Hydrogen-deuterium exchange of aromatic amines and amides using deuterated trifluoroacetic acid

Article abstract of DOI:10.1016/j.tetlet.2014.12.102

The H-D exchange of aromatic amines and amides, including pharmaceutically relevant compounds such as acetaminophen and diclofenac, was investigated using CF₃COOD as both the sole reaction solvent and source of deuterium label. The described method is amenable to efficient deuterium incorporation for a wide variety of substrates possessing both electron-donating and electron-withdrawing substituents. Best results were seen with less basic anilines and highly activated acetanilides, reflecting the likelihood of different mechanistic pathways.

Full text of DOI:10.1016/j.tetlet.2014.12.102

Tetrahedron Letters 56 (2015) 747–749
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Hydrogen–deuterium exchange of aromatic amines and amides
using deuterated trifluoroacetic acid
⇑
Richard Giles, Amy Lee, Erica Jung, Aaron Kang, Kyung Woon Jung
Loker Hydrocarbon Research Institute and Department of Chemistry, University of Southern California, Los Angeles, CA 90089, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The H–D exchange of aromatic amines and amides, including pharmaceutically relevant compounds such
as acetaminophen and diclofenac, was investigated using CF3COOD as both the sole reaction solvent and
source of deuterium label. The described method is amenable to efficient deuterium incorporation for a
wide variety of substrates possessing both electron-donating and electron-withdrawing substituents.
Best results were seen with less basic anilines and highly activated acetanilides, reflecting the likelihood
of different mechanistic pathways.
Received 2 December 2014
Accepted 17 December 2014
Available online 24 December 2014
Keywords:
H–D exchange
CF3COOD
Ó 2014 Elsevier Ltd. All rights reserved.
Acetaminophen
Electrophilic aromatic substitution
Brønsted acid
Introduction
employ a wide variety of acidic catalysts, including heterogeneous
and Lewis acids, homogeneous Brønsted systems are the most
Increasing demand for deuterated compounds, especially due to
a renewed interest in the therapeutic prospects of so-called ‘heavy
drugs’,1 has spurred development of novel methods for hydrogen–
deuterium exchange. One group of compounds with significant
research interest for H–D exchange is the aromatic amines and
amides, due to the presence of these moieties in numerous
clinically important pharmaceuticals, including such mainstays as
acetaminophen (paracetamol), diclofenac, and acebutolol. The for-
mer two compounds are of particular importance with respect to
aromatic isotopic exchange due to their metabolism to potentially
toxic ring-oxygenated metabolites in the liver by cyctochrome
P450 and related enzymes. Several in vitro and in vivo studies per-
formed using deuterated analogs of these compounds have had
profound mechanistic implications.2 As a result, various syntheses
of ring-deuterated variants of acetaminophen3 and diclofenac4
have been reported. However, these methods generally require
several synthetic steps from expensive deuterated starting materi-
als. Efforts have been made toward exploring the direct aromatic
H–D exchange of both of these compounds to varying degrees of
success.5 In particular, a rapid, direct, and metal-free method for
labeling acetaminophen with deuterium has proven elusive.
Acid-catalyzed H–D exchange conditions encompass some of
the most powerful and efficient methods for incorporating
deuterium into an aromatic ring.6 While these methods can
common. Under these conditions, reactivity typically proceeds
through an electrophilic aromatic substitution mechanism and
the aromatic ring is selectively deuterated at the most electron-
rich positions. Commonly-used systems for H–D exchange include
a solvent, usually D2O, as well as a deuterated mineral acid catalyst
such as DCl, DBr, or D2SO4. However, the use of these harsh
reagents has several disadvantages, including unwanted side reac-
tions, low substrate solubility in aqueous solution, and potential
substrate decomposition.
The use of deuterated trifluoroacetic acid (CF3COOD) has several
benefits over other acidic H–D exchange methods, including its
ease of preparation, simple removal in vacuo, low nucleophilicity,
and high solubility properties for a wide variety of substrates. CF3-
COOD can be prepared quantitatively from trifluoroacetic anhy-
dride and D2O in essentially anhydrous fashion. Since the
pioneering efforts of Lauer et al.,7 CF3COOD has found use as a
highly versatile and effective deuterating agent for hormones,8 nat-
ural products,9 and various other biologically relevant aromatic
systems.10 We considered the viability of this reagent for the deu-
teration of aromatic amines and amides such as acetaminophen. A
handful of examples can be found in the chemical literature,
including the CF3COOD-catalyzed H–D exchange of tryptophan
and its derivatives11 along with a single report each regarding H–
D exchange of naphthalene-diamines,12 reserpine,13 and pteroyl-
glutamic acid.14 However, to the best of our knowledge, the H–D
exchange of other aromatic amines and amides in CF3COOD has
not been studied extensively. Herein we report on the generalized
⇑
Corresponding author. Tel.: +1 213 740 8768; fax: +1 213 740 6270.
E-mail address: kwjung@usc.edu (K.W. Jung).
http://dx.doi.org/10.1016/j.tetlet.2014.12.102
0040-4039/Ó 2014 Elsevier Ltd. All rights reserved.

748
R. Giles et al. / Tetrahedron Letters 56 (2015) 747–749
use of CF3COOD as a source of deuterium and reaction solvent for
the preparation of several deuterated aromatic amine and amide
derivatives.
Results and discussion
While investigating palladium-catalyzed H–D exchange of
pharmaceutically relevant compounds, we found CF3COOD be
highly effective at deuterating acetaminophen without any metal
catalyst at 110 °C (Scheme 1). Since efficient and direct H–D
exchange of this compound has only been conducted using the
assistance of rhodium salts5c we decided to investigate this unex-
pected transformation further. At reflux without any additional
catalyst or co-solvent, acetaminophen was deuterated extensively
at the aromatic positions ortho to the hydroxyl substituent and
more slowly at the positions ortho to the amide. The selectivity
of deuteration is opposite that observed in the previously reported
metal-catalyzed example and is unprecedented in the direct H–D
exchange of acetaminophen.
Figure 1. H–D exchange of anilines and acetanilide without ring substituents.15
Encouraged by these results, we endeavored to determine the
scope of this methodology16 and subjected simple aniline and acet-
anilide, as well as aniline derivatives substituted only at nitrogen,
to the same CF3COOD conditions used for acetaminophen (Fig. 1).
Aniline 1 and N-ethylaniline 2 underwent H–D exchange smoothly
under these conditions at the positions ortho and para to the nitro-
gen atom, a result consistent with a standard EAS mechanism.
However, the H–D exchange of N,N-diethylaniline 3 under the
same conditions was poor despite the enhanced induction from
the alkyl substituents on nitrogen. This low rate of exchange was
also observed in previous work on the acid-catalyzed proton
exchange of N,N-diethylaniline17 and attributed to its high basicity
and at least partially to steric interference from the ethyl groups.
Interestingly, 1-phenylpiperazine 4 was even more reactive toward
CF3COOD–catalyzed H–D exchange than aniline, indicating that
dialkylation of the amine is not inherently detrimental toward
the exchange process. Acetanilide 5 was similarly reactive toward
H–D exchange as aniline, suggesting that acetylation of the nitro-
gen does not necessarily deactivate the aromatic ring toward reac-
tion with CF3COOD.
To test the effect of various ring substituents, para-substituted
substrates were then investigated under the reaction conditions
(Fig. 2). The reactivity of p-toluidine 6 was similar but not
enhanced compared to aniline, despite the presence of an activat-
ing methyl substituent on the ring. However, the reactivity of both
p-anisidine 7 and p-aminophenol 8 was significantly lower than
that of p-toluidine, and deuteration of both of these substrates pro-
ceeded with very low selectivity. Interestingly, 4-nitroaniline 9
underwent H–D exchange analogously to p-toluidine despite the
very different electronic properties of these aromatic systems.
Thus a strong correlation between the basicity of the amine and
the rate of H–D exchange can be seen, with more basic anilines
reacting somewhat less efficiently and with lower selectivity.
Acid-catalyzed H–D exchange of aromatic anilines can proceed
via aromatic substitution of either the protonated anilinium ion
(slow) or its free-base counterpart (fast) which exists in equilib-
rium. In strongly acidic solution, electron-rich anilines are more
Figure 2. H–D exchange of para-substituted anilines and acetanilides.15
likely to react through the former pathway, and with deactivated
anilines, the latter pathway predominates.18 The observed selectiv-
ity differences in this experiment likely reflect these mechanistic
differences.
Further support of this hypothesis can be drawn from the reac-
tivity of the N-acetylated derivatives of each compound. Acetyla-
tion of the amine inhibits protonation at nitrogen, essentially
eliminating one of the two possible H–D exchange pathways from
consideration. The reactivity of 40-methylacetanilide 10 was com-
parable to its non-acetylated counterpart 6, reflecting a minor
inductive effect from the methyl substituent. In contrast, the H–
D exchange of 4-methoxyacetanilide 11 was very efficient and
highly selective compared to p-anisidine 7, and the observed H–
D exchange pattern was opposite that of 40-methylacetanilide.
The reactivity and deuteration selectivity of acetaminophen 12
was analogous to that of 4-methoxyacetanilide and similarly
improved relative to 4-aminophenol 8. The reactivity of 11 and
12 revealed the strong directing influence of the hydroxy and
methoxy functional groups in the absence of protonation at nitro-
gen. However, the more-deactivated 4-nitroacetanilide 13 barely
reacted at all compared to 4-nitroaniline 9. Thus, while optimal
H–D exchange results occur with less electron-rich anilines, ace-
tanilides react with greater efficiency with an activating substitu-
ent on the aromatic ring.
We then considered the effect of the substitution pattern of the
ring on H–D exchange by reacting the other anisidine isomers with
CF3COOD (Fig. 3). The influence of the amine group dominated the
H–D exchange selectivity of o-anisidine 14, and the positions ortho
and para to the amine exchanged rapidly under these conditions
while the positions ortho and para to the methoxy group had sig-
nificantly lower deuterium incorporation. A similar pattern was
observed in m-anisidine 15, which underwent rapid exchange
throughout the ring except the single position meta to both the
Scheme 1. CF3COOD-catalyzed H–D exchange of acetaminophen.15

R. Giles et al. / Tetrahedron Letters 56 (2015) 747–749
749
Acknowledgments
We acknowledge generous financial support from the Hydro-
carbon Research Foundation and the National Institute of Health
(S10 RR025432). The authors also acknowledge Joo Ho Lee and
Nima Zargari for insightful discussions.
References and notes
Figure 3. H–D exchange of ortho and meta-substituted anilines.15
1. (a) Timmins, G. S. Expert Opin. Ther. Patents 2014, 24, 1067–1075; (b)
Katsnelson, A. Nat. Med. 2013, 19, 656.
2. (a) Grunwald, H.; Hargreaves, P.; Gebhardt, K.; Klauer, D.; Serafyn, A.; Schmitt-
Hoffmann, A.; Schleimer, M.; Schlotterbecks, G.; Wind, M. J. Pharmaceut.
Biomed. 2013, 85, 138–144; (b) Kozakai, K.; Yamada, Y.; Oshikata, M.; Kawase,
T.; Suzuki, E.; Haramaki, Y.; Taniguchi, H. Drug Metab. Pharmacokinet. 2012, 27,
520–529; (c) Hoffmann, K.-J.; Axworthy, D. B.; Baillie, T. A. Chem. Res. Toxicol.
1990, 3, 204–211; (d) Forte, A. J.; Wilson, J. M.; Slattery, A. J.; Nelson, S. D. Drug
Metab. Dispos. 1984, 12, 484–491.
3. (a) Johnston, D.; Elder, D. J. Labelled Compd. Radiopharm. 1988, 25, 1315–1318;
(b) Freed, C. R.; Murphy, R. C. J. Labelled Compd. Radiopharm. 1978, 15, 637–643.
4. (a) Wu, K.; Tian, L.; Li, H.; Li, J.; Chen, L. J. Labelled Compd. Radiopharm. 2009, 52,
535–537; (b) Leroy, D.; Richard, J.; Godbillon, J. J. Labelled Compd. Radiopharm.
1993, 33, 1019–1027.
5. (a) Atzrodt, J.; Blankenstein, J.; Brasseur, D.; Calvo-Vicente, S.; Denoux, M.;
Derdau, V.; Lavisse, M.; Perard, S.; Roy, S.; Sandvoss, M.; Schofield, J.;
Zimmermann, J. Bioorgan. Med. Chem. 2012, 20, 5658–5667; (b) Tuck, K. L.;
Tan, H.-W.; Hayball, P. J. J. Labelled Compd. Radiopharm. 2000, 43, 817–823; (c)
Lockley, W. J. S. J. Labelled Compd. Radiopharm. 1985, 22, 623–630.
6. Junk, T.; Catallo, W. J. Chem. Soc. Rev. 1997, 26, 401–406.
Scheme 2. H–D exchange of diclofenac.15
amine and methoxy substituents. The effect of an electron-with-
drawing substituent nitro substituent in place of the methoxy
was pronounced, as 3-nitroaniline 16 exchanged at a much lower
rate than m-anisidine.
7. Lauer, W. M.; Matson, G. W.; Stedman, G. J. Am. Chem. Soc. 1958, 80, 6433–
6437.
8. (a) Stack, D. E.; Ritonya, J.; Jakopovic, S.; Maloley-Lewis, B. Steroids 2014, 92,
32–38; (b) Kiuru, P. S.; Wähälä, K. Tetrahedron Lett. 2002, 43, 3411–3412.
9. (a) Betts, J. W.; Kitney, S. P.; Fu, Y.; Peng, W.-M.; Kelly, S. M.; Haswell, S. J. Chem.
Eng. J. 2011, 167, 545–547; (b) Kamounah, F. S.; Christensen, P.; Hansen, P. E. J.
Labelled Compd. Radiopharm. 2010, 54, 126–131; (c) Jordheim, M.; Fossen, T.;
Songstad, J.; Andersen, Ø. M. J. Agric. Food Chem. 2007, 55, 8261–8268.
10. (a) Fan, D.; Taniguchi, M.; Lindsey, J. S. J. Org. Chem. 2007, 72, 5350–5357; (b)
Taylor, P. J. M.; Bull, S. D. Tetrahedron: Asymmetry 2006, 17, 1170–1178.
The applicability of this method to more complex pharmaceuti-
cal compounds was investigated using diclofenac as an example
(Scheme 2). Like acetaminophen, diclofenac is an aromatic amine
or amide-containing NSAID with wide-ranging therapeutic appli-
cations. Upon subjecting the sodium salt of diclofenac to the
described experimental conditions, cyclization was observed to
the amide derivative 18. This transformation was precedented in
another acid-catalyzed H–D exchange experiment of diclofenac,
and the original sodium salt can be recovered using a simple
base-catalyzed procedure.5a Significant H–D exchange was
observed in the ring annulated to the newly-formed lactam at
the positions ortho and para to nitrogen. However, little H–D
exchange was observed at the other aromatic positions.
´
11. (a) Winnicka, E.; Kanska, M. J. Radioanal. Nucl. Chem. 2009, 279, 675–678; (b)
Matthews, H. R.; Matthews, K. S.; Opella, S. J. Biochim. Biophys. Acta 1977, 497,
1–13; (c) Norton, R. S.; Bradbury, J. H. Mol. Cell. Biochem. 1976, 12, 103–111; (d)
Bak, B.; Dambmann, C.; Nicolaisen, F. Acta Chem. Scand. 1967, 21, 1674–1675.
12. Twum, E. A.; Woodman, T. J.; Wang, W.; Threadgill, M. D. Org. Biomol. Chem.
2013, 11, 6208–6214.
13. Lounasmaa, M.; Berner, M.; Tolvanen, A.; Jokela, R. Heterocycles 2000, 52, 471–
474.
14. Hachey, D. L.; Palladino, L.; Blair, J. A.; Rosenberg, I. H.; Klein, P. D. J. Labelled
Compd. Radiopharm. 1978, 14, 479–486.
15. Deuterium incorporation in all experiments was verified using 1H, 2H, and 13C
NMR and a 1,3,5-trimethoxybenzene internal NMR standard. The bracketed
numbers adjacent to each aromatic position in the figures represent deuterium
incorporation at that position, and are averaged in the case of a symmetrical
structure. Yields are given in parentheses.
Conclusion
Using deuterated trifluoroacetic acid, the rapid and efficient H–
D exchange of a wide variety of aromatic amines and amides was
achieved without the need for metal salts or other co-catalysts.
Direct H–D exchange of valuable pharmaceutically relevant com-
pounds such as acetaminophen and diclofenac was conducted,
engendering the possibility of applying this technique toward deu-
teration of other biologically active compounds. The exchange
reaction generally proceeded according to typical EAS patterns,
but was inhibited by strongly basic amines or highly deactivated
acetanilides.
16. General conditions: To the aniline or acetanilide (0.2 mmol) in a 2 dram vial
was added CF3COOD (1.0 mL, 12.98 mmol). A stir bar was added and the sealed
vial was heated to 110 °C for 16 h. The solvent was evaporated in vacuo and the
residue was stirred in 2 M KOH solution (0.5 mL) for 0.1–4.0 h. The deuterium-
labeled substrate was extracted using CH2Cl2, Et2O, or EtOAc and purified using
flash column chromatography with silica gel (hexanes/EtOAc).
17. Tice, B. B. P.; Lee, I.; Kendall, F. H. J. Am. Chem. Soc. 1963, 85, 329–337.
18. Bean, G. P.; Katritzky, A. R. J. Chem. Soc. B 1968, 864–866.