Zirconium and hafnium catalyzed C–C single bond hydroboration

Selective cleavage and subsequent functionalization of C−C single bonds present a fundamental challenge in synthetic organic chemistry. Traditionally, the activation of C−C single bonds has been achieved using stoichiometric transition-metal complexes. Recently, examples of catalytic processes were developed in which use is made of precious metals. However, the use of inexpensive and Earth-abundant group IV metals for catalytic C−C single-bond cleavage is largely underdeveloped. Herein, the zirconium-catalyzed C−C single-bond cleavage and subsequent hydroboration reactions is realized using Cp2ZrCl2 as a catalytic system. A series of structures of various γ-boronated amines are readily obtained, which are otherwise difficult to obtain. Mechanistic studies disclose the formation of a N–ZrIV species, and then a β-carbon elimination route is responsible for C–C single bond activation. Besides zirconium, hafnium exhibits a similar performance for this transformation.

our knowledge (The use of heterogeneous zirconium catalysis for C−C bond cleavage was reported by Basset [82][83][84] ).Consequently, activating C−C single bonds with zirconium catalysis for chemical transformation remains a significant challenge.It is of considerable scientific and practical interest to synthetic organic chemistry to address this.Herein, we report the development of an unprecedented catalytic system that resulted in the realization of the zirconium-and hafniumcatalyzed C−C single-bond hydroboration (Fig. 1d).Mechanistic studies support the formation of N−Zr IV species and then a β-carbon elimination route for C−C single bond activation.Our work provides an alternative catalytic method for C−C single bonds hydroboration, and establishes the bond activation models and catalytic application of group IV transition metals.

Catalytic reaction investigations
As a synthetically significant transformation in organic chemistry, hydroboration of C=C bonds is well-studied.However, catalytic hydroboration of C−C single bonds remains underdeveloped.Only two systems using Ir and Rh are known for the hydroboration of cyclopropanes, as developed by Yamaguchi 85 and Shi 86,87 .Besides making use of precious metals and Nor P-ligands, it is noticed that for the Ir system, the careful choice of a chiral tBu-Quinox ligand is crucial for the C−C bond hydroboration over the C−H boration [88][89][90] .For the Rh system, the PPh 3 ligand is essential in inhibiting side reactions such as the formation of alkenes.Furthermore, in the former case, cleavage of C β −C β' bond is observed, while the latter cleavages C α −C β bond.Thus, it is still highly desirable to develop a facile and inexpensive catalytic system for the hydroboration of C−C single bonds.
We commenced our investigation using 0.2 mmol of N-Piv-cyclopropylamines (1a) with 1.5 equiv.pinacolborane (HBpin) in 1 mL toluene at 120 °C as the model reaction, using 5 mol% Cp 2 ZrCl 2 as catalyst (Table 1).Our preliminary investigations unveiled that the addition of 1 equiv. of base is the key for the zirconium-catalyzed C−C bond hydroboration (Supplementary Table 1); K 2 CO 3 was the optimal choice (Table 1, entry 1, 81% yield of 2a).Various other zirconium complexes were then tested but were unsuccessful.There was no reaction with the sterically bulkier Cp* 2 ZrCl 2 (  ).An attempt was then made to reduce the amount of K 2 CO 3 .Pleasingly, even with 0.1 equiv.K 2 CO 3 already had a 68% yield of 2a (Table 1, entry 8).Finally, a much higher yield of 2a (91%) was obtained with just 0.3 equiv.K 2 CO 3 (Table 1, entry 9).In addition, 95% yield of 2a was obtained using 5 mol% Cp 2 ZrH 2 as catalyst without K 2 CO 3 (Table 1, entry 10) .However, considering the simplicity of using readily available and inexpensive Cp 2 ZrCl 2 as the catalyst, we conducted the following studies using Cp 2 ZrCl 2 /K 2 CO 3 system (Table 1, entry 9).The results with Cp 2 ZrH 2 gave us some clues for the subsequent mechanism studies (vide infra).

Substrates scope studies
Having the reaction conditions for the Zr-catalyzed hydroboration of cyclopropylamines in hand (Table 1, entry 9), various cyclopropane rings were investigated to establish the generality of our methodology (Fig. 2).
First, the tolerance of substituents on the para-position of the phenyl ring was studied.We found that electron-neutral, electrondonating, and electron-withdrawing groups are tolerated; moderate to good yields were obtained (2a-2k, 40-82% yields).In general, electrondonating groups (−Me, −OMe, − t Bu, −SMe, 2c−2e, 2k) gave better results than electron-withdrawing groups such as −CF 3 (2f).Halide substituents −F, −Cl, −Br, which likely undergo competitive hydrodehalogenation or boration reactions, are untouched in our system (2g−2i).We found that steric effects have some influence on the results, as changing the substituents from the para-position to the metaand ortho-position, led to slightly decreased yields or the need for higher reaction temperatures (2l−2o).Significantly, naphthyl, benzodioxole, alkyne, and heteroaromatic rings such as furyl and thiophene substituents are all compatible in our system, with yields in the range of 71 − 78% (2p−2t).Similarly, good results were obtained when the phenyl group is at the β' position (2u).Changing R 1 from an aryl to an alkyl group was also successful, with both acyclic and cyclic alkyl substituents (up to 77% yield, 2v−2ac).Pleasingly, products 2q, 2ab and 2ac were obtained in 69%, 77%, and 67% yields, respectively, with no double or triple bond interference.Moreover, cyclopropane rings with two substituents on the R 1 and R 2 positions were also suitable (2ad−2ae).
The effect of the substituent on the amide groups was then studied.When R is 1-methylcyclohexyl, the hydroboration product 2af was obtained in 76% yield.Changing R to a sterically bulkier adamantly group resulted in a slightly lower yield (2ag, 56%).Substrates with 2,2dimethylbutyl and 1-methylcyclopropyl, and substituents containing chloride are all converted to their corresponding hydroboration products 2ah−2aj in yields of up to 64%.Sulfonamide is also tolerated in our system, which get the hydroboration product 2ak in 66% yield.The reaction also proceeded with thioamide (2al).Finally, we found that cyclopropylamines derived from Oleanolic Acid and Gemfibrozil also reacted well in our system; the corresponding products, 2am and 2an, were obtained in yields of 49% and 59%, respectively.

Hafnium-catalyzed reaction
Compared with zirconium, hafnium has received less attention as a homogeneous catalyst in organic reactions.To our knowledge, reactivity toward C−C single bonds activation is also unknown.After successfully establishing zirconium-catalyzed C−C single bond activation of cyclopropylamines and their subsequent hydroboration, we further explored the reactivity of a hafnium complex towards C−C single bonds.We established that the base plays an essential role in tuning the reactivity.Eventually, Cs 2 CO 3 was found to be the optimal base (Supplementary Table 5).Then, under the optimal reaction conditions, we conducted substrate scope generality studies (Fig. 3).We found that the hafnium system is not only suitable for substrates that work in the zirconium system but also for substrates that do not work there.For example, substrates with a −CN group do not react with the zirconium catalyst, but a 50% yield of product 2ao was obtained with hafnium.Product 2ap, with two fluorides on the phenyl ring, was also obtained in 51% yield.Additionally, cyclohexyl-(2aq) and phenylsubstituted substrates (2ar) were also applicable in the hafnium system.

Synthetic derivation
The practical utilization of our system was then demonstrated on a gram scale (Fig. 4).When we subjected 10 mmol of 1a to our standard reaction conditions, we obtained 2a in 73% yield (1.97 g).Furthermore, the synthetic derivatization of 2a was demonstrated.Using the aminoazanium of DABCO as an amination reagent 91 , and then protecting the amine with TFAA, the corresponding TFA-amide 3a was obtained in 66% yield.γ-Boronated amine 3b, which is otherwise difficult to obtain 92 , was obtained in 72% yield by reducing the amide functional group to an amine and then protecting it with TsCl.Product 2a can be transformed into potassium trifluoroborate salt 3c using KHF 2 (82% yield).Treating 2a with furan-2-yllithium followed by NBS afforded the arylated product 3d in 74% yield.Finally, Pdcatalyzed Suzuki−Miyaura coupling of 2a with Estrone-derived triflate gave 3e in 45% yield.

Mechanistic studies
To shed light on the reaction mechanism, several control experiments were performed (Fig. 5).The possible formation of an alkene intermediate via ring-opening of cyclopropanes followed by hydroboration was studied.However, no alkenes were detected after 3 or 12 h under standard reaction conditions with or without HBpin (Fig. 5Aa, Supplementary Fig. 1).Utilization of alkenes 1a' and 1a" afforded less than 6% 2a (Fig. 5Ab).When enantioenriched substrate (1S, 2R)-1as was applied, the desired product (R)-2as was obtained without erosion of the enantioselectivities (Fig. 5Ac, Supplementary Figs. 2, 3).Those results excluded a consecutive cyclopropane ring openinghydroboration process.Then, the possibility of a reaction pathway that involved a radical species was investigated.TEMPO (2,2,6,6-tetramethylpiperidinyloxyl) (1−2 equiv) had almost no effect on the results.However, upon increasing the amount thereof (4 equiv.) the yields of 2a decreased to 44% (Fig. 5B).At this point, it should be borne in mind that TEMPO inhibition experiments can sometimes provide ambiguous results 93 .Thus, additional experiments with the addition of 9,10-dihydroanthracene (DHA) were conducted; no effect on the yield of 2a was detected (Fig. 5B).The results with TEMPO and DHA excluded a radical mechanism.Then, the active zirconium catalytic species was studied.Upon the combination of Cp 2 ZrCl 2 and K 2 CO 3 in d 8 -Tol heated at 120 °C for 12 h, a new species appeared around 6.0 ppm in the 1 H NMR spectrum (Fig. 5Ca).With 2 equiv.K 2 CO 3 and heating for a longer reaction time, the Cp 2 ZrCl 2 was fully converted to this new species (Fig. 5Cb).Then, the isolated new species was characterized by IR spectroscopy and was currently assigned to Cp 2 ZrCO 3 by comparation with literature data (Supplementary Fig. 4) 94,95 .Nevertheless, upon further adding HBpin to the above solution, we could detect the formation of Zr-H species in the 1 H NMR spectrum by trapping with acetone (Fig. 5Cc, Supplementary Fig. 5).This finding, together with the fact that Cp 2 ZrHCl or Cp 2 ZrH 2 can Fig. 2 | Zirconium-catalyzed hydroboration of cyclopropylamines-generality studies.a,b a Reaction conditions: 0.2 mmol 1, HBpin (1.5 equiv.),K 2 CO 3 (30 mol%), Cp 2 ZrCl 2 (5 mol%) and 1 mL toluene in a 15 mL pressure tube at 120 °C for 24 h; catalyze the C−C bond hydroboration process without K 2 CO 3 (65% and 95% yields, Supplementary Table 2), we concluded that Zr−H species are essentially the active catalysts via the consecutive reactions of Cp 2 ZrCl 2 , K 2 CO 3 , and HBpin (Fig. 5Cd).According to the work from Ganem 96 , Rosenthal 97 , and Cantat 98 , the active Zr−H species can interact with 1a to form N−Zr species via metathesis with N−H bond.This is further proved in our case that Cp 2 ZrHCl reacts with the N−H group of 1a with the release of H 2 or HD when 1a-D was used (Fig. 5D, Supplementary Fig. 6).To add further proof of the importance of the N−H, N-methylated analog substrate 1a-Me, and replace the N−H with CH 2 or O substrates 1a-C, 1a-O were subjected to our reaction conditions.As expected, no corresponding C−C bond hydroboration product were observed (Fig. 5E).
Keep in mind that β-carbon elimination is one of the main pathway for C−C bond cleavage.It is natural to think that after the formation of N−Zr species, a β-carbon elimination may proceed to cleavage the C−C bond to produce an imino propyl zirconium species.This is consistent with the fact that we can observe the presence of the putative imine intermediate both on GC/MS and 13 C NMR when substrate 1 u was used (Fig. 5F, Supplementary Figs. 7, 8).
Based on the above mechanistic study and our DFT calculation results (Supplementary Fig. 9), we conclude the following general reaction pathway for our Cp 2 ZrCl 2 /K 2 CO 3 system (Fig. 6).First, the insitu formed Zr−H species reacts with N−H bonds of the substrates to form N−Zr IV species A via H 2 release (i).Next, the C−C single bond is   (v).The last step is supported by the experiment that when we introduced 1 atm of deuterium gas into the standard reaction, 29% deuterium labeling at the α-carbon adjacent to N−H of 2a could be obtained (Fig. 5G, Supplementary Fig. 10), suggesting that hydrogen metathesis occurred [99][100][101] .In summary, an unprecedented zirconium-and hafnium-catalyzed C−C single bond activation and subsequent hydroboration is realized using a catalytic system based on Cp 2 ZrCl 2 and Cp 2 HfCl 2 .Our catalytic approach applies to various cyclopropylamines.Selective cleavage of the proximal C α −C β single bond was achieved, with the tolerance of multiple functional groups as well as bio-and medicine-derived substrates.Mechanistic studies disclose that the in-situ generated Zr−H  species and the free N−H group of the substrates play key roles in this transformation via Zr−H and N−H metathesis to form N−Zr IV species, and the subsequent C−C single bond activation is realized via a βcarbon elimination route.Our work presents an unprecedented group IV metal-catalyzed C−C single bond activation and hydroboration reaction.The C−C single bond activation model that was well studied for late-transition metals, were also elaborated to be applicable to the group IV metals.

General procedure for the Zr-catalyzed hydroboration of cyclopropylamines
In a nitrogen-filled glovebox, to a 15 mL pressure tube with a magnetic stirrer was added catalytic amount of Cp 2 ZrCl 2 (0.01 mmol, 2.9 mg), K 2 CO 3 (0.06 mmol, 8.3 mg), corresponding cyclopropylamine substrates (0.2 mmol), HBpin (0.3 mmol, 43.5 μL), and toluene (1 mL) in a sequence manner.Then, the pressure tube was taken out of the glove box and allowed to stir at 120 °C for 24 h.Upon completion, all the solvent was evaporated, and the crude product was isolated on silica gel using flash chromatography with dichloromethane/ethyl acetate as the eluent to give the corresponding products.

Fig. 1 |
Fig. 1 | Strategies for C−C single bond activation.a Transition metal mediated C−C single bond activation; b, c stoichiometric amount of zirconium-mediated C−C single bond activation; d zirconium and hafnium catalyzed C−C single bond activation.

Fig. 5 |
Fig. 5 | Experimental mechanism studies.A possible alkene intermediate formation; B possible radical pathway; C 1 H NMR spectra show the formation of Zr−H species; D 1 H NMR spectra show the release of H 2 or HD by reacting Cp 2 ZrHCl with 1a; E Control experiments to show the importance of N−H bond; F Detection of imide intermediate 2u' by crude 13 C NMR spectroscopy; G Deuterium labeling experiment by introducing D 2 in the standard reaction.
b Yields were determined by GC using dodecane as an internal standard.Article https://doi.org/10.1038/s41467-024-45697-yNature Communications | (2024) 15:1846 Reaction in gram scale and further derivatization of 2a.Isolated yields are given; for detailed reaction conditions, please refer to the supplementary information.
cleaved via β-carbon elimination of intermediate A to form the imino propyl zirconium species B (ii).Subsequently, intermediate B reacts with HBpin via C−Zr and H−B bond metathesis to give C and regenerate Zr−H species (iii).In the second catalytic cycle, Zr−H hydride transfer to intermediate C gives intermediate D (iv), which is further reduced by the previously released H 2 to 2a with hydrogenolysis or H 2 metathesis