Dynamic Au-C σ-Bonds Leading to an Efficient Synthesis of [n]Cycloparaphenylenes (n = 9-15) by Self-Assembly.

The transmetalation of the digold(I) complex [Au2Cl2(dcpm)] (1) (dcpm = bis(dicyclohexylphosphino)methane) with oligophenylene diboronic acids gave the triangular macrocyclic complexes [Au2(C6H4) x (dcpm)]3 (x = 3, 4, 5) with yields of over 70%. On the other hand, when the other digold(I) complex [Au2Cl2(dppm)] (1') (dppm = bis(diphenylphosphino)methane) was used, only a negligible amount of the triangular complex was obtained. The control experiments revealed that the dcpm ligand accelerated an intermolecular Au(I)-C σ-bond-exchange reaction and that this high reversibility is the origin of the selective formation of the triangular complexes. Structural analyses and theoretical calculations indicate that the dcpm ligand increases the electrophilicity of the Au atom in the complex, thus facilitating the exchange reaction, although the cyclohexyl group is an electron-donating group. Furthermore, the oxidative chlorination of the macrocyclic gold complexes afforded a series of [n]cycloparaphenylenes (n = 9, 12, 15) in 78-88% isolated yields. The reorganization of two different macrocyclic Au complexes gave a mixture of macrocyclic complexes incorporating different oligophenylene linkers, from which a mixture of [n]cycloparaphenylenes with various numbers of phenylene units was obtained in good yields.

Introduction

Self-assembly of organic ligands and metals, as exemplified in the metal–organic framework, has been recognized as a promising method to synthesize structurally controlled caged molecules and porous materials with controlled size and functions.[1−4] Furthermore, the use of a covalent bond instead of the noncovalent bond has also been actively developed as, for example, the covalent organic framework (COF).[5,6] However, the latter examples are still limited compared to those using noncovalent bonds.[7] This is mainly because of the lower reversibility of covalent bonds than that of noncovalent bonds, necessitating a long reaction time and harsh conditions to reach an equilibrium structure. Therefore, the development of a new dynamic covalent bond would significantly impact the synthesis of COFs and related molecules by the self-assembly process.

Cyclo[n]paraphenylenes ([n]CPPs), where n is the number of phenylene groups, are organic macrocycles that consist of 1,4-linked phenylene units.[7−11] While CPPs were imaginary molecules until 2009, CPPs with various sizes (n = 5–16, 18, 20, 21) are now being synthesized with the development of the innovative synthetic methods by Bertozzi/Jasti,[12] Itami,[13] Yamago,[14,15] and Osakada/Tsuchido[16] (Figure ). Furthermore, these works unveiled unique size-dependent physical properties[17] and applications of CPPs, that is, circularly polarized luminescent materials,[18−21] biological fluorophores,[22,23] gas-adsorption materials,[24−26] and electron-transport materials.[27] In addition, CPPs have also been used to construct unique molecular architectures,[28−30] such as supramolecular host–guest molecules,[31−37] mechanically interlocked molecules,[38−40] and building blocks for tubular nanostructures.[41−44] These results indicate the necessity for a size-selective and high-yielding synthetic method for CPPs.[45]

Figure 1

Synthetic routes to [n]cycloparaphenylenes reported by (a) Bertozzi/Jasti, (b) Itami, (c) Yamago, and (d) Osakada/Tsuchido.

The synthesis of CPPs can be classified into two methods depending on the precursor to induce curvature; one employs masked benzene units (Figure a,b), and the other uses transition-metal complexes (Figure c,d). The latter method resembles COF synthesis as it relies on the assembly of a linear paraphenylene unit by a transition-metal complex, that is, Pt or Au, forming covalent metal (M)–carbon (C) bonds. The presence of equilibrium in the formation of the Pt-complexes has already been reported in the random synthesis of [n]CPPs.[46] However, the control of the rate of equilibrium and the size of the Pt-macrocycles has never been achieved so far.

In 2020, some of the authors of this paper reported the synthesis of [6]CPP from a macrocyclic Au complex (Scheme ).[16] The reaction of 4,4′-diphenyldiboronic acid (L2) with [Au2Cl2(dcpm)] (1) (dcpm = bis(dicyclohexylphosphino)methane) produced the triangular hexagold(I) complex [Au2(C6H4)2(dcpm)]3 (Au-2) isolated in 77% yield. The oxidative chlorination[47−51] of Au-2 by PhICl2 afforded [6]CPP in a total yield of 59% (over two steps from 1). Thus, this synthetic method has the advantage of highly efficient synthesis of CPPs and related nanohoops from three arylene units.[52−56] However, the scope of this synthetic method and the mechanism of the efficient formation of the triangular Au complexes remain to be examined.

Scheme 1

Synthesis of [6]CPP by an Au-Mediated Method (Our Previous Study)[16]

Herein, we report the selective and random synthesis of [n]CPPs (n = 9–15) in high overall yield through macrocyclic Au complexes with oligophenylene linkers. We further clarified the crucial role of the electron-donating tetra-cyclohexyl groups in ligand 1, which significantly promotes the dynamic exchange reaction of Au–C bonds, giving desired cyclic structures under mild conditions. In addition, a detailed mechanism of the exchange reaction is proposed based on kinetic studies and structural analyses of the Au complexes. We believe that the current finding is not only useful for the synthesis of CPPs and related cyclic π-conjugated molecules but also for the development of new COFs.

Results and Discussion

Synthesis and Characterization of Macrocyclic Au Complexes and [n]CPPs (n = 9, 12, 15)

The transmetalation of [Au2Cl2(dcpm)] (1) with an equimolar amount of 4,4″-terphenyldiboronic acid pinacol ester (L3) was conducted in the presence of Cs2CO3 in toluene/ethanol/water at 50 °C. After stirring overnight, the resulting white solid was collected by filtration and characterized as the triangular macrocyclic complex [Au2(C6H4)3(dcpm)]3 (Au-3) in a 72% yield (Scheme a, reaction i). In contrast, the use of [Au2Cl2(dppm)] (1′) (dppm = bis(diphenylphosphino)methane) instead of 1 under the otherwise identical conditions gave a purple solid, which was partially soluble in (Cl2CD)2 and had broad signals in the 1H NMR spectrum (see Figure S9), suggesting the formation of linear and/or undesired macrocyclic oligomers. The results indicate the importance of the dcpm ligand of 1 over dppm to increase the dynamicity of the Au–C bond-exchange reaction, as discussed below.

Scheme 2

Synthesis of [n]CPPs by the Au-Mediated Method; (a) [9]CPP, (b) [12]CPP, and (c) [15]CPP

Reagents and conditions: (i) [AuCl2(R2PCH2PR2)] (1: R = Cy (dcpm), 1′: R = Ph (dppm)) (1.0 equiv), Cs2CO3 (6.0 equiv), toluene/ethanol/water (4:1:1), 50 °C, overnight; (ii) PhICl2 (3.0 equiv), DMF, −60 °C, 0.5 h, and then r.t., overnight. Linker lengths were determined from the molecular structures of oligophenylene simulated using MMFF calculations.

By using 1 as a Au complex, oligoparaphenyl diboronic acid pinacol ester with a quaterphenylene group (L4) or a quinquephenylene group (L5) under otherwise identical reaction conditions afforded the corresponding Au complexes, [Au2(C6H4)(dcpm)]3 (x = 4 from Au-4, 5 from Au-5) in 83 and 72% yields, respectively (Scheme b,c, reaction i). It is worth noting that the triangular complexes were obtained as a sole product in all cases.

Single crystals of Au-3 (Figure a) and Au-4 (Figures b and S95) suitable for X-ray crystallography were obtained via the vapor diffusion of CH3CN into (Cl2CH)2 solutions of each complex. Both molecules adopt a triangular molecular structure similar to that of Au-2,[16] consisting of three oligophenylene linkers and three Au2(dcpm) units. The complex with the terphenylene linker, Au-3, adopts a pseudo-C2-symmetrical structure with PMM- or PPM-helical Au2P2C units at the three corners. Au-4 gave polymorphic crystals with a D3-symmetrical structure and PPP or MMM helicity (Figure b) along with the C2-symmetrical structure (Figure S95).[57] These triangular molecular structures were stabilized by aurophilic interactions[58−61] between the two Au(I) centers in each corner. In the C2-symmetrical structure of Au-3 (Figure a), the distance between the two neighboring gold atoms in two helical corners (3.142(8), 3.118(1) Å) is shorter than that of the other corner (3.297(1) Å). The same phenomenon was observed in the X-ray structures of Au-2(16) and Au-4 (see Figure S95) with C2-symmetry. On the other hand, the X-ray structure of Au-4 with D3-symmetry (Figure b) exhibits Au–Au distances of 3.091(1) Å, which are shorter than those of the C2 isomer. These results indicate stronger aurophilic interactions in the D3 isomer compared to those in the C2 isomer. Additionally, the phenylene linkers adopt a bent conformation in the D3 symmetry, which would be difficult to be adopted in Au-3 and Au-4 form with short oligophenylene linkers.

Figure 2

Molecular structures of (a) Au-3 (C2-symmetry) and (b) Au-4 (D3-symmetry) with thermal ellipsoids at 30% probability. Hydrogen atoms and solvent molecules are omitted for clarity.

The oxidative chlorination[45] of Au-3 occurred upon the addition of 3 equiv of PhICl2 in DMF at −60 °C. The C–C bond formation between two phenylene linkers via reductive elimination gave [9]CPP when the reaction temperature was raised to 25 °C (Scheme a, reaction ii). The 1H NMR spectrum of the crude product showed only one singlet aromatic signal at 7.52 ppm (CDCl3, 25 °C), which was assigned to [9]CPP based on the literature.[17] Purification of the reaction mixture using column chromatography on silica gel afforded the desired product in good yield (78%), along with the Au complex 1, which was also obtained in 78% yield. [n]CPPs (n = 12 for Au-4, 15 for Au-5) were obtained from the corresponding Au complexes in 78% and 88% yields, respectively (Scheme b,c, reactions ii). [n]CPPs (n = 9, 12, 15) have successfully been obtained from oligophenylene diboronates, L3–5, in higher overall yields than other methods (see Tables S1–S3).[62]

Kinetic Studies of the Au–C σ-Bond-Exchange Reaction

To clarify the origin of the unique ligand effect (1 vs 1′) for the formation of Au complexes, we next examined the dynamics of the Au–C bond-exchange reactions using model complexes. There were several reports on the exchange reaction of M–C σ-bonds (M = Pt(II), Pd(II), Au(I), etc.) (see supporting Scheme S3b),[63−69] and the reactions were generally regarded as a slow process compared to those of the noncovalent metal–ligand bonds, that is, Pd–N[1] and Pt–N[2] bonds (see Scheme S3a). However, the exchange reactions involving aryl–Au(I) complexes have never been reported to date.[70−74]

At first, two acyclic dinuclear Au(I) complexes having the dcpm ligand, [Au2Ph2(dcpm)] (Au-HH), and [Au2(C6H4-4-F)2(dcpm)] (Au-FF), were synthesized, and the formation of the unsymmetrical Au(I) complex, [Au2Ph(C6H4-4-F) (dcpm)] (Au–HF), was observed upon mixing equimolar amounts (1.7 mM each) of Au-HH and Au-FF in CDCl3 (Figure a). The 19F NMR spectroscopic analysis indicated that the reaction reached equilibrium after 30 min at 25 °C (Figure b).[75] To our surprise, rapid bond exchange of the Au–C σ-bonds was clearly observed even at or below room temperature. The rate constants of the comproportionation (k1) and disproportionation (k–1) employing a mixture of Au-HH and Au-FF at −20 °C were determined to be k1 = (6.9 ± 0.64) × 10–2 M–1·s–1 and k–1 = (1.4 ± 0.08) × 10–2 M–1·s–1 (Figure c). The values are in good agreement with the reversible second-order reaction model.[76] The temperature dependence of the rate of comproportionation in CDCl3 was obtained to be given in k1 = 0.30 ± 0.039 M–1·s–1, k–1 = (8.6 ± 0.80) × 10–2 M–1·s–1 (at 0 °C), k1 = 0.18 ± 0.025 M–1·s–1, (4.6 ± 0.87) × 10–2 M–1·s–1 (at −10 °C) (Figure a; entry 2, 3). By adding 1 equiv of the ancillary ligand, dcpm, to a mixture of Au-HH and Au-FF in CDCl3 at 0 °C, two doublet peaks of 31P{1H} NMR at δ 45.8 and −10.4 (2JP–P = 80 Hz), corresponding to mononuclear Au complexes,[77,78] [AuAr(dcpm)] (Ar = C6H5, or C6H4-4-F), were obtained, and the reaction was in a resting state (Figure a, entry 4). Even in the presence of a catalytic amount (∼0.05 equiv) of free dcpm, the distinct reduction in the rate of comproportionation of the Au complexes in CDCl3 at 0 °C was observed with the constants of k1 = (2.6 ± 0.35) × 10–2 M–1·s–1 and k–1 = (1.0 ± 0.088) × 10–2 M–1·s–1 (Figure a, Entry 5). Coordinative solvents, such as acetone and DMF, tend to decelerate the aryl exchange (Figure a, entry 6–8). The kinetic rates of the reaction using acetone-d6 at 0 °C were k1 = (6.9 ± 1.6) × 10–2 M–1·s–1 and k–1 = (8.3 ± 1.5) × 10–2 M–1·s–1, which were one-tenth as in CDCl3 at the same temperature. In a DMF solution at 0 °C, the desired aryl exchange reaction did not proceed; thus, the reaction mixture was warmed up to 25 °C to proceed with the reaction with rate constants of k1 = (8.1 ± 1.2) × 10–2 M–1·s–1 and k–1 = (3.9 ± 0.43) × 10–2 M–1·s–1. We believe that these solvents would coordinate with the Au complexes to reconstruct the less reactive compounds, such as [AuArL] (L = coordinative solvent) which is similar to [AuAr(dcpm)]. The putative mechanism is discussed in a later section.

Figure 3

(a) Kinetic rates of comproportionation of (aryl)2Au2 complexes (Au-HH and Au-FF; Au-HH and Au-FF). (b) 19F NMR spectra (376 MHz, CDCl3, 23 °C): pristine Au-FF (top), a mixture of Au-HH and Au-FF 15 min after its preparation (center), and the mixture after 30 min (bottom). (c) Reaction profile of the comproportionation at 0 (circles), −10 (triangles), and −20 (squares) °C (inset: Eyring plot); (d) comproportionation of Au-HH and Au-DD, followed by reductive elimination to yield biphenyls A, B, and C (inset: GC–MS spectra). Reagents and conditions: (i) CH2Cl2, r.t., 2 h; (ii) PhICl2 (2.0 equiv), CH2Cl2, −60 °C, 0.5 h, then r.t., overnight.

To verify the effect of the electronic nature of the aryl groups, we conducted NMR studies of the ligand-exchange reaction using an equimolar mixture of [Au2Ph2(dcpm)] (Au-HH) and [Au2(C6D5)2(dcpm)] (Au-DD) in CDCl3 at −20 °C. However, the isomeric shifts of the deuterated phenyl groups were too small to distinguish deuterated groups from nondeuterated groups or follow the reaction rates determined by NMR. The chemical oxidation of the reaction mixture after 2 h caused the formation of a mixture of C6H5–C6H5, C6H5–C6D5, and C6D5–C6D5 in a 0.18:0.52:0.30 molar ratio, which was determined by GC–MS analysis (Figure d). The results indicated that scrambling of the phenyl group was completed within a short time, similar to the corresponding reactions of Au-HH and Au-FF.

Moreover, the comproportionation of Au complexes having the dppm ligand, Au-HH, and Au-FF did not occur at all even after 1 h at room temperature (Figure a, entry 9). The results clearly suggest that the cyclohexyl groups on the phosphines of the dcpm ligands are essential for the progress of the reaction.

The different dynamic behavior of the Au(I)–aryl complexes with dcpm and dppm can be explained by the structural change of the Au(I) complex due to the electronic of the substituent as estimated from their crystal structures of Au-FF and Au-FF (Figure a). The Au–P bonds of Au-FF (2.2903(17) and 2.2938(16) Å) are shorter than those of Au-FF (2.3013(13) Å) due to the electron-donating effect of the cyclohexyl-substituted phosphine compared to that of the phenyl-substituted one. Then, the strong σ-donation effect leads to a higher trans effect of dcpm compared to that of dppm to place aryl groups opposite to each other. Therefore, the lengths of the two Au–Cipso bonds of Au-FF (2.072(4) and 2.092(5) Å) are slightly longer than those of Au-FF (2.062(5) Å). Namely, the above results suggest that the Au(I)–C σ-bonds in Au-FF are weaker than those in Au-FF.

Figure 4

(a) ORTEP drawings of Au-FF and Au-FF with thermal ellipsoids at 50% probability; hydrogen atoms are omitted for clarity. Selected bond distances (Å) are summarized in the inserted table. (b) Energy levels of the frontier orbitals of Au-HH and Au-HH and a depiction of their LUMOs (M06/6-31G for C, H, P, LANL2TZ(f) for Au, iso value = 0.025); hydrogen atoms are omitted for clarity; (c) putative mechanism for the Au–C bond-exchange reaction between (aryl)2Au2 complexes.

Theoretical calculations further suggested the importance of the cyclohexyl group on the reactivities of the Au–aryl complexes. Thus, the lowest unoccupied molecular orbital (LUMO) of Au-HH is localized at the apical position of the gold atom with E = −0.737 eV, while it is delocalized to the ancillary phosphine ligand in the case of Au-HH (Figure b). In contrast, the lowest unoccupied, localized orbital on the gold of Au-HH was observed at the LUMO + 10 level with E = 0.113 eV. The results indicate that the dcpm ligand enhances the electrophilic reactivity of Au atoms in Au-HH by localizing the electrophilic orbital of the Au atoms and lowering its energy level as compared to the dppm ligand. Since the cyclohexyl group is an electron-donating group, the calculation results that the electrophilicity of the Au atom increases with the dcpm ligand cannot be explained by a simple electronic effect only. Therefore, this seemingly contradictory result is most likely due to the structural difference between Au and Au complexes caused by the trans effect, as discussed in the previous section. These results also suggest the mechanism of the Au–aryl exchange reaction, which involves the nucleophilic attack from a reactive Au–Caryl bond to the electrophilic LUMO orbital on the gold atom.

The activation parameters for the comproportionation between Au-HH and Au-FF were determined to be ΔG⧧ = 17 kcal mol–1, ΔH⧧ = 9.5 kcal mol–1, and ΔS⧧ = −26 cal mol–1 K–1 (at 25 °C) based on the Eyring plot (Figure c, inset). The negative value of ΔS⧧ implies that the bond-exchange process proceeds via an associative mechanism.[63,64,77,79−81] Considering the deceleration of the aryl exchange reaction upon the addition of catalytic amounts of dcpm in the former section, it is plausible that a ligand dissociation precedes an associative transition state. To combine the above results, we could propose the putative mechanism that the bond exchange or metathesis between the metal centers and their organic ligands proceeded via a mechanism with three steps: (i) a ligand dissociation to form reactive Au species [i.e., AuAr or ArAu–AuAr(dcpm)]; (ii) formation of an associative transition state;[65,66,68−70,73,74] (iii) the Au–Caryl bond exchange (Figure c). The complexes with bridging organic ligands, presumed to be the intermediate of organic ligand exchange, have been proposed in kinetic studies in which aryl–Cu(I) and aryl–Au(I) complexes caused trans–cis isomerization of Pd(II) complexes[70,72] and exchange of their aryl and alkynyl ligands bound to Pd(II) and Rh(I) complexes.[71]

We compared the kinetic and thermodynamic parameters of the comproportionation of Au-HH and Au-FF and those of the reaction of [PtPh2(cod)] and [Pt(C6H4-4-F)2(cod)], which we have reported previously (see supporting Scheme S4).[65,82] The comproportionation of Au2 complexes at 0 °C proceeded nearly 105 times as fast as the reaction of [PtPh2(cod)] and [Pt(C6H4-4-F)2(cod)] even at 50 °C with rate constants of k1 = (6.4 ± 0.6) × 10–6 and k–1 = (2.0 ± 0.2) × 10–6 M–1 s–1. The thermodynamic parameters for the above Pt system are ΔG⧧ = 27 kcal mol–1, ΔH⧧ = 23 kcal mol–1, and ΔS⧧ = −11 cal mol–1 K–1. The ΔH⧧ for the Au2 system is smaller than that for the Pt system, suggesting that the Au(I)–C bond dissociation via the formation of the association complex of intermediate can be expected to occur more easily.

These results imply that the highly efficient macrocyclization in this study should be attributed to the highly dynamic, reversible intermolecular exchanges of Au(I)–C σ-bonds (Figure ). In the early stage of the reaction between 1 and L3, a mixture of acyclic and cyclic oligomers can form under kinetic conditions. Then, the triangular complex becomes the major product via the dynamic bond-exchange reaction between these species. As the macrocyclization reaction proceeds in a heterogeneous mixture, the high thermodynamic stability and/or the high crystallinity of the triangular complex could account for the isolation of the triangular complex as the sole product.

Figure 5

Illustration of a plausible mechanism for the formation of the triangular Au(I) complex Au-3 via a self-assembly process between 1 and L3.

Synthesis of CPPs from Two Different Oligophenylene Linkers

To explore the scope of the present dynamic aryl-group exchange reactions, two macrocyclic Au complexes (Au-3/Au-4 or Au-4/Au-5) were mixed in DMF at 50 °C (Figure a, Method A). Then, the resulting reaction mixture was directly treated with PhICl2 without isolating the Au complexes. The 1H NMR spectrum of the products obtained from an equimolar mixture of Au-3 and Au-4 (Figure c) contained four singlet signals that were assigned to [9], [10], [11], and [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPPs. The formation of the [10] and [11]CPPs indicates the formation of isosceles triangular macrocyclic complexes in which two different oligophenylene linkers were reorganized from Au-3 and Au-4 through the dynamic Au(I)–C bond-exchange process. These complexes were experimentally confirmed by Fourier transform ion cyclotron resonance (FT-ICR) MALDI-TOF MS measurement of the crude products after mixing at 50 °C in DMF (Figure e,f).[83] The NMR yields of [10]CPP (11%) and [11]CPP (17%), which were CPPs derived from heteroleptic Au complexes with different linkers, were slightly lower than those of [9]CPP (26%) and [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP (15%), which are CPPs derived from homoleptic Au complexes with identical linkers (Figure c). All of these four CPPs were successfully separated by preparative gel permeation chromatography (PGPC) (see supporting Figure S49). In the Au-4/Au-5 system, the NMR yields of [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP (13%) and [15]CPP (11%) were higher than those of [13]CPP (3.0%) and [14]CPP (3.8%) (Figure d), indicating a similar trend to that of the Au-3/Au-4 system. However, the lower yields of [13] and [14]CPPs suggest that the formation of the heteroleptic macrocyclic Au complexes would be more difficult than the Au-3/Au-4 system. We also conducted the transmetalation of [AuCl2(dcpm)] (1) with a 1:1 mixture of two different oligophenyldiboronic acids (L3/L4 or L4/L5) (Figure a,b, Method B). Oxidation of the product by PhICl2 afforded a mixture of CPPs in much lower NMR yields than Method A. These results indicate that the intermolecular transmetalation based on the dynamic Au(I)–C σ bond exchange could occur between not only the acyclic complexes but also the macrocyclic complexes. The fact that the differences in the ratios of the CPP products depend on the linker length in both methods should be attributed to the thermodynamical stability of the corresponding precursor complexes. Our crystallographic study (Figure ) revealed that the homoleptic Au complexes with long oligophenylene linkers can form D3-isomers with strong aurophilic interactions, resulting in higher thermodynamic stability than that of the heteroleptic Au complexes to regulate the outcome of self-sorting.[92−87]

Figure 6

(a,b) Synthesis of [n]CPPs from two different macrocyclic Au complexes, Au- and Au- [(a) x = 3, y = 4 or (b) x = 4, y = 5] (method A), or two different oligophenyldiboronic acids, Lx and Ly [(a) x = 3, y = 4 or (b) x = 4, y = 5] with [AuCl2(dcpm)] (1) (method B). Reagents and conditions: (i) DMF, 50 °C, overnight; (ii) Cs2CO3 (6.0 equiv), toluene/ethanol/water (4:1:1), 50 °C, overnight; (iii) PhICl2 (3.0 equiv), DMF, −60 °C, 0.5 h, and then r.t., overnight. The yields of the CPPs over two steps were determined from the 1H NMR signal intensities relative to an internal standard (1,2,4,5-tetrabromobenzene, TBB). (c,d) 1H NMR spectra of the reaction mixtures after treatment with PhICl2 (400 MHz, CDCl3, 25 °C). The signals were assigned with reference to reports in the literature.[17,62] (e,f) FT-ICR MALDI-TOF MS spectra of the mixture of different macrocyclic Au complexes, Au- and Au- [(e) x = 3, y = 4 or (f) x = 4, y = 5] after mixing at 50 °C in DMF (X = C2H6O or Me2NH).[83]

In the Pt-mediated method for CPP synthesis (Figure c), the formation of CPPs from two different arylene linkers has been achieved by the transmetalation of [PtCl2(cod)] with a mixture of two different bis(trimethylstannyl)arylenes followed by the bromine-induced reductive elimination in 0.7–9.8% yields over two steps, which is a similar procedure to Method B in this study.[46,88] On the other hand, the Au-mediated CPP synthesis outlined in this study represents a new way to an efficient synthesis of [n]CPPs with numbers of phenylene units other than multiples of three via mixing two “preorganized” macrocyclic Au complexes (Method A). The higher yields observed in the current Au-mediated method can be attributed to the higher reversibility of the intermolecular transmetalation of Au(I)–C bonds than that of Pt(II)–C bonds.

Finally, to demonstrate the substrate applicability of our synthetic method, we have synthesized new nanohoops with condensed aromatic rings. The transmetalation of pinB–C6H4–C16H8–C6H4–Bpin (Lpyr)[55] and [Au2Cl2(dcpm)] (1) afforded macrocyclic Au complex (Au-pyr) in an 83% yield. The oxidative chlorination of Au-pyr with PhICl2 produced a nanohoop containing three 2,7-pyrenylene units (Pyr-3) in a 59% yield (Scheme a). The reorganization method between Au-pyr and Au-4 gave three nanohoops with one to three 2,7-pyrenylene units (Pyr-3, Pyr-2, and Pyr-1) along with [12]CPP in 6.7, 15, 19, and ca. 16% NMR yields (Scheme b), respectively, which were isolated by PGPC. Pyr-2 is a structural isomer of the CPP derivative reported by Itami,[90] in which two 2,7-pyrenylene units are introduced on their opposite sides. These results indicate that the reorganization method developed in this study is effective not only for the synthesis of a series of [n]CPPs but also for the synthesis of a variety of CPP derivatives and nanohoops with functional groups introduced at various numbers of substitution and positions.

Scheme 3

Synthesis of Pyrene-Containing Nanohoops from the Macrocyclic Au Complex; (a) Selective Synthesis of Pyr-3; (b) Synthesis of Pyr-3, Pyr-2, and Pyr-1 by the Reorganization Process from Au-pyr and Au-4

Reagents and conditions: (i) Cs2CO3 (6.0 equiv), toluene/ethanol/water (4:1:1), 50 °C, 18 h; (ii) PhICl2 (3.0 equiv), CH2Cl2, −80 °C, 20 min, then r.t., 14 h; (iii) CHCl3, 50 °C, 2 days; (iv) PhICl2 (3.0 equiv), CHCl3, −60 °C, 30 min, then r.t., 2 days; NMR yields of the CPP derivatives in the reorganization method were determined from the 1H NMR signal intensities relative to an internal standard (1,2,4,5-tetramethylbenzene, TMB). The NMR yield of [12]CPP is not precise because the signal overlaps with other CPP derivatives.

Conclusions

We have demonstrated the highly efficient formation of triangular macrocyclic Au complexes, [Au2(C6H4)dcpm]3 (x = 3, 4, 5) by the self-assembly of [Au2Cl2dcpm] and oligophenylene diboronic acids. The choice of the dcpm ligand instead of the dppm ligand is crucial for achieving the highly selective and high-yielding formation of the triangular complexes. Furthermore, the control experiments and theoretical calculation reveal the role of the dcpm ligand, which increases the Au–aryl exchange reaction by electrophilically activating the gold atom of the Au complex. This is the first example to observe the dynamics of the Au–aryl exchange reaction and its ligand effect. Furthermore, the chemical oxidation of the obtained complexes produced the corresponding [n]cycloparaphenylenes ([n]CPPs) (n = 9, 12, 15) in good overall yields. We have also demonstrated the synthesis of [n]CPPs (n = 10, 11, 13, 14) and pyrene-containing nanohoops by the reorganization of two different macrocyclic Au complexes. Applications of dynamic Au–C bonds, including the syntheses of a variety of functionalized CPPs and related nanohoops,[91] are in progress and will be reported in due course.

Methods

General

All manipulations were carried out under an argon atmosphere using standard Schlenk techniques. Solvents and reagents were purchased from TCI Co. Ltd, Wako Pure Chemical Industries Ltd, Kanto Chemical Co. Inc., Sigma-Aldrich Co., and Tanaka Kikinzoku Kogyo K.K. NMR spectra were recorded on a Bruker Biospin AVANCE NEO 400 (400 MHz), Bruker Biospin AVANCE II DPX-400 (400 MHz), Bruker Ascend 400 (400 MHz), or JEOL ECZ400S (400 MHz). IR spectra were obtained using a JASCO FT/IR-4600 (ATR). Elemental analyses were performed using a J-science JM10. The high-resolution mass spectrometry (HRMS) spectra were measured in positive ion mode on a JEOL JMS-S3000 SpiralTOF (MALDI-TOF) or a Fourier transformation-ion cyclotron resonance-mass spectrometer, Bruker solariX (FT-ICR-MS), equipped with a 7 Tesla superconductive magnet by using a matrix-assisted laser desorption/ionization (MALDI) ion source. The single X-ray structure determination was performed on a Rigaku XtaLAB Synergy-DW diffractometer or a Bruker D8 QUEST diffractometer. All calculations were carried out using the Gaussian 16 Rev. B. Program package.

Synthesis of Oligophenyldiboronic Acids (L3, L4, and L5)

4,4″-Terphenyldiboronic acid pinacol ester (L3) was prepared by the Miyaura borylation of 4,4″-dibromoterphenyl with B2pin2. 4,4‴-Quaterphenyldiboronic acid pinacol ester (L4) and 4,4″″-quinquephenyldiboronic acid pinacol ester (L5) were synthesized via a ring extension from 4,4′-p-biphenyldiboronic acid (L2) and L3 by the Suzuki coupling reaction with 2 equivalents of Br–C6H4–B(dan) (dan = 1,8-diaminonaphthalene), followed by the exchange of the protective groups under acidic conditions (see the Supporting Information for detailed procedures).

Synthesis of the Macrocyclic Au Complex, [Au2(C6H4)(Cy2PCH2PCy2)]3 (X = 3: Au-3, x = 4: Au-4, x = 5, Au-5)

A mixture of oligophenyldiboronic acid pinacol ester, pinB–(C6H4)–Bpin (x = 3: L3, x = 4: L4, x = 5, L5) (1.0 equiv), Cs2CO3 (6.0 equiv), and [Au2Cl2(dcpm)] (1) (1.0 equiv) in degassed toluene/H2O/EtOH was stirred overnight at 50 °C under an argon atmosphere. After the reaction mixture was allowed to cool to room temperature, the precipitates were collected by suction filtration, washed with EtOH, and then dried in vacuo. The Au complex, [Au2(C6H4)(Cy2PCH2PCy2)]3 (x = 3: Au-3, x = 4: Au-4, x = 5, Au-5), was obtained as a white solid. Au-3 (x = 3): 72% yield. 1H NMR (399 MHz, C2D2Cl4, r.t.): δ 7.65 (s, 12H, C6H4), 7.64–7.59 (m, 12 H, C6H4), 7.49 (d, 12H, J = 7.8 Hz, C6H4) 2.30–2.21 (br, 12H, C6H11), 2.14–2.07 (br, 18H, C6H11 and CH2), 1.97–1.91 (br, 24H, C6H11), 1.79–1.74 (br, 12H, C6H11), 1.70–1.59 (br, 24H, C6H11, overlapping with the signal of H2O), 1.57–1.46 (br, 12H, C6H11), 1.38–1.31 (br, 36 H, C6H11). 13C{1H} NMR (100 MHz, C2D2Cl4, r.t.): δ 174.0 (t, J = 57.4 Hz), 140.3 (s), 139.9 (s), 136.6 (s), 126.9 (s), 125.2 (s), 35.2 (t, J = 14.9 Hz), 29.8 (s), 29.0 (s), 26.7 (d, J = 20.4 Hz), 25.8 (s). 31P{1H} NMR (161 MHz, C2D2Cl4, r.t.): δ 48.2 (s). IR (ATR): ν = 2922, 2848, 1442, 1355, 1006, 795, 756, 511 cm–1. Anal. Calcd for C129H174Au6P6: C, 50.10; H, 5.67. Found: C, 50.58; H, 5.77. HRMS (FT-ICR MALDI-TOF, DCTB) m/z: calcd. for C129H174Au6P6+H, 3093.0141 [M + H]+; found, 3093.0170. Au-4 (x = 4): 83% yield. 1H NMR (399 MHz, C2D2Cl4, r.t.): δ 7.70 (s, 24H, C6H4), 7.67–7.61 (m, 12H, C6H4), 7.50 (d, 12 H, J = 7.6 Hz, C6H4), 2.31–2.20 (br, 12H, C6H11), 2.15–2.07 (br, 18H, C6H11 and CH2), 1.98–1.90 (br, 24H, C6H11), 1.82–1.75 (br, 12H, C6H11), 1.71–1.60 (br, 24H, C6H11, overlapping with the signal of H2O), 1.59–1.46 (br, 12H, C6H11) 1.40–1.30 (br, 36H, C6H11). 13C{1H} NMR (100 MHz, C2D2Cl4, r.t.): δ 174.3 (t, J = 57.7 Hz), 140.7 (s), 140.3 (s), 138.5 (s), 136.4 (s), 127.1 (s), 127.0 (s), 125.2 (s), 35.2 (t, J = 13.7 Hz), 29.8 (s), 29.0 (s), 26.7 (d, J = 21.4 Hz), 25.8 (s). 31P{1H} NMR (161 MHz, C2D2Cl4, r.t.): δ 48.3 (s). IR (ATR): ν = 2924, 2849, 1447, 1218, 1160, 1004, 803, 760, 513 cm–1. Anal. Calcd for C147H186Au6P6 6H2O: C, 51.39; H, 6.01. Found: C, 51.48; H, 6.13. HRMS (FT-ICR MALDI-TOF, DCTB) m/z: calcd for C147H186Au6P6+H, 3321.1080 [M + H]+; found, 3321.1061. Au-5 (x = 5): 72% yield. 1H NMR (399 MHz, C2D2Cl4, r.t.): δ 7.76–7.69 (br, 36H, C6H4), 7.67–7.59 (br, 12H, C6H4), 7.55–7.48 (br, 12 H, C6H4), 2.21–2.02 (br, 30H, C6H11, CH2), 1.98–1.91 (br, 24H, C6H11), 1.81–1.75 (br, 12H, C6H11), 1.68–1.56 (br, 24H, C6H11, overlapping with signal of H2O), 1.40–1.29 (br, 48H, C6H11). 31P{1H} NMR (161 MHz, C2D2Cl4, r.t.): δ 48.3 (s). IR (ATR): ν 2924, 2850, 1480, 1446, 1006, 801, 756, 512 cm–1. Anal. Calcd for C165H198Au6P6 3H2O: C, 55.00; H, 5.71. Found: C, 55.12; H, 6.04. HRMS (FT-ICR MALDI-TOF, DCTB) m/z: calcd for C165H198Au6P6+H, 3549.2019 [M + H]+; found, 3549.2084.

Synthesis of [n]CPP (n = 9, 12, 15)

To a suspension of the Au complex, [Au2(C6H4)(Cy2PCH2PCy2)]3 (x = 3: Au-3, x = 4: Au-4, x = 5, Au-5), (1.0 equiv) in degassed DMF was added PhICl2 (6.0 mmol/L in DMF, 6.0 equiv) dropwise with stirring at −60 °C for 5 min under an argon atmosphere. The reaction mixture was stirred at the same temperature for 30 min, and then it was allowed to warm to 25 °C and stirred for overnight. The solvent and iodobenzene (byproduct) were removed under vacuum. The crude product was purified by silica gel column chromatography (eluent; CHCl3) to give [n]cycloparaphenylene (n = 9, 12, 15) and [Au2Cl2(dcpm)]. [9]CPP; 78% yield. 1H NMR (400 MHz, CDCl3, r.t.): δ 7.52 (s, 36H, C6H4). 13C{1H} NMR (100 MHz, CDCl3, r.t.): δ 138.1 (s), 127.5 (s). [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP; 78% yield. 1H NMR (400 MHz, CDCl3, r.t.): δ 7.61 (s, 48H, C6H4). 13C{1H} NMR (100 MHz, CDCl3, r.t.): δ 138.7 (s), 127.5 (s). [15]CPP; 88% yield. 1H NMR (400 MHz, CDCl3, r.t.): δ 7.67 (s, 60H, C6H4). 13C{1H} NMR (100 MHz, CDCl3, r.t.): δ 139.0 (s), 127.5 (s).

General Procedure on NMR Experiments for Determination of the Rate Constants

A solution of [Au2Ph2(dcpm)] (Au-HH) (1.0 mg, 1.0 μmol) in 0.3 mL of CDCl3 (0.001% v/v 1,3,5-tris(trifluoromethyl)benzene) was injected into an NMR test tube and then cooled to −20 °C. To the test tube, a cooled solution of [Au2(4-F-C6H4)2(dcpm)] (Au–HF) (1.0 mg, 1.0 μmol) in 0.3 mL of CDCl3 (0.001% v/v 1,3,5-tris(trifluoromethyl)benzene) was added. The test tube was inserted to an NMR spectrometer which was precooled to −20 °C. The progress of the reaction was monitored by NMR measurements.

Synthesis of CPPs from Two Different Oligophenylene Linkers: Reorganization Approach (Method A)

To a suspension of Au- (1.0 equiv) and Au- (1.0 equiv) (x = 3, y = 4 or x = 4, y = 5) in degassed DMF was stirred for 6 days at 50 °C under an argon atmosphere. After the reaction mixture was allowed to cool to −60 °C, PhICl2 (6.0 mmol/L in DMF, 6.0 equiv) was added dropwise with stirring at the same temperature for 5 min. The reaction mixture was stirred at the same temperature for 30 min, and then it was allowed to warm to 25 °C and stirred for 6 h. The solvent and iodobenzene (byproduct) were removed under vacuum. The crude product was purified by silica gel column chromatography (eluent; CHCl3) to give a mixture of CPPs as a yellow solid. 1H NMR analysis indicated the formation of [n]CPPs. Au-3/Au-4 system: [9]CPP (26%), [10]CPP (11%), [11]CPP (17%), and [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP (15%). Au-4/Au-5 system: [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP (13%), [13]CPP (3.0%), [14]CPP (3.8%), and [15]CPP (11%).

Synthesis of CPPs from Two Different Oligophenylene Linkers: Social Self-Sorting Approach (Method B)

A mixture of 4,4″-p-terphenyldiboronic acid pinacol ester (L3) (42.4 mg, 0.088 mmol), 4,4″′-p-quaterphenyldiboronic acid pinacol ester (L4) (49.0 mg, 0.088 mmol), Cs2CO3 (12 equiv), and [Au2Cl2(dcpm)] (1) (2.0 equiv) in degassed toluene/H2O/EtOH was stirred for 4 days at 50 °C under an argon atmosphere. After the reaction mixture was allowed to cool to room temperature, the precipitate was collected by suction filtration and washed with toluene, H2O, EtOH, and then dried in vacuo. The mixture of Au complexes was obtained as a dark-green solid, which was used in the next reaction without further purification. To a suspension of the mixture of Au complexes (1.0 equiv) in degassed DMF was added PhICl2 (12 mmol/L in DMF, 3.0 equiv) dropwise with stirring at −60 °C for 5 min. The reaction mixture was stirred at the same temperature for 30 min, and then it was allowed to warm to 25 °C and stirred overnight. The solvent and iodobenzene (byproduct) were removed under vacuum. The crude product was purified by silica gel column chromatography (eluent; CHCl3) to give a mixture of CPPs as a yellow solid (Rf = 0.76, 0.95 mg). 1H NMR analysis indicated the formation of [n]CPPs. L3/L4 system: [9]CPP (1.1%), [10]CPP (1.7%), [11]CPP (2.5%), and [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP (1.0%). L4/L5 system: [9]-, [12]-, and [18]cycloparaphenylene: carbon nanohoop structures. J. Am. Chem. Soc.. 2008 ">12]CPP (12%), [13]CPP (1.0%), [14]CPP (1.5%), and [15]CPP (7.9%).

Synthesis of [Au2(C6H4–C16H8–C6H4)(Cy2PCH2PCy2)]3 (Au-pyr)

A mixture of pinB-C6H4-C16H8-C6H4-Bpin (Lpyr)[55] (243 mg, 0.40 mmol), Cs2CO3 (792 mg, 2.4 mmol), and [Au2Cl2(dcpm)] (349 mg, 0.40 mmol) in degassed toluene/H2O/EtOH (16/4/4 mL) was stirred for 2 days at 50 °C under an argon atmosphere. After the reaction mixture was allowed to cool to room temperature, the precipitates were collected by suction filtration, washed with toluene (30 mL), H2O (20 mL), and EtOH (20 mL), and then dried in vacuo. The Au complex, [Au2(C6H4–C16H8–C6H4)(Cy2PCH2PCy2)]3 (Au-pyr), was obtained as an off-white solid (383 mg, 0.11 mmol, 83%).

1H NMR and 31P NMR spectra suggest that the two isomers are exchanged slower than the NMR timescale (C/D = 1:1) (see Figure S18 for detail). Data for the dynamic mixture of C2- and D3-isomers of Au-pyr; 1H NMR (400 MHz, CDCl3, r.t.): δ 8.39 (s, 12H, C16H8 for C), 8.07 (s, 12H, C16H8 for C), 7.84 (s, 12 H, C16H8 for D), 7.76 (s, 24 H, C6H4 for C), 7.46 (br, 24 H, C6H4 for D), 7.10 (s, 12H, C16H8 for D), 2.36–2.04 (br, 54H, C6H11, CH2), 2.02–1.86 (br, 24H, C6H11), 1.84–1.48 (br, 12H, C6H11, overlapping with signal of H2O), 1.45–1.23 (br, 48H, C6H11). 31P{1H} NMR (161 MHz, CDCl3, r.t.): δ 48.1 (s, D), 47.9 (s, C). Anal. Calcd for C159H186Au6P6 5H2O: C, 53.72; H, 5.56. Found: C, 53.52; H, 5.37.

Synthesis of Pyrene-Containing Nanohoop, (Pyr-3)

To a suspension of [Au2(C6H4–C16H8–C6H4)(Cy2PCH2PCy2)]3 (Au-pyr) (350 mg, 0.10 mmol) in degassed CH2Cl2 (100 mL) was added PhICl2 (10 mmol/L in CH2Cl2, 30 mL, 0.30 mmol) dropwise with stirring at −80 °C for 20 min under an argon atmosphere. The reaction mixture was stirred at the same temperature for 30 min, and then it was allowed to warm to 25 °C and stirred for 14 h. After the solvent was removed under vacuum, the crude product was purified by silica gel column chromatography (eluent; CHCl3) to give pyrene-containing nanohoop, (Pyr-3) (Rf = 0.87, 62.1 mg, 59 μmol, 59%) as a pale-yellow solid, and [Au2Cl2(dcpm)] (Rf = 0.15, 220 mg, 25 mmol, 84%) as a white solid. 1H NMR (400 MHz, CDCl3, r.t.): δ 8.26 (s, 12H, C16H8), 7.94 (s, 12H, C16H8), 7.78 (d, 12 H, J = 8.4 Hz, C6H4), 7.61 (d, 12 H, J = 8.4 Hz, C6H4). 13C{1H} NMR (101 MHz, CDCl3, r.t.): δ 139.0 (s), 138.7 (s), 137.1 (s), 131.7 (s), 128.0 (s), 127.8 (s), 127.7 (s), 124.0 (s), 123.8 (s). HRMS (MALDI-TOF, DCTB) m/z: calcd for C84H48, 1056.3751 [M]+; found, 1056.3760.

Synthesis of Pyrene-Containing Nanohoops (Pyr-1, Pyr-2, and Pyr-3) by Reorganization Method

A suspension of Au-4 (133 mg, 40 μmol) and Au-pyr (139 mg, 40 μmol) in degassed CHCl3 (80 mL) was stirred for 2 days at 50 °C under an argon atmosphere. After the reaction mixture was allowed to cool to −60 °C, PhICl2 (24 mmol/L in CHCl3, 10 mL, 0.24 mmol) was added dropwise with stirring at the same temperature for 5 min. The reaction mixture was stirred at the same temperature for 30 min, and then it was allowed to warm to 25 °C and stirred for 2 days. After the solvent was removed under vacuum, the crude product was purified by silica gel column chromatography (eluent: CHCl3) to give a mixture of nanohoops as a yellow solid (52.8 mg) and [Au2Cl2(dcpm)] as a white solid (152 mg, 0.17 mmol, 73%). 1H NMR analysis indicated the formation of [12]CPP, Pyr-1, Pyr-2, and Pyr-3 in ca. 16, 29, 15, and 6.7% yields over 2 steps, respectively (the NMR yield of [12]CPP is not precise because the signal overlaps with other CPP derivatives). A portion of the mixture (19.4 mg) was purified by PGPC (eluent: CHCl3), to give Pyr-1 (2.69 mg, 2.8 μmol) and Pyr-2 (2.51 mg, 2.5 μmol) in a pure form. Data for Pyr-1: 1H NMR (400 MHz, CDCl3, r.t.): δ 8.28 (s, 4H, C16H8), 7.99 (s, 4H, C16H8) 7.79 (d, 4H, J = 8.6 Hz, C6H4), 7.64–7.59 (m, 36H, C6H4). 13C{1H} NMR (101 MHz, CDCl3, r.t.): δ 139.3 (s), 138.7 (m), 137.5 (s), 131.9 (s), 129.2 (s), 128.3 (s), 127.9 (s), 127.8 (s), 127.7 (s), 127.5 (m), 124.1 (s). HRMS (MALDI-TOF, Dithranol) m/z: calcd for C76H48, 960.3775 [M+]; found: 960.3751. Data for Pyr-2: 1H NMR (400 MHz, CDCl3, r.t.): δ 8.28 (s, 4H, C16H8), 8.25 (s, 4H, C16H8) 7.97 (s, 8H, C16H8), 7.80 (d, 4H, J = 8.7 Hz, C6H4), 7.76 (d, 4H, J = 8.6 Hz, C6H4), 7.63–7.54 (m, 24H, C6H4). 13C{1H} NMR (101 MHz, CDCl3, r.t.): δ 139.4 (s), 139.1 (s), 138.9 (s), 138.8 (s), 138.8 (s), 138.5 (s), 138.5 (s), 137.5 (s), 137.2 (s), 131.9 (s), 131.8 (s), 129.2 (s), 128.4 (s), 128.3 (s), 128.2 (s), 128.0 (s), 127.8 (s), 127.8 (s), 127.7 (s), 127.6 (s), 127.4 (s), 125.4 (s), 124.1 (s), 124.1 (s), 124.1 (s), 123.9 (s). HRMS (MALDI-TOF, dithranol) m/z: calcd for C80H48, 1008.3751 [M+]; found, 1008.3748.