A Dendralenic C–H Acid

Published as part of the 30 Years SYNLETT – Pearl Anniversary Issue

Abstract

The design and synthesis of a strong, dendralenic C–H acid is described. Crystal structure analyses confirm the proposed structure. Despite the moderate stability of our motif, an application to Brønsted acid catalysis has been explored.

Key words

dendralenic C–H acids - cross-conjugated acids - triflyl groups - strong acids - non-coordinating anions - Brønsted acids

In contrast to N–H- and O–H-based Brønsted acids, C–H acids enable the incorporation of a greater number of electron-withdrawing groups (EWGs) by virtue of carbon’s higher valency. Experimental and estimated pK _a values of the simple trifluoromethanesulfonyl (triflyl, Tf) containing O–H, N–H, and C–H acids suggest that their acidity directly correlates with the number of electron-withdrawing groups (Scheme [1]). Accordingly, tris(triflyl)methane (1) should be the strongest acid in the series, and indeed it shows a high reactivity in Brønsted and Lewis acid catalysis.[1] Still more electron-withdrawing groups can be introduced by choosing allylic C–H acid frameworks. This notion led to the design of 1,1,3,3-tetratriflylpropene (TTP), which showed a remarkable acidity and catalytic activity.[1d] In the search for still stronger acids, we sought to further increase the number of EWGs, which led to our interest in triene-­derived C–H acids.[2] Purely hydrocarbon-based C–H acid scaffolds on the basis of fluorene and dibenzofluorene have already been realized by Kuhn and the latter showed a ­remarkable pK _a value of 5.9 (in water).[3] Depending on the location of the acidic proton, either trivinylmethane or ­dendralene-derived C–H acids are possible.

Scheme 1 Lead structures for the design of trivinylmethane-derived and highly conjugated dendralene-derived C–H acids

These considerations led to the design of tris(bis(triflyl)vinyl)methane (HTBT).[4] Irrespective of the location of the acidic proton on HTBT, only one anion should be obtained (TBT, after deprotonation) with a highly delocalized negative charge and a possible C ₃-symmetry (Scheme [2]). Furthermore, the peripheral location of the triflyl groups may enable a planar structure of the anion. As a result of this enhanced planarization and the greater number of electron-withdrawing groups, the acidity of HTBT was ­expected to be significantly higher in comparison to the related allylic C–H acid TTP. Synthetic access to HTBT was ­envisaged from triformylmethane and bis(triflyl)methane, as ­Yanai and coworkers[5] have already demonstrated that bis(triflyl)methane reacts with a variety of aldehydes in a self-promoted Knoevenagel-type condensation reaction.

Scheme 2 Design and retrosynthesis of the tris(bistriflylvinyl)methane (HTBT) and its anion TBT. The synthesis of triformylmethane has been previously described[6]

While our first attempts at the synthesis of HTBT led to the formation of a purely organic tricarbanion salt,[2] we found that by condensing triformylmethane[7] with bis(triflyl)methane followed by treatment with 2,2,6,6-tetra­methylpiperidine (TMP), the desired HTMP salt of TBT (­HTMP·TBT) was obtained in poor yield (Scheme [3]).[8] Interestingly, crystal structure analysis of this ion pair revealed that the HTMP cation formed a slightly shorter N–H^…O ­hydrogen bond to solvent water (N^…O, 2.780(4) Å), which was introduced during the crystallization, than to the negatively charged TBT anion (N^…O, 2.971(3) Å). Despite the increased distance between the triflyl groups, the TBT anion adopts a slightly non-planar chiral conformation. We ­assume that this may be due to the short contacts between the vinylic hydrogen atoms and the sulfonyl oxygen atoms. While we observe a local C ₃-symmetry around the central carbon atom with similar bond lengths and torsion angles (see the Supporting Information), no global C ₃-symmetry was observed in the TBT anion.

Scheme 3 Synthesis of HTMP·TBT. HTMP = 2,2,6,6-tetramethylpiperidinium

A work-up with concentrated H₂SO₄ finally delivered HTBT as the free acid (Scheme [4]).[9] NMR spectroscopic investigations and single-crystal structure analysis of HTBT confirmed the location of the acidic proton not on the central carbon atom, as in the crystal of bullvalene, but between two triflyl groups. As a result, HTBT can be considered a cross-conjugated, dendralenic C–H acid. Due to the low stability of HTBT at room temperature and at –25 °C no satisfactory yield could be determined. We would expect the stability of such acids to be increased in a non-coordinating and non-polar solvent, as a degradation pathway via a nucleophilic attack can be prevented. However, we are yet to identify such a solvent system that is also capable of solubilizing HTBT.

Scheme 4 Synthesis of HTBT. Due to the low stability of HTBT no yields could be determined

Despite the inherent low stability of HTBT, we attempted to directly employ freshly prepared HTBT for a benchmark Brønsted acid catalyzed Friedel–Crafts acylation reaction of weakly reactive chlorobenzene with p-fluorobenzoyl chloride (Scheme [5]).[1d] [10] While TTP provided higher yields, HTBT was also able to catalyze this transformation.

Scheme 5 Application of HTBT to the Friedel–Crafts acylation reaction of chlorobenzene with p-fluorobenzoyl chloride and a comparison with TTP

The acidity of HTBT is sufficient to protonate ethers thus allowing its transformation into an etherate salt when an excess of Et₂O was added (Scheme [6]).[11] Single-crystal structure analysis revealed that the TBT anion neither adopts an idealized C ₃-symmetry nor a planar conformation, which is in accordance with our previous findings. ­Interestingly, the oxonium proton prefers to coordinate to the oxygen atom of a second ether molecule rather than to one of the negatively charged triflyl oxygen atoms on the TBT carbanion. We were intrigued to find that the distances between the oxygen atoms of both Et₂O molecules are ­almost identical to those found in BArF etherates with the molecular formula [B(C₆F₅)₄]⁻[H(OEt₂)₂]⁺.[12] Consequently, a similar anion coordination can be assumed, thus classifying the TBT anion as a weakly coordinating anion.

Scheme 6 Conversion of HTBT into the diethyl ether–diethyloxonium salt TBT·H(OEt₂)₂ and structural comparison with related BArF etherates[12]

In summary, we have designed and developed a synthesis of the cross-conjugated dendralenic C–H acid HTBT. Several crystal structures confirmed our design and revealed that the TBT anion adopts a non-planar and chiral conformation. Despite its low stability, HTBT was found to catalyze a Friedel–Crafts acylation reaction of chlorobenzene. A structural comparison with related BArF etherates indicates that the TBT anion may be classified as a C–H-acid-based weakly coordinating anion.

Acknowledgment

Generous support by the Max-Planck-Society, the Deutsche ­Forschungs­gemeinschaft (Leibniz Award to B.L. and Cluster of Excellence RESOLV, EXC 1069), and the European Research Council (Advanced Grant ‘C–H Acids for Organic Synthesis, CHAOS’) is gratefully acknowledged. We also thank Jennifer Arlt for her technical support and the members of our analytical departments for their excellent service. We acknowledge DESY (Hamburg, Germany), a member of the Helmholtz Association HGF, for the provision of experimental facilities. Parts of this research were carried out at PETRA III and we would like to thank Anja Burkhardt, Eva Crosas, and Sebastian Günther for assistance in using the P11 beamline.

• References and Notes

• 1a Kütt A, Rodima T, Saame J, Raamat E, Mäemets V, Kaljurand I, Koppel IA, Garlyauskayte RYu, Yagupolskii YL, Yagupolskii LM, Bernhardt E, Willner H, Leito I. J. Org. Chem. 2011; 76: 391
• 1b Koppel IA, Taft RW, Anvia F, Zhu S.-Z, Hu L.-Q, Sung K.-S, DesMarteau DD, Yagupolskii LM, Yagupolskii YL. J. Am. Chem. Soc. 1994; 116: 3047
• 1c Turowsky L, Seppelt K. Inorg. Chem. 1988; 27: 2135
• 1d Höfler D, van Gemmeren M, Wedemann P, Kaupmees K, Leito I, Leutzsch M, Lingnau JB, List B. Angew. Chem. Int. Ed. 2017; 56: 1411
• 1e Ishihara K, Hiraiwa Y, Yamamoto H. Synlett 2000; 80
• 1f For a description of pentacarboxycyclopentadiene-derived acids, see: Gheewala CD, Radtke MA, Hui J, Hon AB, Lambert TH. Org. Lett. 2017; 19: 4227 ; and references cited therein
• 2 Höfler D, Goddard R, Lingnau JB, Nöthling N, List B. Angew. Chem. Int. Ed. 2018; 57: 8326
• 3 Kuhn R, Rewicki D. Angew. Chem. 1967; 79: 648
• 4a Hansch C, Leo A, Taft RW. Chem. Rev. 1991; 91: 165
• 4b Hendrickson JB, Giga A, Wareing J. J. Am. Chem. Soc. 1974; 96: 2275
• 5a Yanai H, Egawa S, Taguchi T. Tetrahedron Lett. 2013; 54: 2160
• 5b Yanai H, Egawa S, Yamada K, Ono J, Aoki M, Matsumoto T, Taguchi T. Asian J. Org. Chem. 2014; 3: 556
• 6a Arnold Z, Budesinsky M. J. Org. Chem. 1988; 53: 5352
• 6b Krasnaya ZA, Smirnova YV, Krystal GV, Bogdanov VS. Mendeleev Commun. 1996; 6: 17
• 6c Buděšínský M, Fiedler P, Arnold Z. Synthesis 1989; 858
• 7 Triformylmethane was prepared in two steps from commercially available bromoacetic acid following our recently reported procedure, see Ref. 2.
• 8 HTMP·TBTA Schlenk flask was charged with bis(triflyl)methane (1.7 g, 6.0 mmol, 6.0 equiv) and dry CH₂Cl₂ (1 mL) was added under argon. The colorless, clear solution so obtained was cooled to –78 °C in an acetone/dry ice bath. Triformylmethane (0.10 g, 1.0 mmol) was added and a slurry was obtained. Trimethyl orthoacetate (0.47 g, 3.9 mmol, 0.50 mL, 3.9 equiv) and acetic anhydride (1.6 g, 16 mmol, 1.5 mL, 16 equiv) were added and the reaction mixture was allowed to reach RT overnight. A dark red solution was obtained. All volatiles were removed under reduced pressure and the solid mixture so obtained was dissolved in CH₂Cl₂ (5 mL). 2,2,6,6-Tetramethylpiperidine (0.85 g, 6.0 mmol, 1.0 mL, 6.0 equiv) was added and the reaction mixture was subsequently concentrated under reduced pressure. Almost complete removal of all volatiles was achieved by dissolving in CHCl₃ and evaporation to dryness. The solid mixture was transferred to a separation funnel with CHCl₃ (20 mL) and washed with sat. aq NaHCO₃ (20 mL). Both phases were separated and the aq phase was washed with CHCl₃ (20 mL), acidified with aq HCl (conc.) to a pH of 1, and extracted with CH₂Cl₂ (20 mL). The pooled CH₂Cl₂ phases were concentrated under reduced pressure to give the TMP salt of the title compound as an orange solid (0.35 g). As this compound still contained significant amounts of bis(triflyl)methane, the solid mixture was dissolved in CHCl₃, washed with aq HCl (conc.), dried over MgSO₄, filtered, and concentrated under reduced pressure. The red solid so obtained was dissolved in CHCl₃ and all volatiles were removed under reduced pressure. This procedure was repeated once more with CHCl₃ and then with 1,2-dichloroethane. A red solid was obtained, which was triturated with CHCl₃ to give the TMP salt of the title compound as a yellow solid (60 mg, 0.058 mmol, 5.8% yield). ¹H NMR (500 MHz, CDCl₃): δ = 8.31 (s, 3 H), 5.61 (s, 2 H), 1.86–1.81 (m, 2 H), 1.75–1.72 (m, 4 H), 1.49 (s, 12 H). (¹⁴N couplings can be observed in the ¹H NMR spectrum. However, the signals of NH₂ group was not sharp enough for an accurate determination of ¹ J _1H14N.) ¹³C NMR (126 MHz, CDCl₃): δ = 165.51, 119.82 (q, ¹ J _CF = 328 Hz), 114.79, 108.53, 59.94, 35.23, 27.87, 15.95; ¹⁹F NMR (471 MHz, CDCl₃): δ = –73.96; HRMS (ESIneg): m/z [M – H⁺] calcd for C₁₃H₃O₁₂F₁₈S₆ ^–: 884.7667; found: 884.7667. Single crystals suitable for structural analysis were obtained after dissolving the initially obtained orange solid [containing bis(triflyl)methane impurities] in CHCl₃ or 1,2-dichloroethane and slowly evaporating the solvent.
• 9 Synthesis of HTBT as the Free Acid HTMP·TBT (17 mg, 0.017 mmol) was dissolved in CH₂Cl₂ (10 mL) and conc. H₂SO₄ (10 mL) was added. The mixture was stirred at RT for 30 min and the sulfuric acid phase was removed using a Pasteur pipet. BaCl₂ (dry) was added and after stirring for 30 min, the solution was filtered and all volatiles were removed under reduced pressure. A yellowish solid was obtained. NMR spectra (¹H and ¹⁹F) in CDCl₃ were acquired. After 3 d, ¹H and ¹⁹F NMR spectra were acquired once again, but after this time period, all product signals had vanished due to deprotonation and decomposition. ¹H NMR (501 MHz, CDCl₃): δ = 8.80 (t, J = 1.7 Hz, 1 H), 8.59 (t, J = 1.7 Hz, 1 H), 6.52 (d, J = 10.9 Hz, 1 H), 5.36 (d, J = 10.9 Hz, 1 H); ¹³C NMR (126 MHz, CD₂Cl₂): δ = 162.66, 158.32, 136.73, 123.98, 119.52 (q, J = 331 Hz), 76.06. (Due to the fast decomposition of the desired product and the observed ¹³C to ¹⁹F coupling, not all signals in the ¹³C NMR spectrum were obtained.) ¹⁹F NMR (471 MHz, CDCl₃): δ = –71.48, –72.25, –72.52, –72.87, –73.00. The HTMP·TBT salt (0.017 g, 0.017 mmol) was dissolved in CH₂Cl₂ (10 mL) and conc. H₂SO₄ (10 mL) was added. The mixture was stirred at RT for 30 min and the sulfuric acid phase was removed. Treatment with BaCl₂ was omitted. After three months, single crystals of the title compound were obtained which were suitable for structure analysis.
• 10 Höfler D. Ph.D. thesis. Universität zu Köln: Germany 2018
• 11 Conversion of HTBT into the Etherate SaltHTMP·TBT (43 mg, 0.042 mmol) was dissolved in CH₂Cl₂ (10 mL) and conc. H₂SO₄ (10 mL) was added. The mixture was stirred at RT for 30 min and the sulfuric acid phase was removed. All volatiles were removed under reduced pressure and a colorless solid was obtained. Et₂O was added, which afforded a clear, yellow solution. All volatiles were removed under reduced pressure and a yellow solid was obtained. CH₂Cl₂ (10 mL) was added and the formation of a biphasic mixture was noticed. Slow evaporation over 14 d led to the formation of single crystals of the etherate salt, which were suitable for structure analysis.
• 12 Jutzi P, Müller C, Stammler A, Stammler H.-G. Organometallics 2000; 19: 1442