Introduction

The dried pulp of longan (Dimocarpus longan Lour.), which is also known as lumyai for the locals and dragon’s eye for the Chinese, is one of the geographical indicators of Northern Thailand with high economic value. In the production of the dried longan pulp, a large amount of seeds is inevitably and continually disposed leading to environmental and hygienic problems for the locals as they attract flies and have become source of pathogens. To solve the emerging problems, the utilization of the longan seeds have been encouraged. Amidst various possibilities, the use of longan seeds as a supply of natural reagents for the green synthesis of ZnO photocatalyst is appealing since the wastewater pollution introduced from the textile dyeing industry is also prominent in Northern Thailand. In addition, almost all phytochemicals extracted from plants including, for instance, flavonoids, catechins, vitamins, proteins, and sugars can be employed as either the reducing or the capping agents, or both, in the synthesis of metal and metal oxide nanoparticles (Hebbalalu et al. 2013; Nadagouda et al. 2014; Nadagouda and Varma 2008; Varma 2012). Efficiency of the synthesis using these natural reagents can also be enhanced with the use of microwave heating (Baruwati et al. 2009; Baruwati and Varma 2009; Kou and Varma 2012a, 2012b), which also raises the feasibility of further development toward practical implementation (Kharissova et al. 2013).

The employment of the crude water extract of the ground longan seeds in the microwave synthesis of ZnO nanoparticles has been attempted. While the reason for the synthesis of ZnO lies in its nonhazardous properties according to the U.S. Food and Drug Administration (21CFR182.8991) (FDA 2011), polyphenolic acids in the extract of the longan seeds (Soong and Barlow 2005; Sudjaroen et al. 2012) should be effective reagents in the synthesis. The presence of the polar functional groups in the structures of these phytochemicals (Fig. 1), among which the gallic and the ellagic acids are the major components, should additionally enhance the microwave heating in similar fashion to, for instance, glycerol and hydroxylated glycosides (Kou et al. 2013; Kou and Varma 2013). The biowaste would serve as a sustainable supply for the synthesis. Influences of important synthesis parameters on the formation of ZnO and the yielded particle sizes are presented. Performance of the synthesized ZnO nanoparticles in photocatalyzed decolorization of methylene blue (MB), malachite green (MG), methyl orange (MO), and orange II (OII), all of which have been used extensively by the local textile dyeing industry, was also evaluated in comparison with the commercialized TiO2 (Arroxide®P-25).

Fig. 1
figure 1

Molecular structures of phytochemicals found in crude water extract of ground Edor longan seeds

Full size image

Experimental

Synthesis and characterization of ZnO nanoparticles

Longan fruits of Edor cultivar were harvested from Pa Sang area, Lamphun Province, Thailand (18° 31′ 33″ N 98° 56′ 21″ E). The seeds were separated from peels and pulps right after the harvesting and were washed and dried at 75 °C for 24 before being ground to dark brown fine powder using a hammer mill (Armfield Hammer Mill FT2-A). The powder was stored in dark containers and kept at ca. 4 °C for further used throughout the project. The crude water extract of the powder was freshly prepared before each experiment using the procedure reported in the literature (Rangkadilok et al. 2007).

As a typical procedure (Somsri et al. 2016), Zn (CH3COO)2·2H2O (≥ 99.0%, Aldrich) was dissolved in 10.00 mL of the extract to varying concentrations, i.e., 0.010 to 10.00 g·L−1. The reaction was performed using a household microwave oven (Samsung Model MS23F301E), which was operated at various powers (450–800 W) and numbers of 1 min-on-1 min-off irradiation cycle (1–30 cycles). UV-vis spectroscopy (Shimadzu UV-1800) was employed to follow the synthesis reaction. The synthesized samples were characterized by the powder X-ray diffraction (PXRD; Bruker D8 Advance, Cu Kα, λ = 1.54 Å), the FT-IR spectroscopy (Bruker Tensor 27), and transmission electron microscopy (TEM; FEI TECNAI G2 20S-twin microscope1) equipped with selected area electron diffraction (SAED), the thermogravimetric analysis (TGA; NETZSCH STA 409 PC/PG), and the Brunauer­Emmett­Teller experiments (BET; Micromeritics/ASAP2460) conducted at 77 K with a prior degas at 300 °C for 12 h.

Photocatalytic dye decolorization experiments

Photocatalytic activities were evaluated from photocatalyzed decolorization of methylene blue (MB), malachite green (MG), methyl orange (MO), and orange II (OII). As a general procedure, 0.0240 g of the ZnO nanoparticles were suspended in 100.0 mL of 0.0160 g·L−1 dye solution. After leaving in the dark for 30 min, the reaction was exposed to artificial daylight (VeriVide F20 T12/D65 18 W, 300–830 nm), which was positioned at 12 cm above the suspension surface. An aliquot of 5.00 mL was taken from the reaction at every 15 min for the UV-vis spectroscopic measurement (Somsri et al. 2016). With reference to the calibration curve method, concentrations of the remaining dyes were determined. Compared with the initial concentration, efficiencies of the decolorization process were then calculated based on the reduced dye-concentration compared with the initial (Cai et al. 2017; Thennarasu et al. 2012). Photocatalytic decolorization of MB was additionally performed using different amounts of the ZnO nanoparticles ranging from 0.0100 to 0.3000 g. For comparison, Arroxide®P-25 (≥ 99.5%, Aldrich; particle size 21 nm, 36–65 m2·g−1; pHzpc = 7.5) and the as-synthesized samples were tested under the same experimental conditions.

Results and discussion

Microwave synthesis and ZnO nanoparticles

Based on characteristic absorptions of ZnO (Fig. 2), its formation was insignificant when the Zn (CH3COO)2·2H2O concentration in the extract was lower than 1.00 g·L−1, at which the characteristic absorption first appeared at the λmax of 280 nm. As the precursor concentration was increased, this absorption feature became broadened with a small shoulder at 400 nm, suggesting a wide range of particle size distribution. The effect grew substantial when the precursor concentration was 10.0 g·L−1, at which the multi-λmax feature was apparent. An excessive concentration of the precursor in the extract is therefore detrimental to particle size homogeneity which should attribute to the particle growth. A severe diversity in the yielded particle size observed at 10.00 g·L−1 turned, however, silent when either the applied microwave power was ≤ 600 W or the irradiation cycle was ≤ 10 cycles.

Fig. 2
figure 2

UV-vis spectra of the reaction mixtures under the following conditions: a varied precursor concentration, 800 W, 30 cycles; b varied microwave power, 30 cycles, 10.0 g L−1 precursor; and c varied number of irradiation cycles, 800 W, 10.0 g·L−1 precursor, with the corresponding λmax in the inset ( = extract, ●= ZnO)

Full size image

The phytochemicals in the as-synthesized samples, denoted as ZnOamor henceforth was manifested through the samples’ chestnut-brown color, which was affirmed by the FT-IR spectra of the yielded samples compared with the dried extract (Fig. 3). The primary vibrational features, including ν (O-H; 2985–3670 cm−1), ν (C=O; 1585 cm−1), ν (C-OH; 1415 cm−1), and δ (C-OH; 1041 cm−1), correspond well to the functional groups of the phytochemicals found in the extract among which gallic and ellagic acids are the major components (De Goes Sampaio et al. 2019; Sudjaroen et al. 2012). The capping phytochemicals were also explicit in the TEM images showing the expansive light area of the phytochemicals and the dark areas of the ZnOamor (Fig. 4). The PXRD patterns collected on these powders suggested them to be amorphous, which is not unexpected as the long-range structural order of the derived ZnOamor should be disrupted by the capping phytochemicals.

Fig. 3
figure 3

a TGA diagram of ZnOamor with the FT-IR spectra (inset) of the dried extract compared with ZnOamor and ZnO, and b the PXRD patterns of ZnOamor and ZnO

Full size image
Fig. 4
figure 4

TEM images of a ZnOamor and b ZnO, with the corresponding c SAED pattern and d particle size distribution of ZnO

Full size image

Upon the heating of the ZnOamor from room temperature to 900 °C under the nitrogen gas flow (Fig. 3), a gradual weight loss of over 70% occurred resulting in a white powder. The temperature at which the significant weight loss commenced is in a typical range where the organic molecules in most coordination compounds decompose. This is well supported by the FT-IR spectrum of the white powder, exhibiting only ν (Zn-O) at 440 cm−1 and the ν (O-H) of possibly the surface water centering at 3418 cm−1 (Nosrati et al. 2012; Wu et al. 2016). The complete removal of the capping phytochemicals by a simple heat treatment can therefore be assumed, which is ascertained by the TEM images displaying only the well-defined ZnO (Fig. 4). This can be regarded as one of the merits of the longan seeds extract since the complete removal of the capping phytochemicals to generate an active surface of the ZnO nanoparticles is the inherent drawback of the phytosynthesis process. The remaining of the phytochemicals can obstruct access to the surface of the particles and nullify their activity (Somsri et al. 2016).

The yielded white powder was characterized to be a pure phase of hexagonal ZnO (JCPDS 36–1451; a = 3.247(8) Å, c = 5.202(6) Å) by the PXRD (Fig. 3) and the SAED (Fig. 4) experiments. The heat treatment can thereby be regarded also as the calcination step. According to the TEM images, shapes of the ZnO particles were mostly irregular although some clearly manifested the inherent hexagonal symmetry. Sizes of the particles were distributed from 10 to 100 nm of which 50% and 78% were in the ranges of 40–60 nm and 40–80 nm, respectively (Fig. 4). The BET surface area of these ZnO nanoparticles was found to be 35 m2·g−1. The band gap energy of 3.42 eV was determined from the Kubelka-Munk plot and is in the common range reported for the ZnO nanoparticles (Cai et al. 2017; Janotti and Van de Walle 2009; Reynolds et al. 1999; Srikant and Clarke 1998).

Photocatalytic activities of ZnO nanoparticles in decolorization of organic dyes

In accordance with the decolorization of the cationic MB and MG, and the anionic MO and OII, performances of the ZnO nanoparticles and the commercially available TiO2 (Arroxide®P-25) were approximately on par (Fig. 5 and Fig. 6). Efficiency of the ZnO nanoparticles was substantial in the case of MB of which over 70% of the dye was decolored in only half an hour and almost 100% in approximately an hour. The particles became, however, less efficient for the decolorization of MG, OII, and MO, successively. Such disparity can be accounted for by different accessibility of these dyes onto the surface of the ZnO nanoparticles, which is well acknowledged as being crucial in determining efficiency of the catalyst (Abdel-Khalek et al. 2018; Thirumavalavan et al. 2013). The assumption is supported by the silent activities of the chestnut-brown ZnOamor where the access to the nanoparticles’ surface should be blocked by the capping phytochemicals.

Fig. 5
figure 5

Efficiencies of ZnO in photocatalyzed decolorization of a MB, MG, MO, and OII, compared with those of TiO2 (Arroxide®P-25), and b of MB only but with different amounts of the catalysts

Full size image
Fig. 6
figure 6

Photos showing time-dependent decolorization of a MB, b MG, c MO, and d OII solutions catalyzed by ZnO (upper row) and TiO2 (Arroxide®P-25; lower row)

Full size image

Accessibility onto the surface of the photocatalyst depends however on the surface charge of the organic dyes at the experimental pH. Based on the pH drift method (Intarasuwan et al. 2017), the pH at the zero-point charge (pHzpc) estimated for the yielded ZnO nanoparticles was 7.7. Compared with the pH at which the experiments were performed (pH = 8.5), the negatively charged surface could be assumed for the ZnO. The preference toward the adsorption of the cationic dyes (MB and MG) and better catalytic efficiency than the negative dyes (MO and OII) can thus be accounted for by an electrostatic interaction. The better performance of the cationic MB bearing a positive charge at pH > 3.8 (Kim et al. 2013) than that of MG of which the pKa is 6.9 (Culp and Beland 1996) can be rationalized by the feasible co-existence of both the cationic and the neutral carbinol forms of MG (Culp and Beland 1996; Cuong et al. 2012). In concert with the decolorization of MB (Fig. 5), the increase in amount of the ZnO nanoparticles has been revealed to shorten the reaction time required for the complete decolorization. The amount of the ZnO nanoparticles over 0.1000 g however did not introduce any improvement, suggesting the reaching of its limitation under the employed conditions.

Conclusions

In summary, the crude water extract of ground Edor longan seeds has been used in the facile and rapid microwave synthesis of ZnO nanoparticles although the calcination was required. The capping phytochemicals could be completely removed to activate the photocatalytic activity of the particles during the calcination, which can be regarded as the major merit of the longan seed extract. The use of 10.0 g·L−1 zinc acetate, 800 W, and 30 cycles of the microwave irradiation with successive heating at 900 °C for 1.5 h provided the pure hexagonal phase of ZnO nanoparticles of 10–100 nm in sizes and the BET surface area of 35 m2·g−1. At the pH of 8.5, the efficiencies of these ZnO nanoparticles (band gap energy = 3.42 eV and pHzpc = 7.7) toward the decolorization of some commonly used organic dyes were approximately on par with the commercially available TiO2 (Arroxide®P-25). Based on this study, an alternative utilization of the longan seeds which are readily and sustainably available in Northern Thailand is demonstrated. The ZnO nanoparticles synthesized using the water extract of this biowaste can promisingly be employed in the treatment of wastewater released from the local textile dyeing industry.