Copper(II)-Mediated Iodination of 1-Nitroso-2-naphthol

The 3-Iodo-1-nitrosonaphthalene-2-ol (I-NON) was obtained by the copper(II)-mediated iodination of 1-nitroso-2-naphthol (NON). The suitable reactants and optimized reaction conditions, providing 94% NMR yield of I-NON, included the usage of Cu(OAc)2·H2O and 1:2:8 CuII/NON/I2 molar ratio between the reactants. The obtained I-NON was characterized by elemental analyses (C, H, N), high-resolution ESI+-MS, 1H and 13C{1H} NMR, FTIR, UV-vis spectroscopy, TGA, and X-ray crystallography (XRD). The copper(II) complexes bearing deprotonated I-NON were prepared as follows: cis-[Cu(I-NON–H)(I-NON)](I3) (1) was obtained by the reaction between Cu(NON-H)2 and I2 in CHCl3/MeOH, while trans-[Cu(I-NON–H)2] (2) was synthesized from I-NON and Cu(OAc)2 in MeOH. Crystals of trans-[Cu(I-NON–H)2(THF)2] (3) and trans-[Cu(I-NON–H)2(Py)2] (4) were precipitated from solutions of 2 in CHCl3/THF and Py/CHCl3/MeOH mixtures, respectively. The structures of 1 and 3–4 were additionally verified by X-ray crystallography. The characteristic feature of the structures of 1 and 3 is the presence of intermolecular halogen bonds with the involvement of the iodine center of the metal-bound deprotonated I-NON. The nature of the I···I and I···O contacts in the structures of 1 and 3, correspondingly, were studied theoretically at the DFT (PBE0-D3BJ) level using the QTAIM, ESP, ELF, NBO, and IGM methods.


Introduction
Nitrosonaphthols, typically existing in the quinone monoxime tautomeric form (see ESI, Scheme S1) [1], represent a group of compounds that found a wide range of useful applications. These species have been applied as precursors in organic synthesis [2], they were used for the analytical determination of bioorganic species [3,4], and they also exhibit some biological activity modes (e.g., carcinogenic) as iminoxyl and hydronitroxide radical source [5]. Nitrosonaphthols with ortho-positioned functional groups, i.e., NO(H) and O(H), exhibit chelating properties and they found application as analytical reagents for determination [6][7][8], extraction [7][8][9], and sorption [10] of metal ions and also can be applied as sequestering agents for nanoparticle preparation [11]. The o-nitrosonaphthols were also used for the preparation of immobilized reagents for optical metal ion sensing [12] and polymer-supported catalysts [13].
A search of synthetically significant routes for the modification and functionalization of nitrosonaphthols could be an important task for the improvement of the analytical properties of these and structurally related species. In particular, initial modification of nitrosonaphthols can include introduction of a halogen atom, e.g., by direct halogenation. The direct chlorination and bromination of 1-nitroso-2-naphthol by the corresponding dihalogens to afford 3-halo-1-nitroso-2-naphthols have been reported [14,15], while the The direct chlorination and bromination of 1-nitroso-2-naphthol by the corresponding dihalogens to afford 3-halo-1-nitroso-2-naphthols have been reported [14,15], while the iodination is a more complicated task and the iodinated product was obtained only by the exchange reaction of the corresponding brominated derivative [15]. Indeed, the halogenation of aromatic species by X2 strongly depends on the identity of a halogen and, in the case of less reactive I2, this reaction commonly requires Lewis acid catalysis, e.g., application of copper(II) salts, and it was conducted mainly for electron-rich arenes [16][17][18]. In this work, we report on a novel metal-mediated reaction that proceeds in the system Cu II /1-nitroso-2-naphthol/I2 and leads, after liberation of the ligand, to uncomplexed 3iodo-1-nitrosonaphthalene-2-ol. We also succeeded in obtaining several copper(II) complexes with the iodinated product, and all our results and findings are detailed in sections that follow.  (1) complex bearing the iodinated ligands (for detailed characterization and relevant discussion, see below). We applied this reaction as a starting point to develop a synthetic route to I-NON (Scheme 1). On the one hand, I-NON, as a representative of the group of nitrosonaphthols, can be useful as a chelator for transition metal complexes and this sequestering can be applied for analytic or extraction purposes [19,20]; on the other hand, the iodine functionality in I-NON, acting as a halogen bond (HaB) donor, might be useful for HaB-based crystal engineering [21][22][23][24][25].

Scheme 1. Synthesis of I-NON via Cu II -mediated iodination of NON.
Our procedure consists, on the first step, in one-pot reaction of NON with a copper(II) salt and I2 followed by, on the second step, the liberation of I-NON by the treatment of the reaction mixture with an excess Na2S2O3. Aiming to develop the two-step synthetic procedure to I-NON, we performed the optimization of the reaction conditions (Table 1) by varying copper(II) and some other metal salts, load of copper(II) salts, and variation of molar ratios between the reactants. We tested several copper(II) sources (Table 1, entries 1-5) for the system Cu II /NON/I2 (1:2:2 molar ratio) and found that the application of CuCl2·2H2O gave the highest yield of I-NON (38%), while the involvement of Cu(OAc)2·H2O produced the best ratio between the product and the unreacted starting materials (2:1). Therefore, for further optimization, we chose Cu(OAc)2·H2O. Variation of the I2 load (Table 1, entries 5-8) revealed that the highest yield was achieved with 1:4 NON/I2 molar ratio. We also attempted the reaction under the catalytic conditions (Table  1, entries 5 and 9, 7 and 10), but observed the I-NON yield drop on decreasing the load of Cu(OAc)2·H2O. This experiment indicates that the reaction required the stochiometric amount of copper(II). Other tested metal salts were less effective ( Our procedure consists, on the first step, in one-pot reaction of NON with a copper(II) salt and I 2 followed by, on the second step, the liberation of I-NON by the treatment of the reaction mixture with an excess Na 2 S 2 O 3 . Aiming to develop the two-step synthetic procedure to I-NON, we performed the optimization of the reaction conditions (Table 1) by varying copper(II) and some other metal salts, load of copper(II) salts, and variation of molar ratios between the reactants. We tested several copper(II) sources (Table 1, entries 1-5) for the system Cu II /NON/I 2 (1:2:2 molar ratio) and found that the application of CuCl 2 ·2H 2 O gave the highest yield of I-NON (38%), while the involvement of Cu(OAc) 2 ·H 2 O produced the best ratio between the product and the unreacted starting materials (2:1). Therefore, for further optimization, we chose Cu(OAc) 2 ·H 2 O. Variation of the I 2 load (  [11][12][13] in the reaction. Notably, the iodination of NON can proceed under metal-free conditions ( Table 1, entry 15), but, in contrast to the optimized Cu II -mediated reaction, this synthesis gave only ca. 24% conversion to I-NON. An increase of the reaction time to 3 d did not affect the conversion of the starting materials. In summary, the results of these tests demonstrated that the optimal conditions for this copper(II)-involving reaction included the usage of Cu(OAc) 2 ·H 2 O and 1:2:8 Cu II /NON/I 2 molar ratios between the reactants. Although the reaction, as we found, could be carried out without the metal source (ca. 25% conversion, as verified by GC-MS monitoring; the isolation was not performed), the employment of copper(II) led to nearly quantitative conversion of NON to I-NON (94%; entry 8). Based on all these observations, one can conclude that the iodination of NON proceeded in the coordination sphere of Cu II , while the optimal reactant ratio Cu II /NON (1:2) corresponded to the coordination of NON to give {Cu(NON-H) 2 } species, which then converted to the iodination product, bearing the {Cu(I-NON-H) 2 } moiety. Therefore, Na 2 S 2 O 3 on the second step of the synthetic scheme can play the dual role: It was required for the neutralization of the excess of I 2 and, simultaneously, it liberated I-NON from the formed complex via the Cu II -to-Cu I reduction, followed by the complexation of the latter with Iand the precipitation of CuI.
After the optimization, we also attempted to extend this copper(II)-mediated reaction to other substrates, such as 2-nitroso-1-naphthol or 1,2-nitrosophenol; however, only mixtures of yet unidentified products were obtained and any iodination product was not detected by 1 H NMR or HRESI + -MS.
We verified the role of metal center control on the reactivity of the deprotonated NON ligand as compared to uncomplexed NON, which is, at least, dual. Firstly, it was protecting the oxime nitrogen from the well-known oxidative deoximation (ref. [30] and ref erences therein) and, secondly, the copper site stabilized the oxo(1-)-substituent in the benzene ring directing the iodination in the ortho-position. Conventional activation of I by a Lewis acid should also be taken into account.  Table 2).  Thus, we developed a simple and facile approach to I-NON starting from NON. We found only one study focused on the synthesis of I-NON. The latter included the reaction of 3-bromo-1-nitrosonaphthalen-2-ol (Br-NON) with excess KI to give I-NON in 60% yield [28]. In turn, Br-NON can be obtained by the bromination of NON with Br 2 [29]. Our procedure gave a significantly higher yield of I-NON (94%) and did not require the usage of the harmful Br 2 .

Copper(II) Complexes
We verified the role of metal center control on the reactivity of the deprotonated NON ligand as compared to uncomplexed NON, which is, at least, dual. Firstly, it was protecting the oxime nitrogen from the well-known oxidative deoximation (ref. [30] and references therein) and, secondly, the copper site stabilized the oxo(1-)-substituent in the benzene ring directing the iodination in the ortho-position. Conventional activation of I 2 by a Lewis acid should also be taken into account.  Thus, we developed a simple and facile approach to I-NON starting from NO found only one study focused on the synthesis of I-NON. The latter included the re of 3-bromo-1-nitrosonaphthalen-2-ol (Br-NON) with excess KI to give I-NON in 60% [28]. In turn, Br-NON can be obtained by the bromination of NON with Br2 [29]. Ou cedure gave a significantly higher yield of I-NON (94%) and did not require the us the harmful Br2.

Halogen Bonding Interactions Involving Ligated I-NON
Organoiodine atoms from I-NON-H and I-NON ligands form HaBs [21] with iodine atoms of the I 3 − ligands. In 3, I···O HaBs were detected between I-NON-H ligands. In all structures, several intermolecular hydrogen bonds, e.g., H aryl ···O, H aryl ···I, were also recognized. In view of the importance of HaB interactions in crystal engineering and to verify s-hole properties of the ligated I-NON, we studied HaBs in the structures of 1 and 3 in more detail by computational methods.
The nature of the HaBs in 1 and 3 was studied theoretically at the DFT (PBE0-D3BJ) level using the quantum theory of atoms in molecules (QTAIM), electrostatic potential (ESP), electron localization function (ELF), natural bond orbital (NBO), and independent gradient model (IGM) analyses. The calculations were carried out for bimolecular clusters BM1 1 , BM1 2 , BM3 1 , BM3 2 , and BM3 3 with geometries corresponding to the XRD structures of 1 and 3 ( Figure 5, see Computational Details). In 1, two I···I HaBs formed by the organoiodine atoms of the I-NON-H and I-NON ligands were recognized, i.e., HaB(I···I)1 and HaB(I···I)2 ( Table 3). The former bond clearly belongs to type II [27] interactions, with the I4-I3···I1 and C-I1···I3 angle being 135.9 and 178.8 • , respectively. Classification of HaB(I···I)2 based on geometrical parameters is not so straightforward since the C-I2···I4 angle is 150.6 • . However, the ESP, ELF, and NBO analyses clearly indicate that this bond is also of type II with the I2 atom serving as an HaB donor. The low value of the C-I2···I4 angle is accounted for by packing effects, which force the deviation of this HaB from the optimal directionality. In 3, four I···O bondings were detected, i.e., HaB(I···O)1-HaB(I···O)4 (the C-I···O and N-O···I angles being 138.1-153.0 • and 142.7-160.9 • , respectively). Geometry of these interactions resembles the halogen···halogen type I [27] interactions.  The QTAIM analysis revealed bond critical points (BCPs) for all these contacts (Figure 6). The calculated values of electron density, ρb, its Laplacian,  2 ρb, potential, and kinetic energy densities (Vb and Gb) at the BCP are typical for halogen bonds of weak-tomedium strength [36][37][38][39] (Table 4). These parameters are lower for HaB(I···I)1 and HaB(I···O)4 correlating with the longer I···I and I···O contacts in these bonds. Meanwhile, all these bondings have an attractive nature, as indicated by the negative sign of the second eigenvalue of the Hessian matrix at the BCP, λ2,b (Table 4), and confirmed by the IGM plots of the sign(λ2)ρ(r) function mapped on the isosurface of the δg inter descriptor ( Figure  6). Table 4. Halogen bond lengths, d (in Å), calculated electron density, ρb, its Laplacian,  2 ρb, potential and kinetic energy densities, Vb and Gb, second eigenvalue of the hessian matrix, λ2,b, at BCPs (in a.u.) and interaction energies, Eint (in kcal/mol).  The QTAIM analysis revealed bond critical points (BCPs) for all these contacts ( Figure 6). The calculated values of electron density, ρ b , its Laplacian, ∇ 2 ρ b , potential, and kinetic energy densities (V b and G b ) at the BCP are typical for halogen bonds of weak-to-medium strength [36][37][38][39] (Table 4). These parameters are lower for HaB(I···I)1 and HaB(I···O)4 correlating with the longer I···I and I···O contacts in these bonds. Meanwhile, all these bondings have an attractive nature, as indicated by the negative sign of the second eigenvalue of the Hessian matrix at the BCP, λ 2,b (Table 4), and confirmed by the IGM plots of the sign(λ 2 )ρ(r) function mapped on the isosurface of the δg inter descriptor ( Figure 6).  Figure 7A). Despite the C-I2···I4 angle in 1 is significantly lower than 180°, a belt of the negative ESP around the I4 atom is still directed towards the -hole of I2 providing the electrostatic interpretation of this bond. The ELF analysis also demonstrates that the maximum function value for the monosynaptic basin of the I4 atom is directed towards the ELF minimum at the I2 atom ( Figure 7B The interaction energies of the I···I and I···O bonds were estimated using two approaches (Eint(BSSE) and Eint(S); see Computational details). The Eint values of the individual bonds are within the range of −4.0 to −10.4 kcal/mol that is typical for HaBs of weak to medium strength, with the I···I bonds being slightly more stable than the I···O ones (Table  4).  Table 4. Halogen bond lengths, d (in Å), calculated electron density, ρ b , its Laplacian, ∇ 2 ρ b , potential and kinetic energy densities, V b and G b , second eigenvalue of the hessian matrix, λ 2,b , at BCPs (in a.u.) and interaction energies, E int (in kcal/mol).

Cluster
HaB   Figure 7A). Despite the C-I2···I4 angle in 1 is significantly lower than 180 • , a belt of the negative ESP around the I4 atom is still directed towards the σ-hole of I2 providing the electrostatic interpretation of this bond. The ELF analysis also demonstrates that the maximum function value for the monosynaptic basin of the I4 atom is directed towards the ELF minimum at the I2 atom ( Figure 7B).

Discussion
We developed a high-yielding facile route toward I-NON: This formed via copper(II)-mediated iodination of NON with I2. Among v metal salts, copper(II) salts, in particular Cu(OAc)2·H2O, were found to b tors of this reaction. The metal center played multiple roles in this iodin stabilized NON in a deprotonated nitrosophenolate form by the coordi venting oxidative deoximation; the anionic aromatic NON-H − form is p matic electrophilic substitution involving such halogen as iodine. In ad activated I2 as an electrophilic agent [16]. Unlike Cu II -mediated iodinatio matics, the NON iodination product was formed in the coordination sp fore, an additional step of the formed ligand liberation, by the reductio was required.
We also employed the prepared I-NON as a chelator and four ne copper(II) complexes were obtained. The geometry of these complexes presence of I-NON ligands in deprotonated form and the ligand environ

Materials and Instrumentations
The 1-Nitroso-2-naphthol, copper(II) salts, I2, Na2S2O3, and solven from a commercial source and used as received. The HRESI mass spect on a Bruker micrOTOF spectrometer equipped with an electrospray ioniz MeOH was employed as the solvent. The instrument was operated in p using an m/z range of 50-3000. The capillary voltage of the ion source w (ESI + MS) and the capillary exit ±(70-150) V. In the isotopic pattern, th The interaction energies of the I···I and I···O bonds were estimated using two approaches (E int (BSSE) and E int (S); see Computational details). The E int values of the individual bonds are within the range of −4.0 to −10.4 kcal/mol that is typical for HaBs of weak to medium strength, with the I···I bonds being slightly more stable than the I···O ones (Table 4).

Discussion
We developed a high-yielding facile route toward I-NON: This compound was formed via copper(II)-mediated iodination of NON with I 2 . Among various transition metal salts, copper(II) salts, in particular Cu(OAc) 2 ·H 2 O, were found to be the best promotors of this reaction. The metal center played multiple roles in this iodination: Copper(II) stabilized NON in a deprotonated nitrosophenolate form by the coordination, thus preventing oxidative deoximation; the anionic aromatic NON-H − form is preferable for aromatic electrophilic substitution involving such halogen as iodine. In addition, copper(II) activated I 2 as an electrophilic agent [16]. Unlike Cu II -mediated iodination of various aromatics, the NON iodination product was formed in the coordination sphere and, therefore, an additional step of the formed ligand liberation, by the reduction using Na 2 S 2 O 3, was required. We

Materials and Instrumentations
The 1-Nitroso-2-naphthol, copper(II) salts, I 2 , Na 2 S 2 O 3 , and solvents were obtained from a commercial source and used as received. The HRESI mass spectra were obtained on a Bruker micrOTOF spectrometer equipped with an electrospray ionization source and MeOH was employed as the solvent. The instrument was operated in positive ion mode using an m/z range of 50-3000. The capillary voltage of the ion source was set at −4500 V (ESI + MS) and the capillary exit ±(70-150) V. In the isotopic pattern, the most intensive peak was reported. Infrared spectra were recorded using a Bruker FTIR TENSOR 27 instrument in KBr pellets. The absorption spectra were recorded on a Shimadzu UV 1800 spectrophotometer in MeCN. The 1 H and 13 C{ 1 H} NMR spectra were measured on a Bruker Avance III 400 spectrometer at ambient temperature. Residual solvent signals were used as the internal standard. Microanalyses (C, H, N) were carried out on a Euro EA3028-HT instrument. The TGA studies were performed on a NETZSCH TG 209 F1 Libra thermoanalyzer and MnO 2 powder was used as a standard. The initial weights of the samples were in the range of 1.1-2.3 mg. The experiments were run in an open aluminum crucible in a stream of argon at a heating rate of 10 K/min. The final temperature was 610 • C. Processing of the thermal data was performed with Proteus analysis software. The powder diffraction experiments for 1 were carried out using D2Phaser diffractometer (Bruker), Cu anode, at 30 kV and 10 mA, and CuKα 1+2 radiation λ CuKα1 = 1.54059 Å and λ CuKα2 = 1.54443 Å, from 6 • to 60 • on the 2θ scale at 1 s/step with step size 0.02 • .

X-ray Structure Determinations
The XRD experiments were carried out using Oxford Diffraction "Xcalibur" diffractometer with monochromated MoKα radiation. The crystals were thermostated at 100 K throughout the all-experiment time. The structures were solved by ShelXT [40] and Superflip [41] structure solution programs using Intrinsic Phasing and Charge Flipping methods, respectively, and refined using ShelXL [40] minimization program incorporated in Olex2 [42] program package. Empirical absorption correction was accounted by CrysAl-isPro (Agilent Technologies, 2013) using spherical harmonics, implemented in SCALE3 ABSPACK scaling algorithm. The crystallographic data were deposited in the Cambridge Crystallographic Data Centre under the deposition codes CCDC 2076962-2076965 and can be obtained free of charge via the Internet, URL http://www.ccdc.cam.ac.uk/structures/ (accessed on 17 September 2021).

[Cu(I-NON-H)(I-NON)](I 3 ) (1)
. Cu(C 10 H 6 NO 2 ) 2 (0.025 mmol) was added to a solution of I 2 (0.2 mmol) in CHCl 3 /MeOH (10 mL, 1/1 v/v) placed in a 20-mL, round-bottomed flask. The reaction mixture was stirred at RT under ultrasonic treatment until the homogenization was complete and then left to stand at RT for slow evaporation.

Computational Details
The calculations for the QTAIM, ESP, NBO, and IGM analyses were carried out using the crystallographic coordinates at the DFT level of theory with the PBE0 functional [43,44] and the atom-pairwise dispersion correction with the Becke-Johnson damping scheme D3BJ [44,45]. The Gaussian 09 program package [46] was used. Cartesian d and f basis functions (6d, 10f) were used in all calculations. The DZP-DKH basis set for non-iodine atoms [47][48][49][50][51] and the ADZP-DKH basis set for the iodine atoms were applied. The latter basis set was constructed from the DZP-DKH basis set by addition of diffuse functions taken from the ADZP basis set. The Douglas−Kroll−Hess second-order scalar relativistic correction (DKH) [52,53] was applied. An ultrafine integration grid was used for numerical integrations.
Five bimolecular clusters (BM1 1 , BM1 2 , BM3 1 , BM3 2, and BM3 3 ) with geometries corresponding to the XRD structures of 1 and 3 and triplet spin state were used in the calculations. The C-H bonds were fixed at 1.09 Å and the O-H bonds were fixed at 1.0 Å.
The structure of the monomeric complex M1 was optimized at the PBE0-D3BJ/DZP(I-ADZP) level and the energies and ESP values were refined at the PBE0-D3BJ/DZP-DKH(I-ADZP-DKH) level.
The topological analyses of the electron density distribution with the help of the AIM method of Bader [54] were performed using the program AIMAll [55] while the IGM analysis [56,57] was performed using the Multiwfn 3.8 [58] and VMD [59] software. The bond orbital nature was analyzed by using the natural bond orbital (NBO) partitioning scheme [60].
The interaction energies of the I···I and I···O bonds were estimated using two approaches. In the first one, the interaction energy [E int (BSSE)] was calculated as the difference of total energy of dimer and the sum of the energies of monomers with unrelaxed geometries with the corresponding basis set superposition error (BSSE) correction: E int (BSSE) = (E(dimer) − E(monomer1) − E(monomer2) + BSSE)/n where n = 1 for the I···I bonds and n = 2 for the I···O bonds. BSSE was estimated using the counterpoise method [61,62].
In the second approach, the interaction energy [E int (S)] was calculated as the energy difference of two dimers: In dimer1, the HaB donor I atom (I1 or I2) forming the halogen bond was substituted by the hydrogen atom and the length of this C-H bond was fixed at 1.09 Å. In dimer2, the equivalent I atom not forming the HaB was replaced by the H atom (see Figure S1 in Supplementary Information for the structures of dimer1 and dimer2).

Supplementary Materials:
The following are available online: scheme with tautomeric forms of 1-nitrosonaphthalene-2-ol; characterization of I-NON, 1 and 2; HRESI-MS, IR, UV-vis absorption, 1 H and 13 C{ 1 H} NMR spectra, TG and DTG curves, and powder and single-crystal X-ray diffraction data for synthesized species; crystal data and structure refinement for I-NON, 1, 2, and 4; views of the fragment of molecular packing of I-NON, 1, and 3, demonstrating noncovalent contacts.