Development of Metal Nanoparticle Catalysis toward Drug Discovery

Abstract

Transition-metal nanoparticles (NPs) catalysts supported on solid material represent one of the most important subjects in organic synthesis due to their reliable carbon–carbon or carbon–heteroatom bond-forming cross-coupling reactions. Therefore methodologically and conceptually novel immobilization methods for nonprecious transition-metal NPs are currently required for the development of organic, inorganic, green, materials, and medicinal chemistry. We discovered a self-assembled Au-supported Pd NPs catalyst (SAPd(0)) and applied it as a catalyst to Suzuki–Miyaura coupling, Buchwald–Hartwig reaction, Carbon(sp² and sp³)–Hydrogen bond functionalization, double carbonylation, removal of the allyl protecting groups of allyl esters, and redox switching. SAPd(0) comprises approximately 10 layers of self-assembled Pd(0) NPs, whose size is less than 5 nm on the surface of a sulfur-modified Au. The Pd NPs are wrapped in a sulfated p-xylene polymer matrix. We thought that the self-assembled Au-supported Pd NPs could be made by in situ metal NP and nanospace simultaneous organization (PSSO). This methodology involves 4 kinds of simultaneous procedures: i) reduction of a higher valence metal salt, ii) growth of metal NPs with appropriate size, iii) growth of a matrix with appropriate pores, and iv) wrapping of the metal NPs by matrix nanopores. This methodology is different from previously reported metal NPs-immobilizing methods, which use solid supports with preformed pores or coordination sites. We also applied the in situ PSSO method to prepare various immobilized transition-metal NPs, including base metals. For example, the in situ PSSO method can be applicable to easily prepare Ni, Ru, and Fe NPs with good recyclability and low metal leaching for use in organic synthesis.

1. Introduction

The development of metal nanoparticles (NPs) catalyst supported on solid support as catalysts for carbon–carbon bond or carbon–heteroatom bond forming cross-coupling reactions is one of the most important areas in medicinal chemistry as well as organic synthesis. These catalysts are totally different from homogeneous catalytic systems, which usually use ligands such as phosphine and nitrogen-containing heterocyclic carbene (NHC), in the meaning of recyclability and low-metal contamination.^1–4) Conventional NPs catalysts supported on solid support have usually been prepared using various supports with preformed pores or coordination sites, such as functionalized polymers, silica materials, metal oxides, ionic liquids, and dendrimers.²⁾ The pores or coordination sites trap metal NPs, and act as templates to immobilize and stabilize NPs by coordination, adsorption, ionic bonding, or encapsulation, along with preventing aggregation (Fig. 1). Immobilized Pd NPs catalysts have been investigated for various carbon–carbon and carbon–heteroatom bond-forming cross-coupling reactions.^5–7) On the other hand, although non-Pd catalyzed cross-coupling is important to synthesize functional molecules from the viewpoints of minimizing this precious metal and catalyst cost, the use of heterogeneous catalysts containing other metal NPs for cross-coupling reactions is limited.^8–10) One reason for this problem is that methodology for immobilizing these metal NPs has not yet been developed. Hence we envisioned that investigation of methodologically and conceptually novel methods for non-precious transition-metal NPs immobilization is necessary for the future development of organic, inorganic, green, materials, process, and medicinal chemistry.

Fig. 1. Methods to Immobilize NPs on Solid Supports via: a) Adsorption, b) Ionic Bonding, c) Coordination, and d) Encapsulation

In this review, we summarize our recently developed novel Pd NPs catalyst, named self-assembled Au-supported palladium (SAPd(0)), which can repeatedly catalyze Suzuki–Miyaura coupling, Buchwald–Hartwig reaction, carbon(sp² and sp³)–hydrogen functionalization, double carbonylation, removal of the allyl protecting groups of allyl esters, and redox switching without any ligand.^11–22) We also introduce the chemical and structural properties of SAPd(0) by spectroscopic analysis.²³⁾ Transmission electron microscopy (TEM) analyses and extended X-ray absorption fine structure (EXAFS) experiments were performed to clarify the geometrical properties of Pd catalysis in SAPd(0). X-ray absorption near-edge structure (XANES) analysis²⁴⁾ at S and C K-edge were used to determine the chemical states of sulfur and carbon in SAPd(0). As a consequence, we found that SAPd(0) has approximately ten layers of self-assembled Pd(0) NPs, a size of less than 5 nm, on Au surface. These Pd NPs are wrapped in a sulfated p-xylene polymer matrix. We speculated that the self-assembled Pd NPs were constructed by in situ metal NPs and nanospace simultaneous organization (PSSO), which is illustrated in Fig. 2. It involves four simultaneous procedures: i) reduction of a high-valence metal source, ii) growth of transition-metal NPs of appropriate size, iii) growth of a matrix with appropriate pores, and iv) encapsulation of the metal NPs in matrix nanopores. This method is conceptually different from conventional NPs-immobilizing methods using solid supports with preformed coordination sites or pores. We also successfully applied the “in situ PSSO method” to prepare other metal NPs catalysis, such as self-assembled Au-supported Ni NPs catalyst (SANi(0)),²⁴⁾ self-assembled Au-supported Ru NPs catalyst (SARu(0)),²⁵⁾ and self-assembled Au-supported Fe NPs catalyst (SAFe(II)).²⁶⁾

Fig. 2. The in Situ Metal NPs and Nanospace Simultaneous Organization (PSSO) Method

2. Self-assembled Au-Supported Palladium NPs Catalyst (SAPd(0))

2.1. Preparation of SAPd(0) and Its Application to Suzuki–Miyaura Coupling^{11, 12, 16)}

Research involving self-assembled monolayers (SAMs) of alkylthiols (RSH) on Au surfaces has rapidly expanded due to the ease of their preparation, the potential applications of these materials, and their facile connection between organic moieties and metal surfaces. However, since the role of the solvent in the formation of SAMs is not yet well understood, and studies on the thermal stability of SAMs are limited, questions still remain concerning the mechanism of monolayer formation. Piranha solution had been traditionally used to clean Au surfaces, prior to the emergence of SAMs. In 2010, we offered new insight into the treatment of Au with Piranha solution to induce sulfur modification. We showed the application of this process to the development of novel Pd materials that exhibited both remarkably low Pd-leaching and high recyclability in the Suzuki–Miyaura coupling.¹¹⁾

In our continuous research to develop an environmentally benign catalyst toward drug discovery,^27–34) we unexpectedly found in a Synchrotron radiation hard X-ray photoelectron spectroscopy (SR-HXPS) analysis that Piranha treatment can place a sulfur atom on the Au surface. We analyzed Piranha-treated Au(111)/mica using SR-HXPS and found, to our surprise, that the S 1s peak appeared near 2478 eV in the spectrum as shown in Fig. 3 line 5. On the other hand, no S 1s peak was observed on Au(111)/mica before Piranha treatment. These results indicate that Piranha treatment had modified the surface of Au(111)/mica. Although Piranha treatment has been traditionally and frequently used to remove impurities from the Au surface, our results strongly suggest that this treatment could also be employed for sulfur modification of the Au surface. To the best of our knowledge, this is the first report that Piranha treatment attaches S atoms to the Au surface.

Fig. 3. Sulfur 1s Core-level Photoemission Spectra

(Color figure can be accessed in the online version.)

As shown in line 4, after the Pd adsorption, the S 1s peak appears at 2470 eV, which is almost identical to the binding energy of the S 1s peak from the S-modified GaAs.³²⁾ Significant changes in S 1s binding energies between those in the Piranha-treated Au(111)/mica and those in the sample A indicate that the sulfur on Au was reduced and had chemically bonded with Pd during Pd adsorption. In Pd 2p core-level photoemission spectra, the peaks from A appear close to the metallic Pd peak. These results suggest that, in the case of A, Pd(dba)₂ molecules were changed to zerovalent molecules or metallic Pd, as summarized in Fig. 4.

Fig. 4. Palladium 2p_3/2 Core-Level Photoemission Spectra

(Color figure can be accessed in the online version.)

Single crystal Au(111)/mica is not versatile. Eventually, instead of Au(111)/mica, we used Au foil and mesh to prepare B and C according to the Pd-adsorption procedure for A. When A, B, or C was subjected to the Suzuki–Miyaura coupling of 1a and 2a, the yields of the product 3a for runs 1 to 10 were excellent to quantitative in all cases, as shown in Table 1, entries 1–3. Entry 4 indicates that Material D, which was prepared from Au(mesh) with Pd(OAc)₂ as a Pd-source, was also highly active for the Suzuki–Miyaura coupling. These results show that the Pd materials A, B, C, and D are highly reactive and recyclable in this reaction.

Table 1. Yields of Suzuki–Miyaura Coupling Using Pd Materials A–D

Entry	Pd material	Support	Pd source	Yield of 3a (%) a)
				1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
1	A	Au(111)/mica	Pd(dba)2	96	95	97	96	98	99	>99	96	96	99
2	B	Au(foil)	Pd(dba)2	97	96	92	89	97	97	97	99	98	95
3	C	Au(mesh)	Pd(dba)2	>99	>99	>99	>99	>99	>99	>99	>99	>99	>99
4	D (SAPd(0))	Au(mesh)	Pd(OAc)2	>99	98	>99	>99	>99	96	>99	97	99	>99

Conditions of Suzuki–Miyaura coupling: 1a (102 mg, 0.5 mmol), 2a (1.5 eq.), K₂CO₃ (2 eq.), EtOH (3 mL), Pd material, 80°C, 12 h. a) Determined by HPLC.

Next, we measured the amount of Pd on the Pd materials B–D using inductively coupled plasma mass spectrometry (ICP-MS) analysis. These analyses showed that in B, C, and D Pd comprises 95, 80, and 38 µg, respectively. Then, we measured the amount of Pd released into the reaction mixtures of each Suzuki–Miyaura coupling run by ICP-MS. The amount of Pd in the reaction mixture using B–D is far lower than the U.S. government-required value of <5 ppm residual metal in product streams.³⁵⁾ In particular, in the case of D, the leached Pd into the reaction mixture was only 76–38 ng for 1 mmol scale preparation (12.7–6.3 ppb in 3 mL of solvent, 0.2–0.1% of Pd from D), the average being 26 ng for 10 runs. Due to its extremely low Pd leaching levels, we employed D, i.e., SAPd(0) (self-assembled Au-supported Pd material) for further studies.

We then continued our experiments to elucidate the scope and limitation of SAPd(0) in the Suzuki–Miyaura coupling using other aryl iodides and arylboronic acids. As shown in Table 2, the corresponding products were obtained in excellent yields in all entries. It should be noted that when control experiment of entry 9 was carried out using SAPd(0) without Piranha treatment, isolated yields (%) of 3i between 1st run and 4th run were 54, 46, 8, and 5, respectively. From these control experiments, we know that Piranha treatment is necessary and sulfur is needed to create an active catalyst and retain Pd on the surface.

Table 2. Suzuki–Miyaura Coupling of Various Substrates Using Pd Material D, SAPd(0)

Entry	Substrate [Ar-X]			Boronic acid			Yield of 3 (%)a,c)
	Ar		X	Y			1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
1b)	1a	C6H5	I	2a	Cl	3a	>99 (36)	98 (26)	>99 (25)	>99 (28)	>99 (24)	97 (28)	>99 (19)	97 (30)	99 (25)	>99 (23)
2			I	2b	Me	3b	94 (55)	>99	92	97	99	94	90	92	>99	94 (47)
3			I	2c	H	3c	92 (43)	92	92	90	91	91	85	87	91	89 (36)
4			I	2d	Ac	3d	99	98	>99	98	97	99	99	98	97	99
5	1b	4-NO2C6H4	I	2a	Cl	3e	96 (1307)	96	90	93	>99	96	94	>99	99	>99 (295)
6			I	2b	Me	3f	93	96	99	96	88	99	94	95	>99	95
7			I	2c	H	3g	97	99	90	>99	98	96	96	98	99	95
8			I	2d	Ac	3h	96	>99	89	87	95	92	98	98	94	98
9b)	1c	4-MeOC6H4	I	2a	Cl	3i	96 (641)	97	98	96	93	97	95	97	96	92 (54)
10			I	2b	Me	3j	>99	>99	>99	>99	95	>99	>99	>99	>99	96
11			I	2c	H	3k	99	99	96	91	98	95	>99	98	96	93
12b)			I	2d	Ac	3l	97	94	99	99	99	95	94	96	97	92
13	1d	4-NO2C6H4	Br	2a	Cl	3e	>99 (482)	94	>99	>99	>99	>99	>99	>99	>99	99 (116)
14	1e	4-Ac-C6H4	Br	2a	Cl	3m	95	97	99	99	99	93	81	95	86	80
15	1f	2-Me-C6H4	I	2b	Me	3n	96	97	93	94	92	92	91	99	95	89
16	1g	2-HO-C6H4	I	2a	Cl	3o	96	97	93	94	92	92	91	99	95	89

a) Isolated yields. b) Yields were determined by HPLC. c) Numbers in parenthesis indicate the amount of releasing-Pd(ng) in the solution.

Hence in the SR-HXPS measurement of Piranha-treated Au(111)/mica, we found that the Au surface underwent sulfur modification during this treatment, which was traditionally believed to have only removed impurities from the Au surface. We also successfully developed a practical Pd material, SAPd(0), whose Pd was immobilized on Au. This is one of the best Pd materials thus far developed, with high recyclability and lowest Pd-releasing levels as well as excellent catalytic activities.

Furthermore, in 2013, we developed a safer heat generation-controllable preparative method of SAPd(0).¹⁶⁾

2.2. Buchwald–Hartwig Reaction Using SAPd(0)¹³⁾

Arylamines are widely used in the synthesis of pharmaceutical intermediates, natural products, agrochemicals, artificial dyes, and polymers in industrial and academic laboratories. Also, arylamines are useful as ligands for transition metal catalysis.

Pd-catalyzed cross-coupling of tin amides and aryl halides to generate arylamines in homogeneous systems was first studied by Migita and colleagues.³⁶⁾ Buchwald and Hartwig groups were developed by this Pd-methodology under relatively mild conditions.³⁷⁾ Until now, a variety of effective ligands have been reported for the development of Pd-catalyzed aromatic aminations. These ligands are designed to prevent aggregation of the highly active Pd(0) species, to accelerate oxidative addition, to enhance reductive elimination, and occasionally to prevent β-hydride elimination.³⁸⁾ Nonetheless, the use of ligands has some obvious disadvantages. Ligands sometimes hamper the purification and isolation of the desired products and they are toxic and water and/or air sensitive. To solve these issues and meet environmental, safety, and economical standards, the development of ligand-free Pd-catalyzed aromatic aminations is desired.

We performed many experiments to set up the SAPd(0) catalyzed reaction conditions for the cross-coupling between bromobenzene 1h and morpholine 4a. Finally, as shown in Table 3, by setting the exact completion of the reaction time to 7 h, a high yield of the product 5a was repeatedly obtained from the 1st to 10th cycles.³⁹⁾ Thus optimization of ligand-free SAPd(0)-catalyzed coupling of 1h (1.0 equiv) and 4a (1.4 equiv) has been succeeded by the use of KOt-Bu (1.4 equiv) as the base in xylene (1.0 mL) at 130°C for 7 h.

Table 3. Optimization of SAPd(0)-Catalyzed Amination Reaction

5a (Yield %)a)
1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
93	93	92	93	92	92	92	92	91	91

a) Isolated yields.

We next measured the leached Pd in the reaction mixture and also the amount of immobilized Pd in SAPd(0) by ICP-MS. The immobilized Pd in SAPd(0) was measured before and after the 10th cycle of the optimized reactions. The immobilized Pd in SAPd(0) before and after the reactions was 79 ± 11³⁹⁾ and 68 ± 18 µg,⁴⁰⁾ respectively. The amount of released Pd after cooling in each cycle was extremely low (61–588 ng [0.06–0.6 ppm⁴¹⁾] for a 0.32 mmol scale reaction)⁴²⁾ and the amount of Pd is far lower than the U.S. government-required value of residual metal in product streams.³⁵⁾ Moreover, SAPd(0) is highly recyclable (more than 10 cycles). Thus as far as we know, being a Pd-catalyzed aromatic amination, this is not only the first ligand-free example but also the best example of a highly recyclable (more than 10 cycles) and low-leaching (less than 0.6 ppm) catalyst using immobilized Pd.

The above results show that SAPd(0) is an excellent Pd reservoir during the amination to release traces of highly reactive Pd (less than 0.6 ppm) in the reaction mixture.⁴³⁾ Therefore the low Pd concentrations suppress the aggregation of Pd and allow the reaction to be effectively catalyzed. Thus in Fig. 5, a general scheme for the SAPd(0)-catalyzed ligand-free Buchwald–Hartwig amination can be postulated.

Fig. 5. General Scheme for Ligand-Free Aromatic Aminations Catalyzed by SAPd(0) without Aggregation

(Color figure can be accessed in the online version.)

It should be noted that SAPd(0) catalyzed the aromatic amination of 1h with various amines without losing catalytic activity from the 1st to 10th cycles, as shown in Table 4. For instance, when 1h was coupled with cyclic amines, i.e., morpholine 4a, 1,4-dioxa-8-azaspiro[4.5]decane 4b and piperidine 4c (entries 1–3), the average yields of products 5a, 5b, and 5c were 92, 91, and 96%, respectively. In entries 4 and 5, the reactions of an acyclic secondary amine, n-dibutylamine 4d and an aromatic secondary amine, N-methylaniline 4e with 1h yielded the corresponding coupling products 5d and 5e in average yields 97 and 92%, respectively. Next, we explored the reaction of 1h with primary amines, such as benzylamine 4f and cyclohexylamine 4g (entries 6 and 7), in which the corresponding monoarylated products 5f and 5g were obtained in average yields 88 and 91%, respectively.

Table 4. Aromatic Aminations of Various Aryl Bromides with a Variety of Amines Using SAPd(0)

a) Isolated yields. b) The results described in Table 3. c) Monoarylated product was isolated. d) 3.6 equiv. of amine and 4.0 equiv. of KOt-Bu were used to give the corresponding triaminated product.

We next examined the coupling reactions among other aryl bromides with 4a. When 4-bromoanisole 1i, having an electron-donating methoxy group, was treated with 4a, the corresponding product 5h was isolated in an average yield 91%, as shown in entry 8. 4-Bromobenzonitrile 1j, with an electron-withdrawing cyano group, also coupled with 4a, where the average yield of the product 5i was 88%, as shown in entry 9. In the reaction of a fused aryl bromide, 2-bromonaphthalene 1k and 4a, the coupled product 5j was isolated in 91% average yield (entry 10). Interestingly, the aryl tribromide, 1,3,5-tribromobenzene 1l underwent the coupling reaction with 4a to give the triaminated product 5k in an average yield 85%, as shown in entry 11. Heterocyclic bromides also coupled with amine to give the corresponding products in excellent yields (entries 12–13). Under our reaction conditions, the reaction of aryl iodides gave the corresponding product in lower chemical yield than that of aryl bromide.

Although a few heterogeneous Pd-catalyzed aminations of aryl chlorides as the substrate have been reported,^44–46) these reactions are not ligand-free, and the catalysts used are not highly recyclable (>10 times) nor low leaching (<1 ppm). Thus as shown in Table 5, we next examined the ligand-free cross-coupling of the less reactive chlorobenzene 1o using SAPd(0) repetitively for 10 cycles. When 1o was treated with 4a or 4b, according to the optimized conditions (Table 3), after 12 h, the corresponding coupling products 5a and 5b were successfully obtained with average yields 88 and 93%, respectively (entries 1 and 2).

Table 5. Aromatic Aminations of Chlorobenzene 1o with Amines, 4a and 4b Using SAPd(0)

SAPd(0) has a much smaller surface area and could absorb many fewer organic compounds, including starting materials or products, than those of polymer-supported Pd, due to its low affinity for organic molecules. Moreover, the Au mesh, used as the support for Pd in SAPd(0), is malleable and easy to handle with a pair of tweezers. Considering the above-mentioned advantages, we tried liquid-phase combinatorial synthesis with SAPd(0) by changing the aromatic chloride derivatives or the amines as the substrates.

As shown in Table 6, when SAPd(0) was used in the Buchwald–Hartwig reaction of phenyl chlorides 1 and three different amines 4 with tBuOK in xylene, the yields for runs 1 to 7 were excellent to quantitative in all cases. In these reactions, pure products were obtained after the usual aqueous workup. Therefore SAPd(0) was demonstrated an effective Pd reservoir for this kind of synthesis without any contamination of other products or starting materials.

Table 6. Example of Liquid-Phase Combinatorial Chemistry; Aromatic Aminations of Chlorobenzenes 1 with Amines 4 Using SAPd(0)

We have presented a new and efficient method for ligand-free aromatic aminations catalyzed by SAPd(0), an excellent reservoir of Pd, where various aryl bromides and chlorides were effectively coupled with a variety of amines. In the reactions, only a trace amount of the active Pd species (less than 0.6 ppm for a 0.32-mmol scale reaction, one of the lowest recorded in the literature) was leached from SAPd(0). Moreover, the low leaching property of SAPd(0) makes it recyclable for more than 10 cycles without any loss of catalytic activity.

2.3. Carbon(sp²)–Hydrogen Bond Activation Using SAPd(0)¹⁸⁾

The metal-catalyzed direct functionalization of aromatic C–H bonds is one of the most common and straightforward methods for the formation of carbon–carbon and carbon–heteroatom bonds in organic synthesis.⁴⁷⁾ Larock and colleagues⁴⁸⁾ and Gallagher and colleagues⁴⁹⁾ reported the development of a 1,4-Pd migration reaction for the synthesis of fused polycycles. The key step in this process involves the Pd-catalyzed activation of a C–H bond via a five-membered palladacycle intermediate.⁵⁰⁾ In 2011, Ren’s group⁵¹⁾ reported a 1,7-Pd migration/cyclization/dealkylation sequence to give benzotriazoles 7 (Chart 1), which exhibit inhibitory activity towards indoleamine 2,3-dioxygenase (IDO). IDO is an important new therapeutic target for the treatment of cancer.⁵²⁾

Chart 1. 1,7-Pd Migration Reaction Reported by Ren and Colleagues⁵¹⁾

Immobilized Pd NPs have been used effectively for the activation of C–H bonds. Although there have been several reports on the use of Pd-NPs as catalysts for activation of C(sp²)–H bonds,⁵³⁾ there have been none describing the use of Pd-NPs for Pd-catalyzed activation of C–H bonds with the activation occurring via a key five-membered palladacycle intermediate.

Against this background, we found for the first time Pd-NP-catalyzed ligand-free 1,7-Pd migration/cyclization/dealkylation sequence for the regioselective synthesis of N-substituted benzotriazoles.

To establish appropriate ligand-free reaction conditions using the Pd-NP catalyst SAPd(0) instead of Pd(OAc)₂, as shown in Table 7, we screened SAPd(0) as a catalyst for the conversion of 6a to 7a in the presence of various oxidants, because most C–H activation reactions are catalyzed by a Pd(II) species, and the Pd-NPs in SAP(0)d are in the Pd(0) oxidation state. When a solution of 6a (0.16 mmol), KOAc (1.2 eq.) and SAPd(0) (11 × 14-mm mesh, approx. 50 µg of Pd-NPs immobilized on the surface) in N,N-dimethylformamide (DMF) was heated at 110°C for 1 d in the presence of air, oxygen, AgOAc, or benzoquinone as an oxidant, the reaction did not proceed at all (Table 7, entries 1–4). However, when the reaction was performed with PhI(OAc)₂ as the oxidant at 110°C for 1 d, compound 7a was isolated in 14% yield, with 63% of the starting material 6a (Table 7, entry 5). Subsequent optimization of the reaction conditions revealed that increasing the amount of base to 3 eq as well as increasing the reaction temperature and reaction time to 120°C and 3 d, respectively, led to an increase in the isolated yield of product 7a to 92% (Table 7, entries 5–9).

Table 7. Ligand-Free Synthesis of Benzotriazoles via 1,7-Pd Migration Reaction of Triazine 6a Using SAPd(0) as Catalyst

Entry	Oxidant	KOAc (eq.)	Temp. (°C)	Time (d)	Yield of 7a (%)a)	Unreacted 6a (%)a)
1	Airb)	1.2	110	1	0	99
2	O2b)	1.2	110	1	0	99
3	AgOAc	1.2	110	1	0	99
4	Benzoquinoline	1.2	110	1	0	99
5	PhI(OAc)2	1.2	110	1	14	63
6	PhI(OAc)2	1.2	120	1	32	65
7	PhI(OAc)2	1.2	120	3	39	27
8	PhI(OAc)2	3.0	120	3	92	4
9	PhI(OAc)2	5.0	120	3	89	trace

a) Isolated yield. b) An excess of the gas oxidant was used.

To examine the scope of this SAPd(0)-catalyzed 1,7-Pd migration reaction, we repeated the reaction using triazenes 6b–i (Table 8) as substrates under optimized conditions described in Table 7, entry 9. All these reactions successfully proceeded as expected to give the corresponding products 7b–e. The aryliodide substrates 6f–j were more reactive than the arylbromides 6a–e, because the corresponding products were formed over a shorter reaction time. Also, the arylbromide substrates having a Cl or H at the para-position (i.e., 6d or 6c) gave the corresponding products in low yields (49 and 18%, respectively), whereas the corresponding aryliodide substrates 6i and 6j gave the same benzotriazole products in higher yields (73 and 67%, respectively), over a shorter reaction time.

Table 8. Scope of 1,7-Pd Migration Reaction Using SAPd(0) as Catalyst

Entry	Substrate			Time (h)	Yield of 7 (%)a)
		R	X
1	6a	CO2Et	Br	72	7a (89)
2	6b	CN	Br	36	7b (80)
3	6c	F	Br	96	7c (70)
4	6d	Cl	Br	96	7d (49)
5	6e	H	Br	96	7e (18)
6	6f	CO2Et	I	26	7a (79)
7	6g	CN	I	10	7b (76)
8	6h	F	I	36	7c (67)
9	6i	Cl	I	26	7d (73)
10	6j	H	I	17	7e (67)

a) Isolated yield.

With a novel series of 1-aryl-benzotriazoles 7a–e in hand, we studied their activity using a colorimetric in vitro IDO inhibition assay.⁵⁴⁾ The results of compounds 7a–e at 100 µM revealed that none of these compounds inhibited IDO at more than 30% at this concentration. Among all compounds tested, 7a showed the highest inhibitory activity (Fig. 6).

Fig. 6. Inhibition of IDO at 100 µM by Compounds 7a–e

Hence we have developed a new strategy for the synthesis of benzotriazoles using a Pd-NP-catalyzed 1,7-Pd migration reaction. This reaction proceeded via activation and functionalization of a carbon(sp²)–hydrogen bond. It should be noted that this reaction required the addition of an oxidant and provided good to excellent yields of the desired 1-aryl benzotriazole products. Also, the inhibitory activities of these compounds towards IDO were evaluated in vitro. This procedure could be used in several other systems for the activation of carbon(sp²)–hydrogen bonds, because it involves the unique combination of Pd-NPs with a hypervalent iodine reagent.

2.4. Carbon(sp³)–Hydrogen Bond Activation Using SAPd(0)¹⁷⁾

Among various pathways, Pd-catalyzed carbon(sp³)–hydrogen bond activation has potential applications and advantages due to its controllable selectivity and good reactivity compared with other transition-metal-catalyzed carbon(sp³)–hydrogen bond activations.⁵⁵⁾

The use of heterogeneous catalysts such as Pd NPs supported on solid material is a significant development in carbon–hydrogen bond activation, because the catalysts can be recovered and reused several times. When we started this project, there were no reports Pd NPs catalyzed carbon(sp³)–hydrogen bond activation.

In 2011, under homogeneous conditions, Chatani’s group⁵⁶⁾ reported the Pd-catalyzed direct ethynylation of carbon(sp³) atoms positioned β to the amide carbonyl bonds in aliphatic carboxylic acid derivatives (Chart 2).

Chart 2. Pd-Catalyzed Direct Ethynylation of Carbon(sp³) Atoms

Our research target was to accomplish the same carbon(sp³)–hydrogen activation for the same conversion using Pd NPs instead of Pd(OAc)₂.

In the beginning, to achieve Pd-NP-catalyzed unactivated C(sp³)–H bond functionalization, we examined the reaction of aliphatic amide 8a and bromoalkyne 9 (Table 9). When the reaction was carried out under the conditions of conventional system using SAPd(0) instead of Pd(OAc)₂, after 66 h of reaction time, the desired ethynylated product 10a was isolated in 41% yield along with 53% of unreacted 8a (entry 1). Next, solvent screening was performed to determine the appropriate reaction conditions. Only a trace of product 10a was isolated, with EtOH, dichloroethane, or mesitylene as the solvent. We then performed the reaction in xylene at 135°C for 72 h to give 10a in 44% isolated yield; the dialkynylated product was generated in 11% isolated yield and 40% of 8a was recovered (entry 2). We found that when the amount of xylene was increased to 1.5 or 2 mL, the isolated yield of product 10a increased slightly (entries 3 and 4). The isolated yield of 10a decreased, when 3 mL of xylene was used (entry 5). At higher temperature, 135°C (entries 2–5), the corresponding dialkynylated product was obtained as a byproduct. We carried out the reaction at lower temperatures to minimize dialkynylated product formation, but the isolated yield of 10a gradually decreased (entries 6 and 7). When we used xylene (1.5 mL) as the solvent at 135°C, the isolated yields and yields calculated based on recovered amides for product 10a were excellent (entry 8).

Table 9. Effects of Solvent, Solvent Amount, and Temperature

Entry	Solvent (mL)	Additive (equiv)	Temp. (°C)a)	Time (h)	Yield (%)b)
					10a	Dialkynylated product	8a
1	Toluene (1.0)	LiCl (1.0)	110	66	41	—	53
2	Xylene (1.0)	LiCl (1.0)	135	72	44 (89)	11	40
3	Xylene (1.5)	LiCl (1.0)	135	24	51 (71)	12	32
4	Xylene (2.0)	LiCl (1.0)	135	24	52	24	14
5	Xylene (3.0)	LiCl (1.0)	135	24	44	16	35
6	Xylene (2.0)	LiCl (1.0)	120	44	28	—	58
7	Xylene (2.0)	LiCl (1.0)	110	41	20	—	60
8	Xylene (1.5)	LiCl (2.0)	135	24	51 (90)		43
9	Xylene (1.5)	LiCl (3.0)	135	24	40 (91)		55

a) Bath temperature. b) Isolated yields, and yields calculated based on recovered amides 8a in parentheses.

In the presence of AgOAc as the oxidant, SAPd(0) released too much Pd in the reaction mixture; therefore we divided the reaction conditions into two steps, to control Pd leaching from SAPd(0) to the amount needed for catalyzing carbon(sp³)–hydrogen bond activation. The first step was Pd leaching from SAPd(0) in contact with amide 8a and bromoalkyne 9; the second step was carbon(sp³)–hydrogen bond activation and ethynylation (Table 10). After the first step, we removed the SAPd(0) from the system, and it was washed and used for subsequent cycles. 8a (0.206 mmol), 9 (0.75 equiv), and SAPd(0) in xylene (1.5 mL) were heated at 135°C for 6 h then SAPd(0) was removed. To the reaction mixture, AgOAc (1.2 equiv), LiCl (2.0 equiv), and bromoalkyne 9 (0.75 equiv) were added, and the mixture was stirred for 18 h to give 10a in 53% isolated yield (entry 1). The reaction in the second step finished within 10 h, but the isolated yield of 10a decreased after the third cycle (entry 2). When other conditions were kept the same but the reaction mixture in the first step was heated for 4 and 5 h, 10a was obtained in 39 and 52% isolated yields, respectively (entries 3 and 4). In entry 4, heating for 5 h caused too much Pd leaching from SAPd(0) in cycles one to three, which is why the isolated yields of 10a decreased after the 4th cycle. Finally, we optimized the conditions. When a solution of 9 (0.90 equiv), 8a (0.206 mmol), and SAPd(0) was heated at 135°C for 4 h in the first step, and the conditions in the second step remained the same, we successfully recycled SAPd(0) 10 times, although the isolated yields gradually decreased after the fifth cycle (entry 5). We obtained moderate-to-low isolated yields of the desired alkynylated products, but yields calculated based on recovered amides in 10th repeated use were almost the same as those under homogeneous conditions.⁵⁶⁾

Table 10. Recycling of SAuPd from First to 10th Cycles

Entry	Conditions			Yield of 10a (%)a)
	9 (equiv)	Reaction time (h)		1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
		1st step	2nd step
1	0.75	6	18	53 (89)
2	0.75	6	10	52 (89)	53 (91)	42 (82)	23 (70)
3	0.75	4	10	39 (89)	31 (84)
4	0.75	5	10	52 (88)	52 (88)	52 (88)	48 (86)	30 (88)	17 (86)
5	0.90	4	10	53 (91)	52 (89)	52 (87)	51 (92)	50 (91)	42 (85)	37 (78)	33 (83)	25 (86)	16 (84)

a) Isolated yields, and yields calculated based on recovered amides 8a in parentheses.

To know the real loading of Pd catalyst used in these reactions, we measured the amount of Pd in SAPd(0) and the leached Pd in the reaction mixture, using ICP-MS. The amounts of Pd in SAPd(0) were determined before and after the 10th reaction cycle, under the optimized conditions. The amounts of immobilized Pd in SAPd(0) before and after the reactions were 78 ± 41 and 7 ± 4 µg, respectively. The amount of Pd released after cooling in each cycle was low; the turnover numbers (TONs) for carbon(sp³)–hydrogen bond activation were high (233–1710 ng [1–10 ppm] for 0.206 mmol scale reaction. TON = 6790 − 26045, calculated based on the amount of leached Pd in the reaction mixture and isolated yield of 10a).

To generalize the above SAPd(0)-catalyzed direct carbon(sp³)–hydrogen bonds functionalization, we examined the catalytic activities of SAPd(0) in the reactions over 10 cycles, using amides 8b–8i (Table 11). We used the strategy used for SAPd(0) recycling as shown in Table 10, entry 5. All substrates gave beta-functionalized products. Substrates with sterically demanding cyclohexyl and phenyl groups, 8b and 8c, were successfully converted to the corresponding product (entries 1 and 2). Furthermore, the alkynylations tolerated a wide range of functional groups such as ethers 8d and 8h (entries 3 and 7), esters 8e (entry 4), and halides 8f (entry 5). The carbon(sp³)–hydrogen bonds adjacent to the benzylic position (8g and 8h) similarly took part in alkylation (entries 6 and 7). The γ-aminobutyric acid derivative 8i was also converted to the corresponding alkynylated product, and the N-Boc and N-benzyl protecting groups were intact (entry 8).

Table 11. Direct Ethynylation of Amides 8b–8i Using SAPd(0), via C(sp³)–H Bond Functionalization

a) Isolated yields, and yields calculated based on recovered amides 8 in parentheses.

Hence we developed a new strategy for Pd-NP-catalyzed direct ethynylation of aliphatic carboxylic acid derivatives via carbon(sp³)–hydrogen bond functionalization at the beta positions of amides, having 8-aminoquinoline as a directing group. We obtained moderate yields of the desired alkynylated products, but yields calculated based on recovered amides were almost the same as those obtained under homogeneous conditions. Because of the low leaching properties of SAPd(0), the catalyst could be recycled more than 10 times. This protocol shows that Pd NPs are effective catalysts for carbon(sp³)–hydrogen bond activation.

2.5. Double Carbonylation Using SAPd(0)

α-Ketoamides are found as valuable fragments in biologically active molecules⁵⁷⁾ as well as important synthons for the synthesis of nitrogen-containing compounds,⁵⁸⁾ and a number of preparative methods for these compounds have been developed.⁵⁹⁾

In 1982, independently, Yamamoto and colleagues⁶⁰⁾ and Kobayashi and Tanaka⁶¹⁾ reported the first Pd-catalyzed double carbonylation of aryl halides with amines giving α-ketoamides in high yields (Chart 3). Since then, double carbonylation has been applied to the synthesis of a number of molecules, and demonstrated a promising and indispensable methodology for the synthesis of α-ketoamides.⁶²⁾

Chart 3. Pd-Catalyzed Double Carbonylation

(Color figure can be accessed in the online version.)

Pd-catalyzed double carbonylation proceeds efficiently under high pressure of carbon monoxide (CO). Recently, several groups reported double carbonylation under an atmospheric pressure of CO.^63–67) In these reactions, however, they had to use some specific additives, for example, a copper co-catalyst, nucleophilic amine base^64–67) or a bulky trialkylphoshine.⁶⁵⁾ When we started this project, ligand- and additive-free double carbonylations under such mild conditions had not been reported.⁶⁸⁾

We investigated synthetic applications of SAPd(0) catalyst, and found that SAPd(0)-catalyzed double carbonylation of aryl iodides with amine proceeded under an atmospheric pressure of CO to afford an α-ketoamide with good chemoselectivity. That is to say, a solution of p-iodoanisole (1c) and morpholine (4a) in DMF was heated at 80°C for 24 h in the presence of SAPd(0),¹⁶⁾ which under an atmospheric pressure of CO without stirring, gave α-ketoamide 11ca in 40% yield accompanied by the formation of amide 12ca in 9% yield and recovery of 1c (Chart 4). As mentioned above, when we started this project, efficient double carbonylation under atmospheric pressure of CO in the absence of any ligands had not been reported. We therefore decided to conduct further investigation of ligand-free double carbonylation of aryl iodides under atmospheric pressure of CO using SAPd(0) catalyst.

Chart 4. Reaction of p-Iodoanisole (1c) and Morpholine (4a) under a CO Atmosphere (1 atm) in the Presence of SAPd(0)

(Color figure can be accessed in the online version.)

When the reaction in Chart 4 was carried out, Pd-black was precipitated in the reaction vessel, indicating that Pd NPs were aggregated under the reaction conditions. This result also means that SAPd(0) cannot be reused for further double carbonylations by this procedure. Therefore we decided to consider a two-step protocol which was previously reported in SAPd(0) catalyzed C–H alkynylation¹⁷⁾; the first step is for leaching of a sufficient amount of Pd from SAPd(0) by contact with p-iodoanisole and the second step is the carbonylation process (Chart 5). Thus a DMF solution of 1c in the presence of SAPd(0) and K₂CO₃ was heated at 80°C for 2 h without stirring. After SAPd(0) was removed from the reaction vessel, 4a was added to the reaction mixture then the resulting reaction mixture was stirred at 80°C for 24 h under a CO atmosphere (1 atm). As a result, p-iodoanisole (1c) was completely consumed, and double carbonylation product 11ca was produced in 91% yield along with amide derivative 12ca in 9% yield.

Chart 5. Optimization of the Reaction Procedure

With the optimal procedure in hand, we continued to conduct studies on the scope and limitation of the double carbonylation using SAPd(0). First, we investigated reactions of p-iodoanisole 1c with several aliphatic amines (Table 12). Reactions of 1c and secondary cyclic amines 4b–4f gave the corresponding α-ketoamides 11cb–11cf in high yields (runs 1–5). Acyclic primary and secondary amines 4g–4j were also applicable to the double carbonylation, and the desired coupling products 11cg–11cj were obtained in good yields (runs 6–9).

Table 12. Reactions of p-Iodoanisole (1c) and Various Amines

Next, the reactions of various aryl iodides and morpholine (4a) were investigated (Table 13). When m-iodoanisole 1c was used, α-ketoamide 11ca was obtained with good chemoselectivity (run 1). On the other hand, by the reaction using o-iodoanisole 1q as a substrate, mono carbonylation product 12qa was obtained as a major product (run 2). Reactions of 4a with 3,5-dimethyliodobenzene (1r) and 3,4,5-trimethoxyiodobenzene (1s) gave the corresponding double carbonylation products 11ra and 11sa, respectively, in high yields (runs 3 and 4). α-Ketoamide 11aa was obtained in 65% yield accompanied by the formation of benzamide derivative 12aa in 21% yield (run 5), when iodobenzene (1a) was subjected to the double carbonylation conditions. On the other hand, the reactions of aryl iodides 1t–1v with an electron-withdrawing group on the aromatic ring resulted in increases in the formation of mono carbonylation products 12ta–4va compared with the amounts of products obtained from reactions of aryl iodides having an electron-donating group (runs 6–8). Double carbonylation of heteroaromatic compounds 1w–1y proceeded in a chemoselective manner to give the corresponding α-ketoamides 11wa–11ya in high yields (runs 9–11).

Table 13. Reactions of Various Aryl Iodides and Morpholine (4a)

Furthermore, the potent anti-human immunodeficiency virus (HIV) agent 11zo could be synthesized in one step via double carbonylation of 2-iodonaphthalene (1zo) with N-benzoylpiperazine (2o) (Chart 6).

Chart 6. One-Step Synthesis of anti-HIV Agent 11zo

(Color figure can be accessed in the online version.)

Even after a sufficient amount of Pd NPs for progress of the double carbonylation had leached into the reaction mixture, a sufficient amount of Pd NPs seemed to remain on the SAPd(0). Therefore we inspected the reusability of SAPd(0) and amounts of the leached Pd in the reaction mixtures. After the solution of 1a, K₂CO₃, and SAPd(0) in DMF had been heated at 80°C for 2 h, SAPd(0) was removed and reused in the next reaction. SAPd(0) could be reused for at least five reaction cycles without significant loss of catalytic activities, and double carbonylated products were isolated in good yields and with good chemoselectivities in all cycles. Measurements of the amounts of leached Pd in the reaction mixtures were also conducted by ICP-MS analysis, and a catalyst loading in each reaction was estimated 0.009−0.18 mol %.

Hence we demonstrated double carbonylation of aryl iodides and amines catalyzed by Pd NPs leached from SAPd(0) material gave α-ketoamides with good chemoselectivities. It was noteworthy that the reaction proceeded even under atmospheric pressure of CO without any specific additives or ligands.

2.6. Removal of Allyl Protecting Group on Allyl Ester Using SAPd(0)²¹⁾

The allyl group is one of the most commonly used protecting groups in organic chemistry, especially for the protection of carboxylic acids as the corresponding allyl esters,⁶⁹⁾ due to its stability towards both acidic and basic conditions. Allyl protecting groups are usually removed by transition metal-catalyzed methods, including Pd catalysis. However, development of more efficient, milder and safer protocols for the general cleavage of allyl groups are required, because there are several limitations to the conventional procedures described above. One of the biggest limitations of Pd-mediated cleavage of allyl groups is removal of residual Pd species from the product material and/or the reaction mixture,⁷⁰⁾ because this material can have an adverse impact on subsequent reactions and can also pose a serious risk to human health if it is not removed to a suitably low level from pharmaceutical products.

We aimed to develop a new method for the removal of allyl protecting groups from allyl esters using a safe, low-leaching, recyclable, and ligand-free Pd NPs catalyst.

We initially examined the deprotection of benzoic acid allyl ester (13a), to achieve the Pd-NP-catalyzed removal of allyl protecting group on allyl esters, by treating a solution of this material (0.2 mmol) in acetonitrile (0.1 M) with SAPd(0) (Au 100 mesh, 12 × 14 mm²) under conventional conditions instead of a homogeneous Pd catalyst⁶⁹⁾ (Table 14). No reaction proceeded or only trace amount of 14a was yielded under some typical conventional conditions (entries 1–3). The desired benzoic acid product 14a was obtained in 96% yield; the reaction was performed with a combination of formic acid and Et₃N as scavenger agents at 80°C for 2 h (entry 4). We then screened a variety of solvents, temperatures, and reaction times to determine the appropriate conditions for the reaction. Several different solvents were screened against the reaction, including acetonitrile, dioxane, EtOH, and N,N-dimethylformamide, and the desired product 14a was isolated in good-to-excellent yield in all cases; however, acetonitrile gave the best result and was selected for further optimization work (Table 14, entries 4–7). The reaction was then conducted at lower temperatures such as 60 and 40°C, which give 14a in 90 and 63% yields, respectively (Table 14, entries 8 and 9). However, from these reactions we noticed that the starting material 13a was recovered in 6 and 25% yields, respectively. As shown in Table 1, entries 10 and 11, control experiments were performed in the absence of SAPd(0) and in the presence of Au-mesh instead of SAPd(0), and both experiments gave only a trace of 14a even when the reaction was conducted for a longer time. The conventional reaction conditions using Pd/C gave the desired product 14a in 82% yield (entry 12). The amounts of Pd leached into the reaction mixture by SAPd(0) and Pd/C were determined by inductively coupled plasma resonance spectroscopy, and the results of these experiments will be described later in this article. A proposed reaction mechanism was shown for the removal of the allyl-protecting group of allyl ester using a combination of a Pd catalyst and a formic acid (Fig. 7).

Table 14. Optimization of Reaction Conditions for Removal of Allyl-Protecting Group of 13a Using SAPd(0)

Entry	Scavenger	Solvent	Temp. (°C)	Time (h)	Yield (%)a)
					14a	14a
1	N-Methylaniline	CH3CN	80	6	NR	—
2	N,N-Dimethylbarbituric acid	CH3CN	80	6	NR	—
3	Ph3SiH	CH3CN	80	6	trace	—
4	Formic acid and Et3N	CH3CN	80	2	96e)	—
5	Formic acid and Et3N	dioxane	80	1	80e)	—
6	Formic acid and Et3N	EtOH	80	2	62e)	—
7	Formic acid and Et3N	DMF	80	2	85e)	—
8	Formic acid and Et3N	CH3CN	60	3	90	6
9	Formic acid and Et3N	CH3CN	40	24	63	25
10	Formic acid and Et3Nb)	CH3CN	80	2	2	82
11	Formic acid and Et3Nc)	CH3CN	80	2	2	82
12	Formic acid and Et3Nd)	CH3CN	rt	3	82	—

a) HPLC yield. b) Without SAPd(0). c) Au-mesh was used instead of SAPd. d) 10% w/w of 10% Pd/C was used instead of SAPd(0). e) Isolated yield.

Fig. 7. Proposed Reaction Mechanism for Removal of Allyl-Protecting Group of Allyl Ester Using a Combination of Pd Catalyst and Formic Acid

We continued our experiments, with the optimized conditions in hand (Table 14, entry 4), to discover the scope and generality of this SAPd(0)-catalyzed method for removal of protecting groups of esters using a series of different benzoic acid esters 13b–i (Table 15). Under the optimized conditions, prenyl and benzyl groups were easily cleaved from the corresponding ester (Table 15, entries 2 and 3). On the other hand, n-propyl and propargyl groups remained intact (Table 15, entries 4 and 5). All allyl groups on allyl esters 13f–o, including an amino acid (entry 13), were pleasingly converted to the corresponding carboxylic acids in good-to-excellent yields, except 13i, which gave a much lower yield. This was because the tert-butyldiphenylsilyl group on the alcohol of this substrate was partially cleaved under these reaction conditions, even at a lower reaction temperature of 60°C.

Table 15. Removal of Allyl Protecting Groups of Different Esters Using SAPd(0) as Catalyst

A sufficient amount of Pd NPs seemed to remain on the SAPd(0) even after deallylation, and the reusability of the SAPd(0) material was therefore evaluated together with the amount of Pd that leached into the reaction mixtures. After a solution of 1a (0.2 mmol) in acetonitrile (0.1 M) had been heated at 80°C for 2 h in the presence of formic acid (5 eq.), Et₃N (5 eq.), and SAPd(0), the SAPd(0) was removed and reused in the next reaction. It was found that SAPd(0) could be used for at least ten reaction cycles without any significant loss in its catalytic activity, and the deallylated products were obtained in excellent yields in all cycles. By ICP-MS analysis, measurements for the amount of Pd leached into the reaction mixtures were also conducted, which revealed that 0.09–0.92 µg of Pd was present in the reaction mixture. This small amount of Pd was estimated 0.04–0.46 ppm in the whole mixture. In a similar reaction using Pd/C instead of SAPd(0), a much larger amount of Pd was leached into the reaction mixture (1.50 ± 0.37 µg).

Kinetic studies/filtration tests were conducted on three different reactions and the time conversion plots of these reactions compared so as to develop a deeper understanding of the nature of the catalytic species involved in this SAPd(0)-catalyzed process. These investigations were conducted to confirm whether the leached Pd-species exhibited any catalytic activity towards removal of the allyl group (Fig. 8). Under the optimized conditions described in Table 1 (entry 4), the reaction with SAPd(0) was performed to give the deprotected product 2a in 93% yield after 40 min. Another reaction was also performed under the same conditions, except that the SAPd(0) was removed from the mixture 10 min after starting the reaction. In this particular case, the yield of product 2a was determined to be only 49% after 10 min and the chemical conversion stopped completely following removal of SAPd(0), with the yield of 2a remaining at 49% after 120 min. Therefore these results indicate that the active Pd species involved in the removal of the allyl-protecting group from the allyl ester 1a using SAPd(0) as catalyst is indeed the Pd NPs on the SAPd(0) surface, even though some Pd was released from the SAPd(0) during the course of this reaction. This result was particularly interesting when compared with that of our previous experiment, where Pd cross-coupling, Suzuki–Miyaura coupling, Buchwald–Hartwig, and carbon(sp³ and sp²)–hydrogen bond activation reactions using halogen compounds were catalyzed by an active Pd species that was leached from SAPd(0).

Fig. 8. Removal of Protecting Group on Ester 13a Using SAPd(0) as a Catalyst, Filtration Test

(Color figure can be accessed in the online version.)

Hence we have developed a new safe method for removal of allyl protecting groups of allyl esters using SAPd(0) as a Pd-NP catalyst, where the reaction proceeded on the SAPd(0) surface. It should be noted that SAPd(0) could be recovered and reused at least 10 times without any discernible decrease in its activity and that much less Pd leached from the catalyst into the reaction mixture than that observed in conventional methods using Pd/C.

2.7. Redox Switching Using SAPd(0)²²⁾

Various optical switching technologies of π conjugated molecules have been reported. For instance, photochromism, thermochromism, electrochromism, piezochromism, solvatochromism, and harochromism are known as representative technologies. Also, redox-mediated chromism using electricity or chemical reagents such as DDQ, NaBH₄, or Na₂S₂O₃ have been reported.

To achieve the redox system using orthoquinone and hydroquinone moieties, oxidizing and reducing reagents are required. However, side products released from these reagents are potential problems for reversible redox system. For achievement of clean and reversible redox switching, hydrogen and oxygen gas were inspired. Thus hydrogen gas could be employed for the hydrogenation of orthoquinone to hydroquinone in the presence of metal catalyst such as Pd, Pt, Rh, or Ni. In contrast, oxygen gas could be employed for oxidation of hydroquinone to orthoquinone.^72–75) In this proposed redox system using gas energy, main side product is environmentally friendly and removable water, and this system could be reversible without purification. In addition, these technologies possibly apply to photoluminesence materials switching by gas energy or gas sensor for detection of hydrogen or/and oxygen by monitoring of fluorescence. To our best knowledge, only one example that contains redox switching of coumarin derivative as optical material using the combination of oxygen gas and photo energy for oxidation and hydrogen gas in the presence of Pd catalyst for reduction has been reported.⁷⁶⁾

We found the unique redox chromism of newly proposed orthoquinone containing pentacyclic aromatic compounds using hydrogen and oxygen gas in the presence of self-assembled Au-supported Pd catalyst (SAPd(0)).

After completion of the calculation studies, picene-13,14-dione (15) was synthesized (Fig. 9).

Fig. 9. Synthesis of Picene-13,14-dione (15)

(Color figure can be accessed in the online version.)

The redox switching of picene-13,14-dione (15) using gas energy was investigated. Generally, the heterogeneous catalyst such as Pd/C was used in the reduction of orthoquinone 15 to hydroquinone 16. However, for monitoring the optical properties, the heterogeneous condition is not suitable. Therefore SAPd(0) was employed to solve this problem. SAPd(0) is supported on Au mesh sheet. Thus the optical properties such as UV-Vis and fluorescence spectra in situ can be observed clearly. In addition, it can be easily taken out from the reaction vessel and also it is recyclable. When the reduction of orthoquinone 15 in the presence of SAPd(0) under hydrogen gas bubbling in CHCl₃ at ambient temperature was carried out, desired hydroquinone 16 was yielded (Fig. 10). (As an explanatory note, the reaction time of the reduction could be controlled by surface area of SAPd(0).) The yellow color of 15 disappeared along with the progress of the reaction. On the other hand, clear blue fluorescence under the irradiation of UV light (365 nm) appeared at the end of reaction. Indeed, hydroquinone 16 absorbed only UV region in UV-Vis spectrum (Fig. 11). In fluorescence spectrum of 16, maximum fluorescence wavelength of 419 nm was observed as blue emission (Fig. 12). These results suggest that emission of fluorescence of 16 came from the extension of the π conjugated system by the transformation to hydroquinone from orthoquinone. In addition, the hydroquinone 16 had large Stokes shift (129 nm, ν = 10617 cm⁻¹). This confers great advantage for application to fluorescence materials. Notably, all optical properties of 15 and 16 were in agreement with the estimation of in silico study. The subsequent reverse oxidation to the orthoquinone 15 from hydroquinone 16 smoothly proceeded by simple exchange of hydrogen gas for oxygen gas, and the colorless solution was changed to yellow. As a result, reversible redox switching using hydrogen and oxygen was established. The side product in the redox system is mainly environmentally friendly water. In addition, ethylacetate and tetrahydrofuran (THF) could be employed as solvent for this reversible redox switching.

Fig. 10. Redox Switching of 13 Using Hydrogen and Oxygen Gases

(Color figure can be accessed in the online version.)

Fig. 11. UV-Vis Spectra of Picene-13,14-dione (15) and Picene-13,14-diol (16)

(Color figure can be accessed in the online version.)

Fig. 12. Fluorescence Spectrum of Picene-13,14-diol (16)

(Color figure can be accessed in the online version.)

Next, the details of oxidation condition were investigated (Table 16). In our established standard conditions described above, oxygen gas was employed for oxidation in the presence of SAPd(0) (Table 16, entry 1). The aerobic oxidation for conversion to 16 from 15 could be applied in quantitative yield, even though longer reaction time was required (Table 16, entry 2). On the other hand, when the oxidation reaction was conducted in the absence of SAPd(0), the reaction time was more than twice longer than standard condition (entry 3 vs. entry 1). These results suggest that SAPd(0) acted as not only catalyst for hydrogenation but also catalyst for oxidation. Furthermore, this is the first example that SAPd(0) is employed in redox reactions.

Table 16. Investigation of Oxidation Conditions in Redox System

Entry	Bubbling	Catalyst	Time (h)	Result
1	O2	SAPd(0)	2	quant.
2	Air	SAPd(0)	5	quant.
3	O2	—	5	quant.

The redox switching of pentaphene-6,7-dione (17, Fig. 13) using hydrogen and oxygen gas energy was investigated (Fig. 14). When the reduction of orthoquinone 17 in the presence of SAPd(0) under hydrogen gas bubbling in CHCl₃ at ambient temperature was carried out, desired hydroquinone 18 was provided quantitatively, and the color of the solution changed to yellow from orange. At the same time, the fluorescent color under the irradiation of UV light (365 nm) changed to green from red. In fluorescence spectrum, the maximum fluorescence wavelength of hydroquinone 14 was 495 nm which shifted 107 nm shorter than 17 (the maximum fluorescence wavelength of 17 was 602 nm, Fig. 15). In addition, the compound 17 and 18 also had large Stokes shift as with the compound 4. Particularly, the Stokes shift of 17 was over 200 nm (17: 231 nm (ν = 10343 cm⁻¹), 18: 140 nm (ν = 7967 cm⁻¹)). The reverse oxidation from 18 to 17 with oxygen gas succeeded by our established condition smoothly, then reversible redox switching of pentaphene derivative 17 was completed.

Fig. 13. X-Ray Crystal Structure Analysis of Pentaphen-6,7-dione (17)

(Color figure can be accessed in the online version.)

Fig. 14. Redox Switching of 17 Using Hydrogen and Oxygen Gases

(Color figure can be accessed in the online version.)

Fig. 15. Fluorescence Spectra of Pentaphen-6,7-dione (17) and Pentaphen-6,7-diol (18)

(Color figure can be accessed in the online version.)

2.8. Semi-Flow Synthesis Using SAPd(0)¹⁵⁾

Because the microwave technique sometimes allows for reduced reaction times and increased yields than those obtained using conventional heating methods, microwave-assisted chemical reactions are becoming a popular technique in organic synthesis.^77–79)

Our idea was to adopt SAPd(0) as a new solid-support for the Suzuki–Miyaura reaction and use the microwave technique to make the system much more efficient. After many experiments using SAPd(0) under commonly used multi-mode microwave irradiation conditions and consideration of the reaction mechanism of SAPd(0) in Pd cross-coupling,¹²⁾ we planned to use two microwaves with different irradiation modes: a single-mode⁸⁰⁾ for leaching matters and a multi-mode⁸¹⁾ for the catalytic reaction cycle.

The combined microwave systems would solve some problems of the SAPd(0)-catalyzed Suzuki–Miyaura reaction observed using the conventional heating process. The limitations of a longer reaction time might be attenuated, and less reactive arylbromides might be successfully coupled to give the desired products^3a. Here, we report ligand-free SAPd(0)-catalyzed Suzuki–Miyaura cross-coupling reactions with a variety of aryl halides and boronic acids under two kinds of microwave irradiation, in which the SAPd(0)-catalyst was efficiently utilized repeatedly (more than 10 runs) and a trace of Pd (less than 1 ppm) was the active catalytic species.

2.8.1. Design of the Microwave-Assisted Process

SAPd(0) released a trace of highly active Pd in the reaction mixture,^12,16) which allowed us to utilize it in the microwave-assisted Suzuki–Miyaura reactions. In this regard, we planned to use two chambers, namely, a leaching chamber, where a trace of active catalytic Pd would leach from SAPd(0), and a catalytic reaction chamber, where the catalytic cycle would proceed for the desired transformations. To achieve this new strategy, we used two different types of microwaves: one microwave suitable for SAPd(0) with single-mode irradiation, which would operate in the leaching chamber, and another microwave for multi-mode irradiation of the whole reaction mixture in the catalytic reaction chamber.

Based on our hypothesis, as shown in Chart 7, first, the starting solution (Ph-X and leached active Pd) was irradiated by single-mode microwave, MW (S) and poured using a Pasteur pipette into the reaction chamber including boronic acid, base, and solvent; a second irradiation with the multi-mode microwave MW (M) was then applied to obtain the coupling biaryl product.

Chart 7. Microwave-Assisted Process for the Suzuki–Miyaura Reaction in the Presence of SAPd(0)

(Color figure can be accessed in the online version.)

2.8.2. Optimization Using Iodobenzene 1a

On the basis of Chart 7, we first performed the Suzuki–Miyaura coupling reaction using iodobenzene 1a and 4-chlorophenyl boronic acid 2a in the presence of SAPd(0) to obtain the biaryl product 3a. We examined many conditions to achieve a successful coupling reaction by changing the bases, solvents, and the operating conditions of the two microwaves operating. Finally, we observed optimized conditions using K₂CO₃ as a base with single-mode microwave conditions, MW (S): solvent: EtOH (2.0 mL), temperature: 80°C, time: 60 min, power: 200 W, and multi-mode microwave conditions, MW (M): solvent: EtOH (2.0 mL), temperature: 82°C, time: 60 min, and power: 500 W (Table 17). These conditions were maintained from 1st to 10th cycle with excellent reproducible results. Thus the reaction time of the Suzuki–Miyaura coupling reaction using iodobenzene 1a by the microwave-assisted process was decreased compared with that using a conventional heating process from 12 to 2 h.^11,12)

Table 17. Microwave-Assisted Suzuki–Miyaura Reaction for Iodobenzene 1a with SAPd(0)^a)

Yield (%)b)
1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
99	>99	98	99	97	99	99	97	99	99

a) Reaction conditions: Iodobenzene 1a (0.50 mmol), 4-chlorophenyl boronic acid 2a (0.75 mmol), K₂CO₃ (1.00 mmol); single-mode microwave settings, MW (S): temp: 80°C, time: 60 min, power: 200 W; multi-mode microwave settings, MW (M): temp: 82°C, time: 60 min, power: 500 W; b) HPLC yield using Ph-NO₂ (0.25 mmol) as an internal standard.

2.8.3. Optimization Using Bromobenzene 1h

We next performed the Suzuki–Miyaura coupling reaction using bromobenzene 1h with SAPd(0) according to the strategy described in Chart 7. To optimize the bromobenzene 1h coupling with 4-chlorophenyl boronic acid 2a, we examined various conditions by changing the bases and solvents, as well as the operating conditions of the two microwaves. We achieved the optimized conditions using K₂CO₃ as base with single-mode microwave conditions, MW (S): solvent: DMF (2.0 mL), temperature: 90°C, time: 50 min, and power: 300 W; and multi-mode microwave conditions, MW (M): solvent: toluene/H₂O (2 : 1, 2.0 mL), temperature: 104°C, time: 60 min, and power: 500 W (Table 18). These conditions were maintained from the 1st to 10th cycle with excellent reproducible results of the desired biaryl product 3a. Bromobenzene 1h was unreactive to give the coupled product even under the best conditions for aryliodide couplings under the conventional heating technique with SAPd(0), whereas we successfully obtained the coupling product under the microwave-assisted process in the presence of SAPd(0).

Table 18. Microwave-Assisted Suzuki–Miyaura Reaction for Bromobenzene 1h with SAPd(0)^a)

Yield (%)b
1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
>99	>99	99	99	99	>99	99	99	99	99

a) Reaction conditions: Bromobenzene 1h (0.50 mmol), 4-chlorophenylboronic acid 2a (0.75 mmol), K₂CO₃ (1.00 mmol), MW (S) settings: temp: 90°C, time: 50 min, power: 300 W; MW (M) settings: temp: 104°C, time: 60 min, power: 500 W; b) The HPLC yield using Ph-NO₂ (0.25 mmol) as an internal standard.

2.8.4. Measurement of Pd Amount in SAdP(0) and in Reaction Mixture

We measured the amount of immobilized Pd in SAPd(0) and the leached Pd in the reaction mixtures by ICP-MS. The immobilized Pd in SAPd(0) was measured before and after the 10th cycle of the two optimized reactions, namely, 1a + 2a and 1h + 2a (Tables 17 and 18, respectively). During the reaction of 1a with 2a, the amount of immobilized Pd in SAPd(0) before and after the 10th cycle reaction was 81 ± 27 and 93 ± 30 µg, respectively,¹¹⁾ and the amount of released Pd after each cycle was extremely low (56–170 ng [0.02–0.05 ppm] for a 0.50 mmol scale reaction). During the reaction of 1h with 2a, the amount of immobilized Pd in SAPd(0) before and after the 10th cycle reaction was 81 ± 23 and 49 ± 13 µg,⁸²⁾ respectively, and the amount of released Pd after each cycle was also low (204–1321 ng [0.08–0.29 ppm] for a 0.50 mmol scale reaction). Based on the reactivity between 1a and 1h, for successful coupling using 1h, it is reasonable that the leaching amount of Pd in the reaction of 1h with 2a is higher than that in the reaction of 1a with 2a. Although the released Pd is higher compared with that using the conventional heating method,^11,12) for both reactions, the amount of Pd is far lower than the U.S. government-required value of residual metal in product streams.⁸³⁾ It is also important to note that during conventional heating, it is possible to redeposit the active Pd-species to the support, but in microwave-assisted process, redeposition of active Pd-species to the support is not possible. SAPd(0) is highly recyclable (more than 10 cycles) in the microwave-assisted process, however, and our new technique allows for a scale-up reaction.

2.8.5. Necessity of Aryl Halide in Contact with SAPd(0) in MW (S)

Our strategy was to obtain the released active catalytic Pd-species from SAPd(0) by the effects of aryl halide in the MW (S); thus SAPd(0) and aryl halide would be irradiated under a single-mode microwave, MW (S), condition to release the active catalytic Pd-species for the oxidative addition step and the catalytic cycle would then be completed under a multi-mode microwave, MW (M), condition. To determine the applicability of this strategy, we irradiated the same SAPd(0) in MW (S) with and without iodobenzene (Chart 8). We first used only SAPd(0) and EtOH in the MW (S) condition, and then the whole solution containing all coupling reagents (1a, 2a, K₂CO₃) was irradiated in the MW (M) condition; the yield of the desired product was only 16%. When the same SAPd(0) was irradiated in the MW (S) condition with iodobenzene 1a, the yield of the desired product was 94%, consistent with our hypothesis. Therefore aryl halide is important in the MW (S) irradiation condition for leaching the active Pd from SAPd(0) for the oxidative addition step.

Chart 8. Necessity of Aryl Halides in MW (S)

2.8.6. Scope and Limitation

Our next aim was to generalize the microwave-assisted concept for SAPd(0)-catalyzed Suzuki–Miyaura coupling reactions using various arylhalides (1) with a variety of phenylboronic acids (2) to yield the coupling products (3) (Table 19). For example, when 4-bromoanisole 1i and p-bromotoluene1aa, having electron-donating methoxy and methyl groups, were treated with 4-chlorophenylboronic acid 2a, the corresponding products were obtained in mean yields of 98 and 94%, respectively (entries 1 and 2). 4-Bromonitrobenzene 1ab and 4-bromobenzonitrile 1j, with electron-withdrawing nitro and cyano groups, also coupled successfully with 2a, and the mean yields of products were 99 and 99%, respectively (entries 3 and 4). We then explored the reactions of bromobenzene 1h with two different boronic acids, phenylboronic acid 2c and 4-methylphenylboronic acid 2b, from which the corresponding products were obtained in mean yields of 96 and 98%, respectively (entries 5 and 6). We examined the Suzuki–Miyaura coupling-reactions using a variety of aryl iodides with 2a. When 4-iodoanisole 1c and o-iodotoluene 1f were treated with 2a, coupled products were obtained in mean yields of 97% and 95%, respectively (entries 7 and 8). Reactions of 4-acetyliodobenzene 1e and 4-nitroiodobenzene 1d with 2a yielded the corresponding coupling products in 99% mean yield for both (entries 9 and 10). Finally, iodobenzene 1a also successfully coupled with phenylboronic acid 2c and 4-methylphenylboronic acid 2b, leading to the coupling products, in a mean yield of 96% for both (entries 11 and 12).

Table 19. Suzuki–Miyaura Cross-Coupling Reactions of Various Substrates Using SAPd(0)^a)

Entry	Ar-X	Ar′-B(OH)2	Yields of 3 (%)b)										Average yields (%)
		Ar′	1st	2nd	3rd	4th	5th	6th	7th	8th	9th	10th
1	4-MeOC6H4Br	4-ClC6H4	98	99	98	99	98	97	98	98	97	98	98
	(1i)	(2a)
2	4-MeC6H4Br	4-ClC6H4	95	93	95	94	93	95	95	95	93	94	94
	(1aa)	(2a)
3	4-NO2C6H4Br	4-ClC6H4	98	99	99	>99	98	98	99	99	99	98	99
	(1ab)	(2a)
4	4-CNC6H4Br	4-ClC6H4	99	>99	98	99	97	98	99	99	98	99	99
	(1j)	(2a)
5	C6H5Br	C6H5	97	98	97	97	96	97	93	95	94	95	96
	(1h)	(2c)
6	C6H5Br	4-MeC6H4	99	98	96	98	94	98	96	99	98	97	98
	(1h)	(2b)
7	4-MeOC6H4I	4-ClC6H4	97	98	98	97	96	96	97	97	98	97	97
	(1c)	(2a)
8	2-MeC6H4I	4-ClC6H4	96	95	95	97	95	96	95	95	95	95	95
	(1f)	(2a)
9	4-AcC6H4I	4-ClC6H4	>99	>99	98	99	99	98	99	97	98	99	99
	(1e)	(2a)
10	4-NO2C6H4I	4-ClC6H4	99	>99	99	95	98	98	99	99	98	99	99
	(1d)	(2a)
11	C6H5I	C6H5	95	95	96	96	96	95	96	96	95	95	96
	(1a)	(2c)
12	C6H5I	4-MeC6H4	96	97	96	96	96	95	96	94	95	96	96
	(1a)	(2b)

a) Reaction conditions for arylbromides: arylbromides (0.50 mmol), boronic acids (0.75 mmol), K₂CO₃ (1 mmol), single-mode microwave settings, MW (S): solvent: DMF (2.0 mL), temp.: 90°C, time: 50 min, power: 300 W, multi-mode microwave settings, MW (M): solvent: toluene/H₂O (3 : 1, 2 mL), temp.: 102°C, time: 60 min, power: 500 W; Reaction conditions for aryliodides: aryliodides (0.50 mmol), boronic acids (0.75 mmol), K₂CO₃ (1 mmol), single-mode microwave settings, MW (S): solvent: EtOH (2.0 mL), temp.: 80°C, time: 60 min, power: 200 W, multi-mode microwave settings, MW (M): solvent: EtOH (2.0 mL), temp.: 82°C, time: 60 min, power: 500 W. b) The isolated yields.

Hence we developed a new strategy for a microwave-assisted process of Suzuki–Miyaura cross-coupling reactions catalyzed by a solid-supported Pd-catalyst, SAPd(0), which has very low leaching properties. The combined use of two microwaves with different irradiation approaches makes the protocol very efficient, because the leaching can be easily controlled during the process. The great achievements of this protocol are that fewer reactive arylbromides were successfully coupled to the desired products, which has not been achieved under conventional heating conditions by SAPd(0), and the reaction time for aryliodide couplings was reduced from 12 to 2 h. Moreover, the low leaching properties of SAPd(0) makes it recyclable for more than 10 cycles and thus the reaction can be easily up-scaled. Based on our initial results, we hope this technology could also be applicable to other solid-supported metal catalysts.

2.9. Medicinal Chemistry Using SAPd(0)¹⁹⁾

Fragment-based drug discovery (FBDD)^84,85) is a method for hit finding; in essence, it bases its strength on the competent binding of small chemical entities to their targets. FBDD involves identifying small fragments, which, because of their small size, tend to bind with relatively low affinity, and then developing these to produce larger, higher-affinity ligands. The major advantage of FBDD over traditional high-throughput screening methods is that FBDD provides a more rapid and effective method to identify ligands for a protein target. By “linking,” “merging,” or “growth,” identified fragment hits can be optimized. Among them, “fragment growth” is believed more effective than linking or merging, because additional features can be added through iterative cycles to the hit fragment, leading to more-potent compounds.⁸⁶⁾ Privileged structures,⁸⁷⁾ such as indoles, purines, quinolines, and benzimidazoles, are useful to identify valuable optimized compounds effectively by FBDD, and historically yield several biologically active molecules that selectively modulate the activities of enzymes, nuclear receptors, ion channels, G-protein coupled receptors, and so on. Here we report our original “fragment growth” method for FBDD, based on the three-dimensional diversity-oriented strategy using cyclopropane as the key unit to restrict the conformation of privileged molecules.

Although liquid-phase combinatorial synthesis is frequently used in medicinal chemical research, until our report there have been no known Pd catalysts that work repeatedly with low Pd-leaching (less than 1 ppm in the reaction mixture without purification).

Here, a conformationally restricted privileged structure library having not only functional diversity but also three-dimensional diversity was designed, as our original “fragment growth” method for FBDD, and successfully synthesized using the ligand-free Suzuki–Miyaura coupling of vinyl iodides with SAPd(0) as the key strategy. SAPd(0), throughout the study, proved an effective catalyst for liquid-phase combinatorial synthesis, which is repeatedly used for synthesizing structurally diverse fragment molecules, by Suzuki–Miyaura coupling of cyclopropylvinyl iodides with privileged heterocyclic boronic acids and without contamination. Furthermore, preliminary evaluation of the inhibitory effects of the library compounds on a panel of kinases suggested that the library for FBDD growth would be useful for finding hit with three-dimentional active structure information.

We used cyclopropane as the key unit to design conformationally restricted analogs of bioactive ligands, based on a spatial screening concept to search for unknown bioactive conformations of ligands.⁸⁸⁾ These cyclopropane-containing compounds have three-dimensional structural diversity, which has allowed us to identify the appropriate conformations that produce bioactive compounds targeting a variety of biomolecules.

A small molecular library consisting of compounds with three-dimensional diversity would be expected to display a broader range of biological activities. However, most libraries used to date have been limited to aromatic heterocycles with an underrepresentation of chiral sp³-rich compounds. We considered that combinational use of advantage of the small and rigid structural features of cyclopropane and the privileged structures might provide the three-dimensionally diverse sp³-rich compounds with the privileged structures very effective in the fragment growing method.

Thus as shown in Fig. 16, we designed a general structure having a vinylcyclopropane backbone. In the structure for FBDD growth, a functionalized carbamoyl fragment, a privileged heterocyclic vinyl fragment, and a hydroxymethyl fragment were attached to a vinylcyclopropane backbone, in which the relative location of these three fragments was effectively restricted due to the rigid cyclopropane ring. The structure for FBDD growth has the following advantages: 1) by changing the regiochemistry of these three groups on the cyclopropane ring, the library compounds can have remarkable three-dimensional diversity; 2) a variety of privileged heterocyclic fragments can be introduced by Suzuki–Miyaura coupling; 3) the hydroxymethyl group increases the solubility of the molecule in aqueous medium; 4) the hydroxymethyl group can be effectively used for further derivatization in the next “fragment growth” stage; 5) the use of the easy-forming amide linkage to introduce functional groups provides fragments with functional group diversity; and 6) the molecular weights of the library compounds are rather small. In fragment growing approach, it is often difficult to find the site in hit fragments effective for the growing. The hydroxymethyl moiety can be extremely useful, when a hit is identified, because its presence hints to promote the facile elaboration or modifications of the hit for growing into a lead.

Fig. 16. General Scheme for Our Designed Fragment Structure and a Medicinal Example with Cyclopropane

In this research project, we planned to construct a racemic small library for FBDD growth comprising Type-a and b structures, which are two regioisomers included in the general structure. Synthetic strategies for the library compounds for FBDD growth are shown in Fig. 16. Introduction of the key privileged heterocycles was planned by SAPd(0)-promoted ligand-free Suzuki–Miyaura coupling between the cyclopropylvinyl iodides 19 as substrates, which were prepared from a known cyclopropane derivative, and the privileged heterocyclic boric acids 20. Then, various functional groups were introduced by condensation with the amines X-NH₂ (Chart 9).

Chart 9. General Scheme for Synthesis of Designed Privileged Library Compounds

The reaction suffers from contamination problems and troublesome washing is needed due to the strong absorption of the starting material and/or product to the polymer, when polymer-supported Pd is used. By sharp contrast, SAPd(0) has a much smaller surface area and absorbs many fewer organic compounds, including products or starting materials, compared with polymer-supported Pd, due to its low affinity for organic molecules. Moreover, the Au mesh, used as the support for Pd in SAPd(0), is malleable and easy to handle with a pair of tweezers. There has been no practical solid-supported Pd that could be repeatedly used for the synthesis of different kinds of coupled products, although combinatorial synthesis is an important methodology in medicinal chemistry and is widely used for drug development.⁸⁹⁾ Considering the above-mentioned advantages, SAPd(0) may indeed be an ideal solid catalyst for combinatorial Suzuki–Miyaura coupling for medicinal chemical research studies. Eventually, we tried liquid-phase combinatorial Suzuki–Miyaura coupling of cyclopropylvinyl iodides with SAPd(0) by changing the privileged heterocyclic boric acids 20.

According to Chart 10, the coupling substrates, cyclopropylvinyl iodides 19a and 19b, were synthesized. Thus the known cyclopropane derivative 26⁹⁰⁾ was converted into the cyclopropane aldehyde 27a or 27b via protecting group manipulation, reduction, and oxidation, of which Takai coupling gave the cyclopropylvinyl iodides 19a and 19b.

Chart 10. Preparation of Vinyl Iodides 19a and 19b

Fig. 17. Diversity Reagents 20a–e

a; Boc group is removed in the final product 25.

Fig. 18. Diversity Reagents 23a–i

a; tert-Butyl group is removed in the final product 25. b; Boc group is removed in the final product 25.

When SAPd(0) was used as the catalyst in the Suzuki–Miyaura coupling of the vinyl iodide 19a or 19b and boronic acids 20a with K₂CO₃ in EtOH under our previously reported best conditions,^11,12) the corresponding product was not yielded at all. Then, considering the reaction mechanism of SAPd(0) in Suzuki–Miyaura coupling,¹²⁾ we hypothesized that the actual active Pd species were not sufficiently released from SAPd(0) to promote the reaction under the above conditions and that microwave irradiation might accelerate the release of the Pd species. We examined the use of two kinds of microwaves, one for releasing the actual active Pd species (MW1) and another for promoting Suzuki–Miyaura coupling (MW2). After several experiments, we found the optimized conditions for MW1 and MW2. A mixture of the vinyl iodide 19a or 19b (0.17 mmol) and SAPd(0) (14 × 12 mm) in ethanol (1 mL) and DMF (1 mL) was irradiated with weaker microwaves (300 W, single-mode) at 90°C for 45 min, then the resulting mixture was added to a solution of a privileged heterocyclic boronic acid 20a, b, c, d, or e (1.5 eq.) and K₂CO₃ (2 eq.) in toluene (1.5 mL) and irradiated with a stronger microwave (500 W, multi-mode) at 100°C for 1 h. After the usual aqueous workup, the corresponding product 21 was obtained in high purity as shown in Table 20. SAPd(0) was removed from the reaction mixture and used repeatedly for the reaction with substrates 19a or 19b and various boronic acids 20. These results demonstrate that SAPd(0) is an effective Pd catalyst for this kind of combinatorial synthesis without any contamination of other products or starting materials.

Table 20. SAPd(0) Catalyzed Suzuki–Miyaura Coupling of 19 and 20^a)

Entry	Vinyl iodide 19	Boric acid 20	Product 21	Yield (%)b)
1	19a	20a	21aa	78 (48)
2	19a	20b	21ab	40c,d) (33)
3	19a	20c	21ac	88c) (53)
4	19a	20d	21ad	82 (73)
5	19a	20e	21ae	50c,d) (7)
6	19b	20a	21ba	87 (48)
7	19b	20b	21bb	7 (39)
8	19b	20c	21bc	66 (58)
9	19b	20d	21bd	55 (41)
10	19b	20e	21be	57 (37)

a) MW1 conditions; 19a or 19b (0.084 mmol), SAPd(0) (14 × 12 mm), ethanol (1 mL), DMF (1 mL) microwave heating (300 W, single-mode), 90°C, 45 min, then MW2 conditions; boronic acid 20 (1.5 eq.), K₂CO₃ (2 eq.), toluene (1.5 mL) microwave heating (500 W, multi-mode), 100°C, 1 h. b) Numbers in parenthesis indicate the yield using conventional homogeneous catalyst; boronic acid 20 (1.5 eq.), Pd(OAc)₂(10 mol%), PPh₃ (20 mol%), CsCO₃ (2 eq.), DMF, rt - reflux, 3–24 h. c) 0.25 mL of H₂O was added in multi-mode microwave heating. d) DMF was used as a solvent and reaction time was 30 min in single-mode microwave heating.

Successive oxidations of 21 with Dess–Martin periodinane (DMP) and the subsequent Pinnich conditions gave the corresponding carboxylic acids 22 (Table 21).

Table 21. Preparation of Carboxylic Acids 22 from Alcohol 21 by Oxidation

Entry	Alcohol 21	Yield in DMP oxidation (%)	Yield in Pinnich oxidation (%)
1	21aa	90	82
2	21ab	81	77
3	21ac	64	91
4	21ad	67	83
5	21ae	70	26a)
6	21ba	95	82
7	21bb	80	95
8	21bc	72	91
9	21bd	97	91
10	21be	84	90

a) The carboxylic acid produced was unstable and decreased the yield.

With carboxylic acids 22 in hand, we prepared the corresponding fragments by liquid-phase combinatorial chemistry. Thus amine 23, 1-hydroxybenzotriazole (HOBt) and polymer-bound carbodiimide resin were added to a solution of 22 in dichloromethane (Chart 11), and the whole mixture was stirred at ambient temperature overnight. After scavenging the excess amine with PS-isocyanate then HOBt, and any remaining carboxylic acid with macroporous polymer-bound carbonate resin (MP-carbonate), filtration and evaporation of the resulting mixture yielded the desired products 24 in high yields and purities. Acidic treatment of 24 led to production of the corresponding library compounds 25 in good-to-quantitative yields. Therefore we successfully constructed a small library comprising 90 library compounds for FBDD growth (45 Type-a compounds and 45 Type-b compounds), with molecular weights ranging from 314 to 422.

Chart 11. Synthesis of a Privileged Structure Library with Conformational Restriction

We selected five Type-a library compounds for FBDD growth 25aae, 25abe, 25ace, 25ade, and 25aee, with different heterocycles and the same cyclohexylcarbamoyl group, and the corresponding five Type-b library compounds 25bae, 25bbe, 25bce, 25bde, and 25bee for preliminary evaluation with a 20-kinase^91,92) assay panel at 10-µM concentrations. Kinase activities were inhibited by the library compounds for FBDD growth, as expected (Table 22). It should be noted that the inhibitory effects changed depending on the regiochemistry (Type-a or b), i.e., three-dimensional positioning of the three substituents on a cyclopropane ring. For example, although both 25ade (Type-a) and 25bde (Type-b) have the same quinoline-3-yl privileged structure and a cyclohexylcarbamoyl group, the two compounds for FBDD growth inhibited a variety of kinases with different selectivities: typically, 25ade (43.8%) > 25bde (17.4%) against JAK3, and 25ade (19.9%) < 25bde (73.0%) against KDR. The compound 25abe (Type-a), having Cl-quinoline, selectively inhibited TRKA among the 20 kinases, but the corresponding Type-b regioisomer 25bbe had insignificant inhibitory effects on the kinase.

Table 22. Inhibitory Effects of 10 Selected Compounds on a Panel of 20 Different Kinases

Hence conformationally restricted privileged structure library for FBDD growth, having not only three-dimensional diversity but also functional diversity, was designed. SAPd(0)-catalyzed ligand-free Suzuki–Miyaura coupling of vinyl iodides with heteroaromatic boronic acids, which was effectively promoted by microwave irradiation, was developed, resulting in the construction of the designed library for FBDD growth comprising 90 compounds. Biological evaluation with a kinase panel showed that the library is useful for finding hits to provide a further “fragment growth” stage. (Color figure can be accessed in the online version.)

As a result of this study, 25ade, 25bde, and 25abe were identified as hits as inhibitors of JAC3, KDR, and TRKA, respectively. Thus the three-dimensional diversity-oriented conformational restriction strategy using our Pd NPs catalyst, SAPd(0), works effectively in the privileged structure library.

2.10. Structual Characterization of SAPd(0)²³⁾

The structure of SAPd(0) clarified by the present analysis is indicated in Fig. 19. Two independent experiments using TEM and XAFS analysis provided complementary information. The effective catalysts in SAPd(0) are the Pd NPs, which allow for high chemical transformations due to their large surface area. An especially interesting finding was that the Pd NPs are very crowded. The conventional NPs used in carbon–carbon coupling reactions are usually formed by a metal salt reduction and generally stabilized by polymeric molecules, tetraalkylammonium salt, and ionic liquids. The use of an appropriate stabilizing agent is therefore critical not only to obtain NPs of a suitable size to form highly active catalysts, but also to stabilize the surface so that leaching is minimized and NPs aggregations are prevented. Therefore the above results are important toward the assertion that sulfur-modified Au provides another example of an appropriate support for self-assembled multi-layers of Pd NPs (less than 5 nm) in high density.

Fig. 19. Drawing of the SAPd(0) Structure

(Color figure can be accessed in the online version.)

2.11. Self-assembled Glass-Supported Palladium (SGlPd(0))⁹³⁾

One of the big disadvantages of using SAPd(0) as a catalyst is that it contains Au as key ingredient, making it particularly expensive. In this article, we report the development of a novel self-assembled glass-supported Pd (SGlPd(0)) catalyst as a cheaper alternative to the Au-containing SAPd(0) catalyst described above. It is noteworthy that this new SGlPd catalyst holds the Pd NPs on its surface in the same way as SAPd(0) and exhibits equivalent catalytic activity towards the Suzuki–Miyaura coupling reaction.

To begin with, we screened a wide range of different materials as potential low cost alternatives to Au. A variety of different semiconductor, insulator, and metal materials was used to prepare immobilized Pd catalyst in the same way as SAPd(0). Among all materials tested, we found glass could be used as a suitable replacement for Au.

Sulfur-modified alkali-free glass was prepared by the treatment of glass with a Piranha solution (Chart 12). This material was treated with Pd(OAc)₂ in xylene at 100°C for 12 h to allow for the adsorption of Pd onto its surface to give Pd material E. We also investigated the use of several other inexpensive glasses, including quartz glass, white glass, and blue glass, which gave materials F–H under the same conditions as those used for the preparation of E. All four of these materials were subsequently evaluated as catalysts for the Suzuki–Miyaura coupling of iodobenzene (1a) (0.5 mmol) and 4-chlorophenylboronic acid (2a) (117.0 mg, 1.5 eq). Pleasingly, we found that the yields of 3a achieved using catalysts E, F, G, and H over 10 catalytic cycles were almost identical to those achieved using SAPd(0) (Table 1). This initial result therefore indicated that these new Pd-containing materials (E–H) possessed similar catalytic activities to SAPd(0). Furthermore, these results indicated that blue glass would be the optimal support material because of its low cost compared with the other glasses tested (Table 23, entry 4). Material H shall therefore be referred to hereafter as a self-assembled glass-supported Pd (SGlPd(0)) catalyst²⁹⁾ (Chart 12).

Chart 12. Preparation of Self-assembled Glass-Supported Pd

(Color figure can be accessed in the online version.)

Table 23. Glass-Supported Pd Materials E–H and Their Catalytic Activities towards Suzuki–Miyaura Coupling of 1a and 2a to Give 3a

2.11.1. X-Ray Absorption Fine Structure (XAFS) Measurement

To obtain information pertaining to the actual active edge structures of the SAPd(0) and SGlPd(0) catalysts, we subjected them to X-ray adsorption near edge structure (XANES) spectroscopy and compared their spectra with those of several standard materials, including Pd foil, PdO, PdSO₄, PdS and Pd(PPh₃)₄. The Pd K-edge XANES spectra of SAPd(0) and SGlPd(0) were the same and analogous to that of Pd foil. These results therefore suggest that the Pd species present in SAPd(0) and SGlPd(0) are the same. Furthermore, the Pd K-edge XANES spectrum of the SGlPd(0) material collected after 10 catalytic cycles of Suzuki–Miyaura coupling reaction (Table 23) was almost identical to that of fresh SGlPd(0).

2.11.2. Liquid-Phase Combinatorial Experiment

It is well known that reactions involving the use of polymer-supported Pd suffer from significant contamination problems and the catalyst has to be washed very carefully because of the strong absorption of starting materials and/or product to the polymer. As reported in our previous research,¹¹⁾ SAPd(0) absorbs very few organic compounds and can therefore be used in combinatorial synthesis because of its small surface area. It was envisaged that SGlPd(0) would perform in a similar manner to SAPd(0). Thus we investigated the use of SGlPd(0) as a catalyst for the liquid-phase combinatorial synthesis of different biaryl systems by varying the nature of the iodobenzene derivative used in the Suzuki–Miyaura coupling.

SGlPd(0) was evaluated as catalyst in Suzuki–Miyaura coupling of phenylboronic acid 2c with ten different aryl iodides using K₂CO₃ as base in EtOH (Table 24). All 10 reactions proceeded smoothly to give the desired products in near-quantitative yields. Furthermore, all 10 products were obtained in high purity after standard aqueous work-up. Notably, this catalyst worked well for both aryl (runs 1–8), and heteroaryl (runs 9 and 10) iodides. Similar results were observed when the boronic acid derivative was changed (Table 25). Taken together, these results demonstrate that SGlPd(0) possesses similar catalytic activity to SAPd(0) and that it can therefore be used as an effective Pd reservoir for synthesis without being contaminated by products or starting materials. Furthermore, SGlPd(0) was repeatedly used in the Suzuki–Miyaura couplings described above without any discernible decrease in its activity. In a typical procedure, a solution of SGlPd(0) (approx. 1 cm²), phenylboronic acid 2c (1.5 equiv.), aryl iodide (0.5 mmol), and K₂CO₃ (2 equiv.) was heated in EtOH (3 mL) at 80°C for 12 h. The mixture was then cooled and subjected to standard aqueous work-up to give the corresponding product 3 in high purity. SGlPd(0) was readily removed from the reaction mixture by filtration prior to the work-up and used repeatedly.

Table 24. Liquid-Phase Combinatorial Experiments of SGlPd(0) (Part 1)

^a) The same SGlPd was used for all the reaction in this table. ^b) Yield in parenthesis indicates the isolated yield by using SAPd. ^c) This reaction was carried out for 5 h. ^d) This reaction was carried out for 12 h. (Color figure can be accessed in the online version.)

Table 25. Liquid-Phase Combinatorial Experiments of SGlPd(0) (Part 2)

^a) Yield in parenthesis indicates the isolated yield using SAPd. (Color figure can be accessed in the online version.)

2.11.3. Measurement of Pd Amount in SGlPd(0) and in Reaction Mixture

ICP-MS was used to measure the amount of immobilized Pd on the SGlPd(0) catalyst, as well as the amount of Pd leaching into the reaction mixture. The amount of immobilized Pd on the SGlPd(0) catalyst was measured both before and after the application of 10 catalytic cycles. The amount of released Pd after cooling in each cycle was extremely low (0.2–0.7 µg [0.06 − 0.23 ppm] for a 0.5 mmol scale reaction). A comparison of the leaching properties of the SAPd(0) and SGlPd(0) catalysts revealed that the latter released more active Pd into the reaction mixture (less than 0.17 versus less than 0.01 ppm). Almost half immobilized Pd on SGlPd(0) was released in 10 times of Suzuki–Miyaura coupling. Although the amount of Pd leached into the reaction mixture during each cycle was slightly higher for the SGlPd(0) catalyst, this material still allowed for the Suzuki–Miyaura coupling to proceed efficiently from overall 10 cycles under ligand-free conditions.

Hence we successfully developed a self-assembled glass-supported Pd material (SGlPd(0)) using blue glass as a much cheaper alternative to Au. The catalytic activity of SGlPd(0) was examined in a ligand-free Suzuki–Miyaura coupling reaction, with the reaction only requiring 3 h to achieve high yields, which was much shorter than the same reaction using SAPd(0) (12 h). Although SGlPd(0) leached larger amounts of active Pd species (less than 0.17 ppm for a 0.5 mmol scale reaction) into the reaction mixture compared with SAPd(0) (less than 0.01 ppm for a 0.5 mmol scale reaction), this value is still much lower than the value required by US Food and Drug Administration (i.e., <5 ppm residual metal in product streams). Based on its low leaching, our newly developed SGlPd catalyst could also be recycled up to 10 times without any discernible decrease in its catalytic activity.

3. Self-assembled Gold Supported Nickel (SANi(0))²⁴⁾

After we obtained the above results, we decided that the in situ PSSO method could be used to immobilize various transition-metal NPs, including base metals. Our in situ PSSO method can be used easily to produce Ni NPs with low Ni leaching and good recyclability for use in organic synthesis. Homogeneous Ni-catalyzed cross-coupling reactions have contributed greatly to advances in C–C cross-coupling reaction chemistry.^94,95) However, only a few studies of immobilized Ni NP catalysts for C–C cross-coupling reactions have been reported.^96,97) This is attributed to difficulties in the preparation and immobilization of Ni NPs, which are easily oxidized because of their high reactivity.

Against this background, we successfully developed self-assembled Au-supported Ni NP (SANi(0)) catalysts using a combination of the in situ PSSO method and an appropriate reducing agent. We found that SANi(0) repeatedly catalyzed Kumada coupling and Negishi coupling reactions under ligand-free conditions. The chemistry described here is useful from a synthetic point of view because reactions, including liquid-phase combinatorial synthesis, can be repeatedly performed under ligand-free conditions and with low Ni leaching.

In the beginning, we developed a suitable method for preparation of a SANi(0) catalyst that can be repeatedly used to catalyze ligand-free Kumada coupling reactions. Our three-step method for SANi(0) preparation, which is based on our reported method for SAPd(0) preparation, is shown in Chart 13. The first step is the same as that for SAPd(0) preparation. Sulfur-modified Au (S-Au) was prepared by treating Au mesh (12 × 14 mm, 100 mesh) with a Piranha solution, i.e., a mixture of H₂SO₄ and Na₂S₂O₈¹⁶⁾ (Chart 13, step 1). The second step is important and involves immobilization of Ni NPs on S-Au using the in situ PSSO method. We first tried to immobilize Ni NPs on S-Au under various conditions similar to those used for Pd NP immobilization in SAPd preparation. However, when S-Au was treated with Ni(acac)₂ instead of Pd(OAc)₂ in p-xylene at 100°C, the catalyst obtained after the third step, i.e., heating and washing, did not catalyze typical Kumada coupling reactions. It has been reported that a high temperature, i.e., >200°C, is usually needed for Ni NP production by thermal decomposition. On the basis that the solvent would form a matrix of self-assembled multi-layers through benzyl radical species, as in the case of SAPd(0), we selected 1,2,4,5-tetramethylbenzene (Me₄-benzene; bp = 197°C) as the solvent and prepared the catalyst at 200°C using Ni(acac)₂. However, the resulting material did not catalyze Kumada coupling reactions. We then considered the oxidation/reduction levels of Ni and Pd, and assumed that Ni reduction is more difficult than Pd reduction. Several Ni NP preparation methods have been reported in the literature, e.g., thermal or ultrasonic decomposition of organonickel precursors, chemical reduction, electrochemical reduction, and vapor-phase deposition.⁹⁷⁾ We next focused on the conditions for Ni NP synthesis via chemical reduction of Ni(II). It is noteworthy that the reducing agent must be able to reduce Ni(II) to Ni(0) at the same rate as the solvent forms the matrix that encapsulates the Ni NPs. The use of conventional inorganic reductants such as LiBEt₃H or NaBH₄ failed, and a large amount of Ni black was formed in the Ni immobilization step. After screening various conditions and other reducing agents, we identified suitable immobilization conditions: treatment of S-Au with Ni(acac)₂ (0.034 mmol) as a Ni source and 4-methoxybenzyl alcohol (0.515 mmol) as the Ni(II) to Ni(0) reductant in Me₄-benzene (2 g) at 200°C (step 2). The third step consists of heating, and washing in p-xylene at 135°C for 12 h. The resulting SANi(0) catalyzed the Kumada coupling reaction of 4-iodoanisole (1c) (0.25 mmol) and p-tolylmagnesium bromide (1.3 equiv) in THF at 75°C for 3 h under ligand-free conditions to give the corresponding biaryl compound 3a in 95% yield (Chart 14). In the absence of SANi(0), only a trace amount of 3a was detected in this Kumada coupling reaction. ICP- MS) of SANi(0) showed that 509 µg (the average of three samples) of Ni was immobilized on S-Au. These results encouraged us to investigate recycling of SANi(0).⁹⁸⁾

Chart 13. Preparation of SANi(0)

Chart 14. SANi(0)-Catalyzed Ligand-Free Kumada Coupling Reaction of 1a with 2

The yields of 3j in the first 10 runs in recycling experiments are shown in Table 26. The SANi(0)-catalyzed ligand-free Kumada coupling reactions proceeded smoothly to afford the products in 82–97% yields for runs 1 to 10; no significant loss of catalytic activity was observed. We next measured the amount of Ni released into the reaction mixture in each cycle, using ICP-MS. The results showed that the amount of Ni released in each run was 12–23 µg (1.8–3.3 ppm) (Table 26).

Table 26. Yields of 3j and Amounts of Ni Released in Kumada Coupling Reactions Using Recycled SANi(0)

	Yield of 3j [%]a)	Amount of released Ni [µg]b)
1st	91	23
2nd	84	13
3rd	82	17
4th	92	17
5th	90	12
6th	86	12
7th	96	12
8th	97	12
9th	88	12
10th	84	14
Total releasing		144
Used SANi(0)c)		272

a) Yields were average values of three sets of reactions. b) The amount of released Ni into reaction mixture were average values of three sets of reactions. c) The amount of Ni of SANi(0) used for Kumada coupling reaction in 10-cycle experiment.

We next investigated the scope of SANi(0) use by performing liquid-phase combinatorial synthesis using various iodobenzene derivatives and Grignard reagents (Table 27). The reactions of 2-iodonaphthalene (first cycle) and 2-iodotoluene (second cycle) with 4-methoxyphenylmagnesium bromide provided the corresponding biaryl products in high yields. The reaction of 1-chloro-4-iodobenzene with 4-methoxyphenylmagnesium chloride gave the corresponding product in 47% yield in the third reaction. In the next six cycles, the reaction proceeded smoothly to give the corresponding products in 77–93% yields. In the 10th cycle, the biaryl product was obtained in 52% yield. When similar combinatorial syntheses are conducted with polymer-supported metal catalysts, the catalysts have to be washed carefully to prevent contamination because of strong adsorption of starting materials and/or products on the polymers. In contrast, no contamination was observed after simple washing of SANi(0) because of its low affinity for organic compounds.

Table 27. Liquid-Phase Combinatorial Synthesis Using SANi(0)

We next examined SANi(0)-catalyzed ligand-free Negishi coupling reactions of 3-iodoanisole (1ab) with p-tolylzinc chloride, prepared from 1z via a commonly used procedure (Chart 15). We optimized the reaction conditions and found that the conditions 1c (0.25 mmol), p-tolylzinc chloride (1.5 equiv), LiCl (1.5 equiv), THF/NMP (1 : 1, v/v), 100°C, and 12 h provided the coupling product 3p in 95% yield.

Chart 15. SANi(0)-Catalyzed Negishi Coupling Reaction of 1ab with p-Tolylzinc Chloride

With the optimized reaction conditions in hand, we next investigated the scope and limitations of SANi(0)-catalyzed Negishi coupling reactions. The results are summarized in Table 28. The reaction of 4-iodoanisole with p-tolylzinc chloride in the presence of SANi(0) gave the coupling product in 75% yield (entry 1). The reaction of 4-bromoanisole also proceeded smoothly to provide the same product in 75% yield (entry 2). However, the use of 4-chloroanisole resulted in a 10% yield of the coupling product (entry 3). Sterically hindered substrates, i.e., 2-iodotoluene and 1-iodonaphthalene, provided the corresponding products in 74 and 77% yields, respectively (entries 4 and 5). The more sterically hindered 2,6-dimethylbromobenzene was converted to the biaryl product in 37% yield (entry 6). Trisubstituted 3,4,5-trimethoxybromobenzene provided the product in 77% yield (entry 7). A fluorine-substituted aryl bromide, 1-bromo-4-fluorobenzene, was converted to the corresponding product in 77% yield (entry 8). Ethyl 4-iodobenzoate, bearing a nucleophilic functional group, was tolerated in this reaction to give the corresponding coupling product in 55% yield (entry 9). We also investigated sp²–sp³ cross-coupling using an alkylzinc chloride; n-butylzinc chloride reacted with 1ab in the presence of SANi(0) to afford the alkylated product 3q in 32% yield (Chart 16).

Table 28. Substrate Scope of SANi(0)-Catalyzed Negishi Coupling Reaction

Entry	Substrate 1		Product 3
	X	R	Yield [%]
1	I	4-OMe	75
2	Br	4-OMe	75
3	Cl	4-OMe	10
4	I	2-Me	72
5	I	1-Naphthyl	77
6	Br	2,6-Me2	37
7	Br	3,4,5-(OMe)3	77
8	Br	4-F	78
9	I	4-CO2Et	55

Chart 16. SANi(0)-Catalyzed Negishi Coupling Reaction of Alkylzinc Chloride

Recycling experiments were performed and ICP-MS was used to determine the amounts of Ni released during the Negishi coupling reactions (Table 29). When SANi(0) was reused five times for Negishi coupling reactions of 1ab with p-tolylzinc chloride, the yields in runs 1–5 were 79, 78, 72, 64, and 80%, respectively. No significant loss of SANi(0) catalytic activity was observed in the Negishi coupling reactions. The amounts of Ni released were 2–3 µg (1.5–3.8 ppm). These amounts are slightly lower than those released in the SANi(0)-catalyzed Kumada coupling reactions.

Table 29. Yields of 3e and Amounts of Ni Released in SANi(0)-Catalyzed Negishi Coupling Reactions

	Yield of 3p [%]a)	Amount of released Ni [µg]b)
1st	79	2
2nd	78	2
3rd	72	3
4th	64	2
5th	80	2
Total releasing		11
Used SANic)		320

a) Yields were average values of three sets of reactions. b) The amounts of released Ni into reaction mixtures were average values of three sets of reactions. c) The amount of Ni of SANi(0) used for Negishi coupling reaction after 5-cycle experiment.

From the above results, we can conclude that SANi(0) has some advantages over conventional catalytic systems: 1) ligand-free conditions, 2) low leaching of Ni, 3) high recyclability, and 4) suitability for liquid-phase combinatorial synthesis.

3.1. SANi(0) Characterization

The structure and chemical composition of SANi(0) were investigated using some spectroscopic methods. We examined the SANi(0) Ni K-edge EXAFS spectra before the Kumada coupling reaction and after the 10th cycle of the SANi(0)-catalyzed reaction, and of Ni foil and Ni(PPh₃)₄ as standard samples (Fig. 20). The Ni K-edge EXAFS spectra of SANi(0) before and after use, and of Ni foil were identical. However, the K-edge X-ray absorption near-edge structure (XANES) spectrum of Ni(PPh₃)₄, an organonickel compound containing zerovalent Ni, showed different chemical shifts. These results indicate that the Ni species in SANi(0) are zerovalent Ni or metallic Ni, and are probably Ni NPs, as in the previously reported case of SAPd(0). The sizes of the metallic Ni species were precisely determined via theoretical curve fitting of the EXAFS oscillations for the Ni–Ni peak by Fourier transforms of EXAFS oscillation spectra. The Ni K-edge Fourier transforms of the EXAFS oscillation spectra in the k range 2.0–12.0 Å⁻¹ for Ni foil, Ni(PPh₃)₄, and SANi(0) before and after reaction, are shown in Fig. 21. The curve-fitting analysis showed that before the reaction SANi(0) had an average coordination number (N) of 10.5, which is lower than that for Ni foil. From these results, the average size of Ni species with an N of 10.5 was estimated approximately 3 nm (Table 30). These results and XANES analyses show that the Ni source in SANi(0) is Ni(0) NPs. We also estimated the average size of the Ni NPs in SANi(0) after the reaction; it was still approximately 3 nm after the 10th cycle. It should be noted that the XANES spectrum and SANi(0) size before the reaction were almost the same as those after the reaction. These results suggest that the Ni(0) NPs in SANi(0) are stable and tolerate C–C coupling reaction conditions.

Fig. 20. Ni K-Edge XANES Spectra of SANi(0) before and after Use, Ni Foil, and Ni(PPh₃)₄

(Color figure can be accessed in the online version.)

Fig. 21. Ni K-Edge Fourier Transforms of EXAFS Oscillation Spectra of SANi(0) and Ni Foil

(Color figure can be accessed in the online version.)

Table 30. Theoretical Curve-Fitting Results for EXAFS of SANi(0) and Ni Foil

	R (Å)a)	Nb)	D (nm) c)
Ni foil	2.48	12	—
SANi(0) before	2.48	10.5	3
SANi(0) after	2.49	10.5	3

a) Interatomic distance. b) First shell coordination number. c) The average diameter of particles.

Because the metals in SANi(0) and SAPd(0) are both present as NPs, we evaluated these results by comparing the EXAFS spectra of SANi(0) with those of SAPd(0). The EXAFS spectra and data for SAPd(0) are shown in Fig. 22 and Table 31. Figure 22 shows that the spectra of SAPd(0) before use and after use in 10 cycles of a Buchwald–Hartwig reaction (SAPd BH) were the same as that of Pd foil. Curve fitting showed that the average coordination numbers for SAPd(0) before and after the BH reaction were 9.1 and 8.1, respectively, i.e., lower than that for Pd foil. From these results, the average sizes of the Pd NPs before and after reaction were estimated 3 and 2 nm, respectively. The EXAFS results for SAPd(0) were similar to those for SANi(0).

Fig. 22. Pd K-Edge Fourier Transforms of EXAFS Oscillation Spectra of SAPd(0) and Pd Foil

(Color figure can be accessed in the online version.)

Table 31. Theoretical Curve-Fitting Results for EXAFS of SAPd(0) and Pd Foil

	R[Å]a)	N b)	D [nm] c)
Pd foil	2.74	12	—
SAPd(0) before	2.79	9.1	3
SAPd(0) after	2.80	8.1	2

a) Interatomic distance. b) First shell coordination number. c) The average diameter of particles.

We then focused on the sulfur in SANi(0), and recorded the sulfur K-edge XANES spectra of SANi(0) before and after use in Kumada and Negishi coupling reactions, and of standard samples, i.e., NiSO₄, Na₂S₂O₈, Na₂SO₃, Na₂S₂O₃, and PdS (Fig. 23). The XANES spectra of SANi(0) before and after reaction have a clear resonance peak at 2478 eV. A comparison of the XANES spectrum of SANi(0) with those of standard materials suggests that the sulfur in SANi(0) is strongly oxidized and present as SO₄²⁻.²³⁾ The chemical shift of the sulfur in SANi(0) is similar to that previously reported for SAPd(0). These results strongly indicate that SANi(0) consists of a matrix containing SO₄²⁻ and is similar to SAPd(0), which consists of Pd(0) NPs implanted in a matrix formed by complexation of SO₄²⁻ and p-xylene. Based on these spectroscopic analyses, we conclude that SANi(0) consists of Ni NPs (of size approximately 5 nm) and a matrix formed by complexation of SO₄²⁻ and Me₄-benzene as the solvent, and that SANi(0) forms a matrix of self-assembled multi-layers of Ni NPs.

Fig. 23. Sulfur K-Edge XANES Spectra of SANi(0) before and after Use, and Standard Samples

(Color figure can be accessed in the online version.)

Hence we successfully developed self-assembled Au-supported Ni NPs, i.e., SANi(0). SANi(0) was easily prepared by a three-step procedure involving Ni immobilization, using Ni(acac)₂ as Ni source and 4-methoxybenzyl alcohol as a reductant to produce Ni(0) from Ni(II), via an in situ PSSO method. SANi(0) catalyzed C–C bond-forming Kumada and Negishi cross-coupling reactions under ligand-free conditions, and was recycled without loss of catalytic activity. Detailed spectroscopic analyses show that SANi(0) consists of self-assembled multi-layers of Ni(0) NPs with diameters approximately 3 nm. The results also show that the Ni NPs in SANi(0) are stably immobilized.

4. Self-assembled Gold Supported Ruthenium (SARu(0))²⁵⁾

We used the PSSO method to produce Ru NPs with low Ru leaching and good recyclability, which could be a good catalyst for liquid-phase combinatorial organic synthesis. We successfully developed self-assembled Au-supported Ru NPs (SARu(0)) catalysts using a combination of the in situ PSSO method and an appropriate reducing agent. We found that SARu(0) can repeatedly catalyze Suzuki–Miyaura coupling under ligand-free conditions. The chemistry described here is useful from a synthetic point of view because reactions, including liquid-phase combinatorial synthesis, can be repeatedly performed under ligand-free conditions with low Ru leaching.

4.1. SARu(0) Preparation and Catalytic Activity in Suzuki–Miyaura Coupling of Aryl Iodides

In the beginning, we developed a suitable method to prepare a SARu(0) catalyst that can be used repeatedly for ligand-free Suzuki–Miyaura coupling. The three-step method developed to prepare SARu(0), which is based on our reported method for SAPd(0) preparation,^11–23) is shown in Chart 17.

Chart 17. Preparation of SARu(0)

The first step was the same as that for SAPd(0) preparation. The second step involved immobilization of Ru NPs on S-Au using the in situ PSSO method. We first tried to immobilize Ru NPs on S-Au under various conditions similar to those used for Pd NPs immobilization in SAPd(0) preparation. However, S-Au was recovered, when S-Au was treated with Ru(acac)₃ instead of Pd(OAc)₂ in p-xylene at 100°C. Then, we considered the redox potentials of Ru and Pd, and assumed that Ru reduction is more difficult than Pd reduction. Thus we next focused on the conditions for Ru NPs synthesis via chemical reduction of Ru(III). It is noteworthy that the reducing agent must be able to reduce Ru(III) to Ru(0) at the same rate as the solvent forms the matrix that encapsulates the Ru(0) NPs. The use of conventional inorganic reductants such as LiBEt₃H or NaBH₄ failed, and a large amount of Ru black was formed in the Ru immobilization step. After screening various conditions and other reducing agents, we identified suitable immobilization conditions: treatment of S-Au with Ru(acac)₃ as a Ru source and 4-methoxybenzyl alcohol as the Ru reductant in p-xylene at 135°C for 12 h (step 2). The third step involved heating in p-xylene at 135°C for 12 h to remove unnecessary Ru species attached on the catalyst surface. ICP-MS analysis of SARu(0) revealed that 194 µg (average of two samples) of Ru was immobilized on 96.4 mg of S-Au mesh (average of two samples). Also, we detected no Pd on SARu(0) in this ICP-Mass analysis.

The resulting SARu(0) catalyzed the Suzuki–Miyaura coupling of 4-iodobenzene (1a) (0.25 mmol) and phenylboronic acid (2c) (1.5 equiv.) repeatedly (Fig. 24). That is to say, a mixture of SARu(0) (1 sheet, 12 × 14 mm, 100 mesh) and K₃PO₄ (3.0 equiv.) in dimethoxyethane (DME) was heated at 80°C for 6 h under ligand-free conditions. After SARu(0) was removed from the reaction mixture, H₂O (1.0 mL) was added to solve insoluble reagents and the resulting mixture stirred at 100°C for 10 h to give the corresponding biaryl compound 3c in 99% yield. We have to remove SARu(0) in the end of Step 1 and add water to the reaction mixture in the beginning of Step 2, because SARu(0) stands for moisture level of water, but the film, which involves Ru species, on SARu(0) is destroyed by solvent amount of water. Only a trace amount of 3c was detected in this Suzuki–Miyaura coupling reaction in the absence of SARu(0). These results encouraged us to investigate the recyclability of SARu(0) in the same Suzuki–Miyaura coupling reaction.

Fig. 24. SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of 1a with 2a and Its Time–Course Experiments

(Color figure can be accessed in the online version.)

The yields of 3c in the first ten runs in recycling experiments are presented in Table 32. The SARu(0)-catalyzed ligand-free Suzuki–Miyaura coupling proceeded smoothly to afford the products in 91–99% yield for runs 1 to 10; no marked loss of catalytic activity was observed. We next measured the amount of Ru released into the reaction mixture in each cycle using ICP-MS. As shown in Table 32, the amount of Ru released in each run was 0.11–1.93 µg (0.05–0.83 ppm).

Table 32. Yields of 3c and Amount of Ru Released in Suzuki–Miyaura Coupling Using Recycled SARu(0)

Run	Yield of 3c (%)a)	Amount of released Ru (µgb), mmol%c), %d))
1st	99	1.93, 7.64, 0.955
2nd	99	0.48, 1.90, 0.247
3rd	99	0.28, 1.11, 0.144
4th	95	0.25, 0.989, 0.129
5th	97	0.25, 0.989, 0.129
6th	94	0.28, 1.11, 0.144
7th	95	0.15, 0.594, 0.0773
8th	92	0.13, 0.514, 0.0670
9th	91	0.12, 0.475, 0.0619
10th	91	0.11, 0.435, 0.0567
Total releasing		3.95b)
Used SARu(0)c)		145b)

a) Yields were determined by HPLC and are average values of two sets of reactions. b) The amount of Ru released into the reaction mixture is the average value of two sets of reactions. c) mmol% of Ru species in the reaction mixture. d) % of Ru employed in the catalysis. e) The amount of Ru on SARu(0) after ten cycles of Suzuki–Miyaura coupling.

We next investigated the scope of SARu(0) use by performing liquid-phase combinatorial synthesis using various iodobenzene derivatives and boronic acid reagents (Tables 33, 34). The reactions of iodobenzene (first cycle), 4-methoxyiodobenzene (second cycle), 4-acetyliodobenzene (third cycle), 4-nitroiodobenzene (fourth cycle), and 4-iodotoluene (sixth cycle), with phenylboronic acid provided the corresponding biaryl products in quantitative yields. The reaction of 1-chloro-4-iodobenzene with phenylboronic acid gave the corresponding product in 88% yield in the fifth reaction. In the next three cycles, the reaction proceeded smoothly to give the corresponding products in yields of 86–91% (entries 7–9). In the tenth cycle, stilbene was obtained in 86% yield (entry 10).

Table 33. Liquid-Phase Combinatorial Synthesis Using SARu(0) and Aryl Iodides (Part 1)

^a) Isolated Yields

Table 34. Liquid-Phase Combinatorial Synthesis Using SARu(0) and Aryl Iodides (Part 2)

^a) Isolated yields. ^b) 2 equiv. of boronic acid was used.

This was likewise the case when the boronic acid derivative was changed as shown in Table 34. These results demonstrate that SARu(0) is an effective Ru reservoir for this kind of synthesis without any contamination of other products or starting materials. When similar combinatorial syntheses are conducted with polymer-supported metal catalysts, the catalysts have to be washed carefully to prevent contamination because of the strong adsorption of starting materials and/or products on the polymers. In contrast, no contamination was observed after simple solvent washing of SARu(0) because of its low affinity for organic compounds. S-Au is only solid-support and does not behave as catalyst in this Ru NPs catalyzed Suzuki–Miyaura coupling. Au mesh itself is recyclable to make another SARu(0). Also, there is a possibility that Au in SARu(0) can be replaced by cheap glass, because we have succeeded to replace Au in SAPd(0) by glass.⁹³⁾

4.2. SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of Aryl Bromides

We next investigated the scope of SARu(0) use by performing liquid-phase combinatorial synthesis using various bromobenzene derivatives (Table 35). Although we needed a longer reaction time in the first step because of the relatively slow oxidative addition of aryl bromides, the reactions of bromobenzene (first cycle), 4-acetylbromobenzene (third cycle), and 4-nitrobromobenzene (fourth cycle) with phenylboronic acid provided the corresponding biaryl products in quantitative yields. The reactions of 4-methoxybromobenzene (second cycle) and 1-chloro-4-bromobenzene (fifth cycle) with phenylboronic acid gave the corresponding product in 88 and 70% yields, respectively. In the following five cycles, the reaction proceeded smoothly to give the corresponding products in yields of 82–91%. These results indicate that SARu(0) is much more reactive than SAPd(0) in ligand-free Suzuki–Miyaura coupling of aryl bromides, because bromobenzene did not give the coupled product under the conventional heating conditions with SAPd(0).¹⁵⁾ Furthermore, this methodology is very useful to synthesize biaryl compounds with heterocycles (Table 36), because numerous heterocyclic aryl bromides are commercially available at reasonable prices.

Table 35. Liquid-Phase Combinatorial Synthesis Using SARu(0) and Aryl Bromides

^a) Isolated Yields

Table 36. Liquid-Phase Combinatorial Synthesis Using SARu(0) and Heterocyclic Aryl Bromides

^a) Isolated Yields

4.3. SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of Aryl Chlorides

Aryl chlorides are economical and readily available substrates and have rarely been used in Suzuki–Miyaura coupling, because of the high activation barriers associated with oxidative insertion of Ru(0) species into the C–Cl bond. We next investigated the scope of SARu(0) use by performing Suzuki–Miyaura coupling of aryl chloride 1o with phenylboronic acid 2c. Only trace amount of the corresponding coupling product was obtained, as exemplified in Chart 18, although we tried various traditional heating conditions.

Chart 18. SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of 1o with 2c

We previously developed a microwave-assisted strategy for Suzuki–Miyaura cross-coupling reactions catalyzed by SAPd(0), which has very low leaching properties.¹⁵⁾ Because the leaching induced by microwave-assisted physical etching can be easily controlled during the process, the combined use of two microwaves with different irradiation approaches made this protocol very efficient. However, irradiation of the reaction mixture containing 1o and 2c by traditional microwave reactors did not deliver the desired product 3c due to the rapid generation of hot spots (Fig. 25a) on heating. The reaction vessel caused the irradiation to turn off within a very short period when the temperature limit was reached (Fig. 26). Then, we envisioned that the expected reaction might proceed if the whole reaction mixture was continuously irradiated by microwaves. We constructed a special microwave reactor equipped with a cooling unit for this purpose (Fig. 25b), and used it in the first step reaction. The reaction mixture was irradiated at a certain microwave power for 2 h (Table 37, entries 1–4) as the first step and then the whole reaction mixture was treated in the usual second step. We obtained 3c in 31% yield when we used a microwave irradiation power 70 W in the first step. We achieved continuous microwave irradiation of SARu(0), using this novel microwave reactor fitted with a cooling system (Fig. 27). We observed aggregated Ru in the bottom of the flask (Table 37, entry 4), indicating that stronger irradiation leached more Ru than needed for the reaction, while weaker irradiation did not work. Entries 1–4 reveal that a microwave irradiation power of 70 W is suitable. When we shortened the irradiation time to 1 and 1.5 h, the yield of 3c increased to 41 and 38%, respectively. When the reaction time in the second step was prolonged to 24 h, a quantitative yield of 3c was successfully obtained (entries 7 and 8).

Fig. 25. Microwave Reactor Designed for SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of 1o with 2c (First Step)

Schematics of (a) a conventional microwave reactor and (b) our microwave reactor with a cooling system. (Color figure can be accessed in the online version.)

Fig. 26. Microwave Irradiation Data for SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of 1o with 2c (First Step) Using a Conventional Microwave Reactor

(Color figure can be accessed in the online version.)

Table 37. Microwave Assisted SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of 1o with 2c

Entry	Time (h)	MW (W)	Yield (%)a)
1	2	10 W	trace
2	2	50 W	trace
3	2	70 W	31
4	2	100 W	15
5	1	70 W	41
6	1.5	70 W	38
7b	1	70 W	96
8b	1.5	70 W	97

a) Determined by HPLC. b) The reaction was conducted for 24 h in the second step.

Fig. 27. Microwave Irradiation Data for SARu(0)-Catalyzed Ligand-Free Suzuki–Miyaura Coupling of 1o with 2c (First Step) Using Our Microwave Reactor in Fig. 26(b)

(Color figure can be accessed in the online version.)

We next investigated the scope of the combined use of SARu(0) and the special microwave reactor by performing liquid-phase combinatorial synthesis using various chlorobenzene derivatives (Table 38). The reactions of chlorobenzene (first cycle) and 4-nitrochlorobenzene (fourth cycle) with phenylboronic acid provided the corresponding biaryl products in high yields. The reactions of 4-methoxychlorobenzene (second cycle) and 2-methylchlorobenzene (third cycle) with phenylboronic acid gave the corresponding product in 60 and 52% yield, respectively. In the following four cycles, the reactions proceeded smoothly to give the corresponding heterocyclic products in yields of 65–83%. Because a variety of heterocyclic aryl chlorides are commercially available at reasonable prices, this methodology is very useful to make biaryl compounds with heterocycles.

Table 38. Liquid-Phase Combinatorial Synthesis Using SARu(0) and Aryl Chlorides

^a) Isolated Yields

4.4. SARu(0) Characterization

Using several spectroscopic methods, the structure and chemical composition of SARu(0) were investigated. We examined the SARu(0), Ru K-edge X-ray absorption near-edge structure (XANES) spectra before Suzuki–Miyaura coupling (denoted SARu(0) before) and after the 10th cycle of the SARu(0)-catalyzed reaction (SARu(0) after), and of Ru powder, RuCl₃, Ru(acac)₃, [RuCl₂(p-cymene)]₂, and RuO₂ as standard samples. The Ru K-edge XANES spectra of SARu(0) before and after use, and of Ru powder were similar. On the other hand, the K-edge XANES spectra of [RuCl₂(p-cymene)]₂, Ru(acac)₃, RuCl₃, and RuO₂ showed different chemical shifts. We also examined the SARu(0) Ru L-edge XANES spectra before Suzuki–Miyaura coupling and of Ru foil as standard samples. The Ru L-edge XANES spectra of SARu(0) and Ru foil were almost identical. This indicates that the Ru species in SARu(0) are zerovalent Ru or metallic Ru, and are probably Ru NPs, as in the case of SAPd(0).

The sizes of the metallic Ru(0) species were determined precisely via theoretical curve fitting of the EXAFS oscillation for the Ru–Ru peak by Fourier transforms of EXAFS oscillation spectra. The Ru K-edge fourier transforms of the EXAFS oscillation spectra in the k range of 2.0–12.0 Å⁻¹ for Ru powder and SARu(0) before and after reaction are depicted. The curve-fitting analysis revealed that before the coupling reaction, SARu(0) had an average coordination number (N) of 7.0 ± 1.4, which is lower than that for Ru powder. From these data, the average size of Ru species with an N of 7.0 was estimated approximately 1.0 ± 0.3 nm (Table 39). These results and XANES analyses confirmed that the Ru source in SARu(0) is Ru(0) NPs. We also estimated the average size of the Ru NPs in SARu(0) after the reaction, and found that their size was still approximately 2.4 ± 1.6 nm after the 10th cycle. It is worth noting that the XANES results suggest that the Ru(0) NPs in SARu(0) are stable and tolerate C–C coupling reaction conditions.

Table 39. Theoretical Curve-Fitting Results for EXAFS Spectra of SARu(0) and Ru Powder^a)

	CNb)	R[Å]c)	δ2[Å2]d)	E0[eV]e)	Rfactorf)
Ru powder	12g)	2.68 ± 0.01	0.0034 ± 0.0005	2.2 ± 1.0	0.003
SARu(0) before	7.0 ± 1.4	2.67 ± 0.01	0.0081 ± 0.0011	−3.1 ± 1.7	0.014
SARu(0) after	8.7 ± 2.1	2.67 ± 0.01	0.0078 ± 0.0014	−3.5 ± 2.1	0.022

a) k³: k range = 2.0–12.0 Å⁻¹, r range = 1.7–2.9 Å for Ru–Ru. Intrinsic loss factor, S₀² = 0.65 (Ru–Ru from Ru powder data). b) First shell coordination number. c) Interatomic distance. d) Debye–Waller factor. e) Correction of edge energy. f) Goodness-of-fit index. g) The coordination number was fixed as that of a hexagonal close packed lattice.

We then focused on the sulfur in SARu(0), and analyzed the sulfur K-edge XANES spectra of SARu(0) before and after use in Suzuki–Miyaura coupling reactions and of the standard samples Na₂SO₄, Na₂SO₃, Na₂S₂O₈, Na₂S₂O₃, PdS, and SAPd(0). The XANES spectra of SARu(0) before and after reaction display a clear resonance peak at 2481.6 eV. Comparison of the XANES spectrum of SARu(0) with those of standard materials suggests that the sulfur in SARu(0) is strongly oxidized and present as SO₄²⁻. The chemical shift of the sulfur in SARu(0) is similar to that previously reported for SAPd(0). These results strongly indicate that SARu(0) consists of a matrix containing SO₄²⁻ and is similar to SAPd(0), which is composed of Pd NPs implanted in a matrix formed by complexation of SO₄²⁻ and p-xylene. Based on these spectroscopic analyses, we conclude that SARu(0) consists of Ru NPs with a size approximately 1–3 nm and a matrix formed by complexation of SO₄²⁻ and p-xylene as the solvent, and that SARu(0) forms a matrix of self-assembled multilayers of Ru NPs.

4.5. Reaction Mechanism

Now we consider a plausible reaction mechanism from the data of SARu(0) characterization (Chart 19). Via the oxidative addition between arylhalide and Ru(0) nanopartiel on SARu(0), Ru(0) NPs on SARu(0) was released into the reaction mixture in the Step 1 (Chart 18 and Table 37), although continuous irradiation of microwave in Step 1 is critical for arylchloride. After removing SARu(0) in the end of Step 1 and adding water in the beginning of Step 2, insoluble base was dissolved in the mixture and Ru(0) NPs catalyzed ligand-free Suzuki–Miyaura coupling. Soluble base is critical for this reaction to proceed. It is also important to make appropriate Step 1 conditions to release sufficient and minimum amount of Ru(0) into the reaction mixture for both ligand-free Suzuki–Miyaura coupling and the repeated use of SARu(0).

Chart 19. A Plausible Reaction Mechanism: SARu(0) Catalyzed Ligand-Free Suzuki–Miyaura Coupling

(Color figure can be accessed in the online version.)

Hence we successfully developed self-assembled Au-supported Ru(0) NPs, i.e., SARu(0), for recyclable, low-leaching, and ligand-free Suzuki–Miyaura coupling. SARu(0) was easily prepared by a three-step procedure involving Ru immobilization using Ru(acac)₃ as a Ru source and 4-methoxybenzyl alcohol as reductant to produce Ru(0) from Ru(II) via an in situ PSSO method. SARu(0) catalyzed C–C bond-forming Suzuki–Miyaura coupling of not only aryl iodides and aryl bromides but also aryl chlorides under continuous microwave irradiation. Spectroscopic analyses showed that SARu(0) consists of Ru(0) NPs with diameters of approximately 1–3 nm. The results also showed that the Ru(0) NPs, which are known to be easily oxidized to Ru(III), are stably immobilized in SARu(0). These results suggest that our “in situ PSSO method” could be a versatile method to immobilize metal NPs.

5. Self-assembled Gold Supported Iron (SAFe(II))²⁶⁾

Fe is an ideal transition metal, because Fe is enormously earth abundant and relatively safe. In 1971, Kochi and Tamura found the first example of Kumada coupling using Fe catalyst, using alkenyl bromides and alkyl Grignard reagents.⁹⁹⁾ In 2002, Fürstner reported the cross-coupling reactions of aryl chlorides, triflates and tosylates, with alkyl Grignard reagents.¹⁰⁰⁾ Subsequently and independently, using Fe catalyst, Hayashi,¹⁰¹⁾ Fürstner,¹⁰²⁾ Bedford,¹⁰³⁾ and Nakamura¹⁰⁴⁾ have reported the Kumada coupling of alkyl halides with aromatic Grignard reagents. Senanayake¹⁰⁵⁾ used Fe catalyst for Kumada coupling of 2-chloropyrazine and aromatic Grignard reagents. On the other hand, Wang and colleagues¹⁰⁶⁾ performed the reaction between pyrimidine compounds and aromatic and alkyl Grignard reagents. Despite the above extensive research, there are very few reports of ligand-free, environmentally friendly, Fe-catalyzed cross-coupling reactions of easily accessible aryl halides with aromatic Grignard reagents.¹⁰⁷⁾

Therefore we focused on the preparation of Fe(II) NPs via chemical reduction of Fe(III), considering an appropriate reductant. Use of 4-methoxybenzyl alcohol, which is effective in SANi(0) and SARu(0) preparation, failed. However, after screening other reductants under various conditions, we finally identified a suitable preparative method: treatment of sulfur-modified Au with Fe(acac)₃ (35 mg) and ethylene glycol (2.5 mL), as not only the Fe(III) to Fe(II) reductant but also solvent, at 190°C for 12 h (step 2). The third step consists of heating and washing in THF at 75°C for 6 h (Chart 20). After ICP-MS analysis of SAFe(II), we found that the final, fabricated composite mesh has 983 ± 252 µg (mean over three samples) of Fe-based material.

Chart 20. Preparation of SAFe(II)

(Color figure can be accessed in the online version.)

5.1. Ligand-free Kumada Coupling Using SAFe(II)

The above SAFe(II) composite chip was used in the catalytic ligand-free Kumada coupling between 4-iodoanisole (1c) (0.25 mmol) and p-tolylmagnesium bromide (1.3 eq.), and its use was found to give the corresponding biaryl compound 3j in 96% yield, as shown in Table 40, run 1.

Table 40. Reusability of SAFe(II) in the Ligand-Free Kumada Coupling of 1c and p-Tolylmagnesium Bromide; Data Signal the Retention of Conversion Using Recovered SAFe(II) in Cycles 2–5, with ICP-MS Detecting the Entry of Fe into the Reaction Solvent and the Fe Retention by Used SAFe(II)

Run	1St	2nd	3rd	4th	5th	Used SAFe
Yield of 3j (%)a,b)	96	97	95	85	90
Leached iron (µg)a)	121	66	117	47	60	217c)
ppma)	108	59	104	42	53

a) Average of 3 sets of reactions. b) The yield was determined by HPLC yield. c) The amount of Fe immobilized on SAFe(II) after 5 reaction cycles by ICP-MS.

Next, using the same SAFe(II) chip, we repeated the ligand-free Kumada coupling. That is, after the first reaction between 1c and p-tolylmagnesium bromide, the SAFe(II) chip was recovered from the reaction vessel using a pair of tweezers before washing it with THF and using it in the second reaction. By repeating this test-recovery cycle several times (Table 40, runs 2–5), SAFe(II) was established to demonstrate excellent and easy recyclability. The yields of 3j obtained over the first 5 runs in these recycling experiments are shown in Table 40. Thus without systematic or significant loss of catalytic activity, in the presence of SAFe(II), ligand-free Kumada coupling proceeded efficiently to afford 3j in 85–97% yields for runs 1 to 5. We performed ICP-MS analysis on aliquots of reaction mixtures 1–5 to answer the question whether the SAFe(II) chip was acting as a pre-catalyst or as a heterogeneous catalyst under the conditions reported. In each case, Fe was detected in the mother liquor, which suggests that SAFe(II) represents an efficient pre-catalyst of active Fe for the reaction. An additional control experiment was conducted to confirm this. Initially, as indicated in Table 40, run 1, the reaction of 1c and p-tolylmagnesium bromide was conducted. After 1.5 h, the SAFe(II) was extracted by a pair of tweezers and further 1c and p-tolylmagnesium bromide added. After the reaction was continued for a further 1.5 h, a final 3j was isolated in 94% yield, which establish that active catalyst should be present in the mother liquor.

Next, we confirmed the reagent scope of SAFe(II), using several aryl iodides and aryl magnesium bromides (Table 41). The reactions of several substituted iodobenzene derivatives (1), including with electron-withdrawing or electron-releasing groups, with p-tolylmagnesium bromide gave the expected biaryl products 3 in high yields. The reaction of 1-iodonaphthalene or 2-iodopyridine with p-tolylmagnesium bromide gave the corresponding product in 92 or 96% yields, respectively. The reaction of aryl iodides, having heterocyclic motifs like thiophene, with phenylMgBr or 4-fluorophenylMgBr proceeded smoothly to give the corresponding products in 98–83% isolated yields. The reaction of aryl iodides and 4-fluorophenylMgBr also proceeded very well, to give the corresponding products in 95% and 83% yields, respectively.

Table 41. Ligand-Free Kumada Coupling between ArI and ArMgBr Using SAFe(II) as Pre-catalyst

^a) 1 mmol scale raction. (Color figure can be accessed in the online version.)

To explore the scope of the above ligand-free Kumada coupling of aryl halides using SAFe(II) as pre-catalyst, we used aryl bromides and aryl chlorides, which are cheaper and more commercially available than the corresponding iodides. The reactions of several aryl bromides with ArMgBr gave the corresponding products in 94–78% yields (Table 42), although longer reaction times were needed due to the only relatively slow oxidative addition of the bromides. Under microwave irradiation, the reaction of aryl chloride and p-tolylmagnesium bromide proceeded to give the corresponding biaryl compound in 13% yield. Furthermore, we also performed analogous ligand-free Kumada coupling of aryl triflates, which are easily prepared from the corresponding carbonyl compound, to give steroid derivatives.

Table 42. Ligand-Free Kumada Coupling of Aryl Bromide, Triflate, Chloride Reagents with ArMgBr, Using SAFe(II) as a Pre-catalyst

^a) The reaction mixture was performed under continuous irradiation of microwave. ^b) 2.5 eq. of Grignard reagent was used. (Color figure can be accessed in the online version.)

5.2. SAFe(II) Characterization

5.2.1. X-Ray Absorption Near-Edge Structure (XANES) Spectroscopy

SAFe(II) was subjected to spectroscopic and microscopic analysis to reveal its structure and chemical composition. We used Fe K-edge X-ray absorption near-edge structure (XANES) analysis to the SAFe(II) composite chip, before using it in the Kumada coupling of 1c and p-tolylmagnesium bromide to give 3j, and on the same SAFe(II) after using it in five cycles of the reaction. For comparison, Fe foil, FeSO₄, Fe₃O₄, and Fe(acac)₃ were analyzed at the same time as standard samples. The Fe K-edge XANES spectra of SAFe(II) before and SAFe(II) after use were almost identical, which improves the essential stability of SAFe(II). Of the standards, the K-edge XANES spectrum of FeSO₄ was most similar to that of SAFe(II), although the chemical shift was somewhat different. On the other hand, more significantly different chemical shifts were observed on the K-edge XANES spectra of Fe foil, Fe₃O₄, and Fe(acac)₃. Considering these data, we concluded that the Fe component in SAFe(II) is mostly attributable to the presence of Fe(II).

Aiming to know the sulfur component of SAFe(II), we next recorded the S K-edge XANES spectra of SAFe(II) before and after use in 5 cycles of the same Kumada coupling reaction as above. The results were compared with those obtained using the standard samples Na₂SO₄, Na₂S₂O₈, Na₂SO₃, Na₂S₂O₃, PdS, and SAPd(0). The XANES spectra of SAFe(II) before and after reaction displayed a clear resonance at 2481.6 eV. Comparison of these XANES spectra with those of the standards suggests that the sulfur in SAFe(II) is strongly oxidized and present as SO₄²⁻. The chemical shift of the sulfur in SAFe(II) is similar to that previously reported for SAPd(0).

We also performed XANES spectra of the Fe L_2,3-edge in SAFe(II) before and after 5 cycles of Kumada coupling of 1c and p-tolylmagnesium bromide alongside those of relevant standard samples. We measured data by total electron yield (TEY) methods, to obtain information reflecting the chemical state of the samples approximating to their surfaces. The spectra of SAFe(II) before and after use were almost identical, indicating that the chemical state of the Fe did not significantly change during the repeated reaction. We also compared the spectra of SAFe(II) to those of FeCl₂ and FeCl₃, as standard samples of divalent and trivalent Fe, respectively, and found that the spectra of SAFe(II) were clearly most similar to that of FeCl₂. These data suggest that the Fe content at the surface of SAFe(II) is dominated by Fe(II) both before and after Kumada coupling. Furthermore, X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) experiments were conducted to investigate further the nature of pristine SAFe(II). Whilst XRD was inconclusive, XPS supported the view that the surface of a pristine sample of SAFe(II) was dominated by Fe(II) by revealing a binding energy (BE) of 709.2 eV in the Fe 2p3/2 region of the SAFe(II) spectrum. This correlates with BEs of 709.6 and 710.9 eV documented for the Fe2p3/2 regions of FeO and Fe₂O₃ respectively.

Hence we successfully developed a self-assembled Au-supported Fe-based pre-catalyst, i.e., SAFe(II), for recyclable and ligand-free Kumada coupling. SAFe(II) was easily prepared by a three-step procedure involving Fe immobilization, using Fe(acac)₃ as the Fe source, and ethylene glycol as not only solvent but also reductant. SAFe(II) has been used effectively for C–C bond-forming Kumada coupling of not only aryl iodides but also aryl triflate and aryl bromide reagents without the need for ancillary ligands. At the same time, it demonstrated both good-to-near-quantitative conversions and straightforward recyclability. Overall, the results suggest that our in situ PSSO method could represent a versatile method for the immobilization of metal-based NPs.

Acknowledgments

I would like to thank Prof. Satoshi Shuto (Hokkaido University) and Prof. Hiromichi Fujioka (Osaka University) for their continuous advice and support. I also thank all of my coworkers, postdoctoral fellows, graduate students, and undergraduate students, whose names are listed in references. This research was partially supported by a Grant-in-Aid from JSPS KAKENHI for Precisely Designed Catalysts with Customized Scaffolding (Grant No. JP A16H010260 and 18H04260), Molecule Activation Directed toward Straightforward Synthesis (Grant No. JP 23105502 and 25105702), 16689001, 17659001, 21790003, 23246013, A262880510, T15K149760, and T15KT00630, by Platform Project for Supporting Drug Discovery and Life Science Research (Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS)) from AMED under Grant Number JP18am0101084, from JST ACT-C Grant Number JPMJCR12YM, Japan, and from AMED “Dynamic Alliance for Open Innovation Bridging Human, Environment and Materials” from the Ministry of Education, Culture, Sports, Science and Technology of Japan under Grant Number 17am0101084j0001. I am also grateful for financial support received from the Research Foundation for Pharmaceutical Sciences, the Canon foundation, Shorai Foundation for Science and Technology, TEPCO Memorial Foundation, Kobayashi International Scholarship Foundation, the Novartis Foundation, the Iketani Foundation, Yazaki Memorial Foundation for Science and Technology, the Fujisawa Foundation, Mitsubishi Chemical Corporation Fund, a Takeda Chemical Industries Ltd. Award in Synthetic Organic Chemistry, and a Mitsui Chemical Ltd. Award in Synthetic Organic Chemistry. The XAFS measurements were performed at beamlines BL14B2 and BL27SU of SPring-8, with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal Nos. 2011A1835, 2011B1761, 2011B1952, 2012A1621, 2012A1770, 2012B1751, 2013A1792, 2013A1322, 2014A1786, 2014B1247, 2015A1917, 2016A1678, 2016B1745, 2017A1793, 2017B1931, 2017B1732, 2018A1727, 2018A1793, and 2018B1864).

Conflict of Interest

The author declares no conflict of interest.

Note

This review of the author’s work was written by the author upon receiving the 2017 Pharmaceutical Society of Japan Award for Divisional Scientific Promotion.

References and Notes

• 1) Biffis A., Centomo P., Zotto A. D., Zecca M., Chem. Rev., 118, 2249–2295 (2018).
• 2) Astruc D., Lu F., Aranzaes J. R., Angew. Chem. Int. Ed., 44, 7852–7872 (2005).
• 3) Yasukawa T., Miyamura H., Kobayashi S., Chem. Soc. Rev., 43, 1450–1461 (2014).
• 4) Wang D., Astruc D., Chem. Rev., 114, 6949–6985 (2014).
• 5) Molnár Á., Chem. Rev., 111, 2251–2320 (2011).
• 6) Balanta A., Godard C., Claver C., Chem. Soc. Rev., 40, 4973–4985 (2011).
• 7) Deraedt C., Astruc D., Acc. Chem. Res., 47, 494–503 (2014).
• 8) Na Y., Park S., Han S. B., Han H., Ko S., Chang S., J. Am. Chem. Soc., 126, 250–258 (2004).
• 9) Salam N., Kundu S. K., Roy A. S., Mondal P., Ghosh K., Bhaumik A., Islam S., Dalton Trans., 43, 7057–7068 (2014).
• 10) Daneshvar A., Moghadam M., Tangestaninejad S., Mirkhani V., Mohammadpoor-Baltork I., Khalili A., Organometallics, 35, 1747–1755 (2016).
• 11) Hoshiya N., Shimoda M., Yoshikawa H., Yamashita Y., Shuto S., Arisawa M., J. Am. Chem. Soc., 132, 7270–7272 (2010).
• 12) Hoshiya N., Shuto S., Arisawa M., Adv. Synth. Catal., 353, 743–748 (2011).
• 13) Al-Amin M., Honma T., Hoshiya N., Sato T., Shuto S., Arisawa M., Adv. Synth. Catal., 354, 1061–1068 (2012).
• 14) Yokoyama M., Ishii A., Arisawa M., Shuto S., J. Surf. Sci. Nanotech., 10, 630–632 (2012).
• 15) Al-Amin M., Akimoto M., Tameno T., Ohki Y., Takahashi N., Hoshiya N., Shuto S., Arisawa M., Green Chem., 15, 1142–1145 (2013).
• 16) Al-Amin M., Arai S., Hoshiya N., Honma T., Tamenori Y., Sato T., Yokoyama M., Ishii A., Takeuchi M., Maruko T., Shuto S., Arisawa M., J. Org. Chem., 78, 7575–7581 (2013).
• 17) Al-Amin M., Arisawa M., Shuto S., Ano Y., Tobisu M., Chatani N., Adv. Synth. Catal., 356, 1631–1637 (2014).
• 18) Takagi K., Al-Amin M., Hoshiya N., Wouters J., Sugimoto H., Shiro Y., Fukuda H., Shuto S., Arisawa M., J. Org. Chem., 79, 6366–6371 (2014).
• 19) Arisawa M., Sato T., Hoshiya N., Al-Amin M., Kogami Y., Shuto S., ACS Combi. Sci., 16, 215–220 (2014).
• 20) Saito N., Taniguchi T., Hoshiya N., Shuto S., Arisawa M., Sato Y., Green Chem., 17, 2358–2361 (2015).
• 21) Takagi K., Fukuda H., Shuto S., Otaka A., Arisawa M., Adv. Synth. Catal., 357, 2119–2124 (2015).
• 22) Urakawa K., Sumimoto M., Arisawa M., Matsuda M., Ishikawa H., Angew. Chem. Int. Ed., 55, 7432–7436 (2016).
• 23) Arisawa M., Al-Amin M., Honma T., Tamenori Y., Arai S., Hoshiya N., Sato T., Yokoyama M., Ishii A., Takeguchi M., Miyazaki T., Takeuchi M., Maruko T., Shuto S., RSC Adv., 5, 676–683 (2015).
• 24) Hoshiya N., Fujiki K., Taniguchi T., Honma T., Tamenori Y., Xiao M., Saito N., Yokoyama M., Ishii A., Fujioka H., Shuto S., Sato Y., Arisawa M., Adv. Synth. Catal., 358, 2449–2459 (2016).
• 25) Akiyama T., Taniguchi T., Saito N., Doi R., Honma T., Tamenori Y., Ohki Y., Takahashi N., Fujioka H., Sato Y., Arisawa M., Green Chem., 19, 3357–3369 (2017).
• 26) Akiyama T., Wada Y., Jenkinson K., Honma T., Tsuruta K., Tamenori Y., Haneoka H., Takahara T., Suzuki T., Murai K., Fujioka H., Sato Y., Wheatley A. E. H., Arisawa M., ACS Appl. Nano Mater., 1, 6950–6958 (2018).
• 27) Arisawa M., Tsukamoto S., Shimoda M., Pristovsek M., Nishida A., Jpn. J. Appl. Phys., 41 (Part 2, No. 11A), L1197–L1199 (2002).
• 28) Takamiya I., Tsukamoto S., Shimoda M., Miyashita N., Arisawa M., Arakawa Y., Nishida A., Chem. Lett., 33, 1208–1209 (2004).
• 29) Takamiya I., Tsukamoto S., Shimoda M., Arisawa M., Nishida A., Arakawa Y., Jpn. J. Appl. Phys., 45, L475–L477 (2006).
• 30) Arisawa M., Hamada M., Takamiya I., Shimoda M., Tsukamoto S., Arakawa Y., Nishida A., Adv. Synth. Catal., 348, 1063–1070 (2006).
• 31) Gulam R. G., Hamada M., Takamiya I., Shimoda M., Shuto S., Nishida A., Tsukamoto S., Arisawa M., Bull. Chem. Soc. Jpn., 81, 1012–1018 (2008).
• 32) Hoshiya N., Isomura N., Shimoda M., Yoshikawa H., Yamashita Y., Iizuka K., Tsukamoto S., Shuto S., Arisawa M., ChemCatChem., 1, 279–285 (2009).
• 33) Ishii A., Asano H., Yokoyama M., Tsukamoto S., Shuto S., Arisawa M., Phys. Status Solidi., C Curr. Top. Solid State Phys., 7, 359–361 (2010).
• 34) Shimoda M., Konishi T., Tateishi K., Toujyou T., Tsukamoto S., Nishiwaki N., Arisawa M., Hoshiya N., Shuto S., Isomura N., Yokota H., Furukawa Y., Iizuka K., Ogiwara T., Isozaki Y., Yamashita Y., Yoshikawa H., Ueda S., Kobayashi K., J. Appl. Phys., 108, 024309 (2010).
• 35) Flahive E. J., Ewanicki B. L., Sach N. W., O’Neill-Slawecki S. A., Stankovic N. S., Yu S., Guinness S. M., Dunn J., Org. Process Res. Dev., 12, 637–645 (2008).
• 36) Kosugi M., Kameyama M., Migita T., Chem. Lett., 12, 927–928 (1983).
• 37) Guram A. S., Rennels R. A., Buchwald S. L., Angew. Chem. Int. Ed. Engl., 34, 1348–1350 (1995).
• 38) Hartwig J. F., “Organotransition Metal Chemistry,” University Science Books, Sausalito, California, U.S.A., 2010
• 39) This does not mean SAPd(0) cannot be used for the repeated reactions after 10 cycles. Control experiments of 1a and 2a using Au-mesh or Pd-foil gave 3a in 7 or 5%, respectively.
• 40) Pd concentration in hot filtration after 30 minutes, 2 hours and 7 hours were 30 ± 0 ng, 205 ± 148 ng, 2920 ± 1952 ng, respectively (n = 2). There might be the equilibrium between atomic Pd and particle.
• 41) 731 nmol of Pd, 0.229 mol% for 0.32 mmol scale reaction.
• 42) 630 nmol of Pd, 0.197 mol% for 0.32 mmol scale reaction.
• 43) Ppm was calculated based on weight/weight points of Pd in the whole reaction mixture, including substrates, all reagents and solvent, after SAPd(0) was removed.
• 44) Komáromi A., Novák Z., Adv. Synth. Catal., 352, 1523–1532 (2010).
• 45) Monguchi Y., Kitamoto K., Ikawa T., Maegawa T., Sajiki H., Adv. Synth. Catal., 350, 2767–2777 (2008).
• 46) Hirai Y., Uozumi Y., Chem. Commun., 46, 1103–1105 (2010).
• 47) Rouquet G., Chatani N., Angew. Chem. Int. Ed., 52, 11726–11743 (2013).
• 48) Zhao J., Yue D., Campo M. A., Larock R. C., J. Am. Chem. Soc., 129, 5288–5295 (2007).
• 49) Masselot D., Charmant J. P. H., Gallagher T., J. Am. Chem. Soc., 128, 694–695 (2006).
• 50) Ma S., Gu Z., Angew. Chem. Int. Ed., 44, 7512–7517 (2005).
• 51) Zhou J., He J., Wang B., Yang W., Ren H., J. Am. Chem. Soc., 133, 6868–6870 (2011).
• 52) Dolušić E., Larrieu P., Blanc S., Sapunaric F., Pouyez J., Moineaux L., Colette D., Stroobant V., Pilotte L., Colau D., Ferain T., Fraser G., Galleni M., Frère J.-M., Masereel B., Van den Eynde B., Wouters J., Frédérick R., Eur. J. Med. Chem., 46, 3058–3065 (2011).
• 53) Chng L. L., Zhang J., Yang J., Amoura M., Ying J. Y., Adv. Synth. Catal., 353, 2988–2998 (2011).
• 54) Uyttenhove C., Pilotte L., Theate I., Stroobant V., Colau D., Parmentier N., Boon T., Van den Eynde B., J. Nat. Med., 9, 1269–1274 (2003).
• 55) He J., Wasa M., Chan K. S. L., Yu J.-Q., J. Am. Chem. Soc., 135, 3387–3390 (2013).
• 56) Ano Y., Tobisu M., Chatani N., J. Am. Chem. Soc., 133, 12984–12986 (2011).
• 57) Blackburn E. A., Walkinshaw M. D., Curr. Opin. Pharmacol., 11, 365–371 (2011).
• 58) Hughes C. C., Fenical W., J. Am. Chem. Soc., 132, 2528–2529 (2010).
• 59) Zhang C., Zong X., Zhang L., Jiao N., Org. Lett., 14, 3280–3283 (2012).
• 60) Ozawa F., Soyma H., Yamamoto T., Yamamoto A., Tetrahedron Lett., 23, 3383–3386 (1982).
• 61) Kobayashi T., Tanaka M., J. Organomet. Chem., 233, C64–C66 (1982).
• 62) Wu X. F., Neumann H., Beller M., Chem. Rev., 113, 1–35 (2013).
• 63) Satoh T., Kokubo K., Miura M., Nomura M., Organometallics, 13, 4431–4436 (1994).
• 64) Uozumi Y., Arii T., Watanabe T., J. Org. Chem., 66, 5272–5274 (2001).
• 65) Iizuka M., Kondo Y., Chem. Commun., 2006, 1739 (2006).
• 66) Takács E., Varga C., Skoda-Földes R., Kóllar L., Tetrahedron Lett., 48, 2453–2456 (2007).
• 67) de la Fuente V., Godard C., Zangrando E., Claver C., Castillón S., Chem. Commun., 48, 1695–1697 (2012).
• 68) Lin Y.-S., Alper H., Angew. Chem. Int. Ed., 40, 779. 8. (2001).
• 69) Wuts P. G. M., Greene T. W., “Protective Groups in Organic Synthesis (4th ed.),” Wiley, New York, 2007.
• 70) Garrett C., Prasad K., Adv. Synth. Catal., 346, 889–900 (2004).
• 71) Ganguly N. C., Dutta S., Datta M., Tetrahedron Lett., 47, 5807–5810 (2006).
• 72) Mayer F., Chem. Ber., 45, 1105–1113 (1912).
• 73) Dallacker F., Leidig H., Chem. Ber., 112, 2672–2679 (1979).
• 74) Modler-Spreitzer A., Fritsch R., Mannschreck A., Collect. Czech. Chem. Commun., 65, 555–560 (2000).
• 75) Enders D., Niemeier O., Synlett, 12, 2111–2114 (2004).
• 76) Lin C.-H., Yang D.-Y., Org. Lett., 15, 2802–2805 (2013).
• 77) “Microwave Heating as a Tool for Sustainable Chemistry,” ed. by Leadbeater N. E., CRC, Boca Raton, 2011.
• 78) Kappe C. O., Dallinger D., Mol. Divers., 13, 71–193 (2009).
• 79) Moseley J. D., Kappe C. O., Green Chem., 13, 794–806 (2011).
• 80) Loupy A., Petit A., Hamelin J., Texier-Boullet F., Jacquault P., Mathé D., Synthesis, 1998, 1213–1234 (1998).
• 81) Raner K. D., Strauss C. R., Trainor R. W., Thorn J. S., J. Org. Chem., 60, 2456–2460 (1995).
• 82) The standard deviation was calculated from 4 sets of samples.
• 83) Flahive E. J., Ewanicki B. L., Sach N. W., O’Neill-Slawecki S. A., Stankovic N. S., Yu S., Guinness S. M., Dunn J., Org. Process Res. Dev., 12, 637–645 (2008).
• 84) Schulz M. N., Hubbard R. E., Curr. Opin. Pharmacol., 9, 615–621 (2009).
• 85) Leeson P. D., St-Gallay S. A., Nat. Rev. Drug Discov., 10, 749–765 (2011).
• 86) Congreve M., Chessari G., Tisi D., Woodhead A. J., J. Med. Chem., 51, 3661–3680 (2008).
• 87) Evans B. E., Rittle K. E., Bock M. G., DiPardo R. M., Freidinger R. M., Whitter W. L., Lundell G. F., Veber D. F., Anderson P. S., Chang R. S., Lotti V. J., Cerino D. J., Chen T. B., Kling P. J., Kunkel K. A., Springer J. P., Hirshfield J., J. Med. Chem., 31, 2235–2246 (1988).
• 88) Watanabe M., Hirokawa T., Kobayashi T., Yoshida A., Ito Y., Yamada S., Orimoto N., Yamasaki Y., Arisawa M., Shuto S., J. Med. Chem., 53, 3585–3593 (2010).
• 89) Uozumi Y., Nakai Y., Org. Lett., 4, 2997–3000 (2002).
• 90) Kato H., Ishigame T., Oshima N., Hoshiya N., Shimawaki K., Arisawa M., Shuto S., Adv. Synth. Catal., 353, 2676–2680 (2011).
• 91) ABL, ABL tyrosine kinase. CSK; tyrosine–protein kinase CSK. EGFR; the epidermal growth factor receptor (EGFR) subfamily of tyrosine kinases, ephrin type-A receptor 2. EPHB4; ephrin type-B receptor 4. FGFR1; fibroblast growth factor receptor 1. FLT3; Fms-like tyrosine kinase 3. IGF1R; Insulin-like growth factor 1 receptor. ITK; IL2-inducible T-cell kinase. JAK3; Janus kinase 3. KDR; kinase insert domain receptor. LCK; lymphocyte-specific protein tyrosine kinase. MET; Mesenchymal–epithelial transition factor. PDGFR alpha; platelet-derived growth factor receptor A. PYK2; protein tyrosine kinase 2. SRC; Src protein–tyrosine kinase. SYK; Spleen tyrosine kinase. TIE2; endothelium-specific receptor tyrosine kinase. TRKA; neurotrophic tyrosine kinase receptor type 1. TYRO3; tyrosine–protein kinase receptor TYRO3.
• 92) Tückmantel S., Greul J. N., Janning P., Brockmeyer A., Grütter C., Simard J. R., Gutbrod O., Beck M. E., Tietjen K., Rauh D., Schreier P. H., ACS Chem. Biol., 6, 926–933 (2011).
• 93) Xiao M., Hoshiya N., Fujiki K., Honma T., Tamenori Y., Shuto S., Fujioka H., Arisawa M., Chem. Pharm. Bull., 64, 1154–1160 (2016).
• 94) Rosen B. M., Quasdorf K. W., Wilson D. A., Zhang N., Resmerita A.-M., Garg N. K., Percec V., Chem. Rev., 111, 1346–1416 (2011).
• 95) Tasker S. Z., Standley E. A., Jamison T. F., Nature (London), 509, 299–309 (2014).
• 96) Wu L., Zhang X., Tao Z., Catal. Sci. Technol., 2, 707–710 (2012).
• 97) Soulé J.-F., Miyamura H., Kobayashi S., J. Am. Chem. Soc., 135, 10602–10605 (2013).
• 98) We have not succeeded in finding what the mass of support for 509 µg of Ni is based on because the amount is too small to calculate.
• 99) Tamura M., Kochi J. K., J. Am. Chem. Soc., 93, 1487–1489 (1971).
• 100) Fürstner A., Leitner A., Angew. Chem. Int. Ed., 41, 609–612 (2002).
• 101) Nagano T., Hayashi T., Org. Lett., 6, 1297–1299 (2004).
• 102) Fürstner A., Leitner A., Méndez M., Krause H., J. Am. Chem. Soc., 124, 13856–13863 (2002).
• 103) Bedford R. B., Bruce D. W., Frost R. M., Goodby J. W., Hird M., Chem. Commun., 2004, 2822–2823 (2004).
• 104) Ilies L., Okabe J., Yoshikai N., Nakamura E., Org. Lett., 12, 2838–2840 (2010).
• 105) Buono F. G., Zhang Y., Tan Z., Brusoe A., Yang B.-S., Lorenz J. C., Giovannini R., Song J. J., Yee N. K., Senanayake C. H., Eur. J. Org. Chem., 2016, 2599–2602 (2016).
• 106) Xing T., Zhang Z., Da Y.-X., Quan Z.-J., Wang X.-C., Asian. J. Org. Chem., 4, 538–544 (2015).
• 107) Hatakeyama T., Nakamura M., J. Am. Chem. Soc., 129, 9844–9845 (2007).