Detection of weak hydrogen bonding to fluoro and nitro groups in solution using H/D exchange

Article abstract of DOI:10.1039/c5ob02360b

Hydrogen/deuterium (H/D) exchange can be a sensitive technique for measuring the strength of hydrogen bonding to neutral organic nitro and fluoro groups. The slower rates of reaction in comparison to suitable controls suggest that hydrogen bonding is present, albeit rather weak.

Full text of DOI:10.1039/c5ob02360b

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Detection of weak hydrogen bonding to fluoro
and nitro groups in solution using H/D exchange†
Cite this: Org. Biomol. Chem., 2016,
14, 2223
C. R. Shugrue, J. R. DeFrancisco, A. J. Metrano, B. D. Brink, R. S. Nomoto and
B. R. Linton*
Received 16th November 2015,
Accepted 11th January 2016
Hydrogen/deuterium (H/D) exchange can be a sensitive technique for measuring the strength of hydro-
gen bonding to neutral organic nitro and fluoro groups. The slower rates of reaction in comparison to
suitable controls suggest that hydrogen bonding is present, albeit rather weak.
DOI: 10.1039/c5ob02360b
www.rsc.org/obc
bond acceptor.12 Evidence for the strength of hydrogen bonds
to organic fluorine in solution, however, remains elusive.
Introduction
There has been a history of uncertainty about the nature of
Solution-phase hydrogen/deuterium (H/D) exchange using
solution–phase interactions with weak hydrogen bond accep- 1H-NMR provides an additional method for analyzing
tor groups, such as halides or as neutral nitro groups.1 Hydro- hydrogen bonding and can serve as a sensitive technique for
gen bonding has been observed in solution with anionic nitro investigating weak interactions. H/D exchange is a kinetic
compounds,2 anionic fluoride,3 and in metal-fluoride com- analysis of the rate at which a hydrogen atom is replaced by
plexes,4 but as these interactions become weaker, the situation deuterium using a deuterated solvent.13 This technique has
becomes less clear. Failure to detect a hydrogen bond could proven to be quite illustrative in the field of protein folding,
suggest that either there is no interaction, or that the method where amides making strong hydrogen bonds show significant
of detection is not sensitive enough for an accurate determi- protection from chemical exchange, while solvent-accessible
nation. A third possibility is that these weak interactions could amides exchange more rapidly. Our group and others have
exist, but as is often the case, a stronger hydrogen bond accep- applied this technique to look at the intramolecular hydrogen
tors is also present and predominates in any structural bonding that is present in small-molecules in organic
determination.
solvents.14 This approach permits the analysis of the hydrogen
Hydrogen bonds to neutral nitro groups appear in a variety bonding in comparison to similar controls that lack the ability
of examples both in solution5 and in designed solid-state to form hydrogen bonds.
crystal structures.6 On the other hand, hydrogen bonding
interactions with C–F have been described as rare.1 An analysis
of crystal structure databases has suggested that organic
Results and discussion
fluorine hardly ever accepts hydrogen bonds,7 while another
presents evidence for their existence.8 Of course this does not
mean that these weak hydrogen bond acceptors cannot form
hydrogen bonds, but rather that they are more transient inter-
actions and difficult to discern with certainty. Earlier
vibrational spectroscopy9 and modern computational efforts10
have supported the delicate interactions between alkyl halides
and a variety of hydrogen bond donors. Some recent crystal
structures document C–F⋯H–N hydrogen bonds and probe
the changes between positional isomers.11 Foldamers have
been designed that rely on organic fluorine as a hydrogen
N-Phenylbenzamide 1 was used to position a single hydrogen
bond donor within a rigid scaffold that might maximize hydro-
gen-bonding interactions. This scaffold has been used pre-
viously to show the application of H/D exchange to hydrogen
bonding in small molecules.15 Additional derivatives (2–13)
positioned cyano, nitro and halide groups in a manner that
may or may not be able to form intramolecular hydrogen
bonds.16 Placement of the nitro group on the aniline side, and
the halides on the benzoyl side permits comparative analysis
of six-membered ring hydrogen bonds, which have been
shown to be more robust than similar five-membered rings on
this scaffold.15 To separate any inherent electronic effects from
potential hydrogen bonding, the H/D exchange rates for both
ortho and para derivatives for each functional group were
determined by following the disappearance of the amide N–H
in the 1H-NMR. In an effort to ensure we were looking at intra-
Department of Chemistry, College of the Holy Cross, Worcester, MA, USA.
E-mail: blinton@holycross.edu; Tel: +1 508 793 3560
†Electronic supplementary information (ESI) available: Raw data for H/D
exchange and synthetic details. See DOI: 10.1039/c5ob02360b
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2223–2227 | 2223

View Article Online
Paper
Organic & Biomolecular Chemistry
molecular hydrogen bonding instead of aggregation effects, withdrawing effect is slightly greater in the ortho position than
amide chemical shifts were measured at a series of concen- in the para position. By contrast, the slowest rates of H/D
trations and all H/D exchange experiments are performed at exchange were observed with the positioning of either a nitro
concentrations below any observed signs of intermolecular or fluoro group in the ortho position. Both were even slower
assembly.
than the rate of unsubstituted N-phenylbenzamide 1, despite
the presence of a significant electron-withdrawing group. This
selective protection from H/D exchange is consistent with the
formation of an attractive interaction such as a hydrogen bond
to nitro and fluoro, but not to cyano, chloro or iodo. A steric
effect must be considered regarding the ortho substituent.
However, positioning a methyl groups in either ortho positions
results in rates that are almost identical to the parent N-phenyl-
benzamide (1),15 suggesting that steric effects are not a major
factor in the exchange reactions involving this scaffold.
The fluoro derivative 8 has the slowest rate of exchange, but
this does not mean that it makes the strongest hydrogen bond.
The exchange data for the cyano, chloro and iodo derivatives
would suggest that in the absence of hydrogen bonding, ortho
substitution leads to a slightly faster rate of H/D exchange
than para substitution. It is also obvious that the effect of the
para nitro is much greater than that of the para fluoro, and it
seems likely that there is also an inherent electronic effect
occurring with the ortho positioning that is overwhelmed by
the presence of hydrogen bonding. One advantage of H/D
exchange in the analysis of hydrogen bonds is the ability to
correlate the rate of the reaction with hydrogen bond stability.
This has been described for strongly hydrogen bonding
methoxy groups,15 and can also be applied here to the more
weakly interacting halides and nitro. Englander has detailed
that if the overall rate of H/D exchange is known, and the
inherent rate of exchange of the functional group itself has
also been determined, then the relative strength of the hydro-
gen bond can be calculated.18 This complete treatment is not
possible in this case, since there is obviously an inherent
electronic effect from the electron-withdrawing group in the
ortho position. If we use the para-derivatives as a crude
control, we can obtain at best a rough estimate of hydrogen
bond strength. This analysis would suggest the hydrogen bond
The hydrogen/deuterium exchange rates for 1–13 using 1%
CD3OD/CDCl3 are shown in Table 1. The unsubstituted N-phe-
nylbenzamide 1 showed a rate of H/D exchange that was
slower than all para-substituted derivatives (3, 5, 7, 9, 11, 13).
The more strongly withdrawing groups (cyano 3, 7 and nitro 5)
showed the greatest effects. While several mechanisms of H/D
exchange are possible, the inherently electron-withdrawing
ability of all of these groups would best stabilize any increase
in electron density that results from the removal of the hydro-
gen. However, this does not rule out a concerted mechanism
of exchange.17 Molecules containing an ortho-functional group
fell into two categories. The derivatives that contained either
chloro (10), iodo (12), or cyano (2 and 6) groups all showed H/D
exchange rates that were even faster than their para-comp-
lements (11, 13, 3 and 7). This would suggest that the electron-
to the nitro has a strength somewhere around 1.6 kcal mol−1
and the hydrogen bond to the fluoro is around 0.4 kcal mol−1
,
.
In an effort to show that the observed effects are from
hydrogen bonding and not the presence of lone-pair repulsion
between the ortho functional groups and the amide oxygen,
additional data can be considered. Support for the presence of
these hydrogen bonds can be seen in the chemical shifts of
Table 1 H/D exchange half-lives for derivatives 1–13, measured in 1%
CD3OD/CDCl3. Chemical shifts are listed for pure CDCl3, and the
change in chemical shifts after adding 1% CD3OD
t1/2
Chemical shift
in CDCl3 (ppm)
Change in ppm
in 1% CD3OD
Compound
(min)
1
individual amides in the H-NMR. The shift of an N–H signal
1
83
7.79
7.90
7.78
11.37
8.29
7.90
7.78
8.45
7.72
7.85
7.73
7.38
7.84
0.15
0.26
0.30
0.00
0.18
0.17
0.30
0.03
0.20
0.17
0.26
0.27
0.11
to a higher frequency is a typical indication of hydrogen
bonding.19 For the series of molecules presented in Table 1,
most N–H chemical shifts are between 7.3 and 7.9 ppm
(column 3). The notable exceptions are the nitro compounds
(4 and 5) and the ortho-fluoro compound 8. There is certainly a
slight electronic effect of the para-nitro 5, but it is difficult to
explain the additional 3 ppm shift observed in 4 without the
presence of hydrogen bonding. Similarly, the ortho-fluoro
positioning (8) could be an extreme local effect, but is also
consistent with the existence of a hydrogen bond.
2 (CN ortho)
3 (CN para)
4 (NO2 ortho)
5 (NO2 para)
6 (CN ortho)
7 (CN para)
8 (F ortho)
9 (F para)
10 (Cl ortho)
11 (Cl para)
12 (I ortho)
13 (I para)
3.2
9.8
130
8.9
3.4
18
158
70
12
28
12
55
2224 | Org. Biomol. Chem., 2016, 14, 2223–2227
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
A similar pattern is observed when comparing the change
in chemical shift with the addition of CD3OD (Table 1,
column 4). Since methanol can readily form hydrogen bonds,
one would expect any free amide NMR signal to experience a
shift to a higher frequency upon the addition of methanol.
Most derivatives show a shift between 0.10 and 0.30 ppm upon
addition of 1% CD3OD. Derivatives 4 and 8, however, both
show almost no effect with added methanol. This would be
the case if they were already engaged in hydrogen bonding and
did not compete as well for the added methanol. Again, this is
most consistent with hydrogen bonding.
Changes in chemical shift at different concentrations can
also be used as an indicator of hydrogen bonding ability. If a
signal is sensitive to changes in concentration it can indicate
increased intermolecular hydrogen bonding at increased con-
centration, while a similar derivative that shows little change
with concentration may indicate that it is already engaged in
intramolecular hydrogen bonding and therefore experiences
less intermolecular interactions. The ortho-fluoro 8 and ortho-
chloro 10 derivatives were prepared at a concentration of
500 mM and diluted to 0.25 mM (Fig. 1). Unfortunately para-
derivatives were not of sufficient solubility for useful compari-
sons. Over this concentration range, the chloro derivative
experienced a change in chemical shift of 0.40 ppm, while the
fluoro showed a much smaller effect of 0.04 ppm. This insensi-
tivity to changes in concentration is also consistent with the
C–F bond functioning as an intramolecular hydrogen bond
acceptor and the chloro failing to do so. A similar insensitivity
Fig. 2 NMR of para-fluoro 9 (top) and ortho-fluoro 8 (bottom) in CDCl3
(400 MHz, 22 °C). Coupling can be observed in the NH signal at the
bottom left.
to changes in concentration was observed with the ortho-nitro A similar effect has been documented for intramolecular
derivative 4, which did not shift at all in the range of 250 mM F⋯H–N interactions in other substrates.21
to 25 mM.20
All of the evidence presented supports the ability of both
There is also a comparison to be made in the peakshape of nitro and fluoro groups to serve as a hydrogen bond acceptors
1
the N–H signal in the H-NMR for the fluoro derivatives. The in solution. Granted, the molecules involved have been chosen
region between 7–9 ppm is shown for each in Fig. 2. for the express purpose of maximizing both the structural
Para-fluoro 9 showed a characteristic broad N–H signal, but rigidity and the ring-size for promotion of hydrogen bonding.
the amide signal in ortho-fluoro 8 was a broad doublet. This There is also little doubt that the interactions to nitro and
appears to be a result of long-range through-space coupling fluoro are still rather weak. They fail to show clear indications
between the amide hydrogen and the NMR-active fluorine. of hydrogen bonding using variable-temperature NMR or by
comparing infrared spectra,20 further illustrating the useful-
ness of H/D exchange as part of the toolbox to probe hydrogen
bonds. The lack of similar interactions with cyano groups is
consistent with geometric restrictions for the formation of a
hydrogen bond.22 It is unclear if the lack of interactions with
chloro and iodo are due to the increased atom-size affecting
hydrogen bond geometry, or due to differences in electron
density. These results are consistent with crystal structures
where hydrogen bonding to halide of nitro groups has been
used for crystal engineering.6,11 There are numerous examples
in the literature where fluorine either does not form hydrogen
bonds under a given set of conditions or engages in hydrogen
bonding with a stronger hydrogen bond acceptor.23 This does
not mean that it cannot form hydrogen bonds, but rather it
remains possible that competition with either solvent
molecules or better hydrogen bond acceptors in the substrate
itself makes hydrogen bonds to organic fluorine unlikely in
many cases.
Fig. 1 Change in the N–H chemical shifts for ortho-fluoro 8 ( ) and
ortho-chloro 10 ( ) with varying concentration.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2223–2227 | 2225

View Article Online
Paper
Organic & Biomolecular Chemistry
5 T. R. Kelly and M. H. Kim, J. Am. Chem. Soc., 1994, 116,
7072; H. E. Ungnade, E. M. Roberts and L. W. Kissinger, J.
Phys. Chem., 1964, 68, 3225; W. F. Baitinger, P. von
R. Schleyer, T. S. S. R. Murty and L. Robinson, Tetrahedron,
1964, 20, 1635; A. T. Dubis, Z. Lotowski, L. Siergiejczyk,
A. Z. Wilczewska and J. W. Morzycki, J. Chem. Res., 1998,
170; M. West-Nielsen, P. M. Dominiak, K. Wozniak and
P. E. Hansen, J. Mol. Struct., 2006, 789, 81; W. R. Zheng,
J. L. Xu, T. Huang, Q. Yang and Z. C. Chen, Res. Chem.
Intermed., 2011, 37, 31.
Conclusions
Because of the role that hydrogen bonds play in a variety of
chemical processes, it is important to accurately assess their
presence. Unfortunately, as hydrogen bonds get weaker, this
task becomes more difficult. Hydrogen/deuterium exchange
can provide an additional method to investigate hydrogen
bond strength. In these rigid molecules in organic solvents,
there is evidence for both neutral nitro groups and organic
fluorine serving as hydrogen bond acceptors.
6 J. L. Wardell, J. M. S. Skakle, J. N. Low and C. Glidewell,
Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2005, 61,
o634; A. Saeed and J. Simpson, Acta Crystallogr., Sect. E:
Struct. Rep. Online, 2009, 65, o1845; C. V. K. Sharma and
G. R. Desiraju, J. Chem. Soc., Perkin Trans. 2, 1994, 2345;
F. H. Allen, C. A. Baalham, J. P. M. Lommerse, P. R. Raithby
and E. Sparr, Acta Crystallogr., Sect. B: Struct. Sci., 1997, 53,
1017; T. W. Panunto, Z. Urbanczyk-Lipkowska, R. Johnson
and M. C. Etter, J. Am. Chem. Soc., 1987, 109, 7786.
7 J. D. Dunitz and R. Taylor, Chem. – Eur. J., 1997, 3, 89.
8 H.-J. Schneider, Chem. Sci., 2012, 3, 1381.
9 R. West, D. L. Powell, L. S. Whatley, M. K. T. Lee and P. von
R. Schleyer, J. Am. Chem. Soc., 1962, 84, 2221.
Experimental
All H/D exchange kinetics were performed in 1% CD3OD/
CDCl3 and at 22 °C. An initial NMR spectrum was acquired
with 990 μL of an analyte solution with a concentration of
5.0505 mM in CDCl3. Immediately prior to use, the deutero-
chloroform was passed through a small plug of neutral
alumina to remove any acidic impurities and then used
immediately. To this was added 10 μL of CD3OD marking time
= 0, and resulting in a final substrate concentration of 5 mM,
which was well below the concentrations where any evidence
of aggregation has been observed. The added methanol 10 J. Nadas, S. Vukovic and B. P. Hay, Comput. Theor. Chem.,
created a 1% methanol : chloroform solution, with the final 2012, 988, 75.
methanol concentration of 247 mM, which ensured pseudo- 11 P. Mocilac, K. Donnelly and J. F. Gallagher, Acta Crystal-
first-order kinetics. For a typical experiment, spectra were
acquired starting at 3 minutes, then every 1 minute until
10 minutes, every 5 minutes through 100 minutes, at 120, 150
and 200 minutes, and then every 50 minutes thereafter until
the disappearance of the proton signal into the baseline.
logr., Sect. B: Struct. Sci., 2012, 68, 189; K. Donnelly,
J. F. Gallagher and A. J. Lough, Acta Crystallogr., Sect. C:
Cryst. Struct. Commun., 2008, 64, o335; D. Chopra and
T. N. G. Row, CrystEngComm, 2008, 10, 54.
12 Y.-Y. Zhu, C. Li, G.-Y. Li, X.-K. Jiang and Z.-T. Li, J. Org.
Chem., 2008, 73, 1745; Y.-Y. Zhu, J. Wu, C. Li, J. Zhu,
J.-L. Hou, C.-Z. Li, X.-K. Jiang and Z.-T. Li, Cryst. Growth
Des., 2007, 7, 1490; 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.
Acknowledgements
This material is based upon work supported by the National
Science Foundation under CHE-0852232. The authors also 13 For reviews of hydrogen–deuterium exchange see the fol-
wish to thank the American Chemical Society Petroleum
Research Fund for support.
lowing: C. S. Maier and M. L. Deinzer, Methods Enzymol.,
2005, 402, 312; M. M. G. Krishna, L. Hoang, Y. Lin and
S. W. Englander, Methods, 2004, 34, 51; C. K. Woodward,
Curr. Opin. Struct. Biol., 1994, 4, 112; S. W. Englander and
N. R. Kallenbach, Q. Rev. Biophys., 1984, 4, 521;
S. W. Englander, N. W. Downer and H. Teitelbaum, Annu.
Rev. Biochem., 1972, 41, 903.
Notes and references
1 A thorough review of weak hydrogen bonding can be found
in: G. R. Desiraju and T. Steiner, The weak hydrogen bond: 14 L. R. Steffel, T. J. Cashman, M. H. Reutershan and
in structural biology and chemistry, Oxford University Press,
1999.
2 For some early examples see: B. R. Linton, M. S. Goodman
and A. D. Hamilton, Chem. – Eur. J., 2000, 6, 2449;
A. M. Kelly-Rowley, V. M. Lynch and E. V. Anslyn, J. Am.
Chem. Soc., 1995, 117, 3438.
B. R. Linton, J. Am. Chem. Soc., 2007, 129, 12956;
T. J. Cashman and B. R. Linton, Org. Lett., 2007, 9, 5457;
H. Lingard, J. T. Han, A. L. Thompson, I. K. H. Leung,
R. T. W. Scott, S. Thompson and A. D. Hamilton, Angew.
Chem., Int. Ed., 2014, 53, 3650; D. W. Carney, K. R. Schmitz,
J. V. Truong, R. T. Sauer and J. K. Sello, J. Am. Chem. Soc.,
2014, 136, 1922; Z. Shi, Y. Song, F. Lu, T. Zhao, X. Zhao,
W. Zhang and Z. Li, Acta Chim. Sin., 2013, 71, 51;
P. Prabhakaran, V. Azzarito, T. Jacobs, M. J. Hardie,
C. A. Kilner, T. A. Edwards, S. L. Warriner and A. J. Wilson,
Tetrahedron, 2012, 68, 4485; Z. Ke, H.-F. Chow, M.-C. Chan,
3 J. Emsley, Chem. Soc. Rev., 1980, 9, 91.
4 H. D. Selby, B. K. Roland, M. D. Carducci and Z. Zheng,
Inorg. Chem., 2003, 42, 1656–1662; T. Birk, J. Bendix and
H. Weihe, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2008,
68, m369.
2226 | Org. Biomol. Chem., 2016, 14, 2223–2227
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
Z. Liu and K.-H. Sze, Org. Lett., 2012, 14, 394; 20 The complete data is included in the ESI†
A. S. M. Ressurreição, A. Bordessa, M. Civera, L. Belvisi, 21 G. N. Manjunatha Reddy, M. V. Vasantha Kumar,
C. Gennari and U. Piarulli, J. Org. Chem., 2008, 73, 652; Z. Qi,
C. Schlaich and C. A. Schalley, Chem. – Eur. J., 2013, 19, 14867.
15 T. L. Schneider, K. T. Halloran, J. A. Hillner, R. R. Conry
and B. R. Linton, Chem. – Eur. J., 2013, 19, 15101.
T. N. G. Row and N. Suryaprakash, Phys. Chem. Chem. Phys.,
2010, 12, 13232; L. Hennig, K. Ayala-Leon,
J. Angulo-Cornejo, R. Richter and L. Beyer, J. Fluorine
Chem., 2009, 130, 453; R. Dalterio, X. Stella Huang and
K.-L. Yu, Appl. Spectrosc., 2007, 61, 603.
16 D. R. Turner, A. J. Edwards and R. O. Piltz, CrystEngComm,
2012, 14, 6447; N. Ziao, J. Graton, C. Laurence and J.-Y. Le 22 R. Taylor and O. Kennard, Acc. Chem. Res., 1984, 17,
Questel, Acta Crystallogr., Sect. B: Struct. Sci., 2001, 57, 850. 320.
17 S. Campbell, M. T. Rodgers, E. M. Marzluff and 23 Some recent examples where C–F⋯H–N hydrogen bonding
J. L. Beauchamp, J. Am. Chem. Soc., 1995, 117, 12840;
T. Wyttenbach and M. T. Bowers, J. Am. Soc. Mass Spectrom.,
1999, 10, 9.
is not observed: A. G. Dikundwar, U. D. Pete, C. M. Zade,
R. S. Bendre and T. N. G. Row, Cryst. Growth Des., 2012, 12,
4530; W. Zhu, W. Yang, W. Zhou, H. Liu, S. Wei and J. Fan,
J. Mol. Struct., 2011, 1004, 74; J. McMahon, J. F. Gallagher,
F. P. Anderson and A. J. Lough, Acta Crystallogr., Sect. C:
Cryst. Struct. Commun., 2009, 65, o345; N. Shibata,
B. K. Das, K. Harada, Y. Takeuchi and M. Bando, Synlett,
2001, 1755.
18 Y. Bai, J. J. Englander, L. Mayne, J. S. Milne and
S. W. Englander, Methods Enzymol., 1995, 259, 344.
19 L. Yao, A. Grishaev, G. Cornilescu and A. Bax, J. Am. Chem.
Soc., 2010, 132, 10866; G. Wagner, A. Pardi and
K. Wüthrich, J. Am. Chem. Soc., 1983, 105, 5948.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2223–2227 | 2227