Tris(pentafluorophenyl)borane‐Catalyzed Carbenium Ion Generation and Autocatalytic Pyrazole Synthesis—A Computational and Experimental Study

Abstract In recent years, metal‐free organic synthesis using triarylboranes as catalysts has become a prevalent research area. Herein we report a comprehensive computational and experimental study for the highly selective synthesis of N‐substituted pyrazoles through the generation of carbenium species from the reaction between aryl esters and vinyl diazoacetates in the presence of catalytic tris(pentafluorophenyl)borane [B(C6F5)3]. DFT studies were undertaken to illuminate the reaction mechanism revealing that the in situ generation of a carbenium species acts as an autocatalyst to prompt the regiospecific formation of N‐substituted pyrazoles in good to excellent yields (up to 81 %).

Abstract: In recent years, metal-free organic synthesis using triarylboranes as catalysts has become a prevalent research area. Herein we report a comprehensive computational and experimental study for the highly selective synthesis of Nsubstituted pyrazoles through the generation of carbenium species from the reaction between aryl esters and vinyl diazoacetates in the presence of catalytic tris(pentafluorophenyl)borane [B(C 6 F 5 ) 3 ]. DFT studies were undertaken to illuminate the reaction mechanism revealing that the in situ generation of a carbenium species acts as an autocatalyst to prompt the regiospecific formation of N-substituted pyrazoles in good to excellent yields (up to 81 %).
An autocatalytic process is one wherein the products of a reaction act to catalyze the chemical reaction in which they themselves are formed. [1] The fascinating nature of autocatalytic processes has resulted in many investigations to establish their mode and mechanism of action. Indeed, one area which has been thoroughly investigated for autocatalysis is organocatalysis, [2] where such autocatalytic process can often amplify the selectivity of reactions. [3] Uses of boranes in an autocatalytic system remain underexplored. The usage of Lewis acidic boranes, such as tris(pentafluorophenyl)borane, B-(C 6 F 5 ) 3 , as catalysts in organic synthesis has gained unprecedent attention as an alternative to many common transition metals. [4] Recently ourselves, [5] and others [6a] have explored the catalytic activity of B(C 6 F 5 ) 3 in the activation of diazo-esters leading to the formation of carbene intermediates through the elimination of N 2 . The carbene intermediate can subsequently be used as building block for the synthesis of novel organic molecules through a range of different reactions such as O À H/N À H/C À H insertion, [5a, 6b,d,g,h] azide/carbonate transfer, [6e,f] cyclopropanation/cyclopropenation, [5a, 6c] and the ring-opening of heterocyclic compounds. [5a] In this study, we were interested in the B(C 6 F 5 ) 3 -catalyzed synthesis of pyrazoles from vinyl diazoacetates in which the N 2 functionality of the diazo starting material is not released. The metal-free synthesis of nitrogen-containing heterocycles is an important area of research as most of these heterocycles are of biological importance. [7] Pyrazoles are an important class of nitrogen-based heterocycles that are omnipresent in natural products and therefore have a broad impact in medicinal chemistry. [8] Thus, a metal-free synthesis of functionalized pyrazole compounds is desirable as synthetic routes need to avoid trace impurities of toxic metals in the final compounds. Herein we report the reactions between aryl esters and vinyl diazoacetates in the presence of catalytic B(C 6 F 5 ) 3 to afford N-alkylated pyrazoles in a selective manner (Scheme 1 C). The only report of a similar reaction is with gold-based catalysts where selective decomposition of a diazoester in the presence of a second diazoester generates a pyrazole product (Scheme 1 B). [9] In our previous work, we demonstrated that, in the presence of catalytic amounts of B(C 6 F 5 ) 3 , a-aryl a-diazo- esters readily react with aryl esters to afford C=C crosscoupled products with the elimination of N 2 through an alkenylation reaction (Scheme 1 A). [10] Our initial interest was to develop this previous work further to make 1,3-diene ester compounds [11] using vinyl diazoacetates rather than a-aryl adiazoesters. To this end we synthesized the vinyl diazoacetate compound (1 a, Scheme 2) and treated it with an aryl ester (2 a), with catalytic amounts of B(C 6 F 5 ) 3 (10 mol %). Following the reaction at 50 8C in 1,2-C 2 H 4 Cl 2 for 20 h, the resulting product was purified via preparative thin layer chromatography to yield a white solid. The solid however could not be attributed to the product from the alkenylation reaction when observing the multinuclear NMR spectra and high-resolution mass spectrometric data. Slow evaporation of the solid from CH 2 Cl 2 gave a crop of colorless crystals of the product. Singlecrystal X-ray diffraction analysis unequivocally confirmed the product to be a substituted pyrazole ring compound 3 a that was formed in 76 % yield ( Figure 1).
In an attempt to explain the product formation, we hypothesized that the B(C 6 F 5 ) 3 catalyst could either (i) activate the diazoester compound through the ester (1·B(C 6 F 5 ) 3 ) or diazo (1'·B(C 6 F 5 ) 3 ) functionalities, or (ii) activate the aryl ester (2·B(C 6 F 5 ) 3 ) in the first step of the reaction (Scheme 3 A). In the scenario that B(C 6 F 5 ) 3 activates diazo compound 1, one plausible pathway for the reaction could be initial heterocycle 4 formation (formed from the intramolecular attack of the nitrogen atom onto the alkene) followed by B(C 6 F 5 ) 3 catalyzed N-alkylation. To investigate the mechanism for the reaction, we undertook DFT calculations at the SMD/M06-2X-D3/def2-TZVP//SMD/M06-2X/6-31G(d) level of theory. Using the vinyl diazoacetate compound 1 b as an example, the activation barrier to afford the pyrazole product   allylic substituted product. [12] Alternatively, activation of the terminal nitrogen atom of the diazo functionality lowered the activation barrier slightly to 29.4 kcal mol À1 to yield the N! B(C 6 F 5 ) 3 adduct 4'·B(C 6 F 5 ) 3 (Scheme 3 D). Given the high activation barrier for this reaction in all scenarios, we next investigated B(C 6 F 5 ) 3 activation of the ester 2 as the initial step of the reaction (Scheme 4 and Figure 2) as observed in our previous work. [13] Initial coordination of the borane to the ester 2 b was found to be favorable by 5.6 kcal mol À1 yielding 2 b·B(C 6 F 5 ) 3 ( Figure 2). Indeed, the 11 B NMR spectrum of the 1:1 stoichiometric reaction of 2 b and B(C 6 F 5 ) 3 showed initial adduct formation indicated by a broad peak at 4.87 ppm. This promotes the formation of carbenium ion 7 through TS 1 of 15.3 kcal mol À1 . Reaction of the carbenium ion with the vinyl diazoacetate 1 b through the terminal nitrogen atom generates the high energy intermediate 8 through TS 2 (13.9 kcal mol À1 ), which in turn rapidly cyclizes through a low energy transition state (TS 3 = 4.7 kcal mol À1 ) generating the heterocycle 9 in which the pyrazole core has been formed. Alkylated vinyl diazoacetate 8 is much more reactive than vinyl diazoacetate 1 b towards cyclization, supported by a small energy difference between TS 3 and 8 (4.7 kcal mol À1 ). Indeed, the C arylester À N diazoester bond strength in compound 9 is calculated to be 26.7 kcal mol À1 whereas the same bond in 8 is much weaker with DH rxn = 1.3 kcal mol À1 (Figure 2, insert). As such, the relatively low activation barrier to the cyclization via transformation 8!9 can be explained by strengthening of the C arylester À N diazoester bond upon moving from 8 to 9 acting as a driving force to facilitate the process. Heterocycle 9 then undergoes a deprotonation step aided by borate anion 6 to yield the pyrazole compound 10. Notably, pyrazole 10 is the isomer of the obtained product in which the bis(aryl)methyl group is bonded to the nitrogen atom adjacent to the phenyl functionality. The overall energy barrier to generate the pyrazole isomer was calculated to be 21.3 kcal mol À1 .
In cycle 2, the carbenium ion 7 acts as an autocatalyst to convert the kinetic pyrazole product 10 into the thermodynamic pyrazole isomer 3 with a barrier of 20.0 kcal mol À1 .
Finally, we investigated the scope of the reaction to examine if this process was general to other aryl esters and
Highly electron-deficient ester (2 c) was unsuccessful in the reaction due to the reduced basicity of the ester and high activation barrier for the carbenium ion formation. [13] Likewise, the non-aromatic ester 2-isopropyl-5-methylcyclohexyl 4-fluoro benzoate (2 g) failed to react with vinyl diazoesters due to the instability of the carbenium ion formed following B(C 6 F 5 ) 3 activation. Strongly electron-donating groups in both the vinyl diazoacetate (1 c) and diaryl ester (2 d) afforded complicated reaction mixtures and attempts to isolate any pure compound failed.
Given the low activation barrier (20.0 kcal mol À1 ) calculated for the conversion of pyrazole isomer 10 into 3, we wondered if we could observe this isomerization experimentally. To examine this, we took product 3 e and subjected it to the carbocation species 7 (Ar = C 6 H 5 ) which was generated in situ from compound 2 a (1 equiv) using 10 mol % B(C 6 F 5 ) 3 (Scheme 6). The reaction was carried out in 1,2-C 2 H 4 Cl 2 at 50 8C. By 1 H NMR spectroscopy it was found that the starting material 3 e and product 3 d were formed in an approximate 1:1 ratio with a small amount of the minor isomer 10 d being observed. After 22 h 3 e was isolated in 17 % yield, and 3 d was isolated as the major product in 49 % yield. The latter species was formed from the exchange between the two diarylmethylene groups through cycle 2 (Scheme 4, Figure 2). Interestingly, we also isolated the less thermodynamically stable isomer 10 d in 16 % yield. We propose that compound 3 e first reacts with in situ generated carbocation 7 (Ar = C 6 H 5 ) to generate cationic intermediate 11 (Scheme 4xschr4). Loss of carbenium 7 (Ar = p-FC 6 H 4 ) leads to the kinetic isomer 10 d. Reaction of 10 d with carbenium ion 7 (Ar = C 6 H 5 ) generates the thermodynamic isomer 3 d as the major product. This observation supports our DFT-based mechanism to account for the formation of pyrazole compounds.
In conclusion, we have demonstrated a new metal-free synthetic approach for the preparation of regioselective Nalkylated pyrazoles. A new reactivity pattern has been observed whereby catalytic B(C 6 F 5 ) 3 does not decompose the diazo compound allowing the N 2 functionality to be exploited in the generation of N-heterocycles. Detailed mechanistic studies were carried out to explain the mechanism for the reaction. DFT calculations revealed that catalytic amounts of B(C 6 F 5 ) 3 are required to activate the aryl ester generating a carbenium ion. Interestingly, this in situ generated carbenium species subsequently acts as an autocatalyst to promote the regioselective formation of N-alkylated pyrazoles in good to excellent yields. This new reactivity pattern and metal-free synthetic approach for the preparation of novel pyrazoles will have a broad impact in future applications towards the synthesis of biologically important molecules.