Borosilicate Activation of (Difluoroiodo)toluene in the gem-Difluorination of Phenyldiazoacetate Derivatives

Article abstract of DOI:10.1002/ejoc.201600773

A combined experimental and computational investigation was conducted to identify a mild and effective Lewis-acidic activator for TolIF₂in the gem-difluorination of diazo compounds. Computationally, borosilicate, a common constituent of laborat

Full text of DOI:10.1002/ejoc.201600773

DOI: 10.1002/ejoc.201600773
Communication
Fluorination
Borosilicate Activation of (Difluoroiodo)toluene in the gem-
Difluorination of Phenyldiazoacetate Derivatives
Geoffrey S. Sinclair,[a] Richard Tran,[a] Jason Tao,[a] W. Scott Hopkins*[a] and
Graham K. Murphy*[a]
Abstract: A combined experimental and computational inves- vate TolIF2, and an extensive experimental survey confirmed
tigation was conducted to identify a mild and effective Lewis- borosilicate as the most effective activator to date. The key to
acidic activator for TolIF2 in the gem-difluorination of diazo realizing this borosilicate-activated reaction was the use of high
compounds. Computationally, borosilicate, a common constitu- purity TolIF2, which is prepared by a reproducible, multigram-
ent of laboratory glassware, was found to spontaneously acti- scale synthesis.
Introduction
The benefits of incorporating fluorine into pharmaceutical and
agrochemical agents are widely recognized. Fluorine is routinely
employed as a biostere for hydrogen, which alters the lipophil-
icity, metabolic stability and chemical reactivity of biologically
active compounds.[1] Fluorination of organic molecules is a sig-
nificant obstacle that has challenged researchers for decades
and, as evidenced by the relative scarcity of fluorine-containing
natural products, even Nature only has a minimally evolved bio-
chemistry of fluorine.[2] Elemental fluorine is the simplest elec-
trophilic fluorinating agent, though it reacts explosively with
organic compounds. Hypervalent polyfluorides of other
Figure 1. Structure and reactivity profiles of hypervalent sulfur- and iodine-
based fluorination reagents.
strongly electronegative elements, such as XeF2,[3] ClF3,[4]
[6]
BrF3,[5] IF5 and SF4,[7] are also extremely oxidizing and vio-
[8]
lently reactive with water, such that only XeF2 is regularly
TolIF2].[16] Contrary to PhICl2 (7) which has been widely used as
used in synthesis.[9,10]
chlorine-transfer agent for over a century,[17] (difluoroiodo)tolu-
The 1950's discovery of SF4 as a selective deoxygenative
ene has only recently become popular as a fluorination rea-
fluorination reagent[11] was timely due to emerging interest in
gent.[18] While polyfluorides 1–4 react to generate sulfonyl
(and the increasing significance of) novel fluorinating agents.
fluoride waste (Figure 1),[12,19] iodanes 6 and 7 generate iodo-
Less volatile, more controllable derivatives of SF4, [e.g. DAST[12]
benzene and halide ion byproducts. These halide nucleophiles
can be trapped when iodanes 6 or 7 react with ambiphilic di-
(1) and Deoxo-Fluor[13] (2), Figure 1], were significant, but as
fuming liquids they remain challenging to handle owing to
their sensitivity to water and their tendency decompose on
azoesters, giving gem-dihalide products [Equations (1) and
(2)].[20,21] This chemistry parallels the reactivity between diazo
standing. Recently XtalFluor-E (3) and XtalFluor-M (4) were re-
compounds and elemental halogens, and because TolIF2 reacts
ported to be highly effective deoxofluorination salts synthe-
as a surrogate for elemental fluorine, it provides a facile and
economical means to synthesize biologically-relevant difluor-
ides.
In analogy to the activation of sulfur polyfluorides with
BF3·OEt2, Lewis acid activation of TolIF2 was required to bolster
sized by treating DAST with BF3·OEt2.[14]
Iodine (poly)fluorides of increased stability and practicality
over IF5 have also been developed, such as the hypervalent
iodoarene-derived fluorides 5[15] and 6 [(difluoroiodo)toluene,
its reactivity with phenyl diazoacetates (Equation 2). Unfortu-
[a] Department of Chemistry, University of Waterloo,
nately, inconsistent yields and poor functional group compati-
bilities were found for the fluorination reactions employing the
previously reported activator BF3·OEt2.[22] Gouverneur and co-
workers reported similar outcomes in their gem-difluorination
reactions of 11 (Equation 3).[23] The highly variable yields in
these two studies appeared to derive from the poor compatibil-
200 University Ave W. Waterloo, ON N2L3G1, Canada
E-mail: graham.murphy@uwaterloo.ca
shopkins@uwaterloo.ca
https://uwaterloo.ca/murphy-lab/
https://uwaterloo.ca/hopkins-lab/
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/ejoc.201600773.
Eur. J. Org. Chem. 0000, 0–0
1
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
reaction was also observed without BF3·OEt2,[21] we also investi-
gated a borosilicate-like cluster (Figure 2, inset) which predicted
spontaneous fluoride abstraction (ΔG°act = –8.2 kcal mol–1, en-
try 6). This computational result showed that complex forma-
tion derives from shared electron density between the Lewis-
basic F 2p orbital and the Lewis-acidic B 2p orbital (Figure 2),
resulting in an increase in the I–F bond length from 2.04 Å to
2.27 Å.[25,26] This indicates that borosilicate is significantly more
[27]
Lewis acidic towards TolIF2 than BF3·OEt2
Calculations show
that this stems from the local geometry of the boron centre in
borosilicate, where the non-optimal dihedral angles between
donor (O 2p) and acceptor (B 2p) orbitals enhance boron's Le-
wis acidity.[28,29]
ity between BF3·OEt2 and electron-rich functional groups. To
overcome this obstacle we conducted a combined experimen-
tal and computational exploration for milder Lewis-acidic acti-
vators of TolIF2. Here we report the results of the investigation,
and show that borosilicate glassware is a highly effective initia-
tor for the gem-difluorination reaction.
Results and Discussion
The computational studies were based on a two-step activation
of TolIF2: complex formation between 6 and a Lewis acid (LA),
followed by heterolytic fluoride abstraction to generate the io-
donium ion pair (Table 1). The Gibbs' energies for TolIF2 activa-
tion (ΔG°act) were calculated as the sum of the energies of com-
plex formation (ΔG°comp) and fluoride abstraction (ΔG°abs). The
energy for the BF3-mediated iodane activation was calculated
as a reference to facilitate identification of viable Lewis acid
candidates (Table 1, entry 1).[24] Though most gave significantly
higher values for ΔG°act than did BF3 (entries 2–4), TiF4 was
promising (ΔG°act = 24.7 kcal mol–1, entry 5). Because a slow
Figure 2. Highest occupied molecular orbital (HOMO) for the borosilicate·To-
lIF2 complex. Magnification shows the boron atom with substantial orbital
overlap with the Lewis-basic fluorine. Inset: borosilicate (bsg) cluster used in
molecular modelling.
A borosilicate flask-activated gem-difluorination reaction
would be a major improvement over the use of BF3·OEt2 as an
activator. A previous attempt of this reaction only proceeded in
40 % yield in 2 hours (Table 2, entry 1), but the experiment was
Table 1. Gibbs' energies of complex formation (ΔG°comp), fluoride abstraction
(ΔG°abs), and activation (ΔG°act) between TolIF2 and Lewis acidic additives.[a]
Table 2. Control experiments for the gem-difluorination of phenyldiazoace-
tates using various silicates as activating agents for TolIF2.
Entry Substrate
Vessel
Additive (5 wt.-%) Time
Yield[a]
1
2
3
4
5
6
7
8
9
8a R = H
8a R = H
bsg[b]
bsg
–
–
–
–
2 h
10a 40 %
10 min 10a 81 %[c]
8b R = OTs bsg[d]
8b R = OTs PFA
8b R = OTs Nalgene
8b R = OTs PFA
8b R = OTs PFA
8b R = OTs PFA
8b R = OTs PFA
8b R = OTs silanized
8 min
1 h
1 h
10b 81 %[c]
–
–
Entry
Lewis acid[b]
ΔG°comp
ΔG°abs
ΔG°act
[kcal mol–1
]
[kcal mol–1
]
[kcal mol–1
]
–
1
2
3
4
5
6
BF3·OEt2
AlF3
GaF3
FeF3
TiF4
14.9
128.4
64.5
73.4
–21.9
–57.7
44.5
44.8
44.8
63.9
46.6
49.6
59.4
borosilicate
SiO2 gel
Pasteur pipette
quartz
10 min 10b 76 %[c]
173.2
104.9
137.3
24.7
1 h
1 h
1 h
–
–
–
10
–
10 min 10b 67 %[c]
borosilicate
–8.2
borosilicate
[a] Calculations were conducted at the B3LYP/6-311+G(d,p) level of theory
using a 6-311G* basis set for iodine.
[a] Isolated yield. [b] Borosilicate flask. [c] Glyoxylate isolated in ca. 15 % yield.
[d] New glassware used. PFA: 4 mL of PFA (TeflonTM) vial.
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
2
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
repeated and gave 10b in 81 % yield over 10 min (entry 2). transfer reaction (Figure 3). Activation of the iodane gives an
This improvement is a consequence of using pure TolIF2 in the ion pair between the iodonium ion and the borosilicate–fluor-
reaction,[26] indicating the importance of reagent quality to re- ide complex (F-bsg). Attack by the diazoester 8 at the electro-
alizing rapid and efficient activation by the borosilicate glass- philic iodine produces A, which could reductively eliminate io-
ware.[30] To rule out surface impurities being responsible for the dotoluene to give B. Computationally, the initial fluoride ab-
improved reactivity, newly purchased glassware and stir bar straction is irreversible,[31] therefore we propose B to expel ni-
were used in the reaction, and gave 10b in 81 % yield (Table 2, trogen, possibly by anchimeric assistance from the carb-
entry 3). Experiments conducted in Nalgene and PFA vials onyl[15,32] or electron-rich arene groups, and give benzylic cat-
failed, though addition of 5 wt.-% ground borosilicate to a reac- ion C.[33] Should this cation be sufficiently Lewis acidic, it could
tion in PFA gave 10b in 76 % yield (entries 4–6). Reactions per- abstract fluoride from TolIF2 and regenerate the original iodo-
formed in boron-free silicates, such as silica gel, Pasteur pipette nium ion pair.
glass, and ground quartz failed (entries 7–9), which strongly
supports boron's role in the activation process. Use of a silan-
ized borosilicate vial gave 10b in 76 % yield (entry 10), indicat-
ing that passivating the glass surface fails to impede the gem-
difluorination reaction.
Similar improvements in reaction efficacy were also realized
for diazoacetates 8a–f, with the products being recovered in
equal or greater yield than under BF3·OEt2 activation
(Scheme 1). The improvements realized for the aryl ether sub-
strates 8g (R = 4-OMe) and 8h (R = OBn) were less than antici-
pated, presumably because aryl ethers are labile under the reac-
tion conditions. The novel reaction conditions were equally ef-
fective for substrates 8i–n, with a dramatic improvement in
chemoselectivity and reaction efficacy being realized for the
allyl ester (8k) and 3,4-dimethoxybenzyl ester (8m) substrates.
Overall, the reactions carried out under borosilicate catalysis
were rapid, operationally simple, and moderate-to-good yield-
ing. Importantly, borosilicate activation of TolIF2 was sufficiently
mild to enable us to elucidate the possibility of a subtle, though
important discrepancy in the stability of aryl ether substitution
on the arene (8g) and ester (8m) functional groups.
Figure 3. Mechanistic proposal for the borosilicate (bsg)-activated gem-di-
fluorination reaction. Inset: anionic borosilicate·fluoride (bsg-F) counterion.
Conclusion
Computational analysis was a valuable complement to our ex-
perimental search for mild, highly effective activating agents for
the hypervalent iodine reagent TolIF2. Two Lewis acid candi-
dates (TiF4 and borosilicate glass) were identified as viable
leads, and due to the novelty and simplicity of reactions being
initiated by the elemental composition of laboratory glassware,
borosilicate was further investigated. Computational analysis
and control experiments both indicated the boron of borosili-
cate to be crucial to the improvements realized for the gem-
difluorination reaction. These studies have clarified the role of
the Lewis acid as an activating agent, and the need for high
quality TolIF2 in these reactions. We continue to investigate the
nature of the activation process and the source of the oxygena-
tion by-products, and these results will be disclosed in due
course.
Author Contributions: The manuscript was written by G. K. M. and
W. S. H., G. S. S. did the computational work, tested the new Lewis
acids candidates and ran the difluorination reactions on most sub-
strates. R. T. repeated and verified key experiments, and character-
ized various new diazo and difluoride compounds. J. T. synthesized
the bulk of the diazo starting materials, and developed the synthe-
Scheme 1. gem-Difluorination of phenyldiazoacetate using borosilicate activa-
tion. [a] Isolated yield. [b] Yield using 1 mol-% BF3·OEt2.
Based on our computational studies and experimental obser-
vations, a plausible mechanism is proposed for the fluorine sis of TolIF2 that enabled this work.
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
3
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[14] a) A. L'Heureux, F. Beaulieu, C. Bennett, D. R. Bill, S. Clayton, F. LaFlamme,
M. Mirmehrabi, S. Tadayon, D. Tovell, M. Couturier, J. Org. Chem. 2010,
75, 3401; b) L. N. Markovskii, V. E. Pashinnik, E. P. Saenko, Zh. Org. Khim.
1977, 13, 1116.
[15] a) G. C. Geary, E. G. Hope, K. Singh, A. M. Stuart, Chem. Commun. 2013,
49, 9263; b) V. Matoušek, E. Pietrasiak, R. Schwenk, A. Togni, J. Org. Chem.
2013, 78, 6763.
[16] a) R. F. Weinland, W. Stille, Justus Liebigs Ann. Chem. 1903, 328, 132; b)
W. Bockemuller, Ber. Dtsch. Chem. Ges. 1931, 64, 522.
[17] C. Willgerodt, J. Prakt. Chem. 1886, 33, 154.
Acknowledgments
This work was supported by the Natural Sciences and Engineer-
ing Research Council (NSERC) of Canada and Compute Canada.
G. S. S. acknowledges the University of Waterloo for funding.
J. T. and R. T. acknowledge NSERC for graduate studentships.
G. K. M. would especially like to thank Dr. Sonny Lee (UWater-
loo) for helpful discussion.
[18] For uses of TolIF2 in synthesis, see: a) M. Yoshida, K. Fujikawa, S. Sato, S.
Hara, ARKIVOC (Gainesville, FL, U.S.) 2003, 36; b) T. Inagaki, Y. Nakamura,
M. Sawaguchi, N. Yoneda, S. Ayuba, S. Hara, Tetrahedron Lett. 2003, 44,
4117; c) W. B. Motherwell, M. F. Greaney, D. A. Tocher, J. Chem. Soc., Perkin
Trans. 1 2002, 2809; d) M. A. Arrica, T. Wirth, Eur. J. Org. Chem. 2005, 395;
e) P. Conte, B. Panunzi, M. Tingoli, Tetrahedron Lett. 2006, 47, 273; f) J.
Yu, J. Tian, C. Zhang, Adv. Synth. Catal. 2010, 352, 531; for a summary of
its reactivity, see: g) D. M. Lemal, J. Tao, G. K. Murphy, (Difluoroiodo)tolu-
ene, in: e-EROS, 2003; update: 2016.
[19] T. J. Tewson, M. J. Welch, J. Org. Chem. 1978, 43, 1090.
[20] A. Roedig, H. Aman, E. Fahr, Justus Liebigs Ann. Chem. 1964, 675, 47.
[21] J. Tao, R. Tran, G. K. Murphy, J. Am. Chem. Soc. 2013, 135, 16312.
[22] a) For activation of PhIF2 with BF3·OEt2, see: N. S. Zefirov, N. S. Samsoniya,
T. G. Kutateladze, V. V. Zhdankin, Zh. Org. Khim. 1991, 27, 220; b) N. S.
Zefirov, A. S. Koz′min, T. Kasumov, K. A. Potekhin, V. D. Sorokin, V. K. Brel,
E. V. Abramkin, Y. T. Struchkov, V. V. Zhdankin, P. J. Stang, J. Org. Chem.
1992, 57, 2433.
Keywords: Fluorine · Diazo compounds · Hypervalent
compounds · Iodine · Computational chemistry
[1] J.-P. Bégué, D. Bonnet-Delpon, Bioorganic and Medicinal Chemistry of Flu-
orine, John Wiley & Sons, Hoboken, NJ, USA, 2008.
[2] a) K. L. Kirk, Org. Process Res. Dev. 2008, 12, 305; b) K. K. J. Chan, D.
O'Hagan, in: Methods in Enzymology, vol. 516 (Ed.: A. H. David), Academic
Press, 2012, p. 219–235.
[3] a) C. L. Chernick, H. H. Claassen, P. R. Fields, H. H. Hyman, J. G. Malm, W.
M. Manning, M. S. Matheson, L. A. Quarterman, F. Schreiner, H. H. Selig,
I. Sheft, S. Siegel, E. N. Sloth, L. Stein, M. H. Studier, J. L. Weeks, M. H.
Zirin, Science 1962, 138, 136 ; b) R. Hoppe, W. Dahne, K. M. Rodder, H.
Mattauch, Angew. Chem. Int. Ed. Engl. 1962, 1, 599; Angew. Chem. 1962,
74, 903.
[4] O. Ruff, H. Krug, Z. Anorg. Allg. Chem. 1930, 190, 270.
[23] E. Emer, J. Twilton, M. Tredwell, S. Calderwood, T. L. Collier, B. Liegault,
M. Taillefer, V. Gouverneur, Org. Lett. 2014, 16, 6004.
[24] See the supporting information for the computational and experimental
results of all the Lewis acids investigated.
[25] Similar I–F bond elongations were observed for all of the Lewis acid·To-
lIF2 complexes analyzed in silico.
[26] For a crystal structure of TolIF2 see: a) C. F. Ye, B. Twamley, J. M. Shreeve,
Org. Lett. 2005, 7, 3961. For melting points of pure TolIF2, see ref. [26a]
and: M. Sawaguchi, S. Ayuba, S. Hara, Synthesis 2002, 1802.
[27] I. B. Sivaev, V. I. Bregadze, Coord. Chem. Rev. 2014, 270, 75.
[28] a) M. Yasuda, A. Baba, J. Synth. Org. Chem. Jpn. 2013, 71, 1294; b) M.
Yasuda, S. Yoshioka, S. Yamasaki, T. Somyo, K. Chiba, A. Baba, Org. Lett.
2006, 8, 761.
[5] P. Lebeau, Ann. Chim. Phys. 1906, 9, 241.
[6] G. A. Olah, A. E. Pavlath, S. J. Kuhn, J. Inorg. Nucl. Chem. 1958, 7, 301.
[7] a) F. Brown, P. L. Robinson, J. Chem. Soc. 1955, 3147; b) C. W. Tullock,
F. S. Fawcett, W. C. Smith, D. D. Coffman, J. Am. Chem. Soc. 1960, 82, 539;
c) F. S. Fawcett, C. W. Tullock, Inorg. Synth. 1963, 7, 119.
[8] For uses of XeF2 in synthesis, see: a) M. Tramsek, B. Zemva, Acta Chim.
Slov. 2006, 53, 105; b) C. A. Ramsden, ARKIVOC (Gainesville, FL, U.S.) 2014,
109; c) M. A. Tius, Tetrahedron 1995, 51, 6605.
[9] For uses of BrF3 in synthesis, see: a) T. Shishimi, S. Hara, J. Fluorine Chem.
2013, 145, 128; b) S. Rozen, Adv. Synth. Catal. 2010, 352, 2691; c) Y.
Hagooly, M. J. Welch, S. Rozen, Org. Biomol. Chem. 2011, 9, 902; d) U.
Farooq, A. Shah, T. Wirth, Angew. Chem. Int. Ed. 2009, 48, 1018; Angew.
Chem. 2009, 121, 1036.
[29] Glassware acting as a heterogeneous catalyst in fluorination reactions
has been reported previously. For an example, see: A. P. Lothian, C. A.
Ramsden, M. M. Shaw, R. G. Smith, Tetrahedron 2011, 67, 2788.
[30] The original result (40 % yield, 2 h) was reproduced by using TolIF2 that
had been temperature abused and was turning yellow.
[31] R. Minkwitz, W. Molsbeck, H. Oberhammer, I. Weiss, Inorg. Chem. 1992,
31, 2104.
[32] a) W. Carpenter, J. Org. Chem. 1966, 31, 2688; b) T. B. Patrick, J. J. Scheibel,
W. E. Hall, Y. H. Lee, J. Org. Chem. 1980, 45, 4492.
[10] For uses of IF5 in synthesis, see: a) T. E. Stevens, J. Org. Chem. 1966, 31,
2025; b) S. Hara, M. Aoyama, Synthesis 2008, 2510; c) M. Kunigami, S.
Hara, J. Fluorine Chem. 2014, 167, 101; d) N. Yoneda, J. Fluorine Chem.
2004, 125, 7; e) G. A. Olah, J. Welch, Synthesis 1977, 419.
[11] For uses of SF4 in synthesis see: a) W. C. Smith, C. W. Tullock, E. L. Muett-
erties, W. R. Hasek, F. S. Fawcett, V. A. Engelhardt, D. D. Coffman, J. Am.
Chem. Soc. 1959, 81, 3165 ; b) W. R. Hasek, W. C. Smith, V. A. Engelhardt,
J. Am. Chem. Soc. 1960, 82, 543; c) W. C. Smith, Angew. Chem. Int. Ed.
Engl. 1962, 1, 467; Angew. Chem. 1962, 74, 742.
[33] J. Wielgat, Z. Domagala, R. Kolinski, J. Fluorine Chem. 1987, 35, 643.
[12] W. J. Middleton, J. Org. Chem. 1975, 40, 574.
[13] a) G. S. Lal, G. P. Pez, R. J. Pesaresi, F. M. Prozonic, Chem. Commun. 1999,
215; b) M. Yasuda, S. Yoshioka, S. Yamasaki, T. Somyo, K. Chiba, A. Baba,
Org. Lett. 2006, 8, 761.
Received: June 24, 2016
Published Online: ■
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
4
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Fluorination
G. S. Sinclair, R. Tran, J. Tao,
W. S. Hopkins,* G. K. Murphy* ....... 1–5
Borosilicate Activation of (Difluoro-
iodo)toluene in the gem-Difluorin-
ation of Phenyldiazoacetate Deriva-
tives
A computational and experimental most effective activator (to date) in the
search for mild Lewis-acidic activators gem-difluorination of phenyl diazo-
for (difluoroiodo)toluene (TolIF2) has acetate derivatives.
identified borosilicate glass as the
DOI: 10.1002/ejoc.201600773
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
5
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim