Camphor Sulfonimine Compounds: Bottom-Up Design of MOF’s from Organic Frameworks Based on X-rayS and DFT-D3

1. Introduction

The molecular structure of coordination compounds is formed by organic and inorganic moieties that combine to form new metal-ligand interactions. The formation of the new compounds can involve a major change into the pre-existing structural arrangement of the ligand or the metal precursors or, be shaped through binding of one of them, into the pre-organized structures of the other without a major reorganization.

In the case of complexes based on neutral ligands, covalent dative bonds are established involving the ligand heteroatoms lone pair (e.g. N, O, S) without change in the metal oxidation state or modification of the ligand composition. That is the case of camphor sulfonimines (RNC₁₀H₁₃NSO₂) that typically bind the metal through the nitrogen lone pair (=NSO₂) forming dative bonds that in some cases are reinforced by Van der Waals interactions (non-bonding interactions, NBI) involving the imine nitrogen atom (=NR) [1,2,3].

Pre-existing intermolecular NBI in the organic or inorganic scaffolds can be maintained or be broken within the formation of the coordination compound. If the NBI remain, the coordination compound eventually acquires a polymeric character. NBI play a significant role in the structural, reactivity and selectivity of the coordination compounds [4,5] and are referred as being the primary transport mechanism by which thermal conductivity is increased in cross-linked polymers [6].

In some cases a structure/reactivity relationship can be established in complexes [7,8] allowing properties prediction based on the structure, e.g. biological and/or catalytic properties of camphor sulfonimine coordination compounds. [3,9] Unfortunately, the structural analysis of coordination compounds based on X-ray diffraction is sometimes impossible due to no suitable crystal are obtained. However, if X-ray data on the organic moieties exists, it is possible to predict through computational calculations the structure of the derived coordination compounds. Based on the structural arrangement of a set of camphor sulfonimines, some insights into the structural arrangement of several coordination compounds frameworks as well as the electronic or packing effects being predominant in the camphor sulfonimines.

2. Results and discussion

Two new coordination compounds [Ag(NO₃)L₂] (C1: L⁶ and C2: L⁷) were synthesized (see experimental) from reaction of AgNO₃ with camphor sulfonimines: 3,5-Me₂C₆H₃C₁₀H₁₃SO₂ (L⁶) and C₆H₅C₁₀H₁₃SO₂ (L⁷) which enlarge the pool of [Ag(NO₃)L_n] (n=1, 2) camphor sulfonimine complexes with potential biological applications as anticancer or antimicrobial agents.[3,10] Unfortunately, the structure of these two camphor sulfonimine Ag(I) complexes such as that of many others could not be confirmed by X-ray diffraction analysis since no suitable crystals could be obtained. Therefore, calculations based on X-ray data obtained for the ligands can help to get insight into the structural arrangement of the complexes. In one case, data obtained by X-rays showed that [Ag(NO₃)L₂] (R=NMe₂) arranges as a calix (Scheme 1).[3]

2.1. Synthesis of camphor sulfonimines

The camphor sulfonimine compounds (RNC₁₀H₁₃NSO₂: L¹, R=OH; L², R=4-OHC₆H₄; L³, R=4-NH₂C₆H₄; L⁴, R=4-ClC₆H₄; L⁵, R=4-CH₃C₆H₄; L⁶, R=3,5-(CH₃)₂C₆H₃; L⁷, R=C₆H₅) were obtained by condensation (Scheme 2, right) of the convenient primary amine (H₂NR) with (1S)-3-oxo-camphorsulfonyl imine (1). Precursor 1 was obtained from camphor sulfonic acid through a sequential process that involves formation of compound 2 (Scheme 2, left) and ends on the oxidative annulation of a third ring the bicyclic camphor precursor.

From the camphorimine species under study (L¹- L⁷), four are new (L¹, L², L⁴ and L⁶) and were characterized by conventional analytical and spectroscopic techniques (FT-IR, ¹H NMR, ¹³C NMR and 2D NMR). The remaining compounds were just recrystallized to obtain single crystals. The structural arrangement of all compounds was then studied by X-ray diffraction analysis. Based on the X-ray data, the intra and inter molecular NBI were predicted by DFT-D3. Therefore, the supramolecular arrangement of compounds 1, 2 and the camphor sulfonimines L¹- L⁷ was elucidated.

2.2. Structural arrangement predicted by DFT-D3

Compounds 1 and 2

The X-ray data obtained for compound 1 indicates that a 3D helicoidal packing exists around the 3₁ crystallographic axes due to non-conventional oxygen∙∙∙hydrogen (aliphatic) NBI (Figure 1a). DFT-D3 calculations, performed in a subset dimer of 1, confirm the observed NBI and predict them as even stronger than those experimentally observed by X-ray analysis, as illustrated by shorter distances (Figure 2b).

For 2, calculations were based on the X-ray data reported by others.[11] The X-ray shows a zig-zag 1D structural arrangement due to chloride-oxygen intermolecular NBI established by the pendant sulphonyl chloride (SO₂Cl) groups in adjacent molecules (Figure 3). Such type of cooperative intermolecular S-Cl∙∙∙O interaction was previously discussed[12] on the basis of the S-Cl∙∙∙O and S-O∙∙∙Cl angles (θ) and classified (Figure 3, insert) as type I (θ₁=θ₂) or type II (θ₁=180ᵒ, θ₂=90ᵒ).[13]

Such as for 1, DFT-D3 calculations for 2 were carried out in a subset dimer of the structural arrangement found by X-ray diffraction analysis. Once more, the output data shows good agreement between experimental and calculated values (Table 1), in agreement with the electrostatic directionality of the interactions.

2.3. Camphor sulfonimines (L¹-L⁷)

The X-ray data collected for the sulfonimine compounds L¹ to L⁷ show different packing motives according to the imine substituent (R). In L¹ (R=OH) the asymmetric unit is formed by two structurally different molecules as shown in Figure 4 (center) since chirality prevents the existence of an inversion center.

The X-ray data, shows that at the “dimeric” asymmetric unit, the NBI are type I (θ₁=θ₂, i.e. S-O∙∙∙H=C-H∙∙∙O = 108ᵒ) while the NBI established between neighbor asymmetric units (involving the sulfonimine oxygen atom and the hydrogen hydroxyl atom) are of type II (θ₁≠ θ₂, i.e. S-O∙∙∙H=109ᵒ and O-H∙∙∙O=171ᵒ). The DFT-D3 calculations corroborate the type II nature of this NBI (Figure 5, left) but in what concerns the asymmetric unit DFT-D3 calculations predict type II NBI (Figure 5, right) in contrast with experimentally observed by X-ray diffraction. Such results indicate that NBI type I experimentally observed are due to packing constraints rather than to electrostatic effects.

From Figure 4 and Figure 5, it comes out that the asymmetric unit remains unchanged both experimental and in DFT-D3 calculations with the two nitrogen atoms not involved in NBI and thus available for complexation with the metal center AgNO₃. The introduction of the metal center in the calculations (Figure 6) confirms that a MOF can form through coordination of the cyclic imine nitrogen atom to the metal, keeping the 1D chain of the organic framework (OF).

Camphor sulfonimine compounds with aromatic substituents (R, Scheme 2) at the imine group (L² to L⁷), display self-assembly behaviors quite different from L¹ (R=OH).

A 1D assembly exists in L⁴ (R=4-ClC₆H₄), L⁵ (R=4-CH₃C₆H₄), L⁶ (3,5-(CH₃)₂C₆H₄) and L⁷ (R=C₆H₅) which is formed by non-conventional NBI established between the nitrogen atom of the -SO₂N- group and one of the hydrogen atoms of the methylene (CH₂) group of the five membered sulfonimine ring of a neighbor molecule. As a representative example, the X-ray diffraction data is displayed for L⁷ (Figure 7, left). The results from the DFT-D3 calculations agree well with the experimental obtained X-rays data (Figure 7, right). Such as for 1 the predicted NBI distances are shorter than the experimentally measured pointing to stronger electronic interactions.

In the aromatic substituted camphor sulfonimines (L⁴ to L⁷) the cyclic nitrogen imine atom of the 1D organic framework is involved in NBI which means that complexation would break down the self-assembly trends observed in the OF. So, MOFs are expected to have different frameworks than those observed in the ligands L⁴ to L⁷.

At the camphor sulfonimines bearing a protic substituent (R=OH in L² or R=NH₂ in L³) at the -NR imine group (Scheme 2), the characteristics of the NBI differ from those observed for L⁴-L⁷. In L² and L³ the NBI are of the stacking π-π type, established between the aromatic carbon atoms of two distinct asymmetric units occupying the ortho positions relative to the heterogeneous (N or O) atoms, as depicted for L² (Figure 8, left). In L² and L³ NBI involving the sulfonyl imine group (not observed for L⁴-L⁷) are established by a symmetry related chain of molecules (Figure 8, left). These additional interactions were included in the DFT-D3 calculations (Figure 8, right) to get consistency between X-ray and calculated structures. Here again, such as for L¹ the two nitrogen imine atoms are free to complex the AgNO₃ metal center to form a MOF mimicking the self-assembly characteristics of the corresponding OF.

2.4. Redox properties

Packing and electronic parameters have distinct effects on the properties of the molecules. While the electronic characteristics persist no matter the physical state (solid or solution), packing essentially drives the solid-state properties.

The above results show that the structural arrangement in camphor sulfonimines (L² - L⁷) and in compounds 1 and 2 is essentially directed by electronic parameters. In contrast, packing plays an important role in the supramolecular arrangement of L¹, by imposing a type I NBI geometry (see above).

In order, to get insights whether a correlation exists between the structural parameters and the redox potentials and try to ascertain whether NBI interactions remain in solution, the electrochemical behavior of the L¹-L⁷ compounds was studied by cyclic voltammetry.

In acetonitrile, compounds L¹-L⁷ display one irreversible anodic wave and one (L¹, L², L⁵, L⁶, L⁷) or two (L³, L⁴) cathodic waves (Figure 9). In the case there are two cathodic waves, the higher potential wave (I) typically displays a higher intensity than the lower potential one (wave II). The reversibility of wave I decreases upon scan at the potential of wave II as shown (Figure 9, left). The loss of reversibility and emergence of the anodic wave X is consistent with chemical reactivity being induce by electron transfer.

The anodic and cathodic potentials measured for L¹ - L⁷ are displayed in Table 2.

The potential trend observed for wave I (Table 2) suggests that reduction involves the imine >C=N-R group. In fact, an excellent correlation exists (Figure 10, left) between the redox potential and the Hammett sigma parameter (σ_p)[14] which accounts for the electronic characteristics of the R group driving the cathodic process. Compounds L¹ (R=OH) and L³ (4-NH₂C₆H₄) are aside from the linear correlation (R²=0.987) established by the other compounds showing that other than electronic parameters operate.

In what concerns L¹ such behavior was somehow expected due to the non-aromatic character of the R group (unlike the R groups at L²-L⁷) and the distinct type of NBI (see DFT-D3 discussion, above) compared to the other compounds.

In what concerns the relationship between ^I$E_{1/2}^{red}$ and the LUMO (Figure 10, right) the energy calculations for L¹ show that there is a decrease by 0.11 eV in the absolute value of the energy in the H-bonded dimer. Such energy shift may account for the non-fitting of the potential (FIGURE 10, right) pointing to L¹ keeping some dimer character in the electrolyte solution.

L³ is the other compound that does not fit the ^I$E_{1/2}^{red}$ versus σ_p (Figure 10, left) neither fits the ^I$E_{1/2}^{red}$ versus LUMO (Figure 10, right) correlations. The two compounds (L¹, R=OH and L³, R=C₆H₄NH₂) have in common the protic character of the R imine substituent which induces formation of H-bonded dimers (see above).

In what concerns the anodic behavior, just L², L³, L⁵ and L⁶ show oxidation processes. L² and L³ display the lowest oxidation potentials in agreement with the electron releasing characteristics of the aromatic R groups as corroborated through the linear correlation with Hammett σ_p parameters (Figure 11, left). A good correlation with the HOMO was found (Figure 11, right).

The calculated HOMO and LUMO contours for compound L⁷ show that the HOMO is mostly localized at the aromatic ring (all carbons) while the LUMO is localized at the camphor skeleton with an important contribution from the 2, 4 and 6 ring carbon atoms (Figure 12).

The electrochemical behavior of [Ag(NO₃)L₂] (C1, C2) complexes studied by cyclic voltammetry show that a cathodic process exists at a potential ($E_p^{red}$=-1.18 V, C1 and $E_p^{red}$=-1.17 V, C2) similar to that measured for the organic precursors L⁶ and L⁷, respectively (Table 2). Such pattern indicate that the electronic properties of the sulfonimine ligands remain essentially identical in the complexes. As expected, the cyclic voltammograms of C1 and C2 display a cathodic wave attributed to the Ag(I)→Ag(0) reduction at a potential ($E_p^{red}$=0.16 V) that does not differ much of that reported for related complexes.[3]

The redox behavior of 1 was also studied by cyclic voltammetry showing a quasi-reversible cathodic wave ($E_{1/2}^{red}$= -0.92 V; ipc/ipa=1, E_pa-E_pc= 133 mV) but no anodic process. According to the cathodic potential value, reduction of 1 is easier than any of the sulfonimine compounds (L¹-L⁷, Table 1). Therefore, sulfonimines are more electron rich than 1 and consequently better ligands for metal sites. In fact, no Ag(I) complex based on 1 could be experimentally obtained, in contrast with several complexes derived from sulfonimine ligands as for example the new [Ag(NO₃)L₂] (C1 and C2) now reported (see experimental).

3. Conclusions

DFT-D3 calculations based on X-ray data obtained for a series of camphor sulfonimine compounds corroborate that NBI are responsible for their organic framework. On ²L-⁷L the NBI are essentially electrostatic, in contrast with ¹L where packing plays a relevant role in the solid-state arrangement observed by X-rays analysis.

The characteristics of the sulfonimine substituent (R) fine drives the NBI established and direct the ability of the L compounds to bind the Ag(I) metal center forming MOFs such as detected for [Ag(NO₃)(O₂SNC₁₀H₁₃NNMe₂)₂] whose X-ray structure was formerly reported.

It was found that complexation of the metal center always occurs in a monodentate fashion through the imine nitrogen atom in the sulphonyl 5-membered ring. It was predicted by DFT-D3 calculations that the NBI interactions not involving this atom remain after complexation being transferred to the MOF. The conjugation of the x-ray structures of the organic ligands with DFT-D3 calculations proved to be a powerful tool to predict the self-assembly trends of the corresponding MOF structures of the complexes.

The correlation between the redox potentials of ¹L-⁷L (studied by cyclic voltammetry), the Hammett σ-parameter and calculated HOMO and LUMO energies are in excellent agreement with different types of NBI being established by ¹L and ²L compared to the other camphor sulfonimine compounds.

4. Experimental

The camphor sulfonimines L¹, L³, L⁵ and L⁷ were synthesized according to published procedures.[15] The complexes were synthesized under nitrogen using Schlenk and vacuum techniques. The FT-IR spectra were obtained from KBr pellets using a JASCO FT/IR 4100 spectrometer. The NMR spectra (¹H, ¹³C, DEPT, HSQC, and HMBC) were obtained from CDCl₃¸ CD₃CN or DMSO solutions using Bruker Avance II+ (300 or 400 MHz) spectrometers. The NMR chemical shifts are referred to TMS (δ=0 ppm).

The redox properties were studied by cyclic voltammetry using a three compartments cell equipped with Pt wire electrodes (work and secondary) and interfaced with a VoltaLab PST050 equipment. The cyclic voltammograms were obtained from NBu₄BF₄ / CH₃CN (0.10 M) solutions used as electrolyte. The potentials were measured in Volts (±10 mV) versus SCE at 200 mV/s using [Fe(η⁵-C₅H₅)₂]^0/+ ($E_{1/2}^{ox}$= 0.382 V; CH₃CN) as internal reference. The window of potential was established by scanning a fresh solution of the electrolyte to the limits.

4.1. Synthesis

4.1.1. Camphor sulfonimines

(E)-7-(hydroxyimino)-8,8-dimethyl-4,5,6,7-tetrahydro-3H-3a,6-ethanobenzo[c] isothiazole 2,2-dioxide (L¹) - Compound 1 (682 mg; 3 mmol) was dissolved in ethanol (10 mL) acidified with acid acetic (0,4 mL) and the mixture was stirred for ca. 15 minutes before the hydroxylamine (99 mg, 3 mmol) was added. The reaction was stopped after 24 hours stirring by addition of 5 mL of water. The organic phase was then extracted with dichloromethane (3x5 mL) and dried over MgSO₄ for a couple of hours. After filtration the solvent was allowed to evaporate slowly in the exhaust hood. Yield 71%. Elem. Anal. (%) for C₁₀H₁₄N₂O₃S: Found: C, 49.8; N, 11.3; H, 6.0; S, 13.7. Calc.: C, 49.6; N, 11.6; H, 5.8; S, 13.2. FTIR (KBr, cm^-1): 3367, 1668, 1633, 1327, 1158. ¹H NMR (DMSO, δ ppm): 12.8 (s, 1H); 3.59, 3.36 (2d, J=14.0 Hz, 2H); 3.25 (d, J=4.2 Hz, 1H); 2.22 – 2.01 (m, 2H); 1.63 – 1.40 (m, 2H); 1.01 (s, 3H); 0.78 (s, 3H). ¹³C NMR (DMSO, δ ppm): 184.7, 155.3, 64.2, 49.0, 47.9, 47.1, 28.8, 23.6, 19.5, 18.1.

(E)-7-((4-hydroxyphenyl)imino)-8,8-dimethyl-4,5,6,7-tetrahydro-3H-3a,6-methanobenzo[c]isothiazole 2,2-dioxide (L²) - 3-oxo-camphorsulfonimide (1, 682 mg; 3 mmol) was dissolved in ethanol (10 mL) acidified with CH₃COOH (0.4 mL). The mixture was stirred for 15 minutes before adding the 4-aminophenol (329 mg; 3 mmol). A light brown compound was formed upon overnight stirring at 50°C that was filtered off the solution and dried. Yield 81%. Elem. Anal. (%) for C₁₆H₁₈N₂O₃S·H₂O: Found: C, 56.9; N, 8.0; H, 6.2; S, 11.1. Calc.: C, 57.1; N, 8.3; H, 6.0; S, 9.5. FTIR (KBr, cm^-1): 3567, 1653, 1633, 1330, 1160. ¹HNMR (CD₃CN, δ_ppm): 7.35 (s, 1H, OH), 7.00 (d, J= 7.8 Hz, 2H), 6.88 (d, J=7.8 Hz, 2H), 3.48, 3.25 (2d, J= 13.9 Hz, 2H), 3.10 (d, J= 2.6 Hz, 1H), 2.34–2.23 (m, 2H), 1.89–1.75 (m, 2H), 1.06 (s, 3H), 0.84 (s, 3H). ¹³C NMR (CD₃CN, δ_ppm): 187.9, 166.0, 157.3, 141.9, 124.9, 116.7, 64.1, 52.6, 50.5, 47.8, 29.4, 24.4, 19.9, 18.3.

(E)-7-((4-chlorophenyl)imino)-8,8-dimethyl-4,5,6,7-tetrahydro-3H-3a,6-methanobenzo[c]isothiazole 2,2-dioxide (L⁴) - Compound 1 (682 mg; 3 mmol) was dissolved in acidified (acid acetic, 0.4 mL) ethanol (10 mL) and stirred for ca. 15 minutes. Upon addition of 4-chloroaniline (384 mg, 3 mmol) the reaction was stirred overnight at 40°C. A yellow precipitate formed that was filtered off solution and dried. Yield 76%. Elem. Anal. (%) for C₁₆H₁₇ClN₂O₂S·¹/₅H₂O: Found: C, 56.5; N, 8.1; H, 5.3; S, 10.1. Calc.: C, 56.5; N, 8.2; H, 5.2; S, 9.4. FTIR (KBr, cm^-1): 1664, 1636, 1340, 1161. ¹H NMR (CDCl₃, δ ppm): 7.39 (d, J=8.6 Hz, 2H); 6.93 (d, J=8.6 Hz, 2H); 3.42, 3.21 (2d, J=13.4 Hz, 2H); 2.99 (d, J=8.6 Hz, 1H); 2.29 – 2.21 (m, 2H); 2.10 – 2.00 (m, 1H); 1.82 – 1.74 (m, 1H); 1.11 (s, 3H); 0.95 (s, 3H). ¹³C NMR (CDCl₃, δppm): 185.0, 167.5, 147.2, 132.1, 129.4, 122.0, 62.7, 51.4, 50.0, 46.8, 28.4, 24.1, 20.0, 18.4.

(E)-7-((3,5-dimethylphenyl)imino)-8,8-dimethyl-4,5,6,7-tetrahydro-3H-3a,6-methanobenzo[c]isothiazole 2,2-dioxide (L⁶) - Compound 1 (454 mg; 2 mmol) was dissolved in ethanol (7 mL) acidified with acid acetic (0.2 mL) and the mixture stirred for ca. 15 minutes. Then, 3,5-dimethylaniline (0.25 mL, 2 mmol) was added and the reaction stirred for 24 hours at RT. A brown precipitate formed that was filtered off solution and dried. Yield 86%. Elem. Anal. (%) for C₁₈H₂₂N₂O₂S: Found: C, 65.3; N, 8.4; H, 6.7; S, 9.5. Calc.: C, 65.4; N, 8.5; H, 6.7; S, 9.7. FTIR (KBr, cm^-1): 1665, 1636, 1339, 1161. ¹H NMR (CDCl₃, δ ppm): 6.87 (s, 1H); 6.56 (s, 2H); 3.39, 3.18 (2d, J= 13.4 Hz, 1H); 3.01 (d, J= 4.0 Hz, 1H); 2.34 (s, 6H); 2.24 – 2.19 (m, 2H); 2.05 – 1.98 (m, 1H); 1.82 – 1.75 (m, 1H); 1.09 (s, 3H); 0.94 (s. 3H). ¹³C NMR (CDCl₃, δppm): 185.4, 166.5, 149.1, 139.1, 128.1, 118.2, 113.6, 62.9, 51.5, 50.1, 46.8, 28.6, 24.3, 21.5, 21.4, 20.1, 18.5.

4.1.2. Complexes

[Ag(NO₃){(3,5-Me₂C₆H₃)NC₁₀H₁₃NSO₂}₂] (C1) – AgNO₃ (42.5 mg; 0.25 mmol) was added to a solution of L⁶ (165 mg; 0.50 mmol) in CH₃CN (5 mL) and the mixture was stirred for 2 hours. A light suspension formed that was filtered to separate silver residues. The filtered solution was evaporated and dried under vacuum to afford C_B as a dark yellow compound. Yield 72%. Elemental analysis for AgC₃₆H₄₄N₅O₇S₂·¹/₅HNO₃ Exp.: C, 51.2; N, 8.4; H, 5.5; S, 8.8. Calc.: C, 51.3; N, 8.6; H, 5.3; S, 7.6. FTIR (KBr, cm^-1): 1666, 1644, 1384, 1338, 1161. ¹H NMR (400 MHz, CD₃CN, δ ppm): 6.91 (s, 1H); 6.59 (s, 2H); 3.51, 3.28 (2d, J=14 Hz, 2H); 2.96 (d, J=4.8 Hz, 1H); 2.31 (s, 6H); 2.24 – 2.30 (m, 1H); 1.71 – 1.88 (m, 3H); 1.05 (s, 3H); 0.87 (s, 3H). ¹³C NMR (400 MHz, CD₃CN, δ ppm): 186.5, 167.7, 149.2, 139.2, 127.6, 117.8, 63.2, 51.4, 49.7, 46.5, 28.0, 23.6, 20.3, 19.1, 17.1.

[Ag(NO₃)(C₆H₅NC₁₀H₁₃NSO₂)₂] (C2) - Silver nitrate (42.5 mg; 0.25 mmol) was added to a solution of L⁷ (151 mg; 0.50 mmol) in CH₃CN (3 mL) and the mixture stirred for 2 hours.

The suspension was then filtered to remove silver residues. Upon solvent evaporation and drying under vacuum the compound was obtained as a yellow powder. Yield 62%. Elemental analysis for AgC₃₂H₃₆N₅O₇S₂·H₂O Exp.: C, 48.6; N, 9.0; H, 4.7; S, 7.6. Calc.: C, 48.5; N, 8.8; H, 4.8; S, 8.1. FTIR (KBr, cm^-1): 1664, 1635, 1384, 1346, 1165. ¹H NMR (400 MHz, DMSO, δ ppm): 7.46 (t, J=15.7 Hz, 2H); 7.25 (t, J=7.4 Hz, 1H); 7.00 (d, J=8.4 Hz, 2H); 3.74, 3.52 (2d, J=14 Hz , 2H); 2.88 (d, J=4.8 Hz, 1H); 2.34–2.25 (m, 1H); 2.19–2.10 (m, 1H); 1.80–1.70 (m, 2H); 1.01 (s, 3H); 0.84 (s, 3H). ¹³C NMR (400 MHz, DMSO, δ ppm): 186.0, 167.8, 148.8, 129.3, 126.0, 120.1, 63.0, 51.0, 49.3, 46.3, 27.7, 23.4, 19.2, 17.5.

4.2. Computational methods

All theoretical calculations were of the DFT type, carried out using GAMESS-US version R3,[16] using the implemented version of the B3LYP functional and a 6-31G** basis set for organic molecules and SBKJC basis set/ECP in metal complexes. DFT-D3[17] with the standard parametrization of GAMESS-US, was used in all computational work. Optimized geometries were confirmed to be minima by checking for absence of imaginary frequencies in the Hessian.

4.3. X-ray diffraction analysis

X-ray diffraction analysis was performed on mono crystals of 1,3 and L1 to L7. Data was collected on a Bruker AXS-KAPPA APEX II area detector apparatus using graphite-monochromated Mo Kα (λ=0.71073 Å) and were corrected for Lorentz, polarization and empirically for absorption effects. The structure was solved by direct methods using SHELX97[18] and refined by full matrix least squares against F² using SHELX97 all included in the suit of programs WinGX v1.70.01 for Windows.[19] Non-hydrogen atoms were refined anisotropically and H atoms were inserted in idealized positions and allowed to refine riding on the parent carbon atom. Crystal data and refinement parameters are summarized in Table 3. Illustrations of the molecular structures were made with Chemcraft[20] or Mercury.[21]

CCDC 2252538 to 2252543 and CCDC 2252556 contain the supplementary crystallographic data for this paper which can be obtained free of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data Centre, 12, Union Road, Cambridge, CB2 1EZ, UK; fax: +44 1223 336033; or deposit@ccdc.cam.ac.uk).