CF3‐Containing para‐Quinone Methides for Organic Synthesis

A new family of CF3‐containing para‐quinone methides (CF3‐QMs) was systematically investigated for its suitability in organic synthesis. Addition of different nucleophiles gives access to target molecules with a benzylic CF3‐containing stereogenic center straightforwardly. The electrophilicity parameter E of the prototypical CF3‐QM 2,6‐di‐tert‐butyl‐4‐(2,2,2‐trifluoroethylidene)cyclohexa‐2,5‐dien‐1‐one was determined to be –11.68 according to the Mayr scale, making it one of the most reactive quinone methides known so far.


Introduction
para-Quinone methides (p-QMs) have emerged as versatile reagents for a variety of different (asymmetric) transformations over the last years. [1] Most commonly, arylidene-based p-QMs (Ar-QMs) have been utilized as highly electrophilic acceptor molecules, [2] which easily undergo 1,6-addition reactions with a multitude of C-or hetero-atom nucleophiles under a variety of (catalytic) conditions (Scheme 1A). [1,[3][4][5][6] Surprisingly, however, alternative ylidene groups have been much less explored so far. [1] Given the broad interest in CF 3 -containing (chiral) organic molecules, [7,8] we reasoned that the introduction and systematic investigation of CF 3 -based para-quinone methides (CF 3 -QMs) may be a worthwhile task to establish a new platform of versatile prochiral starting materials (Scheme 1B). [8] While racemic trifluoromethylation reactions of arylidene-based Ar-QMs have been reported before, [9] reactions of the CF 3 -containing acceptor compounds CF 3 -QMs with different nucleophiles would result in an unprecedented and complementary synthesis strategy to access a broad variety of (chiral) target molecules straightforwardly. Furthermore, to get a more comprehensive understanding about the reactivity of such novel electrophiles and to predict new reactions thereof as well as to obtain a direct comparison with the well-established Ar-QMs, [2] it would be beneficial to determine the electrophilicity parameter E (according to Mayr's free energy relationship [10] ) of at least one new CF 3 -QM derivative. Scheme 1. (A) General reactivity mode of p-QMs. (B) The herein targeted investigations concerning novel CF 3 -QMs. (C) Our previous report describing the synthesis of CF 3 -QM 1a and its use in a single asymmetric transformation. [13] Interestingly, in 1985 already Umemoto and co-workers described the formation of a CF 3 -containing p-QM [11] and the intermediate formation of CF 3 -containing p-QMs was also observed in reactions of trifluoroethanol-containing phenols some years ago, [12] but apart from these two reports, such potentially useful building blocks have not been reported and utilized for further transformations until very recently, when we described the first synthesis of the CF 3 -QM 1a. [13,14] Starting from the bulk chemical 2, p-QM 1a could be obtained in two steps via a benzylic trifluoromethylation first, [15] followed by oxidation to the QM (Scheme 1C). [13,16] This compound was then successfully used in the asymmetric phase-transfer catalyst (PTC) controlled reaction with pronucleophile 3 to access the masked 2,2 -amino acid derivative 4 with very high levels of stereoselectivity. [13] Considering this very promising initial application of the novel building block 1a, we now became interested in investigating this novel quinone methide-platform more systematically (Scheme 1B). Besides utilizing 1a for a variety of different transformations, we also wanted to install alternative groups R at the phenol part, and we became interested in determining the electrophilicity parameter E for at least one (stable) derivative.

E-Parameter
Arylidene-based p-quinone methides Ar-QMs have been studied by Mayr's group several years ago. They analyzed the second-order rate constants of the reactions of Ar-QMs with several nucleophiles [2] according to the following linear free energy relationship [Equation (1)]. [10] lg k(20°C) = s N (N + E) (1) The availability of electrophilicity parameters E for certain acceptor molecules, such as Ar-QMs or CF 3 -QMs, not only allows for a reactivity comparison between different electrophilic compounds, but also provides a tool for the reliable prediction of novel reactions upon considering the E-parameter of the electrophile and the N-parameter of the nucleophile. [10] In general, if (E + N) > -5 is fulfilled, a certain electrophile/nucleophilecombination can be expected to react at room temperature.
Thus, we started our investigations by determining the Eparameter of the stable CF 3 -QM 1a by studying the kinetics of its reactions with the carbanions 5a-d [2a,17] as reference nucleophiles in DMSO at 20°C (Scheme 2). Scheme 2. Kinetic measurements of the reactions of 1a with the reference nucleophiles 5a (Solvent: DMSO; counteranion for 5a-d; K + ).
The CF 3 -QM 1a has a λ max of 292 nm (in DMSO) and a molar absorption coefficient of 2.5 × 10 4 M -1 cm -1 , which enabled us to follow its reactions with the colorless nucleophiles 5a-d to the hypsochromically shifted phenol derivatives 6a-d photometrically (5a-d were generated by deprotonation of the pronucleophiles 5-H with 0.5 equiv. of KOtBu). [16] As well established for other nucleophile/electrophile combinations before, [2,10,17] the reaction kinetics were determined by employing stopped-flow UV/Vis photometry to follow the fading of the colored 1a upon reaction with a large excess of the carbanions 5 (resulting in absorbance decays that follow first-order kinetics). Based on these measurements, it was then possible to calculate the corresponding experimental second-order rate constants for these four reactions. [2,10,16] By using these rate constants together with the known parameters (N, s N ) of our reference nucleophiles, [2,17] it was finally possible to determine the electrophilicity parameter E for CF 3 -QM 1a being -11.68. [16] By comparing this value with E-parameters for well-known tertbutyl-substituted Ar-QMs [2] (Figure 1), it becomes obvious that the presence of the CF 3 -group significantly boosts the electrophilicity of 1a, which makes it a very promising building block for further transformations: The CF 3 -QM 1a should be capable of reacting with various types of nucleophiles provided that Figure 1. Electrophilicity E of 1a in comparison with established arylidenebased p-QMs. [2] their nucleophilic reactivity parameter N exceeds the value of +7.

Syntheses of CF 3 -QMs
For the synthesis of the tBu-containing 1a our recently developed strategy starting from phenol 2 turned out to be highly reproducible and was easily carried out on several gram scale (Scheme 3A). [13] Hereby, the CF 3 -group was first installed by means of a benzylic trifluoromethylation of 2 using Togni's second generation reagent 7 [18] in analogy to a recently published procedure. [16] Conversion of 8 to the quinone methide 1a was then achieved by employing 2,3-dichloro-5,6-dicyano-benzoquinone (DDQ) as an oxidant. Scheme 3. Synthesis strategies to access CF 3 -QMs 1a-e.
Unfortunately, this route was found to be not suited for other phenol derivatives (e.g. when using mesitylene derivatives to access QM 1b, the radical trifluoromethylation was not site-selective and the subsequent purification and oxidation turned out to be rather problematic). Therefore, we opted for a different strategy to access the alternatively substituted QMs 1b-e next. It was shown before, [12,19] that phenol 9b can be converted into the CF 3 -containing benzylic chloride 12b in a twostep procedure by first carrying out an S E Ar-reaction with trifluoroacetaldehyde hemiacetal 10 (giving alcohol 11b), followed by chlorination using SOCl 2 (Scheme 3B). Interestingly, compound 12b already contains traces of the corresponding QM 1b and we found that treatment with Et 3 N allows for the direct formation of 1b with a reasonable isolated yield of 54 %. It should however be noted that the dimethyl-QM 1b was found to be significantly less stable than the di-tert-butyl-QM 1a, making its purification and isolation a difficult task. The benzylic chlorides 12c-e could be accessed analogously from the corresponding phenols 9c-e, but in neither of these cases was it possible to isolate the final QMs 1c-e upon treatment with base. Interestingly however, as compounds 12c-e degraded quickly under basic conditions, and given the fact that we were able to detect traces of QMs 1 in crude samples of precursors 12 already, it seems very likely that the less-stable QMs 1c-e are actually formed from 12c-e under basic conditions and that it may be possible to generate and utilize them for further transformations in situ (vide infra).
We also tested if the S E Ar strategy outlined in Scheme 3B may be applicable to the corresponding di-tBu-phenol 9a, but surprisingly in this case the yields were rather low and unpractical (compared to the route outlined in Scheme 3A).

Application Scope
As the tBu-QM 1a was found to be the most stable of the so far accessed CF 3 -QM derivatives 1, we started our investigations of the application scope by reacting 1a with a variety of C-and heteroatom-nucleophiles (Scheme 4). Scheme 4. Racemic addition of different nucleophiles to QM 1a. [16] All the racemic products shown in Scheme 4 were obtained straightforwardly under operationally simple conditions, [16] giving access to a variety of trifluoromethylated functionalized phenol derivatives. The moderate yields obtained for products 6 can to some extent be rationalized by a certain sensitivity of these products under basic reaction conditions in the presence of an excess of nucleophile. The synthesis of indole derivative 13a performed well under established Lewis acid (BF 3 ·OEt 2 )-mediated conditions. [20] In sharp contrast, when reacting 1a with indole in the absence of any Lewis acid, no reaction occurs. This result is in perfect accordance with the hypothesis that a given electrophile/nucleophile-combination can be expected to react at room temperature if (E + N) > -5, [10] as the experimentally determined nucleophilicity of indole (N < 5.6) [21] is too low to fulfil this requirement in combination with 1a (E = -11.68) in the absence of any Lewis acid. It should be noted that we also tested the addition of NaBH 4 and NaBD 4 to QM 1a, which resulted in quantitative formation of precursor 8 and the monodeuterated 8-D respectively. [16] Having established that the most stable QM 1a reacts well with a variety of different nucleophiles, we next focused on the use of QM 1b (preformed and/or in situ formed from 12b) and on QM-precursors 12c-e (Scheme 5). [16] First experiments carried out with 1b allowed for the straightforward synthesis of the glycine Schiff base-containing 17b and the indole-based 13b. The later reaction again required the use of BF 3 ·OEt 2 as a Lewis acid, indicating that the dimethyl QM 1b is not significantly more electrophilic than the di-tert-butyl QM 1a (vide supra). We next tested if it may be possible to use QM precursors like 12d directly for the reaction with indole. When 12d was treated with indole and BF 3 ·OEt 2 in the absence of any base, no product 13d was formed. However, when 12d was first stirred with one equivalent of Et 3 N, before adding indole and BF 3 ·OEt 2 , the target 13d was obtained straightforwardly. This strongly supports our hypothesis that the in situ formation and utilization of QMs 1 from precursors 12 is indeed possible. Thus, we next reacted 12b-e with diethyl malonate and two different amines in the presence of Cs 2 CO 3 , which allowed for the direct formation of the products 6e-h and 14d-k in reasonable yields as well.
Scheme 5. Reactions of QM 1b and QM-precursors 12b-e. [16] Finally, as QM 1a could be successfully employed for the highly asymmetric phase transfer-catalyzed synthesis of the masked 2,2 -amino acid derivative 4 (Scheme 6A), [13] we also became interested in testing the two other asymmetric transformations depicted in Scheme 6. First, 1a was treated with the glycine Schiff base 18 [22] under phase-transfer conditions to access the protected α-amino acid derivative 17a. The asymmetric addition of nucleophiles 18 to Ar-QMs was recently reported by the groups of Fan and Deng, [23] and in analogy to Fan's work, the use of Cinchona alkaloid-based chiral PTCs, i.e. ammonium salt B, allowed for reasonable selectivities for the addition of 18 to 1a (Scheme 6B). Unfortunately, it was not possible to assign the relative and absolute configuration of enantioenriched 17a as we were not able to obtain single crystals of sufficient quality for a detailed X-ray analysis. However, it was easily possible to fully deprotect and debutylate 17a under acidic conditions, which gave a direct entry to -CF 3 -αtyrosine 19 (Scheme 6B). Scheme 6. Asymmetric transformations of 1a and debutylation/deprotection strategy for the synthesis of the free amino acid 19. [16] We have recently shown that Ar-QMs can also undergo highly enantioselective spirocyclopropanation reactions with chiral ammonium ylides (formed in situ upon deprotonation of ammonium salts 20). [6e] While the addition of achiral ammonium salts 20 (with R 3 N = Me 3 N) to CF 3 -QM 1a yielded the racemic trans-spirocyclopropanes 21a and 21b straightforwardly (Scheme 6C), the asymmetric synthesis turned out to be more difficult (i.e. compared to the highly asymmetric spirocyclopropanation of Ar-QMs [6e] ). When testing different chiral ammonium ylide-precursors 20, we found that Cinchona alkaloids al-low for some moderate enantioselectivity in the synthesis of 21a only, but unfortunately the synthesis of 21b turned out to be rather sluggish and unselective. While this example demonstrates the sometimes significant reactivity difference of achiral and chiral ammonium ylides in asymmetric cyclization reactions, [24,25] it still serves as another illustrative example about the general use of CF 3 -QMs 1 for asymmetric transformations.

CF 2 H-Quinone Methide
Difluoromethyl (CF 2 H)-containing (chiral) molecules have attracted more and more attention recently, [26] especially because CF 2 H-containing compounds show unique properties like e.g. H-bonding capability, in comparison to analogous CF 3 -derivatives. [27] This results in a powerful tool to alter the properties of a given structural motif by subtle structural changes only, and we were therefore interested to obtain a first proof-of-concept if our herein developed strategy to access and utilize the CF 3 -QMs 1b-e in situ, starting from the benzylic chlorides 12b-e, may also be useful to access analogous and so far unprecedented CF 2 H-QMs.
The difluoromethyl-based hemiacetal 22 is commercially available and we were glad to see that the synthesis of the QM-precursor 12b-CF 2 H could be carried in the same manner (Scheme 7A) as for the CF 3 -analogs (compare with Scheme 3B). NMR of purified 12b-CF 2 H showed around 10-15 % of the target quinone methide 1b-CF 2 H, but unfortunately we were not able to isolate this QM after treatment with base (decomposition). Nevertheless, it was again possible to use the precursor 12b-CF 2 H directly for further manipulations, as demonstrated for the synthesis of the two CF 2 H-phenol derivatives 6e-CF 2 H and 14db-CF 2 H (Scheme 7B). Scheme 7. (A) Synthesis route to access the QM-precursor 12b-CF 2 H and (B) applications thereof. [16] Conclusions A new family of either preformed, or in situ generated, CF 3containing para-quinone methides was developed and systematically investigated for their use as starting materials to access phenol derivatives with a CF 3 -substituted benzylic stereogenic center. In addition, we determined the electrophilicity parame-ter E (according to Mayr's linear free energy relationship) for one derivative being -11.68. This is one of the highest E parameters for a quinone methide so far, making these quinone methides very useful and highly reactive building blocks that undergo 1,6-addition reactions with a variety of C-and heteroatom nucleophiles straightforwardly. In addition, it was also shown that an analogous novel CF 2 H-QM precursor can be accessed in a similar manner and utilized directly for further manipulations.

Experimental Section
General details can be found in the online supporting information. This document also contains detailed synthesis procedures and analytical data of all novel compounds and reaction products, details about the kinetic investigations to determine the E parameter, as well as copies of NMR spectra and HPLC traces.

Synthesis of QM 1a:
Step 1: CuI (190 mg; 1 mmol) and Togni's reagent 7 (4.74 g; 15 mmol) were dissolved in 50 mL of DMF. Then the phenol 2 (2.2 g; 10 mmol) was added and the mixture was stirred at 40°C. After 1 h the reaction mixture was diluted with EtOAc and washed with NaH-CO 3 . The organic layer was dried with Na 2 SO 4 , filtered, and concentrated in vacuo. The crude product was purified by column chromatography on silica gel (Heptanes/EtOAc = 20:1) to yield the product 8 in 91 % (2.62 g; 9.1 mmol).
Step 2: The trifluoroethylated phenol 8 (2.88 g; 10 mmol) was dissolved in 200 mL of MeOH and DDQ (5.67 g; 25 mmol) was added. The reaction mixture was stirred at room temperature for 1 h. After completion of the reaction, the solvent was evaporated, and the crude reaction mixture was purified by column chromatography (heptanes/CH 2 Cl 2 = 10:1) to afford product 1a in 84 % yield (2. The corresponding phenol 9 (30 mmol) and trifluoroacetaldehyde hemiacetal 10 (30 mmol) were mixed and K 2 CO 3 (1.5 mmol) was added. The mixture was heated to 60°C and stirred for 16 h. After cooling to room temperature, the reaction mixture was dissolved with ethyl acetate and washed with ammonium chloride, H 2 O and brine. The organic layer was dried with Na 2 SO 4 and the solvents evaporated to dryness. The crude products 11 were directly used for the next step.
Step 2: A mixture of compound 11 (10 mmol) and SOCl 2 (14 mmol) in 15 mL dry toluene was cooled to 0-5°C and pyridine (10 mmol) was added. After 1 h the mixture was heated to 70°C and stirred for another 2 h. After cooling to room temperature, the mixture was poured on 20 g ice and stirred for another 30 min. Then the organic layer was separated, and the aqueous layer was extracted with ethyl acetate twice. The combined organic layers were dried with Na 2 SO 4 and the solvents evaporated to dryness. The crude reaction mixture was purified by column chromatography (CH 2 Cl 2 / heptanes = 2:1) giving the products 12 in the yields reported in Scheme 3.
Step 3 (Synthesis of 1b): Precursor 12b (1.19 g; 5 mmol) was dissolved in 20 mL CH 2 Cl 2 and triethylamine (0.76 mL; 5.5 mmol) was added. The reaction mixture was stirred at room temperature overnight (completion shown by TLC) and H 2 O was added. The organic layer was washed with 1 N HCl followed by NaHCO 3 (sat.) and brine. After drying the organic layer over Na 2 SO 4 it was evaporated to dryness and the quinone methide 1b was isolated by column chromatography (heptanes/EtOAc = 20:1-1:1) in 54 % yield (0.55 g; 2.7 mmol).