l-Aspartic acid-functionalized magnetic nanoparticles: as a new magnetically reusable bifunctional acid–base catalysts for the synthesis of benzo[b]pyran and pyrano[3,2–c] chromene derivatives

Abstract

“Green chemistry” describes the development of new technologies that reduce or eliminate the need for hazardous compounds or the production of them. In order to accomplish this goal, we have developed a new magnetic recyclable biocatalyst in this study by successfully applying aspartic acid to magnetic nanoparticles. Aspartic acid's molecular makeup made it possible for it to stabilize on magnetic nanoparticles using a straightforward method. We characterized the synthesized catalyst using microscopic and spectroscopic techniques such as Fourier transform infrared (FT-IR), X-ray diffraction (XRD), Field emission scanning electron microscopy (FE-SEM), Energy-dispersive X-ray spectroscopy (EDS), vibrating sample magnetometer (VSM) and transmission electron microscopy (TEM). The catalytic activity of this organocatalyst was evaluated for the synthesis of benzo[b]pyran and pyrano[3,2–c] chromene derivatives, exhibiting excellent efficiency. This protocol offers several benefits, such as using a low-cost biocatalyst, nontoxicity, high product yield, easy separation, short reaction times, catalyst reusability, and H₂O/EtOH solvent. In summary, our research indicates a feasible approach towards developing a novel magnetic biocatalyst suitable for application in organic synthesis.

Introduction

Green chemistry methods have received a lot of interest as sustainable chemical and pharmaceutical synthesis is pursued. Using organocatalysts has become a practical tactic in this situation. These small organic molecules appeal due to their economic feasibility, widespread availability, and inherent stability. The removal of metal-based pollution from catalytic reactions is one of the significant environmental advantages of organocatalysis’s absence of metals. Using organocatalysts instead of traditional metal catalysts is therefore a feasible route toward sustainable chemical synthesis^1–5. Amino acids are one of those natural molecules that could be special due to having both acid and base groups in their structure^6,7. They used the amino acid l-proline as a catalyst for the synthesis of quinolone derivatives⁸ and aspartic acid for synthesizing some pyran-annulated scaffolds⁹ as a homogeneous catalyst are some examples of this. Numerous investigations have examined using amino acids as either homogeneous or heterogeneous catalysts. Chemical processes can be facilitate by the acidic and basic moieties of amino acid structures when they are employed as free molecules. Nevertheless, a significant disadvantage of these systems is their incapacity to effectively separate and discard the catalysts following the reaction or to reuse them. Numerous techniques for immobilizing amino acids onto various substrates have been developed in order to overcome these limitations. Still, there's a lot of interest in creating new, diverse heterogeneous organocatalysts using amino acids. The potential advantages of such systems, including ease of separation, reusability, and enhanced economic and environmental efficiency, have increased the urgency to achieve this goal. Magnetite nanoparticles (MNPs) comprising Fe₃O₄ have garnered significant attention in basic scientific research and catalytic systems due to their nano-scale size, expansive surface area, favourable biocompatibility, exceptional biodegradability, adsorptive properties, and biological activity^10–12. Therefore, MNPS can be a reputable option for catalyst improvement and has been used in recent research.

In this context, Maleki et al.¹³ successfully immobilized l-proline onto nanoparticles by employing N, N′-dicyclohexylcarbodiimide as a substrate. The reaction underwent several stages, which required harsh conditions and the utilization of noxious solvents such as DMF and toluene. Despite these limitations, ongoing efforts seek to identify ideal conditions for this reaction.

Davarpanah et al.¹⁴ also designed SBA-Nicotinic acid as a heterogeneous organocatalyst. One of the study's shortcomings was the non-magnetic catalyst, which was expensive and difficult to make, and the part where the bases were blocked.

On the other hand, benzo[b]pyrans and pyrano[3,2–c] chromenes are heterocyclic compounds present in the structure of different biologically active natural products and numerous medicinal compounds. They have anti-allergic, antioxidant, anti-Alzheimer, anti-coagulant, and anti-anaphylactic properties^11,15–19. Because of the importance of these compounds, several attempts have been made to synthesize these compounds more efficiently and faster. Many methods and strategies have already been reported, including the use of ultrasonic radiation^20,21, microwaves^22,23, and various catalysts and reagents such as NH₂@SiO₂@Fe₃O₄²⁴, Fe₃O₄@SiO₂@GPTMS@guanidine²⁵, L-arginine modified graphene oxide²⁶, MNPs–PhSO₃H¹⁵, Zr@ IL-Fe₃O₄ MNPs²⁷Cu(II) complex of tetradentate Schiff-base supported on silica²⁸, Fe₃O₄@gly@Furfural@Co(NO₃)₂²⁹, [DABCO-C18] Br³⁰, and Hercynite@SiO₂@Tris³¹. Maleki et al.³², introduced Fe₃O₄@SiO₂-creatine as an organometallic biocatalyst for synthesizing 2-amino-4H-chromene. However, Creatine as an organic molecule is a compound that is neither abundant nor affordable. While each of the methods as mentioned above, employed thus far offers certain benefits, several drawbacks have been identified, including the use of hazardous solvents, the production of costly catalysts, low product yield, and prolonged reaction times. Consequently, researchers have redoubled their efforts to explore alternative approaches that address these limitations.

We investigated the development and synthesis of a novel heterogeneous natural nanocatalyst with dual acid–base characteristics in order to overcome the shortcomings of existing techniques. Aspartic acid is a naturally occurring, inexpensive, and easily available chemical that was used to create the catalyst. Aspartic acid, with two carboxylic groups and one amine group in its structure, is classified as an acidic amino acid, with a pK_a of 3.8³³. Here, we developed aspartic acid on nanoparticles as a novel organocatalyst. The use of (3-chloropropyl) trimethoxysilane on the surface of magnetic nanoparticles (MNPs) covered with Fe₃O₄@SiO₂ helped to stabilize aspartic acid quickly by a nucleophilic attack reaction. The catalytic power of this new nanocomposite was also investigated in the synthesis of benzo[b]pyrans and pyrano[3,2–c] chromenes, exhibiting significant results. Noteworthy advantages of our work include the facile separation of the catalyst through an external magnet, the ability to reuse the catalyst up to 7 times, the achievement of a high product yield, and the performance of the reaction under mild conditions and at 70 °C temperature.

Experiment

Materials and method

The chemicals used in this work were purchased from Merck Chemical Company (Germany) and were used without purification. Thin layer chromatography (TLC) on PolyGram SILG/UV254 silica gel plates was used to monitor the progress of the reactions. Using the Buchi B-540 B melting point device, the melting points of the products were measured. Shimadzu FT-IR 8300 spectrophotometer was used to identify the synthesized compounds and catalyst. The XRD pattern was made using Cu and Ca radiation with a Ni-FILTERED filter (λ = 1.54°A) in the range of 2θ from 4 to 70 °C. EDS analysis was also recorded by SAMX MIRA II. The morphology of the synthesized nanocomposite was recorded using a field emission scanning electron microscope (FE-SEM) (MIRA3TESCAN-XMU). The magnetic measurements were performed using a vibrating sample magnetometer Model VSM-4 inh. A thermogravimetric analysis (TGA) was conducted on a STA 504 thermal analysis machine.

General procedure for the preparation of Fe₃O₄@SiO₂(CH₃)₃@Asp

Synthesis of Fe₃O₄@SiO₂ nanoparticles

Fe₃O₄ nanoparticles were synthesized according to previous studies¹. In a 100 mL flask, Fe₃O₄ nanoparticles prepared (1.00 g) were dispersed in deionized water (20 mL) under ultrasound waves; and tetraethyl orthosilicate (TEOS) (3 mL) and were added dropwise. The reaction mixture was allowed to heat under reflux conditions for 2 h. The obtained product was collected with a magnet, washed three times with hot ethanol, and dried in an oven for 6 h.

Synthesis of Fe₃O₄@SiO₂(CH₃)₃–Cl nanoparticles (CMNP)

0.5g of Fe₃O₄@SiO₂ obtained from the previous step was added in a solution containing toluene (40 mL) and ultrasonically dispersed for 30 min, then (3-Chloropropyl) trimethoxysilane (1 mL) was slowly added into the solution under sonication. After that, the mixture was under reflux conditions for 6 h. The obtained product was collected with an external magnet washed several times with hot ethanol, and dried in an oven at 80 °C for 6 h.

Synthesis of nanocatalyst Fe₃O₄@SiO₂@(CH₃)₃@l-aspartic acid (LAMNP)

0.5g of Fe₃O₄@SiO₂@(CH₃)₃–Cl and 50 mL of H₂O solvent were added to the flask and the mixture was ultrasonically dispersed. Then, 0.6 g of l-aspartic acid and 0.6 g of NaOH as a base were added to the reaction container and stirred for 12 h under reflux conditions. After the reaction container was allowed to cool, the deposit of magnetic nanoparticles containing amino acid was separated from the reaction container. Then, it was washed several times with hot ethanol. The obtained magnetic organocatalyst was kept at room temperature for 24 h and dried (Fig. 1).

Fig. 1.

Preparation of Fe₃O₄@SiO₂(CH₃)₃@Asp.

Synthesis of pyrano[3,2–c] chromene derivatives

A mixture of malononitrile (0.066 g, 1.0 mmol), 4-hydroxycoumarin (0.162 g, 1.0 mmol), arylaldehyde (1 mmol), and LAMNP (0.04 g) was stirred in 10 mL EtOH/H₂O (50/50) at 70 °C. The progress of the reaction was monitored on thin-layer chromatography. After completion of the reaction, the heterogeneous catalyst was separated by an external magnet. The result appeared as a pure solid in high yields when water was added to the residue. Lastly, by evaporating the solvent, the desired product is recrystallized in hot ethanol for higher purification (Fig. 2).

Fig. 2.

Synthesis of pyrano[3,2–c] chromene derivatives.

Synthesis of benzo[b]pyran derivatives

In a round-bottomed flask, a mixture of malononitrile (0.06 g, 1.0 mmol), aryl aldehyde (1 mmol), and dimedone (0.14 g, 1.0 mmol), and LAMNP (0.04 g) stirred in 10 mL of H₂O/EtOH (50/50) at 70 °C. Following the TLC-monitored reaction's completion, the mixture was cooled to room temperature. An external magnet was used to separate the heterogeneous catalyst from the reaction solution once the reaction was finished. The result appeared as a pure solid in high yields when water was added to the residue. Finally, with the removal of the solvent, the reaction product was separated. The solid precipitate was recrystallized in ethanol hot solvent, purifying the product (Fig. 3).

Fig. 3.

Synthesis of benzo[b]pyran derivatives.

Hot filtration test

To study the leaching of LAMNP, a hot filtration test was conducted. To test the hot filtration, we performed the model reaction with a LAMNP catalyst. After 30 min, reaction progress was recorded by TLC. Then, the catalyst was removed from the reaction mixture by a magnet. The remaining mixture was stirred for 20 min. The progress of the reaction was not observed, and as a result, no leaching was observed for this catalyst. We employed EDX analysis, to ascertain the catalyst's stabilization. The presence of all elements was observed in this analysis (Fig. 4).

Fig. 4.

EDX analysis of the LAMNP for hot filtration test.

Investigating the catalytic properties of nanocomposite

After the successful synthesis and identification of LAMNP, the catalytic activity of this nanocomposite was investigated in the synthesis of pyrano[3,2–c] chromene derivatives, for this purpose, the reaction of malononitrile,4-nitrobenzaldehyde, and 4-hydroxycoumarin was chosen as a model reaction. Then the reaction of the model was checked in different conditions to optimize the temperature, solvent, and different amounts of catalyst, the results of which are shown in Table 1. At first, to investigate the effect of aspartic acid on improving the progress of the reaction, the model reaction was carried out in the presence of CMNP and LAMNP composite. By comparing how well these two performed in terms of reaction speed and product yield, the effectiveness of the amino acid presence was ascertained. In the following, considering the importance of solvent as a basic factor in industrial chemistry and helping to improve the safety and productivity of the reaction, the reaction of the model was investigated in different solvents and at room temperature. The reaction was carried out in H₂O, chloroform, ethyl acetate, ethanol, and H₂O/EtOH solvents. In addition, no solvent condition was also investigated. Finally, H₂O/EtOH was obtained as a suitable and low-risk solvent for this reaction with an efficiency of 98% in 12 min. Among the suitable characteristics of H₂O/EtOH were its green solvent, availability, appropriate polarity, and ability to dissolve raw materials.

Table 1.

Optimization conditions of Malononitrile, 4-nitrobenzaldehyde, and 4-hydroxycoumarin.

Entry	Catalyst	Catalyst (g)	Temp	Solvent	Time (min)	Yield (%)
1	–	0.4	r.t	EtOH	65	60
2	CMNP	0.4	r.t	EtOH	35	60
3	LAMNP	0.4	r.t	EtOH	30	97
4	LAMNP	0.4	r.t	H2O	45	93
5	LAMNP	0.4	r.t	CHCl3	50	67
6	LAMNP	0.4	r.t	Solvent-free	40	96
7	LAMNP	0.4	r.t	EtOAc	50	73
8	LAMNP	0.4	r.t	H2O/EtOH(50/50)	25	98
9	LAMNP	0.4	60	H2O/EtOH(50/50)	15	98
10	LAMNP	0.4	70	H2O/EtOH(50/50)	12	98
11	LAMNP	0.5	70	H2O/EtOH(50/50)	12	98
12	LAMNP	0.3	70	H2O/EtOH(50/50)	15	85

Significant values are in [bold].

The model reaction was then conducted at various temperatures to determine the ideal temperature. The 70 °C conditions were the best conditions for this reaction.

In the end, various catalyst concentrations were used in the reaction to obtain the optimal amount. By examining the amount of different of the desired catalysts, 0.4 g of LAMNP was found to be the best amount.

As a result, the best conditions for advancing the reaction are using H₂O/EtOH solvent at 70 °C and the presence of 0.4 g of the desired catalyst.

Synthesis of 2-amino-4H-chromene and 2-amino-4H-pyran derivatives under optimized conditions

Then 2-amino-4H-chromene and 2-amino-4H-pyran derivatives were obtained, and synthesized under optimal conditions and the results were summarized in Tables 2 and 3. Different types of aromatic aldehydes were used to investigate the effects of different groups on the aromatic ring ( Electronic supplementary material The online version of this article (10.1038/s41598-024-71901-6) contains supplementary material, which is available to authorized users).

Table 2.

Synthesis of pyrano[3,2–C] chromenes in the presence of LAMNP.

Entry	R	Time (min)	Yield (%)b	m. p. (ref.)	TON (mol)	TOF (mol/h)
4a	4-NO2-C6H4-	12	98	258–25926	725	60.41
4b	2-NO2-C6H4-	14	95	254–25616	703	50.2
4c	C6H5-	20	95	254–25626	703	35.15
4d	4-OH-C6H4-	25	90	261–26234	666	26.64
4e	3-OH-C6H4-	20	93	270–27326	688	34.4
4f	3-pyridine-	12	95	255–25816	703	58.58
4g	2-Cl-C6H4-	14	93	210–21326	633	45.21

Table 3.

Synthesis of benzo[b]pyrans in the presence of LAMNP.

Entry	R	Time (min)	Yield (%)b	m. p. (ref.)	TON (mol)	TOF (mol/h)
5a	4-NO2-C6H4-	14	98	233–23535	725	51.85
5b	3-NO2-C6H4-	12	95	198–20116	703	58.64
5c	2-NO2-C6H4-	14	98	234–23636	725	51.78
5d	4-OH-C6H4-	20	95	246–24726	703	35.15
5e	C6H5-	20	93	225–22616	688	34.44
5f	4-OMe-C6H4-	30	90	204–20526	666	22.22
5g	4-Me-C6H4-	25	92	211–21326	681	27.25
5h	4-Cl-C6H4-	15	95	215–21636	703	46.86

The proposed mechanism for the reaction in the presence of a LANMP catalyst is shown in Fig. 5.

Fig. 5.

The proposed mechanism for the reaction in the presence of LANMP.

The first step is activating the carbonyl group of aldehydes by COOH of the catalyst (Brønsted acidic). In the second step, 1,3-dione interacted with the catalyst and attacked the knoevenagel intermediate benzylidene malononitrile (I). Finally, the product was obtained by the intramolecular reaction from intermediate (III) in the presence of an amino group in the catalyst (Lewis base).

Discussion

Figure 1 shows Our strategy for the synthesis of LAMNP. The magnetic bio-organocatalyst was synthesized in four steps and via a simple procedure (Fig. 1). At First, Fe₃O₄ nanoparticles were synthesized according to previous reports and then covered with silica to get core–shell (Fe₃O₄@SiO₂), facilitating its separation using only an external magnet. Next, with the help of (3-chloropropyl) trimethoxysilane, aspartic acid was placed on the Fe₃O₄@SiO₂ nanocomposite surface, and the desired catalyst was synthesized. The binding of aspartic acid nitrogen to magnetic nanoparticles and the freeness of its two carboxylic acid groups increased the acidic properties of the catalyst. Placing amino acids on the surface of spherical particles and their proper accumulation improved the catalytic properties. The intended nanocatalyst was identified and characterized using FT-IR spectroscopy, XRD analysis, VSM, TGA, FE-SEM, and EDX.

FT-IR analysis

To confirm catalyst synthesis, we use FT-IR analysis. For this purpose, the spectrum of Fe₃O₄@SiO₂, CMNP, and LAMNP is compared in Fig. 6.

Fig. 6.

FT-IR spectra of Fe₃O₄@SiO₂ (a), CMNP (b), LAMNP (c).

Fe₃O₄@SiO₂ magnetic particles have several specific peaks in 3220, 1083, and 579/cm, respectively related to hydroxyl groups, Si–O–Si asymmetric stretching, and Fe–O bonds of iron oxide.

In spectrum CMNP, the intensity of the peaks decreased, and the peak in the 3220/cm regions, related to OH groups, was removed to some extent. From the comparison of this spectrum with the FT-IR spectrum of LAMNP composition, it is well observed that several peaks are created. The broad peak observed in the region 3235/cm is related to the OH groups of carboxylic acid and aspartic acid. The peaks in the region 2921/cm and 2852/cm are related to aliphatic CH groups. The low-intensity peak appearing in 1615 can be attributed to the COO^- in LAMNP. The presence of peaks in the region of 576 and 1090/cm is related to Fe₃O₄@SiO₂ particles, which are well preserved in the structure of the catalyst.

EDX analysis

The EDX analysis was performed to show that aspartic acid was present in LAMNP. As Fig. 7 shows, The EDX spectrum of CMNP shows the presence of O, Si, Cl, and Fe, and the percentage composition is 32.15, 3.61, 0.04, and 58.16% respectively.

Fig. 7.

EDX analysis of the CMNP (a), and LAMNP (b).

Fe, C, O, N, and Si elements were seen in the EDX analysis of the LAMNP organocatalyst, whose percentages are 17.9, 43.61, 33.90, 3.63, and 0.90, respectively. It shows the successful placement of amino acid on the Fe₃O₄@SiO₂ surface.

Figure 8 shows the EDS-mapping analysis, according to the results of EDS mapping analysis in Fig. 8, the distribution of these elements can be seen homogeneously and uniformly on the surface of the nanocatalyst.

Fig. 8.

Elemental mapping images of (a) CMNP, (b) LAMNP.

FE-SEM analysis

Field emission scanning electron microscope (FESEM) analysis was used to investigate the morphology and size of particles (Fig. 9). These images show that LAMNP is regular in shape and mostly spherical. The process and speed of the reaction are enhanced by a spherical and uniform surface because it increases the contact surface between the catalyst and the raw materials. The particle size in was distribution diagram of LAMP be measured by SEM images as 15–25 nm.

Fig. 9.

FE-SEM image of LAMNP.

XRD analysis

Figure 10 shows the XRD patterns of LAMNP. The figure has specific peaks in the region of 2 θ = 18°, 31°, 37°, 43°, 54°, 57°, 63°and 73° which corresponds to the standard pattern of Fe₃O₄ and indicates the presence of Fe₃O₄ nanoparticles without degradation in the catalyst structure³⁷.

Fig. 10.

XRD analysis of LAMNP.

VSM analysis

Vibrating sample magnetometer (VSM) analysis was performed to study the magnetic properties of the catalyst for LAMNP at room temperature (Fig. 11). According to the diagram in Fig. 10, the superparamagnetic property of the catalyst is confirmed, and the magnetization is saturated up to 58 emu/g at an applied field of 9800 Oe, with nearly no important coactivity. As a result, we synthesized a magnetic bio-organocatalyst that can be easily separated from the reaction medium by using an external magnet.

Fig. 11.

Vibrating sample magnetometer analysis of LAMNP.

TGA analysis

In Fig. 12, Thermal analysis of the LAMNP was performed with TGA analysis. According to thermo-gravimetric analysis (TGA), losing weight happens in four stages. The first weight loss (~ 100 °C) pertains to the moisture extraction from the catalyst structure. a weight loss of about 2.2% in the range of 100 °C should be attributed to the evaporation of water molecules. Around 225 °C is when mass loss starts, and it continues until around 425 °C, corresponding to elimination the organic functional groups amino acid on the magnetic nanoparticles surface.

Fig. 12.

TGA curves of LAMNP.

Next, at a temperature between 425 and 500 °C, every organic molecule breaks down on the surface of magnetic nanoparticles. Eventually, at higher temperatures of 500 °C, there is a mass loss that arises from the complete decomposition.

Investigating the reuse of the catalyst

In the end, the reusability and recoverability of the catalyst was checked. For this purpose, the catalyst used in the model reaction was separated and washed several times in hot ethanol solvent. Then, it was dried at room temperature for 24 h. The re-dried catalyst is used in the model reaction. The catalytic capacity did not significantly alter, as predicted, and it demonstrated recoverability and reusability in up to seven reaction cycles (Fig. 13). Achieving this goal is possible by magnetizing an organic catalyst. we also used FT-IR measurement to verify the stability of our catalyst both before and after the chemical reaction. The resulting spectra, which are displayed in Fig. 14, did not exhibit any discernible differences.

Fig. 13.

Recoverability of the LANMP.

Fig. 14.

FT-IR spectrum of LAMNP. (a) Before reaction, and (b) after reaction.

Comparison of catalyst efficiency with previous works

The efficiency of the synthesized catalyst in reaction with benzo[b]pyrans and pyrano[3,2–c] chromenes was compared with the works reported in the past (Table 4). After this comparison, the suitability and effectiveness of the synthesized catalyst could be seen. The use of a cheap biocatalyst and its magnetic nature is a confirmation of economic efficiency and compatibility with the green chemistry of this work. In addition, the product yield in this work and the performance of the reaction in balanced and green solvent conditions confirm the superiority of the work compared to previous works.

Table 4.

Comparison of catalyst efficiency with previous works.

Product	Catalyst/Temp/Time (min)/Solvent	Yield (%)	Product	Catalyst/Temp/Time (min) /Solvent	Yield (%)
	Fe3O4@Dendrimer-NH2-HPA /reflux/5/EtOH	9238		o-Benzenedisulfonimide (OBS)/120 °C/50/Solvent-free	8539
	piperazine-GO/50 °C/15/H2O-EtOH	9540		Zn3(PO4)2∙4H2O/reflux/15/EtOH	8941
	MnFe2O4@SiO2NHPhNH2–PTA./80 °C/ 30/Solvent-free	9539		PS-PTSA/ 80 °C/120/EtOH	9442
	titania-supported sulfonic acid (n-TSA) /80 °C/45/ Solvent-free	9643		[Amb]L-prolinate/reflux/20/EtOH	9044
This work	LAMNP/70 oC /20 /H2O/EtOH	93	This work	LAMNP/70 oC /20 /H2O/EtOH	95

Conclusion

In conclusion, by stabilizing aspartic acid on MNP, a magnetic heterogeneous biocatalyst with dual acid–base characteristics was created that is environmentally beneficial. The ability to use this catalyst for the synthesis of pyrrole derivatives showed acceptable results. The use of amino acid as a cheap and available material, its convenient placement on the surface of magnetic particles, simplicity, cleanliness, and performing the reaction in balanced conditions were some of the salient advantages of this work. Additionally, this proposed biocatalyst is also readily recovered using an external magnet and can be used up to seven times in subsequent reactions without experiencing a significant decrease in catalytic activity. We successfully created and synthesized pyrans and chromones in an H₂O/EtOH solvent at a temperature of 70 °C by designing and creating an appropriate and efficient organocatalyst.

Author contributions

Leila Amiri-Zirtol, Conceptualization, Investigation, Writing- Original. Hanieh Mostashfi, Experimental, Writing- Original. Razieh Sabet, Conceptualization, Writing- Reviewing and Editing. Zahra Karimi, Experimental. Reza Ranjbar-Karimi, Writing- Original and Investigation.

Funding

Financial assistance from the Shiraz University of Medical Sciences by way of grant number 27533 is gratefully acknowledged.

Data availability

All data generated or analyzed during this study are included in this published article.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-71901-6.