Introduction

Natural zeolites have attracted worldwide attention because of their presence in a variety of different structure and chemical composition that depend on the geological conditions of each country. They are eco-friendly, cost-effective, non-toxic, and easily accessible materials.

Mineral prices might vary depending on purity, processing, country of origin, and market development. According to the U.S. Geological Survey1 the annual production of natural zeolites is about 86.000 tons per year. The average production cost is estimated in 2024 at $145, and the price varies depending on the physical and chemical characteristics of zeolites. Since synthetic zeolite demand is substantially higher in the global zeolites market, the price of natural zeolites is tens of millions of dollars, much lower than the billions of dollars of total synthetic zeolite sales2. Due to their low price, ion exchange capacity, high porosity, and thermal stability, they have applications in a variety of domains, including research, building construction, energy storage, medicine, and other industries3,4,5.

Volcanic rocks are the most common source of zeolites, with a widespread abundance. Natural materials, such as zeolitic volcanic tuff, are environmentally friendly and inexpensive, with a three-dimensional structure composed of SiO4 tetrahedra, some of which can be replaced with AlO4. The substitution of silicon by aluminium results in an overall negative charge, which can be balanced by cations, giving zeolites cationic exchange capabilities6,7. Zeolites are a group of minerals usually formed by the dissolution of the parent volcanic glass and the alteration of other minerals during interaction with alkaline solutions8,9. The geochemical fingerprint makes zeolitized volcanic tuff distinct at each location. There are over 50 different natural zeolites in the world5and each mineral differs by location due to the chemistry of different types of alkaline hosts such as phonolite, basanite, tephrite, and leucitite10. The Si/Al ratio is an important indicator of the economic viability of zeolitic-rich tuff. Some studies demonstrated the excellent performance of zeolite-bearing rocks from the Campania region (Southern Italy) with a low Si/Al ratio in the glass, in the formation of lightweight aggregates11.

Volcanic tuff deposits are also widespread in Romania and are characterized by the presence of clinoptilolite, a zeolite mineral with an (Si, Al)O4 tetrahedral structure linked by oxygen atoms12,13,14,15. The volcanic tuff from Bârsana-Iza River (Maramureș County) is the most representative for this study due to its reported high cation exchange capacity of 69 meq/100 g and clinoptilolite content of more than 90%16,17. The mineralogical description of Neogene zeolitic tuff from this area is vitric texture, represented over 90% by the volcanic glass, followed in the proportion of 10% by crystaloclasts (quartz, plagioclase, feldspar, biotite, muscovite) and lithic fragments16,18. The chemical composition of Bârsana zeolitized tuff is as follows: SiO2 – 66.76 wt%; TiO2 – 0.25 wt%; Al2O3 – 10.82 wt%; Fe2O3 T – 1.67 wt%; MgO – 0.74 wt%; CaO – 2.82 wt%; and Na2O – 1.01 wt.%16.

Several investigations have reported high capacity in adsorption and storage of different metal(loid)s and ammonium in the structure of Romanian clinoptilolite-rich tuff13,19,20,21,22,23. Stanić et al.22investigated As (V) adsorption efficiency using modified clinoptilolite-rich tuff from the Maramureș District. Maranon et al.21also investigated the potential of clinoptilolite volcanic tuff in NH4+ removal from aqueous solution, in the presence and absence of Zn and Cd. Other studies confirmed the efficacy of clinoptilolite-rich volcanic tuff from Cluj District in the adsorption of zinc ions and ammonium from aqueous solution13,19 and sorption of ammonia from air24. Romanian zeolites have also been used successfully in soil decontamination processes20,23.

Several techniques have been recommended for wastewater treatment, including biological treatments, adsorption, membrane filtration, chlorination, or advanced oxidation processes (AOPs). As a physicochemical process, the adsorption is easy and cheap to operate, but for some persistent organic compounds requires further treatment25,26,27. AOPs are a promising chemical treatment technique that can be utilized as an alternate approach for eliminating contaminants from wastewater. In the last decade, photocatalysis has an extensive interest for water treatment due to oxygen reactive species (hydroxyl radical (·OH) and hydroperoxyl (HO2·)) known as strong and non-selective chemical oxidizer that effectively degrades many organic pollutants with a high reaction rate28,29,30. Some catalysts based on zeolites with ferrous iron in combination with hydrogen peroxide can also be efficiently used in the oxidation process31.

To our best knowledge, the performance of Romanian volcanic tuff used in dyes elimination from wastewater has not yet been investigated. Thus, the study’s novelty begins with the use of functionalized Romanian volcanic tuff from the Bârsana-Iza River deposits to remove two persistent dyes from wastewater, Methylene Blue (MB) and Reactive Black 5 (RB5). The research proposes to investigate the mechanism involved in dyes elimination by using acid (VT1) and iron-coated (Fe_VT) materials via photo-Fenton-like oxidation, which is technically and economically advantageous in comparison to other AOPs, and very efficient in pollutant removal25,26. The role of iron-doped particles is to modify the potential surface of zeolites from negative to positive to remove anionic contaminants and oxidize organic molecules in the Fenton-like reaction.

The Raman spectra obtained on natural volcanic tuff from the Bârsana region are the second significant aspect that contributes to the novelty of this study. Currently, the volcanic tuff has not been examined using surface-enhanced Raman spectroscopy (SERS). Since the rock consists of more than 90% vitric structure containing volcanic glass, lithic fragments, and other minerals, the use of conventional Raman spectroscopy is restricted by weak intensity and the presence of fluorescence effect. Thus, this study’s novel aspect is applying surface-enhanced Raman spectroscopy (SERS) to increase the signal through a silver colloidal and a gold thin film as SERS active substrates. The spectra obtained can be used as a reference for volcanic tuff, particularly from the Bârsana area.

The research addresses a current concern related to pollutants present in different environments that have a negative impact on flora, fauna, and human lives. To reduce costs and pollution, as well as to avoid unsafe synthetic materials that can release additional pollutants, natural resources of different types of minerals could be a promising catalyst for wastewater treatment.

Results and discussion

Characterization of volcanic tuff-based catalysts

Figure 1 presents the SEM micrographs of natural volcanic tuff before and after treatments, as well as after five reuse cycles with RB5 in the presence and absence of H2O2. The detected particles are often irregular in shape, with tabular crystals representing the volcanic tuff’s clinoptilolite phase16. The surfaces of the raw sample and acid-treated tuff (Figs. 1a, b) are distributed heterogeneously, having a vitroclastic texture13,16. The addition of iron (Fig. 1c) significantly improved the surface’s porosity appearance, as confirmed also by BET investigations. After five reuse cycles of material treated with iron in the presence and absence of hydrogen peroxide (Fig. 1d, e), the particle morphologies remained unchanged, indicating material stability, which was further supported by other investigations.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

SEM images of raw volcanic tuff (VT) (a), treated with acid (VT1) (b) or treated with iron (Fe_VT) (c), as well as Fe_VT after five cycles of reuse with RB5 in the absence of hydrogen peroxide (d) or with 10 mM H2O2 (e).

The presence of crystalline phases has been identified through X-ray diffraction (XRD). Figure 2a depicts the crystallographic effects caused by Fe and acid treatments in volcanic tuff, as well as the impact of photo-Fenton-like degradation in crystalline phases during five reuse cycles [Fe_VT (R) and Fe_VT + H2O2 (R)] in the presence of RB5 (10 mg/L). The most significant peaks that indicate the presence of clinoptilolite are observed in the spectra of volcanic tuff sample (Fig. 2a1) at 2θ of 22.50° corresponding to (330) reflection plan and d-spacing of 3.94 Å, and 2θ of 30.12° at d-spacing of 2.96 Å associated with the (350) reflection plan, respectively32,33,34. The XPS data presented below also indicate the presence of Na, which could be related to the Na-rich clinoptilolite mineral, with the following formula (Na, K,Ca)2−3Al3(Al, Si)2Si13O36·12H2O35. The diffraction spectrum indicates that the raw volcanic tuff consists of 88% clinoptilolite phase (Fig. 2a1), one of the main minerals in the zeolite group. Additionally, the pattern also reveals the presence of quartz (SiO2) in a minor amount of 12%. Since some mineral phases may have overlapped reflection peaks36the mineralogical composition was further confirmed using Raman spectroscopy as an additional method. For the samples treated with acid (Fig. 2a2) and iron (Fig. 2a3), the mineral phases identified are the same, clinoptilolite and quartz. The difference in phase percentage in the treated samples is low, with a variation of about 1% compared to raw volcanic tuff.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

(a) XRD patterns of (1) raw volcanic tuff (VT), (2) acid-treated volcanic tuff (VT1), (3) iron-treated volcanic tuff (Fe_VT), (4) iron-treated volcanic tuff after five reuse cycles with RB5 (Fe_VT R) and (5) iron-treated volcanic tuff after five reuse cycles in the presence of 10 mM H2O2 (Fe_VT + H2O2 R). The C and Q annotations correspond to the clinoptilolite and quartz mineralogical phases. (b) Raman and (c) FT-IR spectra of volcanic tuff and modified materials.

The diffraction spectra of modified samples present some structural changes in crystallinity, observed by differences in the reticular distance (d-spacing) that could affect the lattice parameters of the crystals17,37. The difference between the modified materials and untreated ones suggests that heated samples, as well as the presence of acid and iron, may produce some structural modification. These defects are observed in the clinoptilolite main peak that shifts to a higher 2θ° in VT1 and lower 2θ° in Fe_VT samples.

The spectra obtained of volcanic tuff after five reuse cycles with RB5 (10 mg/L) in the presence and absence of 10 mM H2O2 are shown in Figs. 2a4–5. After photo-Fenton-like catalytic degradation, none of the phases disappeared from the reutilized materials. On the other hand, the fraction of clinoptilolite phase decreased significantly, at 82% in the Fe_VT (R) and 80% in the Fe_VT + H2O2 (R) samples, while the quartz increased from 12% (in volcanic tuff) to 18% and 20%, respectively. The structural difference between raw and reused samples is due to several heating steps in the material recovery process.

The clinoptilolite and quartz phases obtained by XRD were validated using surface-enhanced Raman spectroscopy (SERS). Additionally, titanite CaTi(SiO4)O was also identified in the raw volcanic tuff. It is well-known that it is difficult to obtain suitable Raman spectra signal-noise on zeolites due to the highly dispersed material and the fluorescence effect38. The sensitivity of the Raman signal was improved through the addition of a silver colloidal suspension onto a gold film as a metal substrate. Therefore, the Raman spectra of raw volcanic tuff and modified samples (VT1 and Fe_VT), have been obtained by SERS and presented in Fig. 2b. Clinoptilolite is a zeolite belonging to heulandite group and is characterized by T–O–T (T = Si, Al) atomic vibrations of the SiO4 tetrahedra from tectosilicates structure32. In the obtained Raman spectra, this mineral has been identified in each sample in the region of 408–503 cm− 1 corresponding to ν1 bending vibrations of Si–O–Si or Al–O–Al bonds39. The peaks observed in the region of 600–900 cm− 1 correspond to ν4 symmetric stretching modes of Si–O or Al–O40. The primary peak of clinoptilolite, identified at 408 cm− 1 in raw volcanic tuff, slightly shifted toward a higher wavenumber (415 cm− 1) in acid- and iron-doped samples, respectively. A possible cause could be explained by some disorder that occurs in the mineral structure because of the treatment operations41.

Figure 2c shows the vibrations of functional groups in modified volcanic tuff and raw samples determined by FT-IR spectroscopy. There are no significant differences between the spectra. Small shifts to a lower wavenumber may be due to an increased number of Al atoms in the Al/Si ratio42. The specific bands belonging to the clinoptilolite phase have been observed in the raw volcanic tuff at frequencies of 471, 608, 797, and 1637 cm− 1. The peaks identified below 500 cm− 1 could be assigned to the Si–O or Al–O bending vibrations in tetrahedral groups42. Other studies suggest that the peak detected at 797 cm− 1 is associated with the stretching vibration of T–O–T bonds (T = Si, Al)43. The peaks found in all spectra at frequencies ranging from 1637 to 1639 cm− 1 may be attributed to water’s hydrogen bending mode42,43.

X-ray photoelectron spectroscopy (XPS) was applied to investigate the main chemical elements found in raw and modified volcanic tuff samples (Fig. 3).

The O 1s XPS spectrum was deconvoluted into two peaks corresponding to lattice oxide (Fe-O) and loosely bound oxygen, such as absorbed O2− or adsorbed H2O at B.E. = 530.10 ± 0.2 eV and at B.E. = 532.38 ± 0.2 eV, respectively44. The Si 2p high-resolution spectrum was deconvoluted into one peak, which can be attributed to Si in Al4Si4O10(OH)8 chemical bond45 or Si in SiO2, quartz with B.E. = 102.97 ± 0.2 eV46. The K 2p XPS spectrum show a peak corresponding to K in zeolite chemical bond Al0.041Si0.264Na0.04K0.02O0.635 with a B.E. = 293.9 ± 0.2 eV47. The Al 2p high-resolution spectrum was deconvoluted in one peak having a B.E. = 74.51 ± 0.2 eV corresponding to Al in Al4Si4O10(OH)845. In the case of Fe, the Fe 2p peak was split into two components, Fe 2p1/2 at B.E. = 725.78 ± 0.2 eV and Fe 2p3/2 at B.E. = 712.18 ± 0.2 eV, due to spin-orbit coupling, with their satellite peak. The binding energy values correspond to Fe in Fe(OH)O chemical bond48. The Na 1s XPS spectrum show one peak at B.E. = 1072.95 ± 0.2 eV which can be attributed to Na in Na(AlSi2O6)H2049.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Extensive XPS spectra of raw volcanic tuff – VT (a), VT1 (b), and Fe_VT (c), as well as XPS spectra of elements identified in modified and raw samples.

N2-sorption measurements investigated the porous characteristics of modified and raw volcanic tuff materials. The obtained isotherms are presented in Fig. 4, characterizing each analyzed sample. At first look, it can be concluded that, according to IUPAC classification50all three samples exhibit type IV isotherms (Fig. 4), being characteristic of meso/macroporous materials. All isotherms are accompanied by an H3 hysteresis loop with a knee appearance for the Fe_VT sample, meaning the pores in the material come from non-rigid aggregates of plate-like particles, but also confirm the incompletely filled macropores with condensate molecules.

Brunner–Emmett–Teller theory has been considered for calculations of the specific surface area, and the obtained values are shown in Table 1. As can be observed, the acid-treated sample (VT1–7.339 m2/g) has a decreasing effect on the surface area of raw volcanic tuff material (VT − 12.290 m2/g), while the iron atoms (Fe_VT − 43.511 m2/g) have a positive effect, improving the surface area about four times. The same effect on total pore volume evolution in those materials could be observed, with a small difference in the VT1 increasing effect. The surface area and pore volume of volcanic tuff used in our study are six to eight times lower than those reported in the literature for zeolites31,51,52 which makes them bad adsorbents. The improvement of both textural characteristics of the Fe_VT sample may open new opportunities in its application.

Table 1 Textural characteristics of the synthesized samples.
Full size table
Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Nitrogen adsorption-desorption isotherms registered on the synthesized samples.

Given the outcome of the material characterization and surface data, we may state that the material was successfully functionalized and can be used in further experiments as a catalyst.

Photocatalytic activity

The photocatalytic activity of modified volcanic tuff for the oxidation of RB5 and MB dyes was studied in detail. The effect of catalyst concentration, hydrogen peroxide concentration, and real wastewater effluent was investigated. The experiments have been carried out at pH 5.5 and used without any modifications or control during the process. In our previous studies using commercially available Fe-exchanged Y zeolite as a catalyst, the experiments at an initial pH value close to 5.0 have shown a good catalytic activity31,52. The photodegradation of dyes was performed after adsorption/desorption equilibrium had been reached between dyes and the volcanic tuff.

The influence of the amount of two modified natural volcanic tuffs on dyes degradation is presented in Figs. 5a and b and 6a and b. The preliminary experiments on raw volcanic tuff have shown a low degradation of dyes. Therefore, in Figs. 5a and b and 6a and b the results obtained for the use of acid-treated and iron-impregnated volcanic tuff for oxidation of RB5 and MB dyes are presented. As it is expected, the increase in materials dosage from 0.5 to 2 g/L accelerated the dyes removal. The number of active sites on the surface and respectively of more ·OH radicals increased as the amount of volcanic tuff increased, which caused the removal of the dye solution. As described in the studies of Akpan and Mahamud, at a higher amount of catalyst, the solution becomes turbid and so inhibits UV–vis rays, resulting in a decrease of degradation53,54. The color removal of a higher than 80% were achieved over both volcanic tuffs, with k = 0.0135–0.0227 min− 1, after two hours of irradiation, while the degradation was faster over Fe_VT.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Performance of acid-treated volcanic tuff (VT1) in RB5 and MB removal under UVA at different optimization settings, such as dosage (a,c) and hydrogen peroxide concentration (b, d). Experimental conditions: CRB5 = 10 mg/L; CMB = 10 mg/L; pH of 5.5; dynamic regime.

Similar results have been reported by other authors31,52,55,56,57,58,59,60. Cardona et al.55found that the catalyst with the highest content of both iron (20 wt%) and titanium gives the highest removal values. Tetracycline showed good degradation using all catalytic supports, whereas enhanced removals of 2,6-DCP and BPA have been obtained using Fe2O3/TiO2/Al-PILCAE and Fe2O3/TiO2/Al-PILCBE. Under very mild conditions and solar light irradiation over a 10 wt% Fe-doped TiO2/HY zeolite and CuO-doped synthetic zeolite NaX, a total removal of MB (> 98% and 94%) was achieved56,57. In another recent study, Mehrbakhsh et al.58reported that the photocatalyst obtained by a combination of ZnFe2O4 and CoFe2O4 nanoparticles with zeolite (CZZ) exhibited a significant Cr (VI) reduction, which was 9.27, 5.37 and 3.58 times higher than those of single ZnFe2O4 nanoparticles, CoFe2O4 nanoparticles, and CoFe2O4-ZnFe2O4 respectively58. A high removal of MB (56-82%) and H2 production rate was also observed due to the competitive photocatalytic performance of modified Losod-zeolite59. In a study of Oviedo et al.60using an analcime nanozeolite as the catalytic support (CuO-NPs@nANA), 76.93% MB removal was obtained at pH 10, [MB] = 200 mg/L, and [CuO-NPs@nANA] = 0.5 g/L after 180 min under visible light, with k = 0.0088 min− 160. The experiments performed by Jiang et al.61showed that O2 is essential to the photocatalytic process, and the efficacy of the process is further improved by the addition of h + and ·OH. The photocatalysts contributed to the degradation of 84.6% of TC, and the Bi5O7I/ZIF-8 also showed high degradation stability after 4 cycles.

In our study, it can be decided that the optimum concentration could be between 1 g/L and 2 g/L. To prevent Fe leaching, the concentration of 1 g/L was selected for further experiments.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Photocatalytic activity of iron impregnated volcanic tuff (Fe_VT) in RB5 (a, b) and MB (c, d) retention. The effect of catalyst dosage (a, c) and hydrogen peroxide concentration (b, d). Experimental conditions: CRB5 = 10 mg/L; CMB = 10 mg/L; pH of 5.5; dynamic regime.

In the next investigation, the kinetic studies with different hydrogen peroxide concentrations for the catalytic oxidation of RB5 and MB dyes were performed at the same operational conditions: pH = 5.5, t = 25 °C, and 1 g/L of catalyst, and are presented in Figs. 5c and d and 6c and d. The rise in hydrogen peroxide concentration enhanced dye decolorization at the beginning of the reaction. After 60 min of reaction time over Fe_VT, more than 70% (k = 0.0199–0.0239 min− 1) of both dyes were removed in the presence of all three hydrogen peroxide concentrations. As expected, under the same operation conditions, when the hydrogen peroxide concentration rose from 5 to 20 mM, the color removal efficiency decreased or remained at the same level. This could be explained by the fact that in the presence of a local excess of H2O2, the ·OH radicals contribute to hydroperoxyl radicals (HO2·) formation. The HO2· are much less reactive and do not contribute to the degradation of the pollutant, which takes place only through the reaction with ·OH29,62,63. Therefore, because of hydrogen peroxide costs a concentration higher than 10 mM can be considered as an unprofitable consumption of hydrogen peroxide.

The degradation mechanisms of various zeolites have been extensively studied64,65,66,67,68,69,70,71including in our earlier publications31,52. In solids containing redox-active sites, such as Fe³⁺ in our study, these sites undergo reduction to Fe²⁺ by protons in the reaction medium64. This reduction triggers a Fenton-type reaction, where Fe²⁺ interacts with H2O2 to generate ·OH radicals, which then facilitate electrophilic attacks on the aromatic ring of the organic compound65. The ·OH radicals produced on the inner surface of the microporous material can disperse to the external surface to split the large molecule into smaller fragments, which can then circulate inside the microporous material. The small number of sites on the external surface of the microporous material may be sufficient to break the large molecule into smaller fragments66. Letaief et al.67demonstrated that phenol oxidation in the presence of solid catalysts, such as Fe–Al–PILCs, follows an electrophilic aromatic substitution mechanism. This reaction preferentially occurs at the ortho or para positions due to resonance activation of the phenol’s OH group and the formation of hydroxonium cations via H2O2 interaction with the acid sites. Furthermore, the process is influenced by strong Brønsted acid sites, which can catalyze H2O2 decomposition68,69. This mechanism suggests that Brønsted sites enhance the Fe³⁺ to Fe²⁺ transformation67. Germain et al.70studied the role of acidity in phenol hydroxylation using hydrogen peroxide over FAU catalysts with varying Si/Al ratios. Their findings indicate that electrophilic aromatic substitution is promoted by the acidic functions of the solid. Baldez et al.71reported that by irradiation, the ZnO from the surface of the FAU is activated and e− are transferred from the valence band (VB) to the conduction band (CB), resulting in the formation of h+ in the VB. Thus, the h+ radical, which is highly reactive, can act in the direct degradation of ciprofloxacin or participate in the oxidation process of H2O molecules, resulting in the formation of ·OH radicals, which are highly oxidizing species. The schematic degradation mechanism of dyes removal is presented in Fig. 7.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Schematic representation of photo-Fenton-like degradation mechanism to remove RB5 and MB contaminants.

As it is known, the principal inorganic reactions which are considered common to the photo-Fenton reaction system represent the interactions among various inorganic species including ·OH, HOO·, O2·−, H2O2, Fe2+ and Fe3+ as shown below.

Under UV irradiation, Fe3+ is continuously reduced to Fe2+: Fe3+ + H2O + hν → Fe2+ + ·OH + H+.

$$\begin{aligned} & {\text{Fe}}^{{{\text{2}} + }} + {\text{ H}}_{{\text{2}}} {\text{O}}_{{\text{2}}} \to {\text{Fe}}^{{{\text{3}} + }} + ^{ \cdot } {\text{OH }} + {\text{ HO}}^{ - } \\ & {\text{Fe}}^{{{\text{3}} + }} + {\text{ H}}_{{\text{2}}} {\text{O}}_{{\text{2}}} \to {\text{Fe}}^{{{\text{2}} + }} + {\text{ HOO}}^{\cdot} + {\text{ H}}^{ + } \\ & {\text{Fe}}^{{{\text{3}} + }} + {\text{ HOO}}^{\cdot} \to {\text{Fe}}^{{{\text{2}} + }} + {\text{O}}_{{\text{2}}} + {\text{ H}}^{ + } \\ & {\text{R}}{-}{\text{H }}\left( {{\text{pollutant}}} \right){\text{ }} + ^{ \cdot } {\text{OH}} \to {\text{R}}^{ \cdot } + {\text{ H}}_{{\text{2}}} {\text{O}} \\ & {\text{R}}^{ \cdot } + {\text{ Fe}}^{{{\text{3}} + }} \to {\text{R}}^{ + } + {\text{ Fe}}^{{{\text{2}} + }} \\ & {\text{R}}^{ + } + {\text{ H}}_{{\text{2}}} {\text{O}} \to {\text{ROH }} + {\text{ H}}^{ + } \\ & {\text{Fe}}^{{{\text{2}} + }} + ^{ \cdot } {\text{OH}} \to {\text{Fe}}^{{{\text{3}} + }} + {\text{ HO}}^{ - } \\ & {\text{H}}_{{\text{2}}} {\text{O}}_{{\text{2}}} + ^{ \cdot } {\text{OH}} \to {\text{H}}_{{\text{2}}} {\text{O }} + {\text{ HOO}}^{ \cdot } \\ & {\text{HOO}}^{ \cdot } \to {\text{O}}_{{\text{2}}} ^{{ \cdot - }} + {\text{ H}}^{ + } \\ & {\text{HOO}}^{ \cdot } + {\text{ Fe}}^{{{\text{2}} + }} \left( { + {\text{ H}}^{ + } } \right) \to {\text{Fe}}^{{{\text{3}} + }} + {\text{ H}}_{{\text{2}}} {\text{O}}_{{\text{2}}} \\ \end{aligned}$$

From the practical point of view, it is most important to evaluate the performance of selected volcanic tuff on real wastewater effluents. The results obtained are presented in Fig. 8. The investigated dyes have shown slightly higher removals (more than 90%, k = 0.0242–0.0309 min− 1) over both volcanic tuff in the real wastewater effluent than in MilliQ water (82–90%). This probably can be explained by the occurrence of common water constituents such as nitrate, iron ions, or dissolved natural organic matter (DNOM). It is known that nitrate can form ·OH radicals and produce, in our case, an even larger increase in the degradation rate of dyes72,73. It is generally known that DNOM influences the degradation of pollutants dissolved in water through various reactions. The light absorption by DNOM of natural waters leads to the formation of reactive species such as singlet oxygen1O2), superoxide anion (·O2−) hydroxyl radicals (·OH), excited triplet states, and peroxyl radicals (ROO·), which react with dyes72,73. In a study of Guesh74 the degradation process is slower in the real wastewater sample than in model MO solution. After 9 h of reaction time with the hybrid 10% and 40% TiO2-Zeolite Y photocatalysts 83% and 87% photocatalytic degradation, respectively were achieved.

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

RB5 and MB removal in wastewater effluent (WW). Experimental conditions: UVA irradiation, CRB5 = 10 mg/L; CMB = 10 mg/L; CFe_VT = 1 g/L; pH of 5.5; reaction time = 2 h; dynamic regime.

To evaluate the applicability of these materials, a further investigation of the reuse of catalysts was performed (Fig. 9). The results indicate that the activity of volcanic tuff remains high over several reuse cycles. The removal efficiency for RB5 decreases from 85 to 63% after five cycles, whereas for MB from 93 to 50% after three reusing cycles. After five cycles the decolorization of RB5 is still higher than 60%. High removals are important, but not enough for the catalysts to be considered efficient. To prevent additional pollution caused by the leaching of iron ions, the stability of the catalyst is an important indicator. The presence of iron ions in the solution could initiate additional generation of hydroxyl radicals by the homogeneous Fenton reaction. It is noticeable that after 2 h of reaction at 1 g/L catalyst in the presence of 10 and 20 mM H2O2, the catalyst leaching in our experiments remains low (between 0.133 and 0.195 ppm, respectively). The amount of pollutants degraded by homogeneous catalysis resulting from leached iron, under these conditions, is negligible. Moreover, the crystallographic and morphological investigations have shown that the iron-doped material remains stable after five cycles of dye exposure in the presence and absence of H2O2, as well as its morphology remains unchanged. Furthermore, under different leaching conditions varying from acid to near neutral, the TCLP (toxicity characteristic leaching procedure) tests show that modified volcanic tuff is stable and may be safely disposed after usage. Similar results were reported by Centi et al.75Ai et al.76and Barrault et al.77 for the catalytic wet peroxide oxidation of carboxylic acids, 4-chlorophenol, and phenol, respectively, and in our previous studies on synthetic zeolites31,52 and with similar composite materials54,71,74.

Fig. 9
Fig. 9The alternative text for this image may have been generated using AI.
Full size image

Reuse of iron-doped composites in the absence and presence of hydrogen peroxide. Experimental conditions: UVA irradiation, pH of 5.5; CRB5 = 10 mg/L; CMB = 10 mg/L; CH2O2 = 10 mM; reaction time = 2 h; dynamic regime.

The recyclability study of the zeolite-supported CdS/TiO2/CeO2 composite photocatalyst revealed 86.0% degradation of MB after 4 consecutive cycles, denoting the stability of the composite54. TiO2-ZY zeolite has been shown to yield 84% degradation of real wastewaters from the textile industry, and only a 5% reduction was observed after three cycles74. The results confirm that using natural volcanic tuff, it is possible to operate at near-neutral pH for several runs, obtaining high removal rates, without generating additional pollution. The studied volcanic tuff can be a practical solution for the removal of persistent organic pollutants in the concept of the Circular Economy.

Methods

Materials

Pure-grade chemical reagents were utilized as follows: HNO3 (CAS:7697-37-2); Fe(NO3)3 (CAS: 10421-48-4); Glacial acetic acid – CH₃COOH (CAS: 64-19-7); NaOH (CAS:1310-73-2); Reactive Black 5 (RB5) (CAS: 17095-24-8); Methylene Blue (MB) (CAS: 122965-43-9). The choice of these dyes is based on consumption consideration, their trace-level toxicity (ng/L) for aquatic flora and fauna, persistence in the environment, and stability after municipal WWTP processing. The selection criteria are also based on the high production and consumption volume at a regional scale, and reported aquatic toxicity.

Catalyst functionalization

The natural volcanic tuff was collected in June 2024 from a local deposit in the Bârsana-Iza River area of Maramureș county (NW Romania), at GPS coordinates of 47°48’17.8"N and 24°04’01.5"E. The samples had previously been crushed, sieved to less than 1 mm, and washed with MilliQ water. The first material, VT1 (4 g), was treated with 3 M HNO3, rinsed with DW until it reached neutral pH, and then heated to 105 °C for 6 h. The second material, labelled Fe_VT, was obtained by stirring 2 g of VT1 with a 1 M solution of Fe(NO3)3 for 6 h at 80 °C. The resulting mixture was filtered, washed with ultrapure water to remove any residuals, and dried in an oven at 105 °C for 6 h. The ion exchange method was repeated three times, as described in previous investigations31,52.

Materials characterisation

The morphological properties of raw and treated volcanic tuff, as well as after five cycles of material reuse with/without hydrogen peroxide, were examined using a scanning electron microscope (SEM) VEGA II LSH/TESCAN.

The crystalline phases of volcanic tuff (raw sample) and modified materials (VT1 and Fe_VT), as well as the effect of pollutants (RB5 and MB) on the zeolite structure during reuse cycles, were investigated using X-ray diffraction (XRD). Crystallographic patterns were acquired using a SHIMADZU X-ray 6000 diffractometer equipped with a CuKα radiation source (λ = 1.5418 Å) at 40 kV and 30 mA, with a scan speed of 2°/min, in the 2θ range of 10° – 90°.

The phases detected in the raw and modified samples by XRD were validated by Raman spectroscopy using the surface-enhanced Raman scattering (SERS) technique. Raman spectra were obtained using a Horiba Jobin-Yvon RPA-HE 532 spectrometer with an Nd-YAG incident laser (λ = 532 nm) at 100 mW and a CCD detector. The silver colloid was prepared based on the method proposed by Baia et al.78 and reproduced by Astefanei et al.79 by mixing and heating to boiling 36 mg of AgNO3 with 200 mL of deionized water for 60 min. After this interval, 4 mL of 1% aqueous trisodium citrate was added dropwise and heated for another 60 min. The Raman signal was significantly amplified when the colloidal silver and a gold thin film were used as the SERS substrate80.

The chemical structure of functionalized materials was investigated by Fourier transform infrared spectroscopy (FT-IR). The FT-IR spectra of the solid samples were acquired in the transmittance mode of 500–4000 cm− 1 using a Nicolet Impact 4000 spectrophotometer equipped with a monolithic diamond crystal.

X-ray photoelectron spectroscopy (XPS) has been used to investigate the elemental composition and chemical state of elements present on surface samples. The measurements were obtained using a PHI 5000 VersaProbe photoelectron spectrometer with monochromated Al Kα radiation (1486.7 eV) and C 1s core level signal (B.E. = 284.6 eV). Peak deconvolution of high-resolution XPS spectra was performed using the PHI MultiPak software to determine oxidation states and types of atomic binding. The binding energy measurements were accurate within ± 0.2 eV. The same device has been reported in a previous study27.

To measure the specific surface area of the synthesized solid materials, the N2 adsorption-desorption isotherms were recorded on a Quantachrome NOVA 2200e BET apparatus. Before isotherm acquisition, all samples have been outgassed at room temperature overnight, allowing pores to empty. Afterward, the sample has been weighted and subjected to adsorption of nitrogen molecules at − 196 °C, and the pressure inside the cell analysis is monitored until saturation. The data acquired gave the adsorption-desorption isotherms to which the BET theory has been applied to calculate the BET specific surface area and the total pore volume for all analysed samples.

The concentration of iron was measured in aqueous solutions by Flame Atomic Absorption Spectroscopy (FAAS), using a Continuum Source Atomic Absorption Spectrometer - contrAA 800 (AnalitykJena, Germany) - equipped with an optimized high-resolution Echelle double monochromator and a CCD line detector with high quantum efficiency and increased UV sensitivity. To provide an iron-specific wavelength (248 nm), a light beam from a xenon short-arc lamp with UV arc passed through the hot spot, having an automatic hot-spot tracking and simultaneous drift correction. The mixture is ignited in a flame whose temperature ranges from 2100 to 2800 °C. The spectra were collected, and as a function of the absorbance, the element concentrations were calculated.

Photo-Fenton-like catalytic degradation of RB5 and MB pollutants

Fresh ultrapure water was used to prepare a stock aqueous solution of two pollutants, RB5 and MB, at an initial concentration of 10 mg/L. Experiments were carried out in a dynamic regime under different conditions by varying the amounts of materials and hydrogen peroxide concentrations. The powder samples were added to 25 mL dye solutions at concentrations of 10 mg/L. To determine the adsorption/desorption equilibrium between contaminants and catalyst surfaces, the solutions were kept in the dark with continuous stirring for 30 min. No adsorption could be observed. After that point, the time was considered zero, and H2O2 at different concentrations (5 mM, 10 mM, and 20 mM) was added to the samples under UVA light to enable the reaction. The experiments were conducted in petri dishes with continuous stirring under Ultraviolet A light (λ = 365 nm) using a 15 W Analytik Jena power lamp. At different intervals, aliquots of suspension samples were collected and filtered through a 0.22 μm nitrocellulose membrane. The absorbance of pollutants was measured using a UV-Vis Spectrophotometer (Thermo Scientific Microplate Reader Multiskan GO) at the maximum bands of λ = 597 nm (RB5) and λ = 668 nm (MB).

Wastewater treatment experiments

The RB5 and MB degradation experiments were carried out using natural wastewater effluent from a local wastewater treatment plant (WWTP). The test was performed through previously established optimum conditions that included 10 mg/L initial dye concentrations and 1 g/L of Fe_VT as catalyst dosage, in the presence and the absence of 10 mM H2O2. A small volume of each dye has been added to the effluent to achieve a dye concentration of 10 mg/L. The experiments are conducted in the same manner as in DW. The wastewater characteristics are provided in Table S1.

Recovery and material reuse

After photo-Fenton-like degradation, the Fe-doped volcanic tuff was evaluated through multiple reuse cycles in the presence and absence of 10 mM H2O2 to determine material effectiveness. The samples were filtered, washed with ethanol and MilliQ water till neutral pH, and finally dried at 105 °C for 6 h. Following reuse tests, the stability of material was assessed using XRD.

Solid waste evaluation

Toxicity studies and leaching procedure (TCLP) were used to evaluate the material’s stability and dyes mobility in different environmental conditions from acid to near neutral. The test followed USEPA standard 131181, applying three extraction fluids: E1 at pH 2.88 ± 0.05, E2 at pH 4.93 ± 0.05, and additionally E3 using MilliQ water at neutral pH. The first eluent, E1, was obtained from glacial acetic acid (0.5%), while the second eluent, E2, was adjusted with NaOH 1 N to a pH of 4.93 ± 0.05. The solid samples (VT, VT1, and Fe_VT) resulting from UVA irradiation were extracted with eluents E1, E2 and MilliQ water in batch reactors for 24 h. The absorbance was measured using UV-Vis. No colour could be observed.