Functionalized Magnetic Nanoparticles as Catalysts for Enantioselective Henry Reaction

Abstract

With the aim to easily recover and reuse the catalyst, an efficient amino alcohol catalyst previously tested in the asymmetric addition of diethylzinc to several aromatic aldehydes has been immobilized on proper functionalized superparamagnetic core–shell magnetite–silica nanoparticles and employed in the Henry reaction in the semi-homogeneous phase. The nanocatalyst exhibits a promising catalytic activity that remains unchanged in the three catalytic cycles performed. The results prove that highly efficient catalysts, by being immobilized on suitable magnetic nanosupports, can be easily recovered and reused, maintaining their catalytic behavior.

Introduction

The nitroaldol or Henry reaction is a classical and powerful method to form carbon–carbon bond with the production of a new stereogenic center under mild conditions. The coupling of the nucleophile generated from a nitroalkane with a carbonyl electrophile leads to the formation of a β-nitroalcohol, a versatile intermediate that can easily undergo further transformations such as oxidation, reduction, and dehydration.¹ Considering the potential of the intermediate obtained, several studies have been directed toward the development of asymmetric catalysts for the Henry reaction;² a variety of catalysts have been employed, such as metal-based catalysts,³⁻⁶ organocatalysts,^7,8 and enzymes.⁹

β-amino alcohol motif is typical of numerous chiral auxiliaries, ligands, and catalysts used in asymmetric synthesis;¹⁰ nevertheless, only few examples of their use in the Henry reaction can be found in the literature. A noteworthy example is the catalyst reported by Palomo et al.,¹¹ which generally provides more than 90% ee, using N-methylephedrine (45%), Zn(OTf)₂ (30%), iPrNEt₂ (30%), and low temperatures. More recently, it has been reported that the employment of 2-amino-1,2-diphenylethanol derivatives, forming a chiral complex with Cu^II salts, provides the addition of nitromethane to a variety of aldehydes with more than 90% ee.^12,13

Even if asymmetric catalysis represents a powerful method for the synthesis of enantiopure molecules, its practical applications are extremely limited by the high costs and the severe ecological impact; therefore, the opportunity to recover and reuse the catalysts become a very important factor. The heterogenization of the catalyst facilitates its separation¹⁴ and potentially allows its recycling,^15,16 but unfortunately, the immobilization of chiral catalysts often results in lower activities and enantioselectivities. Recently, the use of magnetic nanoparticles led to the development of new catalysts that combine advantages of both homogeneous and heterogeneous catalysis; nanoparticles, owing to the high surface area to volume ratio, can be used as supports for asymmetric catalysts, with resulting activities close to the homogeneous ones. Specifically, nanoparticles of magnetite 10–20 nm in size exhibit a special form of magnetism, called superparamagnetism,^17,18 which allows them to be extremely dispersible in solvents in the absence of an external magnetic field and to overcome the recovery step by means of an agile magnetic decantation.

So far, only few examples of an asymmetric magnetic nanoparticle-supported amino alcohol catalyst have been reported.^19,20 With the aim to develop a novel versatile, magnetically recoverable, and recyclable nanocatalyst, we have focused on the design and synthesis of β-amino alcohol ligands bearing, in addition to a fine-tunable catalytic site, a functionality (an alkoxysilane group) for their covalent anchoring to magnetite nanoparticles (Figure 1).

Figure 1.

Anchoring strategy.

Before preparing supported nanoparticles, we tested the catalytic activity and enantioselectivity of free ligands 2 with a total length comparable to 1. Earlier studies²¹ proved that ligand 2a led to the best results in the asymmetric addition of diethylzinc to several aromatic aldehydes (Figure 2).

Figure 2.

Catalytic activity of ligand 2a.

Herein, we report the results of the employment in the Henry reaction both of the homogeneous amino alcohol ligand 2a and of its immobilized form onto superparamagnetic core–shell magnetite–silica nanoparticles.

Results and Discussion

Optimization of the Reaction Conditions

Initially, the conditions were optimized in the reaction of nitromethane with 2-chlorobenzaldehyde in the presence of the chiral ligand 2a. On the basis of literature data, the solvent, catalyst loading, and metallic salt were varied, that is, the reaction test was performed in diethyl ether, dichloromethane, ethanol, and 2-propanol with an amount of catalyst between 2.5 and 20% and employing copper acetate or several zinc salts. As inferable from Table 1, the employment of diethyl ether and 2-propanol led to a marked increase in efficiency, the first being slightly more efficient in enantioselectivity while the second in the chemical yield. Regarding metallic salt, zinc dramatically lowered the yields (entries 7–9), although the good results previously obtained²¹ suggested an efficient coordination to ligand 2a.

Table 1. Optimization of the Reaction Conditions.

entry	solvent	salt (mol %)	catalyst (mol %)	yielda (%)	ee (%)
1	Et2O	Cu(OAc)2 (10%)		0
2	Et2O	Cu(OAc)2 (2.5%)	2.5	57	68
3	Et2O	Cu(OAc)2 (5%)	5	73	55
4	Et2O	Cu(OAc)2 (10%)	10	80	75
5	Et2O	Cu(OAc)2 (15%)	15	80	36
6	Et2O	Cu(OAc)2 (20%)	20	68	34
7	Et2O	Zn(OAc)2 (2.5%)	2.5	0
8	Et2O	ZnCl (2.5%)	2.5	0
9	Et2O	ZnOTf (2.5%)	2.5	0
10	CH2Cl2	Cu(OAc)2 (10%)	10	30	0
11	EtOH	Cu(OAc)2 (10%)		20
12	EtOH	Cu(OAc)2 (5%)	5	75	24
13	EtOH	Cu(OAc)2 (10%)	10	88	23
14	2-PrOH	Cu(OAc)2 (10%)		16
15	2-PrOH	Cu(OAc)2 (10%)	10	88	71

^aChemical yields are referred to isolated compounds.

Finally, catalyst loading was found to have a significant effect on the enantioselectivities, as 10 mol % catalyst loading produced the highest value both in terms of yields (80%) and ee (75%) (entry 4).

Having concluded the condition optimization process, the reaction was carried out with a variety of aromatic aldehydes with the purpose of studying the catalyst’s efficiency relative to different ring substituents. All the substrates were treated with nitromethane, 10% ligand, and 10% Cu(OAc)₂ in diethyl ether or 2-propanol for 72 h at room temperature.

As reported in Table 2, in spite of the encouraging results obtained with 2- chlorobenzaldehyde, all the aldehydes tested in diethyl ether showed mediocre yields, around 30% (except for entry 3) and generally weak enantiomeric excesses, except for 2-methoxybenzaldehyde (93%, entry 4) and 3-methylbenzaldehyde (89%, entry 8).

Table 2. Henry Reaction Catalyzed by Ligand 2a.

^aChemical yields are referred to isolated compounds.

^bExperiment performed at −20 °C.

On the contrary, all the aldehydes tested in 2-propanol were converted in good yields for both electron-withdrawing and electron-donating substituents on the ring and, in some cases, with slightly higher values of enantiomeric excesses than those obtained with diethyl ether.

In order to increase the enantioselectivity, we performed the reaction at −20 °C on 3-nitrobenzaldehyde, which had shown, at room temperature, a very high yield but only good ee. Unfortunately, the moderate improvement of the ee obtained in this test was not balanced by an acceptable yield, as that dropped from 91 to 20% (entries 12 and 13).

Afterward, we started studying the diastereoselectivity of ligand 2a in the Henry reaction by reacting o-tolylbenzaldehyde, which had produced the best result of those reported in Table 2, with nitroethane (Figure 3). It resulted in the 83% conversion of the starting aldehyde into the two diastereoisomers 5a and 5b in a 1:1 ratio, the stereochemistry of each being assigned by comparison with reported NMR spectra.²² The enantioselectivities were very different for the two isomers, as 5a was obtained in 75% ee, while 5b in 45% ee.

Figure 3.

Henry reaction using nitroethane and 2-methylbenzaldehyde catalyzed by ligand 2a.

It is assumed that a nitroaldol reaction using a metal-based catalyst predominantly affords the syn product likely because a cyclic transition state is involved,²³ in which both the nitronate and the aldehyde coordinate to one metal cation. Since the authors supposed that a steric hindrance enhancement favors the syn diastereoselectivity, we performed the nitroethane addition using a bulkier aldehyde, that is, cyclohexanecarboxaldehyde. Effectively, products 6a and 6b were collected in an excellent diastereomeric ratio (88:12),²⁴ with good yields and enantioselectivities (Figure 4).

Figure 4.

Henry reaction using nitroethane and cyclohexanecarbaldehyde catalyzed by ligand 2a.

Ligand Immobilization onto Nanoparticles

Having extensively explored different anchoring strategies, different functionalities necessary for the bond with the surface, and different suitable spacers, the nanostructured catalyst 7 was designed (Figure 5) (22). The most important modification of catalyst 2a is the introduction of an aromatic and a triazole ring through a Cu^I-catalyzed azide/alkyne cycloaddition (CuAAC)²⁵ and a magnetic nanoparticle functionalized for anchoring the catalyst. We decided to employ nanoparticles with a magnetite (Fe₃O₄) core coated with a thin silica (SiO₂) layer; the silica shell is reported to preserve the inner Fe₃O₄ from oxidation by air to prevent the tendency to agglomerate and be easily functionalizable, owing to the many silanol groups exposed on the surface (Figure 5).²⁶

Figure 5.

Immobilization strategy for catalyst 7.

Before immobilizing the new ligand onto the nanoparticles, an analogue of catalyst 7, catalyst 14 (Scheme 2), was evaluated in a homogeneous phase catalysis test to verify that the introduced aromatic and triazole groups would not affect the catalytic properties.

Scheme 2. Preparation of Catalyst 14.

To this purpose, the alkyne 8 was synthesized in three steps starting from the acid precursor 10. First, the acid group underwent the condensation with the commercial tyramine to form the amide 11. The hydroxyl group of the tyramine moiety was then alkylated with propargyl bromide, and finally, the oxirane ring underwent the opening reaction leading to the anchorable precursor 8 in good yield (Scheme 1).

Scheme 1. Preparation of the Anchorable Precursor 8.

We performed a click reaction with the (3-azidopropyl)benzene 13(27) in the original conditions²⁸ with CuSO₄ and sodium ascorbate in a THF/H₂O mixture. However, the cycloaddition reaction product 14 was collected in a very low yield (26%) probably due to low solubility of the substrate 8. Consequently, we modified the procedure in order to improve the yields using cuprous iodide as the copper source, N,N-diisopropylethylamine (DIPEA) as the base, and tetrahydrofuran (THF) as the solvent. In these conditions, we obtained the catalyst 14 with a yield of 89% (Scheme 2).

Ligand 14 was then evaluated in the homogeneous phase catalysis test.

Data collected (Table 3) confirmed the range of yields and enantioselectivities obtained with the previous ligand tested in this reaction (ligand 2a).

Table 3. Nitromethane Addition to Aldehydes Catalyzed by Ligand 14.

entry	R	yield (%)a	ee (%)
1	H	90	68
2	2-Me	80	75
3	3-NO2	96	50

^aChemical yields are referred to isolated compounds.

Having confirmed the retained catalytic activity, we immobilized 8 onto azido-modified magnetic nanoparticles 9a and 9b, the latter obtained after treating 9a with hexamethyldisilazane. This is often used as a silylating agent, especially for silica, as it transforms the hydroxyl groups on the nanoparticle surface into trimethylsilyloxide, leaving a nonpolar surface.²⁹ Transmission electron microscopy (TEM) and high-resolution TEM (HR-TEM) images of 7b are shown in Figure 6a,b, respectively. The nanoparticle diameter is around 20 nm (Figure 6a). In Figure 6b, it is possible to distinguish lattice fringes of the magnetic phase.

Figure 6.

(a) TEM and (b) HR-TEM images of catalyst 7b.

The superparamagnetic catalysts 7a and 7b were thus obtained with 0.28 and 0.41 mmol/g, loading respectively. The loading was determined considering the N w/w % values found by elemental analyses. The Fourier transform infrared (FTIR) spectra in fact (Figure S54) show the disappearance of the azido group absorption of 9a and 9b, indicating a quantitative click reaction; in this way, the N w/w % value is diagnostic for the loading as it can be totally related to the organic moiety of 7a and 7b (Scheme 3).

Scheme 3. Preparation of 7a and 7b.

The new superparamagnetic amino alcohol catalysts 7a and 7b were finally evaluated in the addition of nitromethane to benzaldehyde. As reported in Table 4, while 7a led to mediocre values of yields and ee (entries 1 and 2), far more encouraging results came from the silylated catalyst 7b, leading to yields and enantioselectivities totally comparable to those obtained in the homogeneous phase reported in Table 3. This different behavior could be attributed to the better dispersibility of 7b in 2-propanol due to both the silyl ethers on the nanoparticle surface and the higher loading. Moreover, the catalytic activity remained unchanged in the three catalytic cycles performed (entries 3–5).

Table 4. Nitromethane Addition to Aldehydes Catalyzed by Catalyst 7a or 7b.

entry	R	catalyst (%)	yield %	ee %
1	H	7a (10%)	60	35
2	H	7a (20%)	65	36
3	H	7b (10%) I cycle	81	67
4	H	7b (10%) II cycle	75	60
5	H	7b (10%) III cycle	73	60
6	2-Me	7b (10%)	50	71
7	2-OMe	7b (10%)	83	68
8	3-NO2	7b (10%)	91	44
9	2-Cl	7b (10%)	>95	51

Conclusions

With the final aim to develop a new versatile recoverable and recyclable “nanocatalytic” system, we designed 7a and 7b, the immobilized version of 2a, already successfully used in the asymmetric addition of diethylzinc to a variety of aldehydes. We tested the ligand 2a in the asymmetric Henry reaction; the high yields and the good ee suggest a fairly good copper coordination ability of the ligand 2a. The immobilized version 7b exhibited a promising catalytic activity that remained unchanged in the three catalytic cycles performed.

The preliminary results prove that highly efficient catalysts can be easily recovered and reused by being immobilized on suitable nanosupports. In light of the encouraging reported results, we are currently investigating the catalytic efficiency of 2a and its immobilized analogues 7a and 7b in other asymmetric reactions and in the synthesis of molecules of pharmaceutical interest.

Experimental Section

Unless otherwise stated, commercial reagents purchased from Alfa Aesar, Acros, and Aldrich chemical companies were used without further purification. Purification of reaction products was carried out by flash chromatography using Kiesegel F Merck silica gel (230–400 mesh). Thin-layer chromatography (TLC) was performed on Kiesegel F 254 pre-coated silica gel plates, and visualization was achieved by inspection under UV light (Mineralight UVG 11254 nm) followed by staining with phosphomolybdic acid dip [polyphosphomolybdic acid (12 g) and ethanol (250 mL)] or 2,4-dinitrophenylhydrazine dip [2,4-dinitrophenylhydrazine dip (4 g), ethanol (100 mL), water (20 mL), and sulfuric acid]. ¹H NMR spectra were recorded using a Varian Mercury 300 (75 MHz) or a Bruker Avance 400 (100 MHz). Chemical shifts (δ) are reported in parts per million (ppm) relative to the internal standard of a residue solvent peak, chloroform (77.00 ppm). Optical rotations were measured on a digital polarimeter (Jasco DIP-370) with a cell path length of 1 cm; solution concentrations are reported in grams per 100 mL. Enantiomeric excesses were determined by way of an Agilent 1260 Infinity HPLC system equipped with a diode array detector (DAD) and a chiral stationary phase; chiral columns used were Chiralpak IA, Chiralpak IB, and Chiralpack IC.

Attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectra were recorded on a Shimadzu IR Prestige-21. Elemental analyses for C, H, and N were performed on an EA 1110 CHNS-O element analyzer. Morphologic and structural investigations were performed by way of a JEOL JEM 2010 transmission electron microscope.

(E)-3-((2S,3S)-3-Cyclohexyloxiran-2-yl)-N-(4-hydroxyphenethyl)acrylamide (11)

Carboxylic acid 10 (1 mmol) was added to a solution of 1.1 mmol of tyramine in 2 mL of dimethylformamide (DMF). A solution of 1 mmol of HOBt and 1 mmol of N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide (EDC) in 2 mL of DMF were added to the reaction mixture. The reaction was stirred at room temperature for 12 h. The reaction mixture was diluted with AcOEt and washed with water. The aqueous layer was extracted with AcOEt. The organic layer was washed with cold acidic water (pH 4) and then dried over Na₂SO₄. The solvent was removed in vacuo. The crude product was purified by flash chromatography (Hex/AcOEt, 1:1) to give 248 mg (79%) of 11. ¹H NMR (400 MHz, CDCl₃): δ 8.12 (bs, 1H, PhOH), 6.94 (d, J = 8.3 Hz, 2H, Ph), 6.77 (d, J = 8.4 Hz, 2H, Ph), 6.59 (dd, J = 15.3, 6.6 Hz, 1H, CH=CHCONH), 6.41–6.28 (m, 1H, NH), 6.06 (d, J = 15.2 Hz, 1H, CH=CHCONH), 3.48 (d, J = 6.2 Hz, 2H, NHCH₂), 3.20 (dd, J = 6.5, 1.5 Hz, 1H, CHOCH=), 2.70 (t, J = 6.9 Hz, 2H, CH₂Ph), 2.62 (dd, J = 6.7, 1.9 Hz, 1H, c-HexCHO), 1.79–1.60 (m, 5H, c-Hex), 1.27–1.00 (m, 6H, c-Hex). ¹³C NMR (100 MHz, CDCl₃): δ 165.6, 155.3, 141.1, 129.7, 129.7, 125.3, 115.8, 66.2, 55.4, 41.3, 40.1, 34.6, 31.8, 29.5, 28.8, 26.2, 25.5. Anal. Calcd for C₁₉H₂₅NO₃: C, 72.35; H, 7.99; N, 4.44. Found: C, 72.69; H, 8.02; N, 4.64.

(E)-3-((2S,3S)-3-Cyclohexyloxiran-2-yl)-N-(4-(prop-2-yn-1-yloxy)phenethyl)acrylamide (12)

Phenolic compound 11 (1 mmol) was dissolved in 12 mL of CH₃CN,. and 2 mmol (0.22 mL of an 80 wt % solution in toluene) of propargyl bromide and 3 mmol (415 mg) of K₂CO₃ were added under an argon atmosphere. The reaction was refluxed for 12 h. The solid residue was eliminated by filtration, and the solvent was removed in vacuo. The crude product was purified by flash chromatography (Hex/AcOEt, 60:40) to give 293 mg of 12 (83%). ¹H NMR (300 MHz, CDCl₃): δ 7.08 (d, J = 8.5 Hz, 2H, Ph), 6.88 (d, J = 8.5 Hz, 2H, Ph), 6.58 (dd, J = 15.2, 6.6 Hz, 1H, CH=CHCONH), 6.12 (t, J = 5.6 Hz, 1H, NH), 6.03 (d, J = 15.2 Hz, 1H, CH=CHCONH), 4.63 (d, J = 2.3 Hz, 2H, PhOCH₂), 3.56–3.41 (m, 2H, CH₂NH), 3.20 (dd, J = 6.6, 1.6 Hz, 1H, CHOCH=), 2.74 (t, J = 7.0 Hz, 2H, CH₂Ph), 2.62 (dd, J = 6.6, 1.6 Hz, 1H, c-HexCHO), 2.51 (t, J = 2.2 Hz, 1H, C≡CH), 1.74–1.02 (m, 11H, c-Hex). ¹³C NMR (75 MHz, CDCl₃): δ 164.9, 156.2, 149.9, 140.7, 131.8, 129.7, 125.4, 115.0, 78.6, 75.64, 66.0, 55.8, 55.4, 40.9, 40.1, 34.7, 31.9, 29.5, 28.8, 26.2, 25.6. Anal. Calcd for C₂₂H₂₇NO₃: C, 74.76; H, 7.70; N, 3.96. Found: C, 74.91; H, 7.95; N, 4.25.

(4R,5S,E)-5-Cyclohexyl-5-hydroxy-4-morpholino-N-(4-(prop-2-yn-1-yloxy)phenethyl)-pent-2-enamide (8)

330 mg (75%). 1 mmol of 12, LiClO₄ (1.59 g, 15 mmol), and morpholine (10 mmol, excess) in acetonitrile (3 mL) were stirred at 55 °C for 24 h. Then, H₂O was added, and the aqueous layer was extracted with CH₂Cl₂. The combined organic extracts were dried over Na₂SO₄, concentrated in vacuo, and purified by flash chromatography (CH₂Cl₂/MeOH, 95:5) to give 330 mg (75%) of 8. ¹H NMR (400 MHz, CDCl₃): δ 7.08 (d, J = 8.3 Hz, 2H, Ph), 6.87 (d, J = 8.0 Hz, 2H, Ph), 6.68 (dd, J = 15.6, 9.9 Hz, 1H, CH=CHCONH), 6.05–5.93 (m, 1H, NH), 5.88 (d, J = 15.6 Hz, 1H, CH=CHCONH), 4.66–4.59 (m, 2H, PhOCH₂), 3.63 (t, J = 4.1 Hz, 4H, CH₂O x2), 3.54–3.44 (m, 3H, CHOH, CONHCH₂), 3.20 (s, 1H, OH), 2.82 (dd, J = 9.9, 3.4 Hz, 1H, CHN), 2.75 (t, J = 7.0 Hz, 2H, CH₂Ph), 2.57–2.33 (m, 5H, C≡CH, CH₂N_morp x2), 2.03–1.91 (m, 1H, CH_c-Hex), 1.73–1.45 (m, 4H, c-Hex), 1.34–0.78 (m, 6H, c-Hex). ¹³C NMR (100 MHz, CDCl₃): δ 164.9, 156.2, 139.0, 131.8, 129.7, 128.1, 115.0, 78.6, 75.6, 72.4, 68.8, 67.0, 55.8, 51.5, 40.9, 39.4, 34.7, 29.3, 28.0, 26.4, 25.7, 25.6. Anal. Calcd for C₂₆H₃₆N₂O₄: C, 70.88; H, 8.24; N, 6.36. Found: C, 71.12; H, 8.72; N, 6.52.

(3-Azidopropyl)benzene (13)

3-Phenylpropan-1-ol (1 mmol, 136 mg) was dissolved in 1 mL of CH₂Cl₂ under an argon atmosphere in an ice bath, and 0.3 mmol (0.028 mL) of PBr₃ was added. The reaction was stirred at room temperature for 4 h. The reaction mixture was washed with water, and the aqueous layer was extracted with CH₂Cl₂. The organic layer was washed with NaHCO₃ s.s. and with brine and then dried over Na₂SO₄. The solvent was removed in vacuo to give the corresponding bromide derivative. The crude was dissolved in 1.3 mL of DMSO under an argon atmosphere, and 4 mmol (260 mg) of NaN₃ was added. The reaction was stirred at room temperature for 12 h. CH₂Cl₂ was then added, and the organic layer was washed several times with small portions of water. The aqueous layer was then extracted with CH₂Cl₂, and then the organic layer was washed with brine and dried over Na₂SO₄. The solvent was removed in vacuo. The crude product was purified by flash chromatography (Hex/AcOEt, 95:5) to give 48 mg of 13 (31% over two steps). ¹H NMR (300 MHz, CDCl₃): δ 7.39–7.20 (m, 5H, Ph), 3.32 (t, J = 6.8 Hz, 2H, CH₂N₃), 2.75 (t, J = 7.6 Hz, 2H, CH₂Ph), 2.02–1.90 (m, 2H, CH₂CH₂CH₂). ¹³C NMR (75 MHz, CDCl₃): δ 140.9, 128.6, 128.5, 126.2, 50.7, 32.8, 30.5. Anal. Calcd for C₉H₁₁N₃: C, 67.06; H, 6.88; N, 26.07. Found: C, 9; 67.32; H, 7.12; N, 26.19.

(4R,5S,E)-5-Cyclohexyl-5-hydroxy-4-morpholino-N-(4-((1-(3-phenylpropyl)-1H-1,2,3-triazol-4-yl)methoxy)phenethyl)pent-2-enamide (14)

The terminal alkyne 8 (1 mmol, 440 mg) and the azide 8 (1.1 mmol, 177 mg) were dissolved in 4 mL of THF. CuI (0.2 mmol, 38 mg) and DIPEA (13 mmol, 2.2 mL) were then added, and the reaction was stirred for 12 h. The reaction mixture was diluted with AcOEt and washed with water. The aqueous layer was extracted with AcOEt. The organic layer was washed with saturated aqueous NH₄Cl and brine and then dried over Na₂SO₄. The solvent was removed in vacuo. The crude product was purified by flash chromatography (CHCl₃/MeOH, 96:4) to give 535 mg (89%) of 14. ¹H NMR (400 MHz, CDCl₃): δ 7.56 (s, 1H, =CH—N—N=N), 7.32–7.06 (m, 7H, Ph), 6.92 (d, J = 8.3 Hz, 2H, Ph), 6.70 (dd, 1H, J = 15.6, 9.9 Hz, CH=CHCONH), 5.88 (d, 1H, J = 15.6 Hz, CH=CHCONH), 5.82–5.73 (m, 1H, CONH), 5.17 (s, 2H, PhOCH₂), 4.34 (t, 2H, J = 7.1 Hz, CH₂N—N=N), 3.68 (t, 4H, J = 4.5 Hz, OCH₂ x2), 3.58–3.47 (m, 3H, CHOH, NHCH₂), 3.07 (bs, 1H, OH), 2.86 (dd, 1H, J = 9.6, 2.3 Hz, CHN), 2.78 (t, 2H, J = 6.8 Hz, NHCH₂CH₂Ph), 2.64 (t, 2H, J = 7.5 Hz, NCH₂CH₂CH₂Ph), 2.60–2.39 (m, 4H, CH₂N_Morp x2), 2.30–2.21 (m, 2H, NCH₂CH₂CH₂Ph), 2.01 (bd, J = 14.0 Hz, 1H, c-Hex), 1.77–1.61 (m, 3H, c-Hex), 1.52 (bd, 1H, J = 11.8 Hz, c-Hex), 1.34–0.81 (m, 6H, c-Hex). ¹³C NMR (100 MHz, CDCl₃): δ 164.9, 157.1, 144.2, 140.1, 138.9, 131.5, 129.9, 128.7, 128.5, 128.3, 126.5, 122.7, 115.1, 72.5, 69.0, 67.1, 62.2, 51.7, 49.7, 41.0, 39.6, 34.8, 32.5, 31.7, 29.6, 28.1, 26.5, 25.8, 25.7. Anal. Calcd for C₃₅H₄₇N₅O₄: C, 69.86; H, 7.87; N, 11.64. Found: C, 71.15; H, 8.09; N, 11.91.

General Procedure for the Addition of Nitroalkane to Aldehydes Catalyzed by Free Ligands in the Homogeneous Phase

The chiral ligand (0.10 mmol) and Cu(OAc)₂·H₂O (0.10 mmol, 20 mg) were dissolved in 3 mL of 2-PrOH or Et₂O and stirred for 20 min. Nitroalkane (10 mmol, 0.53 mL of nitromethane or 0.71 mL of nitroethane) was added, and the mixture was stirred for 20 min. The aldehyde (1 mmol) was added, and the reaction was stirred at room temperature for 72 h. The solvent was evaporated in vacuo, and the residue was dissolved in diethyl ether and filtered through a 1:1 silica/Celite pad with ether. The solvent was evaporated in vacuo, and the crude product was purified by flash chromatography (Hex/EtOAc, 80:20).

General Procedure for the Addition of Nitroalkane to Aldehydes Catalyzed by Functionalized Nanoparticles

Cu(OAc)₂·H₂O (0.025 mmol, 5 mg) was added to a dispersion of the functionalized nanoparticles (10 mol %) in 3 mL of 2-PrOH, and the mixture was mechanically agitated for 20 min. Nitromethane (2.5 mmol, 0.132 mL) was added, and the mixture was stirred for 20 min. The aldehyde (0.25 mmol) was added, and the reaction was stirred at room temperature for 72 h. The reaction vessel was placed over an external magnet until the reaction mixture became transparent, and the solution was separated from the nanoparticles. The mixture was filtered through a 1:1 silica/Celite pad with diethyl ether. The solvent was evaporated in vacuo, and the crude product was purified by flash chromatography (Hex/EtOAc, 80:20). The recovered nanoparticles were repeatedly washed with diethyl ether and 2-PrOH and then stored in 3 mL of 2-PrOH.

Detailed descriptions of homogeneous phase nitromethane addition products

Absolute configurations of the final alcohols were assigned by comparing the sign of the optical rotation or the retention time on HPLC chromatograms with the literature value. The following data are related to the use of ligand 2a using 2-PrOH.

(R)-1-(2-Chlorophenyl)-2-nitroethanol (4a)²³

ee = 71% (HPLC: Chiralpak IB, hexane/i-PrOH = 99:1, 1.3 mL/min, 220 nm, major 24.2 min and minor 25.8 min). ¹H NMR (300 MHz, CDCl₃): δ 7.66 (d, J = 7.0 Hz, 1H, Ph), 7.48–7.15 (m, 3H, Ph), 5.83 (dd, 1H, J = 9.6, 1.5 Hz, CHOH), 4.66 (dd, 1H, J = 13.4, 2.3 Hz, CH_aH_bNO₂), 4.43 (dd, 1H, J = 13.4, 9.6 Hz, CH_aH_bNO₂), 3.13 (bs, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 135.3, 131.5, 129.9, 129.7, 127.6, 127.5, 79.3, 67.8.

(R)-3-(1-Hydroxy-2-nitroethyl)benzonitrile (4b)²³

Obtained following the procedure in diethyl ether. [α]_D²⁰ – 26.2 (c 2.2, CHCl₃). ee = 54%. ¹H NMR (300 MHz, CDCl₃): δ 7.89–7.45 (m, 4H, Ph), 5.52 (dd, J = 8.2, 3.8 Hz, 1H, CHOH), 4.68–4.40 (m, 2H, CH₂NO₂), 3.40 (m, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 139.8, 132.5, 130.4, 129.8, 129.7, 118.2, 113.0, 80.8, 69.8.

(R)-4-(1-Hydroxy-2-nitroethyl)benzonitrile (4c)²³

ee = 66% (HPLC: Chiralpak IB, hexane/i-PrOH = 97:3, 1.0 mL/min, 220 nm, major 12.9 min and minor 14.2 min). ¹H NMR (300 MHz, CDCl₃): δ 7.71 (d, 2H, J = 8.3 Hz, Ph), 7.56 (d, 2H, J = 8.1 Hz, Ph), 5.57–5.52 (m, 1H, CHOH), 4.64–4.46 (m, 2H, CH₂NO₂), 3.13 (d, J = 4.0 Hz, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 143.1, 132.8, 126.7, 118.1, 112.8, 80.6, 70.1.

(R)-1-(2-Methoxyphenyl)-2-nitroethanol (4d)³⁰

ee = 85% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 0.8 mL/min, 273 nm, major 17.1 min and minor 19.5 min). ¹H NMR (300 MHz, CDCl₃): δ 7.52–7.26 (m, 2H, Ph), 7.10–6.86 (m, 2H, Ph), 5.70–5.55 (m, 1H, CHOH), 4.75–4.46 (m, 2H, CH₂NO₂), 3.88 (s, 3H, OCH₃), 3.47 (bs, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 155.3, 130.9, 130.5, 127.7, 126.0, 120.6, 79.7, 67.5, 55.2.

(R)-1-(3-methoxyphenyl)-2-nitroethanol (4e)

Obtained following the procedure in diethyl ether. ee = 0%. ¹H NMR (300 MHz, CDCl₃): δ 7.34–6.69 (m, 4H, Ph), 5.44 (dd, 1H, J = 9.4, 3.0 Hz, CHOH), 4.64–4.48 (m, 2H, CH₂NO₂), 3.82 (s, 3H, OCH₃), 2.85 (s, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 139.5, 130.1, 118.0, 114.40, 111.4, 110.2, 81.2, 70.9, 55.3.

(R)-1-(4-Methoxyphenyl)-2-nitroethanol (4f)²³

Obtained following the procedure in diethyl ether. [α]_D²⁰ – 33.3 (c 2.1, CHCl₃). ee = 79%. ¹H NMR (300 MHz, CDCl₃): δ 7.31 (d, 2H, J = 8.6 Hz, Ph), 6.91 (d, 2H, J = 8.7 Hz, Ph), 5.40 (dd, 1H, J = 9.5, 3.0 Hz, CHOH), 4.59 (dd, 1H, J = 13.2, 9.5 Hz, CH_aH_bNO₂), 4.47 (dd, 1H, J = 13.2, 3.2 Hz, CH_aH_bNO₂), 3.81 (s, 3H, OCH₃), 2.52 (s, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 160.0, 130.2, 127.2, 114.4, 81.2, 70.7, 55.3.

(R)-2-Nitro-1-(o-tolyl)ethanol (4g)³⁰

ee = 80% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 0.8 mL/min, 220 nm, major 16.2 min and minor 22.9 min). ¹H NMR (300 MHz, CDCl₃): δ 7.52 (d, 1H, J = 4.8 Hz, Ph), 7.28–7.18 (m, 3H, Ph), 5.68 (dd, 1H, J = 9.6, 2.1 Hz, CHOH), 4.55 (dd, 1H, J = 13.0, 9.6 Hz, CH_aH_bNO₂), 4.43 (dd, 1H, J = 13, 2.1 Hz, CH_aH_bNO₂), 2.38 (bs, 1H, OH), 2.39 (s, 3H, CH₃). ¹³C NMR (75 MHz, CDCl₃): δ 137.7, 134.4, 130.9, 128.7, 126.8, 125.6, 80.2, 67.9, 18.9.

(R)-2-Nitro-1-(m-tolyl)ethanol (4h)³⁰

ee = 57% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 1 mL/min, 220 nm, major 14.2 min and minor 15.6 min). ¹H NMR (300 MHz, CDCl₃): δ 7.32–7.18 (m, 4H, Ph), 5.42 (dd, J = 9.5, 3.0 Hz, 1H, CHOH), 4.61 (dd, 1H, J = 13.3, 9.6 Hz, CH_aH_bNO₂), 4.50 (dd, J = 13.3, 3.0 Hz, CH_aH_bNO₂), 2.37 (s, 3H, CH₃). ¹³C NMR (75 MHz, CDCl₃): δ 138.8, 138.02, 129.7, 128.9, 126.5, 122.9, 81.2, 71.0, 21.4.

(R)-3-Nitro-1-(p-tolyl)ethanol (4i)²³

ee = 63% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 1 mL/min, 220 nm, major 15.7 min and minor 18.5 min). ¹H NMR (300 MHz, CDCl₃): 7.29 (d, 2H, J = 8.1 Hz, Ph), 7.21 (d, 2H, J = 8.1 Hz, Ph), 5.43 (dd, 1H, J = 9.6, 3.0 Hz, CHOH), 4.61 (dd, 1H, J = 13.3, 9.6 Hz, CH_aH_bNO₂), 4.49 (dd, 1H, J = 13.3, 3.1 Hz, CH_aH_bNO₂), 2.72 (bs, 1H, OH), 2.33 (s, 3H, CH₃). ¹³C NMR (75 MHz, CDCl₃): δ 138.82, 133.60, 129.68, 125.77, 81.23, 70.76, 20.91.

(R)-2-Nitro-1-phenylethanol (4j)³⁰

ee = 68% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 1 mL/min, 220 nm, major 16.6 min and minor 18.9 min). ¹H NMR (300 MHz, CDCl₃): δ 7.47–7.32 (m, 5H, Ph), 5.47 (dd, 1H, J = 9.5, 3.1 Hz, CHOH), 4.61 (dd, 1H, J = 13.3, 9.5 Hz, CH_aH_bNO₂), 4.52 (dd, 1H, J = 13.3, 3.1 Hz, CH_aH_bNO₂), 2.56 (bs, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 138.3, 129.0, 128.7, 125.9, 81.3, 69.9.

(R)-1-(2-Fluorophenyl)-2-nitroethanol (4k)²³

ee = 46.8% (HPLC: Chiralpak IB, hexane/i-PrOH = 98:2, 1.2 mL/min, 220 nm, major 16.0 min and minor 16.5 min). ¹H NMR (300 MHz, CDCl₃): δ 7.56 (m, 1H, Ph), 7.39–7.32 (m, 1H, Ph), 7.22 (m, 1H, Ph), 7.09 (m, 1H, Ph), 5.75 (dd, 1H, J = 8.5, 3.2 Hz, CHOH), 4.64 (dd, 1H, J = 13.7, 3.6 Hz, CH_aH_bNO₂), 4.62–4.56 (m, 1H, CH_aH_bNO₂), 3.0 (bs, 1H, OH). ¹³C NMR (75 MHz, CDCl₃): δ 130.5, 130.4, 127.5, 124.8, 115.7, 115.4, 79.7, 65.4.

(R)-2-Nitro-1-(3-nitrophenyl)ethanol (4l)³⁰

ee = 66% (HPLC: Chiralpak IB, hexane/i-PrOH = 90:10, 1.2 mL/min, 220 nm, major 15.9 min and minor 17.3 min). ¹H NMR (300 MHz, CDCl₃): δ 8.32 (s, 1H, Ph), 8.22 (d, J = 7.6 Hz, 1H, Ph), 7.77 (d, 1H, J = 7.6 Hz, Ph), 7.61 (t, 1H, J = 7.6 Hz, Ph), 5.61 (dd, 1H, J = 8.1, 4.2 Hz, CHOH), 4.68–4.54 (m, 2H, CH₂NO₂), 2.95 (bs, 1H, OH). ¹³C NMR (100 MHz, CDCl₃): δ 190.1, 140.4, 132.1, 130.3, 123.9, 121.3, 80.8, 70.0.

(R)-1-Cyclohexyl-2-nitroethanol (4m)³⁰

ee = 78.4% (HPLC: Chiralpak IB, hexane/i-PrOH = 99:1, 1.2 mL/min, 220 nm, major 12.5 min and minor 13.0 min). ¹H NMR (400 MHz, CDCl₃): δ 4.48 (dd, 1H, J = 13.1, 2.9 Hz, CH_aH_bNO₂), 4.42 (dd, 1H, J = 13.1, 8.9 Hz, CH_aH_bNO₂), 4.10 (ddd, 1H, J = 8.9, 6.0, 2.9 Hz, CHOH), 2.29 (bs, 1H, OH), 2.02–1.58 (m, 5H, c-Hex), 1.57–0.94 (m, 6H, c-Hex). ¹³C NMR (100 MHz, CDCl₃): δ 79.6, 73.1, 41.7, 29.1, 28.3, 26.4, 26.2, 26.1.

(1R,2R)-2-Nitro-1-(o-tolyl)propan-1-ol (5a)

ee = 74.6% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 1 mL/min, 220 nm, major 8.5 min and minor 9.6 min). ¹H NMR (400 MHz, CDCl₃): δ 7.41–7.35 (m, 1H, Ph), 7.30–7.17 (m, 3H, Ph), 5.37 (d, 1H, J = 9.3 Hz, CHOH), 4.86 (dq, J = 9.3, 6.9 Hz, 1H, CHNO₂), 2.45 (s, 3H, CH₃-Ph), 2.29 (bs, 1H, OH), 1.33 (d, 3H, J = 6.9 Hz, CH(NO₂)CH₃). ¹³C NMR (100 MHz, CDCl₃): δ 136.9, 136.2, 131.4, 129.2, 127.2, 126.9, 89.1, 72.6, 19.9, 16.4.

(1R,2S)-2-Nitro-1-(o-tolyl)propan-1-ol (5b)

ee = 45% (HPLC: Chiralpak IB, hexane/i-PrOH = 95:5, 1 mL/min, 220 nm, major 10.3 min and minor 11.3 min). ¹H NMR (400 MHz, CDCl₃): δ 7.57–7.50 (m, 1H, Ph), 7.30–7.15 (m, 3H, Ph), 5.62 (d, 1H, J = 3.0 Hz, CHOH), 4.64 (dq, 1H, J = 6.8, 3.1 Hz, CHNO₂), 2.49 (bs, 1H, OH), 2.38 (s, 3H, CH₃-Ph), 1.52 (d, J = 6.8 Hz, 3H, CH(NO₂)CH₃). ¹³C NMR (100 MHz, CDCl₃): δ 137.0, 134.6, 131.1, 128.7, 126.8, 126.3, 85.7, 71.2, 19.1, 11.9.

(1R,2R)-1-Cyclohexyl-2-nitropropan-1-ol (6a)³¹

ee = 74.2% (HPLC: Chiralpak IA, hexane/i-PrOH = 99:1, 1.2 mL/min, 220 nm, minor 20.8 min and major 35.1 min). ¹H NMR (400 MHz, CDCl₃): δ 4.71 (p, 1H, J = 6.9 Hz, CHNO₂), 3.65 (dd, 1H, J = 7.0, 4.8 Hz, CHOH), 2.13 (bs, 1H, OH), 1.88–1.61 (m, 5H, c-Hex), 1.56 (d, 3H, J = 6.9 Hz, CH₃), 1.45–1.35 (m, 1H, c-Hex), 1.35–0.98 (m, 5H, c-Hex). ¹³C NMR (100 MHz, CDCl₃): δ 85.8, 77.3, 40.2, 30.2, 26.5, 26.4, 26.4, 26.1, 16.8.

(1R,2S)-1-Cyclohexyl-2-nitropropan-1-ol (6b)²⁴

ee = 42.2% (HPLC: Chiralpak IA, hexane/i-PrOH = 99:1, 1.2 mL/min, 220 nm, major 22.1 min and minor 25.2 min). ¹H NMR (400 MHz, CDCl₃): δ 4.64 (dq, 1H, J = 6.9, 3.2 Hz, CHNO₂), 3.94 (dd, 1H, J = 8.3, 3.1 Hz, CHOH), 2.13 (bs, 1H, OH), 1.88–1.61 (m, 5H, c-Hex), 1.54 (d, 3H, CH₃, J = 7.1 Hz,), 1.45–1.35 (m, 1H, c-Hex), 1.35–0.98 (m, 5H, c-Hex). ¹³C NMR (100 MHz, CDCl₃) δ 84.6, 76.6, 40.5, 29.3, 29.2, 26.4, 26.4, 25.9, 12.2.

Synthesis of Magnetite–Silica Core–Shell Nanoparticles

Mohr salt ((NH₄)₂Fe(SO₄)₂·6H₂O) (3.82 mmol, 1.5 g), Fe₂(SO₄)₃ (3.82 mmol, 1.5 g), and polyvinylpyrrolidone (PVP) (19 mmol, 2.12 g) were dissolved in 200 mL of distilled water. The Fe²⁺/Fe³⁺ solution thus prepared was filtered and added into a filtered solution of 19 mmol (2.12 g) of PVP and 150 mL of NH₃ (33 wt % aqueous solution) in 500 mL of water under mechanical stirring and continuously flowing argon gas. The solution was reacted for 30 min at room temperature, and then it was heated to 80 °C for 30 min. After cooling down the reaction mixture, the product was separated by magnetic decantation and dialyzed overnight. The magnetite nanoparticles were dispersed in 30 mL of water by sonication, and a solution of 0.4 mL of tetraethylortosilicate (TEOS) and 3.25 mL of in NH₃ (33 wt % aqueous solution) in 150 mL of 2-PrOH was added dropwise under vigorous mechanical stirring. The reaction was stirred at room temperature for 3 h, and the product was separated by magnetic decantation and repeatedly washed with 2-PrOH and water. The as-synthesized silica-coated magnetite nanoparticles were stored in 20 mL of water. Obtained: 770 mg. Diameter: 10–15 nm determined by TEM images. HR-TEM images showed lattice fringes of both magnetite (Fe₃O₄) and maghemite (γ-Fe₂O₃) phases. FTIR (neat/ν cm^–1): 3341, 1634, 1049, 965, 813, 551.

Synthesis of 3-Azidopropyltrimethoxysilane

In a flame-dried pear-shaped flask, 4.5 mmol (0.88 mL) of 3-iodopropyltrimethoxysilane was dissolved under an inert atmosphere in 15 mL of anhydrous DMSO. NaN₃ (9 mmol) was added, and the reaction was stirred at 60 °C for 12 h. A 10 mL solution of hexane was injected under argon gas, and the mixture was vigorously stirred for 1 h. The hexane layer was then collected via cannula, and this procedure was repeated twice. The combined hexane layers were concentrated in vacuo under an inert atmosphere to give 766 mg of the product (83%). ¹H NMR (300 MHz, CDCl₃): δ 3.57 (s, 9H, OCH₃ x3), 3.26 (t, 2H, J = 7.0 Hz, N₃CH₂), 1.71 (m, 2H, CH₂CH₂Si), 0.69 (m, 2H, CH₂CH₂Si). ¹³C NMR (75 MHz, CDCl₃): δ 54.0, 50.8, 22.7, 6.5. FTIR (neat/ν cm^–1): 2941, 2840, 2092, 1080, 809.

Surface Modification with Azidosilane (9a)

Silica-coated magnetite nanoparticles (750 mg) were dispersed in 36 mL of deionized water. A solution of 2.6 mL of 3-azidopropyltrimethoxysilane in 80 mL of EtOH was added under mechanical stirring. A 64 μL solution of NH₃ (33 wt % aqueous solution) was added, and the reaction was mechanically stirred for 3 h. The product was separated by magnetic decantation and repeatedly washed with 2-PrOH and water. The azido-modified silica-coated magnetite nanoparticles were stored in 20 mL of water. Obtained: 760 mg. Loading: 0.44 mmol/g calculated by elemental analysis: N, 1.87%; C, 2.16%. FTIR (neat/ν cm^–1): 3394, 2928, 2868, 2100, 1627, 1444, 1030, 807, 551.

End-Capping of the Free Silanols (9b)

Azido-functionalized silica-coated magnetic nanoparticles (390 mg) were dispersed in 10 mL of toluene. A solution of 9.5 mmol (2 mL) of hexamethyldisilazane in 8 mL of toluene was then added under mechanical stirring. The solution was heated to 110 °C and stirred at this temperature for 3 h. The product was separated by magnetic decantation and repeatedly washed with toluene and acetone. The dispersibility of the as-synthesized nanoparticles in organic solvents appeared remarkably increased. The silylated azido-modified silica-coated magnetite nanoparticles were stored in 20 mL of toluene. Obtained: 290 mg. Elemental analysis: N, 1.89%; C, 3.52%. FTIR (neat/ν cm^–1): 3421, 2948, 2094, 1627, 1444, 1260, 1030, 846, 800, 551.

Immobilization via Click Reaction

The azido-functionalized silica-coated magnetic nanoparticles 9a or 9b (150 mg) were dispersed in 10 mL of THF. The terminal alkyne 8 (0.3 mmol, 140 mg) and CuI (0.09 mmol, 19 mg) were added under mechanical stirring. A 0.4 mL solution of DIPEA was injected dropwise into the mixture, and the reaction was mechanically stirred for 48 h at room temperature. The product was separated by magnetic decantation and repeatedly washed with THF and then with NH₄Cl s.s., water, acetone, and toluene. The functionalized silica-coated magnetite nanoparticles were stored in 20 mL of toluene. 7a, obtained: 200 mg. Loading: 0.28 mmol/g calculated by elemental analysis: N, 1.95%; C, 8.69%. FTIR (neat/ν cm^–1): 3421, 2934, 2849, 1510, 1056, 551. 7b, obtained: 280 mg. Loading: 0.41 mmol/g calculated by elemental analysis: N, 2.89%; C, 13.94%. FTIR (neat/ν cm^–1): 3315, 2928, 2849, 1614, 1510, 1450, 1246, 1037, 840, 551.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.9b02683.

• NMR spectra of all compounds, HPLC chromatograms of nitroalcohols, and ATR-FTIR spectra of synthetized nanoparticles (PDF)

Author Present Address

^○ Present address: Di Renzo Regolatory Affairs, via dell’Arco di Travertino 11, 00178, Rome.

Author Present Address

^∇ Present address: LASER LAB s.r.l. Roma, via Camerata Picena 385, 00138 Rome.

Author Present Address

^# Present address: Department of Pharmaceutical, Chemical and Environmental Sciences, Faculty of Engineering and Science, University of Greenwich, Central Avenue, Chatham -maritime ME4 4TB, U.K.

Author Present Address

^⊥ Present address: Bsp Pharmaceuticals S.p.A., Via Appia, 04013 Latina, Italy.

Author Contributions

^◆ C.S. and L.P. contributed equally to this work.

The authors declare no competing financial interest.