Biomass-derived multiatom-doped carbon dots for the photocatalytic reduction of Cr(VI) and precipitation of Cr(III)

Heavy metal contamination of natural water sources has become a critical environmental issue, prompting extensive research into mitigation strategies over the past few decades. Chromium (Cr), which primarily occurs as chromite and crocoite, is a significant element in the Earth’s crust and is widely utilized in various industrial processes. In particular, Cr is extensively employed in the steel industry to enhance the hardness, brittleness, and corrosion resistance of steel. Cr naturally exists in two oxidation states: hexavalent [Cr(VI)] and trivalent [Cr(III)], with both states having widespread applications in catalysis¹, leather tanning², pigment and dye production³, and chrome electroplating⁴. However, due to the extensive industrial applications of Cr, a large amount of waste is produced and released into the environment in solid, liquid, and gas forms, which adversely affects ecological systems⁵ and causes various health issues, such as skin problems and cancer^6,7. Although Cr(III) is less toxic than Cr(VI) in small quantities and is an essential constituent of glucose and lipid metabolism in the human body⁴, excessive levels of Cr(III) can also have harmful effects. Therefore, addressing Cr(VI) contamination in water sources is of paramount importance.

Ion exchange, chemical reduction, electrodialysis, and membrane separation are conventional methods used to treat Cr(VI) contamination⁸. However, these traditional techniques have limitations such as low efficiency, labor intensity, incomplete removal, high operational costs, secondary contamination, limited transportability, and the generation of hazardous sludge⁹. As a result, various metal-based catalysts, including semiconductors¹⁰, organic compounds¹¹, metal composites^12,13, and metal-organic frameworks^14,15, have been employed for Cr(VI) removal. However, these materials tend to be expensive and may pose adverse environmental or health effects. Consequently, achieving complete Cr removal without generating secondary products or harmful effects remains a critical area of ongoing research.

Eco-friendly photocatalytic materials are essential for the complete removal of pollutants from contaminated water. In particular, biomass-derived materials hold significant promise as sustainable and environmentally efficient photocatalysts for pollution remediation. Among these, biomass-derived nanocarbons stand out as a viable alternative to organic compounds and metal composites due to their high solubility, large surface area, eco-friendliness, sustainability, cost-effectiveness, excellent luminescence, biocompatibility, and low toxicity. Consequently, they are considered ideal candidates for photocatalytic applications. Several studies have notably focused on the synthesis of nanocarbons from biomass and their wide-ranging applications¹⁶, such as photocatalysis¹⁷, sensing^17,18, bio-imaging¹⁹, electrochemical sensing²⁰, energy storage²¹, and biological applications²². Mahfouz et al. synthesized nano carbons from date palm leaves using pyrolysis and ball milling, which were then utilized in the fabrication of a high-performance symmetric supercapacitator²¹. Similarly, Su et al. employed expired milk for the subgram-scale synthesis of nitrogen-doped fluorescent carbon dots, applying them in the bioimaging of HeLa cells, Fe(III) ion detection, and solid-state patterning²³. Rajapandi et al. developed green fluorescent nitrogen-doped biogenic carbon dots, applying them as eco-friendly catalysts for the degradation of safranin O dye and for antibacterial applications against four different bacteria²⁴. Sonkar et al. synthesized water-soluble carbon nano-onions from wood wool and utilized them to promote the growth of gram plants²⁵. Avila et al. investigated the fluorescence turn-off sensing of Cr(VI) and Cu(II) and the flocculation-based removal of Cu(II) using avocado-seed-derived carbon dots through carbonization at different temperatures²⁶. These studies collectively highlight the superior performance and promising applications of biomass-derived materials, offering environmental benefits without adverse ecological impacts.

This study involved the synthesis of an eco-friendly multiatom-doped carbon dots (MACDs) photocatalyst using a single-step solvothermal method, with Kalanchoe pinnata leaves as the precursor and ethanol as the solvent. The selection of Kalanchoe pinnata leaves was based on several factors, including the availability of free, natural materials. In addition, these leaves are thicker than many other plant leaves and contain a high concentration of elements that facilitate multiatom doping. Moreover, they are free of toxic and harmful chemicals, ensuring minimal environmental impact. During the solvothermal process, carbon self-assembly results in the formation of MACDs from biomass. At 180 °C, the solvothermal treatment drives self-assembly, thereby promoting the synthesis of spherical MACDs. The use of elements already presents in the precursor material instead of introducing external dopants is referred to as self-doping, which enhances the sustainability and eco-friendliness of the carbon dots synthesis process. In the case of the MACDs, multiatom doping occurred without the addition of external dopants during synthesis. The primary objective of this study was to develop an efficient, cost-effective, and environmentally friendly photocatalyst capable of simultaneously removing Cr(VI) and Cr(III) through photocatalytic reduction and precipitation, respectively. The prepared MACDs were characterized using various techniques to analyze their structural morphology, optical properties, and elemental composition. The photocatalytic reduction performance of the MACDs was evaluated by removing Cr(VI) from contaminated water under direct light irradiation, resulting in the precipitation of final products [Cr(III)/MACDs]. Furthermore, the MACDs successfully removed Cr(VI) without the addition of a hole scavenger. In addition, the effects of catalyst loading, initial solution concentration, coexisting ions, and reaction medium were thoroughly investigated. The recyclability and potential photocatalytic reduction mechanism were also explored. MACDs exhibit significant potential for the complete removal of Cr(VI) without generating harmful by-products. This study provides a new perspective for designing eco-friendly environmental remediation strategies based on photocatalytic processes.

Materials, chemicals, and instrumentation

The Kalanchoe pinnata plant was purchased from a local market flower market in Daejeon, South Korea. The following chemicals were procured from Thermo Fisher Scientific, Inc. (USA) and Sigma-Aldrich Inc. (USA): 1,5-diphenylcarbazide (DPC), acetone (C₃H₆O), sulfuric acid (H₂SO₄), ethyl alcohol (EtOH), potassium dichromate (K₂Cr₂O₇), hydrochloric acid (HCl), nitric acid (HNO₃), iron (III) nitrate, calcium nitrate, copper(II) nitrate, sodium chloride, potassium carbonate, ethylenediaminetetraacetic acid disodium salt (Na₂-EDTA), p-benzoquinone (p-BZQ), tertiary butyl alcohol (t-Bu), potassium persulfate (K₂S₂O₈) and sodium hydrogen phosphate. Milli-Q distilled water was used for all experiments. All chemicals were used without further purification. The entire experiment, including the handling use of hazardous chemicals, was conducted carried in a fume hood while adhering to safety precautions, such as wearing personal protective equipment (PPE) and gloves.

An ultraviolet (UV)–visible (Vis) spectrophotometer (S-3100; SCINCO Co., LTD, Korea) was used to obtain the absorption spectra for the DPC study. X-ray photoelectron spectroscopy (XPS) was conducted using a K-Alpha X-ray photoelectron spectrometer (Thermo Scientific Inc., USA). Micrographs were acquired using transmission electron microscopy (TEM; FEI Tecnai G² F30 S-TWIN, 300 kV, FEI Company, USA). Fluorescence analysis was performed with a Fluorolog-3 spectrofluorometer (HORIBA Ltd., Japan). Diffuse reflectance spectroscopy (DRS) measurements were performed to determine the bandgap using a UV–vis–near-infrared (NIR) spectrophotometer (SolidSpec-3700; Shimadzu Corp., Japan). The photocatalytic experiment was performed with a solar simulator customized with a 300 W Xenon-lamp. Electron paramagnetic resonance (EPR) analysis was conducted using a Bruker BioSpin GMBH, an EPR spectrometer, under the following conditions: center magnetic field of 3350 G, microwave frequency of 9.42 GHz, and power of 6.32 mW.

Synthesis of MACDs

MACDs were synthesized using a simple single-step solvothermal technique. The green leaves of Kalanchoe pinnata were selected as the precursor material due to their greater thickness and higher concentration of useful components, such as chlorophyll, metal ions, cellulose, wax hydrocarbons, and wax alcohols, compared to other plants²⁷. The leaves were thoroughly cleaned with tap water and deionized (DI) water, and then dried using filter paper. Approximately 6 grams of the leaves were then cut into small pieces to facilitate uniform mixing and placed in a 100 mL Teflon-lined stainless-steel autoclave, with 60 mL of ethanol used as the solvent. The autoclave was heated in an oven at 180 °C for 6 h, after which it was allowed to cool to room temperature (25 °C). The resulting solution was filtered using a 0.2 µm nylon Whatman syringe filter, yielding approximately 2 grams ( ~ 33%) of a black carbonized solution, which contained the MACDs. The filtered solution was transferred to a glass beaker and heated in a water bath at 90 °C to evaporate ethanol and water. After drying, the synthesized MACDs were scraped off using a spatula and stored in vials for further experimentation. The synthesized material shows excellent stability. The UV-visible spectra of the same MACDs material were obtained before and after 30 days. As a result, we observed that the spectra were similar, and all-electron transition bands were situated at the same positions (Supplementary Fig. 1a). Moreover, the photocatalytic efficiency of the MACDs material was examined before and after 30 days for 20 ppm Cr(VI) solution. The efficiency was the same except for minimal variation (Supplementary Fig. 1b).

Photocatalytic reduction and kinetic study

The photocatalytic reduction performance of the MACDs was investigated using a synthetically contaminated orange aqueous solution of Cr(VI) under light irradiation. First, 708 mg of K₂Cr₂O₇ was dissolved in 250 mL of deionized (DI) water and sonicated for 10 min to prepare a 1000 ppm stock solution of Cr(VI). The stock solution was stored and diluted as needed to prepare 50 mL Cr(VI) solutions with concentrations of 20, 50, 100, 200, and 500 ppm. Subsequently, 10 mg of MACDs were dissolved in 1 mL of DI water in six separate 2 mL Eppendorf tubes and sonicated for 30 min to ensure thorough dispersion of the MACDs in the aqueous media. The MACDs solutions were then mixed with each of the diluted Cr(VI) solutions at varying concentrations. Each solution was kept in the dark for 30 min to achieve adsorption-desorption equilibrium before exposure to light irradiation. The solutions were then exposed to direct light irradiation (200−1100 nm) for 390 min to complete the photocatalytic reduction of Cr(VI) to Cr(III).

The reaction kinetics of the photocatalytic reduction process under optimum conditions were assessed by collecting 1 mL samples from the Cr(VI) solutions at 1-h intervals. The collected samples were centrifuged at 12,500 rpm for 20 min, and the supernatant was transferred to another Eppendorf tube and diluted to 1 ppm for the DPC assay. UV–vis spectroscopy was used to obtain the absorption data of the Cr(VI) solution during the DPC assay. The Cr(VI) concentration decreased with increasing irradiation time. The color variations observed during the DPC assay are shown in Supplementary Fig. 2. The Cr(VI) concentration at each time interval was determined using the Beer-Lambert equation (A = εcl) based on the absorbance values of the samples²⁸. The concentration versus time data were linearly fitted to a pseudo-first-order kinetic model, and the rate constants and half-lives of the reactions were calculated for each Cr(VI) concentration.

DPC (C₁₃H₁₄N₄O) assay

First, 25 mg DPC was dissolved in 5 mL acetone. Subsequently, 5 mL DI water was mixed with 5 mL H₂SO₄ (8 M) to prepare a diluted H₂SO₄ solution. In a 20 mL glass vial, 7.5 mL of the diluted H₂SO₄ solution was mixed with 5 mL of the DPC solution, covered with aluminum foil to maintain sensitivity, and kept in a dark room until further use. The samples collected at specific intervals were diluted to a concentration of 1 ppm using DI water based on the dilution formula (N₁V₁ = N₂V₂). Subsequently, 48 μL of the DPC solution was added to 2 mL of the diluted samples and allowed to sit for 5 min for color stabilization. The absorbance spectra were measured at a wavelength of 540 nm using UV–vis spectroscopy to determine the absorbance of each sample. The absorption spectra for an initial Cr(VI) concentration of 20 ppm after various intervals are shown in Supplementary Fig. 3.

Fig. 1 illustrates the simple single-step solvothermal approach for the synthesis of multifluorescent MACDs, using the leaves of Kalanchoe pinnata as the precursor material. The synthesized MACDs were evaluated as efficient photocatalysts for the reduction of Cr(VI) and the precipitation of Cr(III). The Cr removal efficiencies of the synthesized MACDs were measured at various concentrations under direct light irradiation.

Schematic of the synthesis of MACDs and their application in Cr(VI) removal.

Full size image

Material characterization

The morphology and size distribution of the MACDs were examined using TEM. Low-magnification images were obtained under low-resolution conditions (Fig. 2a), while high-magnification images were captured using high-resolution TEM (Fig. 2b, c). The TEM images revealed the spherical shape of the MACDs. The particle size distribution showed that the majority of particles were within the 90–95 nm range (Fig. 2b, c), with an average particle size distribution ranging from 20 to 120 nm (Fig. 2d).

a Low and b, c high magnification. d Particle size distributions corresponding to b and c. The red circles indicate particles.

Full size image

The UV–vis absorption spectra (200–800 nm) of the MACDs in water (dark red line) and ethanol (blue line) are presented in Fig. 3a. In both ethanol and water, the π–π* electronic transition peaks at ~277 nm are associated with the C=C bond, clearly indicating the sp² graphitic framework²⁹. Notably, two additional peaks are observed in the ethanol solvent at 412 and 663 nm, which are attributed to C=N/O/Cl bonding and metal-ligand (M–L) bonding, respectively³⁰. These bonds are related to the doping of metals and heteroatoms.

a UV–vis absorbance spectra of MACDs in ethanol (blue line) and water (dark red line) with their excitation (green dashed line) and emission (black dashed line) plots. The cuvette pictures of the MACDs in ethanol under sunlight (left) and 365-nm UV light (right) irradiation are shown in the inset. b Back-excitation spectra with four excitation centers. c Reflectance percentage versus wavelength of the MACDs, as obtained by DRS. d Tauc plot of c, denoting the bandgap (Eg).

Full size image

The MACDs exhibit red fluorescence at a longer wavelength (365 nm) and blue-green fluorescence at a shorter wavelength (254 nm) under UV light. The peaks denoted by the green dashed line at 409 nm and the black dashed line at 670 nm represent the excitation and emission properties, respectively (Fig. 3a). Cuvette photographs under sunlight show the colorless appearance of the MACDs in water at long wavelengths (365 nm). However, in a solution with a 50:50 ratio of water and ethanol, blue-green and light-red fluorescence are observed under short (254 nm) and long (365 nm) wavelengths, respectively (Supplementary Fig. 4a–d). The excitation centers for the 670 nm emission wavelength were analyzed. The back-excitation spectra of the MACDs reveal four excitation centers at 3.03, 2.66, 2.45, and 2.30 eV, as shown in Fig. 3b. The excitation signal at 409 nm is associated with the maximum emission at 670 nm.

The bandgap of the MACDs was determined using UV–vis–NIR DRS through their interaction with light in the 240–2600 nm range. The reflectance percentage versus wavelength plot is shown in Fig. 3c. A corresponding Tauc plot was obtained between the Kubelka–Munk function F(R) and energy (eV) using the Kubelka–Munk function theory, whereby a bandgap of 1.72 eV was obtained³¹ (Fig. 3d). This lower bandgap highlights the excellent performance of the MACDs as both a photocatalyst and a luminescent material.

Subsequently, an XPS analysis was performed to determine the doping percentage and elemental composition, as shown in Fig. 4. In the full survey (Fig. 4a) and magnified (Fig. 4b) spectra, the signals confirm the multiatom doping with C, O, N, Mg, Ca, K, and Cl inside the MACDs. The prominent peaks of C1s (74.41%) and O1s (21.89%) correspond to the carbon framework and surface-functionalized oxygen, respectively, which constitute the primary elemental composition of the MACDs. The remaining smaller peaks indicate the doping of N, Ma, Ca, K, and Cl atoms. In particular, the peaks at 284.77, 399.80, 532.27, 292.90, 347.46, 197.91, and 1303.48 eV correspond to C1s, N1s, O1s, K2p, Ca2p, Cl2p, and Mg1s, respectively (Fig. 4a, b). The elemental contributions are presented in Supplementary Table 1. The deconvoluted C1s spectra in Fig. 4c demonstrate various binding positions at 288.28 eV (C = O/C = N), 286.27 eV (C = O/C–N), and 287.74 eV (C = C/C–C)^32,33. The O1s short-scan spectra were deconvoluted into two peaks at 532.45 and 531.04 eV, corresponding to C = O and C–O binding, respectively^34,35, as shown in Fig. 4d. Similarly, the deconvoluted N1s spectra in Fig. 4e show peaks at 401.78 and 399.78 eV, corresponding to graphite-N and pyridinic-N binding, respectively^32,36. The Mg1s scan revealed a single peak at 1303.58 eV, corresponding to the binding with C (Fig. 4f)³⁷. In the deconvoluted spectrum of K1s, two peaks are observed at 295.78 and 292.93 eV, which are associated with K2p_1/2 and K2p_3/2, respectively (Fig. 4g)^38,39. The short-scan spectra of Cl2p depict two peaks at 199.48 and 197.88 eV, confirming the presence of Cl2p_1/2 and Cl2p_3/2 binding, respectively (Fig. 4h)⁴⁰. The short-scan spectra of Ca2p exhibit two signals at 351.35 and 347.48 eV, which are ascribed to Ca2p_1/2 and Ca2p_3/2, respectively (Fig. 4i)⁴¹.

a Full-survey scan spectra with two intense peaks attributed to C1s and O1s and one small peak attributed to Mg1s. b Magnified spectra of a, with peaks indicating the presence of different elements. c Deconvoluted C1s spectra, with three binding peaks. d Deconvoluted O1s spectra, with two binding peaks. e Deconvoluted N1s spectra, with two binding signals. Short-scan f Mg1s, g K2p, h Ca2p, and i Cl2p spectra.

Full size image

Finally, the comparative FT-IR spectra of the MACDs were analyzed before and after photocatalytic reduction. Various peaks were observed at 3390, 2926, 2856, 1717, 1616, 1385, 1082, 1029, and 547 cm⁻¹, corresponding to the stretching vibrations of N–H/O–H, =C–H, –C–H, C=O, C=N, C–N, C–O, and M–L bonding, respectively. These specific vibrational frequencies represent the functional groups present on the surface of the MACDs. A significant change was observed after photocatalytic reduction, with the disappearance of the peak at 1717 cm⁻¹, corresponding to the carbonyl functional group (C=O), attributed to the electrostatic interaction between carboxylate ions (COO⁻) and C=O with Cr(VI)/Cr(III) (Supplementary Fig. 5a). The Raman spectra of the MACDs were recorded using Raman spectroscopy under a 514 nm laser with a power of 25 W. Remarkably, the G and D bands, corresponding to carbon compounds, were identified at approximately 1580 and 1350 cm⁻¹, respectively⁴². The G band represents the graphitization of sp²-hybridized carbon atoms, while the D band reflects surface defects associated with sp³-hybridized carbon atoms (Supplementary Fig. 5b).

MACDs performance study of Cr removal

Cr(VI) removal studies were conducted using six 50 mL solutions containing varying concentrations of Cr(VI). A standard solution of synthetically polluted Cr(VI) was prepared using K₂Cr₂O₇. The MACDs were used as the photocatalyst for the complete removal of Cr(VI). Equal volumes of Cr(VI) solutions with different initial concentrations were used to analyze the photocatalytic reduction efficiency of MACDs under light irradiation for 420 min. The use of 0.2 mg/mL MACDs resulted in the removal of 19.2, 37.0, 60.0, 91.0, 199.0, and 290.0 ppm of Cr(VI) from solutions with initial concentrations of 20, 50, 100, 200, 500, and 1000 ppm, respectively. The removal amount of 290 ppm is notably high, with removal efficiency increasing as the initial concentration increased from 20 to 1000 ppm. The Cr(VI) removal efficiency was determined from a plot of C/Co (ppm) versus time (min) (Fig. 5a). The intensity variations of the yellow color in the Cr(VI) solutions after the photocatalytic reduction process are shown in Supplementary Fig. 6.

a Photocatalytic reduction of Cr(VI) solutions with initial concentrations of 20–1000 ppm for 420 min. b Ratio of removed and remaining Cr(VI) concentrations after reduction. c Pseudo-first-order linear-fitted plots of Cr(VI) solutions of various concentrations. d Rate constants and half-life of the reaction. The error bars represent the SD (n = 3).

Full size image

As shown in Fig. 5a, adsorption-desorption equilibrium was achieved after 30 min under dark conditions, after which the solutions were exposed to direct light irradiation. Subsequently, 1 mL of photoreduced Cr(VI) solutions were collected at 1-h intervals and centrifuged. The supernatants from the samples were stored for UV–vis analysis. A DPC assay was used to determine the concentrations of the collected Cr(VI) samples over time. The absorbance of each sample was measured at 540 nm using a UV–vis spectrometer. Fig. 5a illustrates the gradual decrease in Cr(VI) concentration, expressed as C/Co, over time under direct light irradiation. Fig. 5b shows the histogram of the removed and final Cr(VI) concentrations following photocatalytic reduction. The data were fitted to a pseudo-first-order kinetic model (Eq. 1) to analyze the kinetics of the photocatalytic reduction process:

where C_o and C are the initial and residual concentrations (ppm) of Cr(VI), respectively; t (min) is the interval; and k (min⁻¹) is the reaction rate constant. The plot of ln (Co/C) versus time is illustrated in Fig. 5c, and the reaction constants (k) and half-life (t_1/2) of the reaction are shown in Fig. 5d. The values of the kinetic parameters, such as regression coefficient (R²), k, and t_1/2, for solutions of various Cr(VI) concentrations are shown in Table 1. A comparison of the photocatalytic reduction performance of MACDs, in terms of Cr(VI) concentration, catalyst amount, light source, and reaction time, with organic compound and metal-composite-based catalysts is provided in Supplementary Table 2. Notably, MACDs demonstrated high performance even without the use of hole scavengers (acid/alcohol).

Next, to assess the influence of different factors on the Cr-removal performance of the synthesized MACDs, an experiment was conducted under various reaction conditions by mixing 0.2 mg of the MACDs in a 20 ppm Cr(VI) solution (50 mL) and allowing the reaction to proceed for 420 min. The effects of these conditions on the photocatalytic reduction activity of the MACDs under both light-irradiated and dark conditions, as well as in the absence of MACDs under light irradiation, are shown in Fig. 6a. Under dark conditions, an initial adsorption-desorption equilibrium was achieved, with approximately 20% of Cr(VI) adsorbed. In contrast, ~91% of Cr(VI) was removed through photocatalytic reduction by the MACDs under light irradiation. When the MACDs were absent, no photocatalytic reduction of Cr(VI) occurred under light irradiation, and the Cr content remained unchanged. The photocatalytic reduction efficiency was investigated under dark conditions (30 min) and light irradiation conditions (390 min) with different MACDs loading quantities (10, 20, 30, and 40 mg) in 50 mL of 500 ppm Cr(VI) solution. The photocatalytic reduction efficiency increased as the MACDs content was raised, reaching a maximum at 0.8 mg/mL (Fig. 6b). These findings indicate a dose-dependent Cr(VI) removal by the MACDs (Supplementary Fig. 7). Notably, even the lowest catalyst loading demonstrated sufficient photocatalytic reduction efficiency.

a Removal efficiency of Cr(VI) under different reaction conditions. b Effect of the variation in the MACDs amounts on the removal efficiency of Cr(VI). The error bars represent the SD (n = 3). c Effect of various coexisting ions on the photocatalytic reduction efficiency. d Recyclability of MACDs and removal performance of Cr(VI) for five cycles. The error bars represent the SD (n = 3).

Full size image

The effect of various coexisting ions on the efficiency of the photocatalytic reduction process was indicated. Solutions of 1 mM coexisting cations, such as Fe⁺³, Ca⁺², and Cu⁺², and anions, such as Cl⁻, carbonate (CO₃²⁻), and hydrogen phosphate (HPO₄²⁻), were prepared with 20 ppm Cr(VI). The control Cr(VI) and ion-mixed solutions were exposed to direct light irradiation under conditions similar to those outlined in Section 2.3.3. All experiments were conducted with 0.2 mg/mL MACDs under light irradiation for 420 min (Fig. 6c).

Negligible changes in photocatalytic reduction efficiency were observed with Cu⁺², Ca⁺², and Cl⁻. In contrast, the photocatalytic reduction efficiency was slightly reduced with CO₃⁻² and HPO₄²⁻ ions. Fe increased the photocatalytic reduction efficiency (~100%) owing to the electrons provided by Fe during Cr(VI) reduction⁴³. Fe also exhibited a Fenton-like reaction toward Cr(VI) reduction, which was responsible for increasing the effectiveness^44,45. Thus, coexisting ions have a minimal effect on the Cr(VI) removal efficiency except Fe.

Recycling experiments were conducted using 0.5 mg/mL MACDs in a 50 ppm Cr(VI) solution under light irradiation. The MACDs were washed three times with 1 M HNO₃ and deionized water before being recycled⁴⁶. Subsequently, the MACDs were recovered by centrifugation at 12,500 rpm for 20 min and dried in an oven at 60 °C for 5 h. The photocatalytic reduction efficiency remained above ~92% after five cycles. The observed 8% decrease in efficiency is attributed to MACDs loss during the washing process. The combined removal efficiency of Cr(VI) and Cr(III) was approximately ~96% over three cycles, with a slight decline thereafter. Therefore, the MACDs demonstrated good recyclability without significant loss of performance after five cycles (Fig. 6d).

Furthermore, the photocatalytic performance of the MACDs was investigated in various reaction media, including ethanol (EtOH), an acidic solution (0.1 M HCl), a basic solution (0.1 M NaOH), and deionized (DI) water. The experiments were conducted using 0.2 mg/mL MACDs and 20 ppm Cr(VI) for 420 min under light exposure.

The photocatalytic reduction efficiency of the MACDs after 60 min was significantly higher for the removal of Cr(VI) in acidic and ethanol (EtOH) media compared to deionized (DI) water and basic media (Supplementary Fig. 8). The removal performance decreased in the basic medium due to the capture of H⁺ ions. In contrast, the increased presence of H⁺ ions in the acidic medium enhanced the reaction rate by promoting the hole-trapping process. However, no precipitation occurred along with the MACDs in the acidic medium. Instead, in situ precipitation of the final products (Cr(III)/MACDs) occurred within the medium. Consequently, in this study, all experiments were conducted in a pure aqueous medium, without the use of hole scavengers (acid/ethanol), to accurately assess the intrinsic photocatalytic efficiency of the catalyst. In contrast, many studies employ hole scavengers to enhance the photocatalytic efficiency of catalysts (Supplementary Table 2). However, the addition of acid or ethanol can be costly at an industrial scale, and these chemicals may pose risks to aquatic organisms and human safety. Therefore, our proposed method for remediating Cr(VI)-contaminated water is both cost-effective and ecologically safe. Next, we analyzed the proposed method using real samples.

Three different water samples—tap water, river water from the Geumgang River, and lake water from Daechung Lake—were collected near the campus of Chungnam National University in Daejeon, Republic of Korea. Initially, the collected water samples were filtered using a 2 µm nylon Whatman syringe filter and subsequently used to prepare the stock solution of K₂Cr₂O₇ and for subsequent dilutions. A typical Cr(VI) photoreduction experiment was performed with a 20 ppm Cr(VI) solution (50 mL) under similar experimental conditions. The removal efficiencies were then investigated for the different water samples. The removal efficiencies for tap, river, and lake water were approximately 94%, 79%, and 84%, respectively (Supplementary Fig. 9). The removal efficiency varied among these samples; for tap water, it was slightly higher, whereas, for river and lake water, it was slightly lower compared to that of deionized (DI) water (91%). The variation in removal efficiencies can be attributed to the presence of interfering substances such as metal ions, soil, pesticides, animal residues, plant residues, and microplastic particles.

Finally, a Cr(VI) removal experiment was conducted once under natural sunlight to examine the variation in the photocatalytic performance of the MACDs. Six distinct concentrations of Cr(VI) were used in the standard experiment to perform photocatalytic removal tests under natural sunlight: 20, 50, 100, 200, 500, and 1000 ppm (50 mL). These conditions were identical to those used in experiments with a 300 W xenon-lamp. The photocatalytic efficiency under natural sunlight was approximately 96% for the 20 ppm solution, which was ~5% higher than that under artificial light irradiation. The overall photocatalytic reduction rate slightly increased under natural sunlight. The plots for the sunlight conditions are shown in Supplementary Fig. 10: (a) C/Co (ppm) versus time (minutes), (b) removed and final remaining concentration of Cr(VI), (c) pseudo-first-order kinetic linear equation plot, and (d) rate constant (k) and half-life (t_1/2) trends for different initial Cr(VI). The k values of 0.0051 (R² = 0.95), 0.0031 (R² = 0.98), 0.0021 (R² = 0.98), 0.0015 (R² = 0.99), 0.0013 (R² = 0.99), and 0.0009 (R² = 0.95) were calculated for 20, 50, 100, 200, 500, and 1000 ppm respectively. The t_1/2 values of 136.96, 223.55, 322.33, 450, 525, and 704.25 were also determined for 20, 50, 100, 200, 500, and 1000 ppm, respectively.

Possible mechanism analysis for Cr(VI) removal

The proposed mechanism for Cr(VI) removal by the MACDs is depicted in Fig. 7. Upon light irradiation, the interaction with the MACDs surface generates holes (h⁺) in the valence band (VB) and electrons (e⁻) in the conduction band (CB) (Eq. 2)⁴⁷. Most of the holes generated in the VB react with H₂O molecules. However, some electrons in the CB recombine with holes in the VB, as not all holes can be occupied by H₂O molecules. In the case of a lower band gap, electrons can easily transition from the VB to the CB, resulting in higher photoreduction efficiency of the material. This is because materials with narrow band gaps can absorb visible light, making them more effective photocatalysts. Conversely, materials with wide band gaps are unable to utilize visible light, limiting their effectiveness as photocatalytic materials⁴⁸. In many Cr photocatalytic reduction studies, small amounts of acid are typically used to reduce the recombination of holes and excited electrons, thereby improving efficiency^{7,14,49,50,51,52,53,54}. Generally, photocatalysts exhibit higher efficiency in acidic media compared to neutral media, as the decrease in pH promotes more effective trapping of holes. This leads to the formation of highly reactive H⁺ ions in the presence of O₂ through the interaction of holes with H₂O molecules (Eq. 3). The CB electrons interact with the nearest Cr(VI) ions, converting them to Cr(III) (Eqs. 4 and 5). The highly reactive intermediate species, including electrons and H⁺ ions, collectively react with Cr(VI), further reducing it to Cr(III). Following the reduction, the nonhazardous Cr(III) precipitates at the bottom (Eq. 6). Cr(OH)₃ has a high adsorption tendency due to its lower adsorption energy compared to other chromium oxidation states, such as (Cr₂O₇²⁻, Cr₂HO₇²⁻, Cr₂O₆²⁻, CrO₃⁻, CrHO₃⁻, and CrH₂O₃⁻)⁷.

Possible mechanism of the photocatalytic reduction of Cr(VI) by MACDs.

Full size image

Additionally, the plausible photocatalytic reduction mechanism of Cr(VI) was investigated using a trapping experiment. During the photocatalytic reduction process, MACDs generated four types of reactive species: electrons (e⁻), holes (h⁺), superoxide radicals (O₂·^•–), and hydroxyl radicals (OH·^•). These species are responsible for the photoreduction of Cr(VI). To explore the distinct roles of each reactive species, a trapping experiment was conducted using four different scavengers: tertiary butyl alcohol (t-Bu) to trap hydroxyl radicals, potassium persulfate (K₂S₂O₈) for electrons, para-benzoquinone (p-BZQ) for superoxide radicals, and disodium ethylenediaminetetraacetate (Na₂-EDTA) for holes. According to the literature, these scavengers are well known for trapping electrons, superoxide radicals, holes, and hydroxyl radicals^55,56,57. A 0.5 mM concentration of each scavenger was added to 20 mL of 20 ppm Cr(VI) solutions. As shown in the histogram (Supplementary Fig. 11a), photoreduction activity decreased significantly by 84.18% and 61.24% in the presence of K₂S₂O₈ and p-BZQ, respectively. In contrast, reductions of 7.20% and 10.67% were observed in the presence of t-Bu and Na₂-EDTA, respectively. The significant decrease in photocatalytic activity in the presence of K₂S₂O₈ and p-BZQ highlights the dominant roles of electrons and superoxide radicals in the Cr(VI) reduction process compared to hydroxyl radicals and holes. Thus, electrons and superoxide radicals are the primary contributors to photoreduction, although hydroxyl radicals and holes also play a role, albeit to a lesser extent.

Furthermore, the EPR spectra of MACDs were recorded at room temperature in the liquid phase to investigate the presence of unpaired electrons and the paramagnetic behavior of the photocatalytic material. The presence of unpaired electrons generates reactive species (free radicals), both of which are responsible for the photoreduction process. The well-resolved EPR spectrum, with an average g-value of 2.02, as shown in Supplementary Fig. 11b, confirms the presence of carbon-based radicals⁵⁸. The proposed Cr(VI) photoreduction mechanism, along with the findings from the quenching experiments, is further supported by the EPR spectrum.

In this study, a green synthesis approach for fluorescent MACDs was adopted, which involved employing a single-step solvothermal method using the leaves of the Kalanchoe pinnata plant. The results emphasized the efficient photocatalytic reduction of Cr(VI) by MACDs, along with the precipitation of the final product. The photocatalytic properties of the MACDs were highly effective for the complete removal of Cr(VI) contamination via photocatalytic reduction and precipitation processes without generating by-products in an aqueous medium. Notably, an extremely small amount of MACDs (0.2 mg/mL) was sufficient for removing a large amount of Cr(VI) (290 ppm) from a 50 mL solution with an initial concentration of 1000 ppm within 420 min. These results demonstrate that MACDs are highly effective and excellent candidates for the treatment of Cr-contaminated water. Thus, this study demonstrated the development of a biomass-derived, effective, and fluorescent multiatom-doped photocatalytic material with the advantages of eco-friendliness, cost-effectiveness, nontoxicity, reusability, and accessibility. Existing photocatalysts are typically synthesized using commercially available hazardous chemicals. In contrast, this study highlights the relevance of a biomass-derived photocatalyst as an alternative to conventional metal and organic compound-based photocatalysts. Further investigations are required to optimize the synthesis conditions for enhancing the performance of Cr removal and mass-producing MACDs.