multi-jet mode [10,11].

electric field strength further increased, multiple unstable jets were formed at the end of the stainless steel capillary known as multi-jet mode [10,11]. The solvent in aerosol droplets evaporated while droplets passed the cylindrical plastic tube. As the solvent evaporated, nanoparticles in the droplet gradually got closer and began to self-assemble The solvent in aerosol droplets evaporated while droplets passed the cylindrical plastic tube. As the solvent evaporated, nanoparticles in the droplet gradually got closer and began to self-assemble by capillary force. When nanoparticles were brought into contact with each other, nanoparticles were fixed to a spherical opal structure by van der Waals force. The self-assembly and fixation process could be completed in a few seconds after droplet generation by electro-hydrodynamic atomization.

by capillary force. When nanoparticles were brought into contact with each other, nanoparticles were fixed to a spherical opal structure by van der Waals force. The self-assembly and fixation process could be completed in a few seconds after droplet generation by electro-hydrodynamic atomization.

nanoparticles. The periodicities of the structures in Figure 3b,d are 239 nm and 224 nm, respectively. The periodicity of the structure was decreased by about 6.3% by the calcination. Because of the defects in Figure 3b, there are slight differences between the average diameter of PS/HEMA nanoparticles

Since the sizes of titania nanoparticles were much smaller than those of PS/HEMA nanoparticles, titania nanoparticles were assembled at the interstitial site of hexagonally packed PS/HEMA nanoparticles. The periodicities of the structures in Figure 3b,d are 239 nm and 224 nm, respectively. The periodicity of the structure was decreased by about 6.3% by the calcination. Because of the defects in Figure 3b, there are slight differences between the average diameter of PS/HEMA nanoparticles and the periodicity of the structure.

A large amount of spherical PS/HEMA opal structure could be obtained rapidly using the electro-hydrodynamic atomization method. As shown in Figure 2c,f, spherical PS/HEMA opal structures were dispersed in water in order to demonstrate optical properties. Reflected diffraction colors could be controlled depending on the size of nanoparticles constituting the spherical opal structure. Since the surface of the spherical opal structure was composed of the (111) plane of the FCC structure, light corresponding to the bandgap of the photonic crystal was reflected. The bandgap position in wavelength (λ) could be estimated by Bragg's law for the (111) plane of the FCC structure: [2,19]

$$\lambda = 2 \text{d}n\_{eff} = \left(\frac{8}{3}\right)^{0.5} \text{D}n\_{eff} \tag{1}$$

where d was the spacing of the (111) plane, *ne f f* was the effective refractive index, and *D* was the diameter of constituent nanoparticles. The *ne f f* could be obtained by

$$m\_{eff} = \left\{\phi\_p n\_p^2 + (1 - \phi\_p)n\_m^2\right\} \tag{2}$$

whereφ*<sup>p</sup>* was the volume fraction of nanoparticles, and *n<sup>p</sup>* and *n<sup>m</sup>* were refractive indices of nanoparticles and matrix, respectively. The bandgap position (λ) could be estimated from the Bragg's law using the volume fraction of nanoparticles in the FCC structure (φ*<sup>p</sup>* = 0.74), refractive index of PS/HEMA particles (*n<sup>p</sup>* = 1.59), and refractive index of water (*n<sup>m</sup>* = 1.33). For spherical opal structures composed of 280 and 210-nm-sized PS/HEMA nanoparticles, bandgap positions (λ) estimated from the Bragg's law were 698 nm and 523 nm, respectively. Since spherical PS/HEMA opal structure could be fabricated in large quantities using the electro-hydrodynamic atomization method, bandgap positions (λ) can be visually confirmed from the digital camera images shown in Figure 2c,f.

The bandgap position (λ) of the spherical titania inverse opal structure could be estimated from Bragg's law using the refractive index of anatase phase titania (*n<sup>m</sup>* = 2.5) and the refractive index of water (*n<sup>p</sup>* = 1.33). Here, we assumed that 74% of the unit cell structure was occupied by water for the case of spherical inverse opal structure. When 280-nm-sized PS/HEMA nanoparticles were used to make the spherical titania inverse opal structure, the bandgap was located in the near infrared region (i.e., λ = 783 nm). The bandgap position could be controlled by changing the size of PS/HEMA nanoparticles used to fabricate spherical titania inverse opal structures. The bandgap position can also be controlled by changing the refractive indices of materials used to make spherical inverse opal structures and solvents used to disperse the spherical inverse opal structure.

#### **4. Materials and Methods**

#### *4.1. Chemicals*

All chemicals and solvents were reagent grade and used without further purification. Styrene (99%) was purchased from Kanto chemical (Tokyo, Japan), and 2-hydroxyethyl methacrylate (reagent grade) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Potassium persulfate (98%) as the initiator was purchased from Kanto chemical (Tokyo, Japan). Titania nanoparticles in methanol dispersion (DH 60) were obtained from Nissan chemicals (Tokyo, Japan). Ethanol (≥ 99.9%) was purchased from Merck (Kenilworth, NJ, USA).

#### *4.2. Synthesis of Monodisperse PS*/*HEMA Nanoparticles*

Monodisperse PS/HEMA nanoparticles were synthesized through batch type surfactant-free emulsion copolymerization in aqueous medium. Detailed synthesis procedures can be found elsewhere [20]. Aqueous PS/HEMA nanoparticle dispersion was dried on a piece of silicon wafer and coated with gold for observation using a field emission scanning electron microscope (FE-SEM, XL305FEG, Philips (Amsterdam, The Netherlands)).

#### *4.3. Fabrication of Aerosol Droplets Using Electro-hydrodynamic Atomization*

PS/HEMA nanoparticles were re-dispersed in ethanol prior to electro-hydrodynamic atomization. To make titania inverse opal, PS/HEMA in ethanol dispersion and titania in methanol dispersion were mixed at 3:1 volumetric ratio. Vortex mixer (Maxi mix II, Thermo Scientific (Waltham, MA, USA)) and ultrasonic cleaner (EW-08893-21, Cole-parmer (Vernon, IL, USA)) were used for re-dispersion and the mixing process.

Aerosol droplets with uniform size distribution were prepared by using the electro-hydrodynamic atomization method. The schematic of the experimental set up for electro-hydrodynamic atomization is illustrated in Scheme 1a. The apparatus for electro-hydrodynamic atomization consists of a colloidal dispersion injection part, an electric field control part, a meniscus observation part, and a colloidal self-assembled structure collection part. In the colloidal dispersion injection part, colloidal dispersion was introduced to a disposable 1-mL syringe which was connected to a stainless capillary needle (metal hub needle, Hamilton (Reno, NV, USA)). The disposable syringe was loaded to a syringe pump (KDS 100, KD Scientific (Holliston, MA, USA)) to maintain a constant flow rate of 0.5 mL/h. In the electric field control part, an arbitrary function generator (DS345, Stanford Research Systems (Sunnyvale, CA, USA)) and a high voltage power amplifier (20/20B, Trek Inc. (Lockport, NY, USA)) were connected to a stainless capillary needle and a ring electrode with circular hole of 1 cm in diameter. The shape of the meniscus formed on the capillary was changed with AC electric field. The AC electric field was maintained in the range of 1.2–1.8 kV/mm with 1–5 kHz frequency for stable generation of droplets in Taylor cone jet mode [18]. In order to precisely observe the shape, a CCD camera with a ring-shaped illuminator was used in the meniscus observation part. At the bottom of the ring electrode, there was a 90-cm-long cylindrical plastic tube for evaporating solvent in aerosol droplets. When the solvent in the aerosol droplet was completely evaporated as shown in Scheme 1b, PS/HEMA nanoparticles inside an aerosol droplet self-assembled to form spherical PS/HEMA opal structures. In addition, when PS/HEMA nanoparticles and titania nanoparticles were used simultaneously as shown in Scheme 1c, titania nanoparticles with smaller size were self-assembled at the interstitial site of hexagonally-packed PS/HEMA nanoparticles. As the solvent evaporated, spherical opal composite structures were formed as shown in Scheme 1c. Spherical opal structures and spherical opal composite structures could be collected by placing a glass petri-dish under the cylindrical plastic tube. As shown in Scheme 1c, spherical titania inverse opal structures could be obtained by removing PS/HEMA nanoparticles from the spherical opal composite structures via 500 ◦C calcination for about 8 hours using a muffle furnace (PK9712150030-9, Isuzu (Tokyo, Japan)). Spherical PS/HEMA opal structures and spherical titania inverse opal structures were observed using a FE-SEM (XL305FEG, Philips (Amsterdam, The Netherlands)).

*Molecules* **2019**, *24*, x FOR PEER REVIEW 7 of 9

**Scheme 1.** (**a**) Schematic of experimental set up of electro-hydrodynamic atomization for generating uniform droplets. Schematics of self-assembly processes of (**b**) spherical PS/HEMA opal structures and (**c**) spherical titania inverse opal structures. **Scheme 1.** (**a**) Schematic of experimental set up of electro-hydrodynamic atomization for generating uniform droplets. Schematics of self-assembly processes of (**b**) spherical PS/HEMA opal structures and (**c**) spherical titania inverse opal structures.

#### **5. Conclusions 5. Conclusions**

Uniform aerosol droplets could be produced by the Taylor cone jet mode of the electrohydrodynamic atomization method using an AC electric field. As the solvent evaporated, PS/HEMA nanoparticles self-assembled in droplets to create a spherical opal structure consisting of PS/HEMA nanoparticles. When PS/HEMA nanoparticles and anatase titania nanoparticles were simultaneously used, spherical opal composite structure could be prepared. After removing PS/HEMA nanoparticles by 500 °C calcination, spherical titania inverse opal structure could be fabricated. Bragg's law was used to estimate the bandgap position (λ) of the spherical PS/HEMA opal structure and the spherical titania inverse opal structure. Since spherical opal structures could be prepared rapidly in large quantities using the electro-hydrodynamic atomization method, the bandgap position of spherical PS/HEMA opal structures could be visually confirmed from digital camera images. The bandgap position could be controlled simply by changing the size of PS/HEMA nanoparticles. The spherical opal structures and the spherical titania inverse opal structures produced by the electro-Uniform aerosol droplets could be produced by the Taylor cone jet mode of the electro-hydrodynamic atomization method using an AC electric field. As the solvent evaporated, PS/HEMA nanoparticles self-assembled in droplets to create a spherical opal structure consisting of PS/HEMA nanoparticles. When PS/HEMA nanoparticles and anatase titania nanoparticles were simultaneously used, spherical opal composite structure could be prepared. After removing PS/HEMA nanoparticles by 500 ◦C calcination, spherical titania inverse opal structure could be fabricated. Bragg's law was used to estimate the bandgap position (λ) of the spherical PS/HEMA opal structure and the spherical titania inverse opal structure. Since spherical opal structures could be prepared rapidly in large quantities using the electro-hydrodynamic atomization method, the bandgap position of spherical PS/HEMA opal structures could be visually confirmed from digital camera images. The bandgap position could be controlled simply by changing the size of PS/HEMA nanoparticles. The spherical opal structures and the spherical titania inverse opal structures produced by the electro-hydrodynamic

reflective mode display, photo catalysis, solar cell electrode materials, and analytical systems.

hydrodynamic atomization method are of practical significance for various applications, including

atomization method are of practical significance for various applications, including reflective mode display, photo catalysis, solar cell electrode materials, and analytical systems.

**Author Contributions:** Conceptualization, J.-M.L.; methodology, J.-M.L.; software, S.J.; formal analysis, J.-M.L. and S.J.; investigation, J.-M.L. and S.J.; data curation, J.-M.L.; writing—original draft preparation, J.-M.L.; writing—review and editing, J.-M.L. and S.J.; visualization, J.-M.L.; supervision, J.-M.L.; project administration, J.-M.L.; funding acquisition, J.-M.L.

**Funding:** This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industry & Energy (MOTIE) of the Republic of Korea (No. 20184030202130). This work was also supported by the Soonchunhyang University Research Fund.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


**Sample Availability:** Samples of the compounds are not available from the authors.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Molecules* Editorial Office E-mail: molecules@mdpi.com www.mdpi.com/journal/molecules

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com ISBN 978-3-0365-1759-9