The intramolecular C-F...HO hydrogen bond of 2- fluorophenyldiphenylmethanol

Article abstract of DOI:10.1039/b902203a

The intramolecular C-F...HO hydrogen bond of 2- fluorophenyldiphenylmethanol was observed by X-ray crystallographic analysis and NMR spectroscopy. The features of the spectra were compared to triphenylmethanol (a fluorine-free analog) and 2-fluorophenyldi

Full text of DOI:10.1039/b902203a

View Article Online / Journal Homepage / Table of Contents for this issue
LETTER
www.rsc.org/njc | New Journal of Chemistry
The intramolecular C–Fꢀ ꢀ ꢀHO hydrogen bond of
2-fluorophenyldiphenylmethanolw
Hiroyuki Takemura,*a Megumi Kaneko,a Katsuya Sakob and Tetsuo Iwanagac
Received (in Gainesville, FL, USA) 3rd February 2009, Accepted 25th August 2009
First published as an Advance Article on the web 7th September 2009
DOI: 10.1039/b902203a
The intramolecular C–Fꢀ ꢀ ꢀHO hydrogen bond of 2-fluorophenyl-
diphenylmethanol was observed by X-ray crystallographic analysis
and NMR spectroscopy. The features of the spectra were
compared to triphenylmethanol (a fluorine-free analog) and
2-fluorophenyldiphenylmethane.
foldamers were found by Li et al.11 Other examples have been
found by our group1 and by Vasella et al. (described below).12
In such a situation, the role of the C–Fꢀ ꢀ ꢀHO hydrogen bond
is somewhat subtle, and thus we must accumulate examples.
In our opinion, because the interaction is very weak, the
C–Fꢀ ꢀ ꢀHO hydrogen bond is influenced by the subtle
conditions of molecular structure; in some cases being visible
and in others invisible.
In a previous report, we discussed the C–Fꢀ ꢀ ꢀHO hydrogen
bond of 9-fluoro-18-hydroxy[3.3]metacyclophane,1 and its
stabilization energy was estimated to be 0.84–3.7 kJ molꢁ1
.
The aim of this study is to observe a C–Fꢀ ꢀ ꢀHO interaction
by some of the most common and familiar tools to organic
chemists, X-ray crystallographic analysis, and NMR and
IR spectroscopy. For this purpose, the molecular design of
the target molecule becomes very important. Similar
approaches were attempted by Vasella et al. using fluorinated
saccharides, and they succeeded in observing the C–Fꢀ ꢀ ꢀHO
hydrogen bond by 1H NMR and IR spectroscopy.12 However,
their X-ray crystallographic analyses revealed that the
OH proton in each compound was not directed toward the
F atom.
C–Fꢀ ꢀ ꢀHO or C–Fꢀ ꢀ ꢀHN hydrogen bonds occur very rarely
because they are weak,2 and other stronger interactions
predominate under typical circumstances. However, several
attempts have continued in order to observe and estimate these
hydrogen bonds under particular conditions (mainly in the gas
phase) using various modern spectroscopic methods or
ab initio calculations.3 In both solution and the solid phase,
there are controversies about the existence of this hydrogen
bond. Shimoni and Glusker concluded that ‘‘C–Fꢀ ꢀ ꢀH–X
(X = C, N, O) interactions cannot be ignored in predictions
of modes of molecular packing in complexes and in crystals’’
based on crystallographic analyses and a CSD (Cambridge
Structural Database) database analysis.4 On the contrary,
Dunitz and Taylor took a negative view about the interaction
after their analysis of the CSD.5 In a recent study, Hulliger
et al. pointed out that X–Hꢀ ꢀ ꢀF contacts are caused not only
by hydrogen bonds but also by crystal packing, and that the
number of O–Hꢀ ꢀ ꢀF and N–Hꢀ ꢀ ꢀF contacts is significantly
lower because O–H and N–H act rather as proton acceptors
than proton donors.6 Another group, however, support the
positive role of the F atom in crystal engineering.7 In bio-
organic chemistry, the isosteric and isoelectronic nature of the
F atom to the OH group were noted, and Hagan and Rzepa
mentioned that the F atom is an OH mimic hydrogen bond
acceptor in desolvated enzyme cavities, although rare.8
However, there are a few observations of O–Hꢀ ꢀ ꢀF and
N–Hꢀ ꢀ ꢀF interactions in solution. Recently, experimental
O–Hꢀ ꢀ ꢀF evidence by Strauss et al.9 was denied by Espinet
et al.10 On the other hand, N–Hꢀ ꢀ ꢀF hydrogen bonding-driven
In our current approach to detect such a weak hydrogen
bond, we chose 2-fluorophenyldiphenylmethanol (1)13 as a
touchstone because it is a simple molecule and its reference
compound analogues are accessible. Furthermore, as shown
in Fig. 1, the F and H(–O) atoms of 1 would form a
six-membered ring through C–C bonds. This is very important
and a preferable condition for the formation of a hydrogen
bond. Triphenylmethanol (2) and 2-fluorophenyldiphenyl-
methane (3) were used as reference compounds, and were
purchased or prepared according to the literature.14,15 A single
crystal of 1 suitable for an X-ray crystallographic analysis was
obtained from n-hexane, and its molecular structure is shown
in Fig. 2. An analysis shows that H(15) and F are adjacent to
each other and that the distance between them (223 pm) is
much shorter than the sum of the van der Waals radii of the
two atoms (267 pm), although rotating C–OH and C(OH)–Ar
bonds is very easy. In this report, the values of the van der
Waals radii used are those reported by Bondi.16 The O(1)–F
distance (283 pm) is also shorter than the sum of the van der
Waals radii of the F (147 pm) and O (152 pm) atoms.
a Department of Chemical and Biological Science, Faculty of Science,
Japan Women’s University, Mejirodai 2-8-1, Bunkyou-ku,
Tokyo 112-8681, Japan
Accordingly, these values are definitive evidence of
a
C–Fꢀ ꢀ ꢀHO hydrogen bond in 1. The Fꢀ ꢀ ꢀH–O angle is
somewhat acute (118.71) for a hydrogen bond, but this is
due to the very nature of the molecular structure.
b Nagoya Institute of Technology, Department of Systems
Management and Engineering, Gokiso, Showa-ku, Nagoya 466-8555,
Japan
On the contrary, four molecules of 2 form an intermolecularly
hydrogen-bonded tetrahedron with each four OH groups.17
This is a distinct difference between 1 and 2 caused by the
presence of the F atom at an appropriate position.
c Department of Chemistry, Faculty of Science, Okayama University
of Science, 1-1 Ridaicho, Okayama 700-0005, Japan
w CCDC 693164. For crystallographic data in CIF or other electronic
format see DOI: 10.1039/b902203a
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
2004 | New J. Chem., 2009, 33, 2004–2006

View Article Online
and the coupling of the OH proton signal are indices of a
C–Fꢀ ꢀ ꢀHO hydrogen bond in this system.19
In the IR spectra, the nHO of 1 appeared as a sharp band at
3569 cmꢁ1 in KBr and that of 2 appeared as a broad band at
3472 cmꢁ1. Because fluorine-free 2 forms an intermolecular
hydrogen bond in the crystal, the broadening and red-shifting
of the band is plausible. In the case of 1, as shown in the
crystallographic analysis, no intermolecular hydrogen bond in
the crystal is formed, and thus the sharp band of 1 is
characterized as an intramolecularly hydrogen-bonded
nC–Fꢀ ꢀ ꢀHO band. Thus, in the solid state spectra, we cannot
compare unimolecular nHO bands between 1 and 2. The IR
spectra in CCl4 were then obtained. In solution, the inter-
molecular hydrogen bond of 2 was easily broken, as depicted
in the 1H NMR spectra (it was negligible, even at the con-
centration of 2 ꢃ10ꢁ2 mol dmꢁ3). The nOH bands of 1 and 2 in
CCl4 (1 ꢃ 10ꢁ3 mol dmꢁ3) appeared at almost the same
wavenumber values, 3612 and 3611 cmꢁ1, respectively, and
these are within experimental error. In solution, each observed
band can be considered as a unimolecular band, but the
difference is very small and can be ignored. Thus, the
effect of the hydrogen bond on the solution spectra is
equivocal. A similar phenomenon was also observed in a
previous report.1
Fig. 1 The structure of 1 and its reference compounds 2 and 3.
Fig. 2 The molecular structure of 2-fluorophenyldiphenylmethanol (1).
In part, the interaction described here can be explained
by a dipole–dipole interaction (B5.6 kJ molꢁ1) that is ca. ten
times larger than O–H rotation (B470 J molꢁ1).z However,
the spin coupling between the OH and F atoms in this
strain-free system is definitive evidence of a redistribution of
In addition to the crystallographic information, the hydrogen
1
bond was also observed in solution. In the H NMR spectra,
the OH signal of 1 appeared as a doublet (5JH,F = 9.2 Hz) at
3.51 ppm in CD2Cl2 (1 ꢃ 10ꢁ2 mol dmꢁ3). The coupling
constant was too large for long-range coupling through
five bonds, because the OH signal of 2-fluorobenzyl alcohol
(when considered to be a partial structure of 1) appears as a
singlet at 2.8 ppm. Bernet and Vasella have described that the
coupling constant is reduced in basic solvents.12a Also, in the
case of 1, the coupling constant was reduced to 2.2 Hz
(5.50 ppm) in THF-d8, and the OH signal appeared as a sharp
singlet at 6.41 ppm in DMSO-d6. The disappearance of the
coupling is due to the strong OHꢀ ꢀ ꢀDMSO hydrogen bond,
which breaks the weak C–Fꢀ ꢀ ꢀHO bond. On the other hand,
the OH signal of 2 appeared as a sharp singlet at 2.88 ppm in
CD2Cl2 (2 ꢃ 10ꢁ2 mol dmꢁ3). Thus, the large difference in
chemical shift (DdOH = 0.71–0.63 ppm) between 1 and its
reference compounds (1, 3.51 ppm; 2-fluorobenzyl alcohol,
2.8 ppm; 2, 2.88 ppm), together with the spin coupling,
represent a hydrogen bond between the F and HO of 1 in
solution.
electrons between the OH and
F orbitals. Therefore,
the electrostatic term in this hydrogen bond should be
evaluated too. Furthermore, the proximity of the F and H
atoms is also evidence of a bond, because molecule 1 has
no steric crowding and the rotation of each bond is free.
If there is no interaction between F and HO, compound 1
would form an intermolecular OHꢀ ꢀ ꢀOH hydrogen bond, just
as for 2.
Therefore, the observed phenomena in this report
appropriately indicate the presence of C–Fꢀ ꢀ ꢀHO hydrogen
bond.
In conclusion, another example of a C–Fꢀ ꢀ ꢀHO hydrogen
bond system has been confirmed by crystallographic
analysis and 1H NMR spectroscopy. As mentioned above,
the C–Fꢀ ꢀ ꢀHO hydrogen bond is very weak; thus, careful
molecular design is very important in order to observe the
interaction using the typical techniques available to organic
chemists.
As a result of previous studies, high field shifts in the 19F
2
NMR signal and a reduced JC,F coupling constant in 13C
This work was supported by a Grant-in-Aid for Scientific
Research (no. 19550052) from the Japan Society for the
Promotion of Science (JSPS).
NMR spectra are acknowledged as specific for C–Fꢀ ꢀ ꢀM+
interactions.18 Since the hydrogen bond can be considered
C–Fꢀ ꢀ ꢀHd+Odꢁ, similar spectroscopic features are expected.
However, in the 19F NMR spectra, the 19F signal of 1
appeared in a rather lower field region (ꢁ110.5 ppm, CFCl3)
than that of 3 (ꢁ116.9 ppm). Also, in the 13C NMR spectra,
2
the JC,F values of 1 and its reference compounds 3, 2-fluoro-
benzylalcohol, fluorobenzene and 2-fluorotoluene, were not in
such a relationship. These results show the weakness of the
C–Fꢀ ꢀ ꢀHd+ interaction in comparison with the C–Fꢀ ꢀ ꢀM+
interaction. As a result, in the NMR spectra, the chemical shift
Fig. 3 A diagram of a dipole–dipole interaction
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 2004–2006 | 2005

View Article Online
r = 2.83 ꢃ 10ꢁ10 m. From these data, we obtained E D 5.6 kJ molꢁ1
.
Experimental
The resultant energy is close to our previously reported value, and thus
it is shown that the interaction is reproducible by simple estimation of
the dipole–dipole interaction. Furthermore, the rotational energy of
the O–H group was estimated to be 467 J (E = h2/8p2I ꢃ J(J + 1),
I = m ꢃ rOH ꢃ cos 14.51 (+H–O–C ꢁ 901)). Thus, the dipole–dipole
interaction is enough to stop the OH rotation.
Melting points: Yanaco MP-500D apparatus in Ar sealed
tubes; values are uncorrected.
NMR: JEOL AL-300 (300.4 MHz for 1H, 75.6 MHz for 13
C
and 283.0 MHz for 19F, with TMS and CFCl3 as internal
references, respectively).
1 H. Takemura, M. Kotoku, M. Yasutake and T. Shinmyozu,
Eur. J. Org. Chem., 2004, 2019–2024.
2 G. R. Desiraju and T. Steiner, The Weak Hydrogen Bond, in
Structural Chemistry and Biology, International Union of Crystallo-
graphy Book Series, Oxford University Press, Oxford, 1999,
pp. 202–205.
IR: JASCO IR-FT/IR 4200 (CCl4, in NaCl cells (0.1 and
0.5 mm) at 25 1C).
FAB MS: JEOL JMS-SX/SX102A.
Elemental analysis: the Service Centre of the Elementary
Analysis of Organic Compounds affiliated with the Faculty of
Science, Kyushu University.
3 (a) M. Zahedi-Tabrizi, F. Tayyari, Z. Moosavi-Tekyeh, A. Jalali
and S. Faramarz Tayyari, Spectrochim. Acta, Part A, 2006, 65,
387–396; (b) Z. Mielke, S. Coussan, K. Mierzwicki, P. Roubin and
M. Saldyka, J. Phys. Chem. A, 2006, 110, 4712–4718;
(c) W. Caminati, S. Melandri, A. Maris and P. Ottaviani, Angew.
Chem., Int. Ed., 2006, 45, 2438–2442; (d) W. Caminati, J. C. Lopez,
J. L. Alonso and J.-U. Grabow, Angew. Chem., Int. Ed., 2005, 44,
3840–3844; (e) R. L. Waterland, M. D. Hurley, J. A. Misner,
T. J. Wallington, S. M. L. Melo, K. Strong, R. Dumoulin,
L. Castera, N. L. Stock and S. A. Mabury, J. Fluorine Chem.,
2005, 126, 1288–1296; (f) I. Rozas, I. Alkorta and J. Elguero,
J. Phys. Chem. A, 2001, 105, 10462–10467; (g) A. Longarte,
J. A. Fernandez, I. Unamuno, F. Basterrechea and F. Castano,
J. Chem. Phys., 2001, 115, 270–276; (h) W. Caminati, S. Melandri,
I. Rossi and P. G. Favero, J. Am. Chem. Soc., 1999, 121,
10098–10101; (i) M. Pham, M. Gdaniec and T. Po"onski,
J. Org. Chem., 1998, 63, 3731–3734; (j) F. Lahmani,
A. Zehnacker, G. Denisov and G. G. Furin, J. Phys. Chem.,
1996, 100, 8633–8639; (k) L.-H. Xu, G. T. Fraser, F. J. Lovas,
R. D. Suenram, C. W. Gillies, H. E. Warner and J. Z. Gillies,
Chem. Phys., 1995, 103, 9541–9548; (l) E. Vajda and I. Hargittai,
J. Phys. Chem., 1993, 97, 70–76; (m) K.-M. Marstokk and
H. Møllendal, J. Mol. Struct., 1977, 40, 1–11; (n) G. L. Carlson,
W. G. Fateley, A. S. Manocha and F. F. Bentley, J. Phys. Chem.,
1972, 76, 1553–1557.
4 L. Shimoni and J. P. Glusker, Struct. Chem., 1994, 5, 383–397.
5 J. D. Dunitz and R. Taylor, Chem.–Eur. J., 1997, 3, 89–98.
6 K. Reichenbacher, H. I. Suss and J. Hulliger, Chem. Soc. Rev.,
2005, 34, 22–30.
7 G. Althoff, J. Ruiz, V. Rodriguez, G. Lopez, J. Perez and
C. Janiak, CrystEngComm, 2006, 8, 662–665.
8 D. O’Hagan and H. S. Rzepa, Chem. Commun., 1997, 645–652.
9 T. J. Barbarich, B. G. Nolan, S. Tsujioka, S. M. Miller, O. P. Anderson
and S. H. Strauss, J. Fluorine Chem., 2001, 112, 335–342.
10 C. Bartolome, P. Espinet and J. M. Martın-Alvarez, Chem. Commun.,
2007, 4384–4386.
11 C. Li, S.-F. Ren, J.-L. Hou, H.-P. Yi, S.-Z. Zhu, X.-K. Jiang and
Z.-T. Li, Angew. Chem., Int. Ed., 2005, 44, 5725–5729.
12 (a) B. Bernet and A. Vasella, Helv. Chim. Acta, 2007, 90,
1874–1888; (b) M. A. Biamonte and A. Vasella, Helv. Chim. Acta,
1998, 81, 695–716.
Chromatography was performed using the YAMAZEN
YFLC-254-GRII medium-pressure liquid chromatography system.
2 was purchased from Tokyo Chemical Industry Co., Ltd.
and used without purification.
2-Fluorophenyldiphenylmethanol (1)
This compound was obtained by the reaction of methyl 2-fluoro-
benzoate and phenyl magnesium iodide, and its physical and
spectral properties were consistent with those in the literature.13
2-Fluorophenyldiphenylmethane (3)
This compound was prepared by the reduction of 1 by Et3SiH in
CH3COOH.14 Compound 1 (100.2 mg, 0.36 mmol) was
dissolved in 1.8 mL of AcOH, and 0.4 mL of Et3SiH was added.
The mixture was heated under reflux for 40 h. The mixture
was then evaporated in vacuo and the resultant brown oil
was chromatographed on silica gel using hexane–CH2Cl2
(50 : 50 volume) as the eluent. Colorless crystals, 28.0 mg
(29.6%). m.p. 84.3–84.7 1C (lit. 85–87 1C15). dH (300 MHz,
CDCl3, Me4Si): 7.31–6.91 (m, 14H, ArH) and 5.83 (s, 1H, CH).
dC (75.4 MHz, CDCl3, Me4Si): 160.7 (d, J = 246.8 Hz), 142.6
(s), 130.9 (d, J = 3.6 Hz), 129.3 (s), 128.4 (s), 128.1 (d, J =
8.2 Hz), 126.5 (s), 123.8 (d, J = 3.7 Hz), 115.3 (d, J = 21.9 Hz)
and 49.4 (s). dF (282.2 MHz, CDCl3, CFCl3): ꢁ116.86 (m).
HRMS (FAB) calc. for C19H15F 262.1158, found 262.1155.
Crystallographic data for 1. C19H15OF, Mr = 278.31 g molꢁ1
,
platelet crystal (grown from n-hexane), size 0.53 ꢃ 0.45 ꢃ
0.37 mm, monoclinic, space group P21/n (#14), a = 8.2379(3),
b = 10.9417(4), c = 15.6797(4) A, V = 1411.10(8) A3, Z = 4,
rcalc = 1.310 g cmꢁ3, mCu-Ka = 0.88 cmꢁ1, F(000) = 584.00,
T = 113 K using the o-2y scan technique to a maximum 2y
value of 54.51. A total of 3160 reflections were collected. The
final cycle of the full-matrix least-squares refinement was
based on 2749 observed reflections (I 4 2s(I)) and 194
variable parameters, and converged with unweighted and
weighted agreement factors of R = 0.0544, Rw = 0.1797
and GOF = 1.409. The maximum and minimum peaks on the
final difference Fourier map corresponded to 1.054 and
13 A. Khvorost, D. Lemenovskii, N. Ustynyu, K. Rufanov and
J. Kipke, Z. Naturforsch., B: Chem. Sci., 2002, 57, 1207–1214.
14 P. Huszthy, K. Lempert and G. Simig, J. Chem. Soc., Perkin Trans. 2,
1985, 1351–1354.
15 E. Bergmann, Recl. Trav. Chim. Pays-Bas, 1939, 58, 863–870.
16 A. Bondi, J. Phys. Chem., 1964, 68, 441–451.
17 (a) G. Ferguson, J. F. Gallagher, C. Glidewell and
S. N. Scrimgeour, Acta Crystallogr., Sect. C: Cryst. Struct.
Commun., 1992, 48, 1272–1275; (b) H. Serrano-Gonzalez,
K. D. M. Harris, C. C. Wilson, A. E. Aliev, S. J. Kitchin,
B. M. Kariuki, M. Bach-Verges, C. Glidewell, E. J. MacLean
and W. W. Kagunya, J. Phys. Chem. B, 1999, 103, 6215–6223.
18 (a) H. Takemura, S. Nakashima, N. Kon, M. Yasutake,
T. Shinmyozu and T. Inazu, J. Am. Chem. Soc., 2001, 123,
9293–9298; (b) H. Plenio, Chem. Rev., 1997, 97, 3363–3384 and
references therein.
ꢁ0.530 eꢁ
A
ꢁ3, respectively.w
References
z Based on the crystallographic analysis data, the dipole–dipole inter-
action energy of the O–H and C–F groups in compound 1 was roughly
estimated (Fig. 3). Here, y1 = 70.31, y2 = 109.11, f = 9.71, m1 = 4.7 ꢃ
10ꢁ30 C m, m2 = 5.0 ꢃ 10ꢁ30 C m, e0 = 8.85 ꢃ 10ꢁ12 C2 Jꢁ1 mꢁ1 and
19 An example in which the reduction of JC–F and a high field shift of
the 19F NMR signal by a hydrogen bond was found in recent
research: H. Takemura, R. Ueda and T. Iwanaga, J. Fluorine
Chem., 2009, 130, 684–688.
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
2006 | New J. Chem., 2009, 33, 2004–2006