On the Reaction of Naphthalene Diimides with Fluoride Ions: Acid/Base versus Redox Reactions

Article abstract of DOI:10.1002/anie.201705223

The characterization of a naphthalene diimide (NDI) radical anion is presented. Its properties allow a reliable comparison point for other systems involving the NDI radical anion, such as systems claimed to perform the oxidation of the fluoride anion. In addition to reiterating the obvious thermodynamic objections to such an unlikely oxidation, we present several observations that show that the fluoride anion does not act as a one-electron reducing agent toward the NDI investigated in this work.

Full text of DOI:10.1002/anie.201705223

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: On the reaction of Naphthalene Diimides with Fluoride ions: Acid/
Base vs Redox reactions
Authors: Guillaume Bélanger-Chabot, Ahmed Ali, and Francois Pierre
Gabbai
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201705223
Angew. Chem. 10.1002/ange.201705223
Link to VoR: http://dx.doi.org/10.1002/anie.201705223
http://dx.doi.org/10.1002/ange.201705223

Angewandte Chemie International Edition
10.1002/anie.201705223
COMMUNICATION
On the reaction of Naphthalene Diimides with Fluoride ions:
Acid/Base vs Redox reactions
[a]
Guillaume Bélanger-Chabot, Ahmed Ali and François P. Gabbaï*
th
Dedicated to Karl Christe on the occasion of his 80 birthday
Abstract: The characterization of a napthalenediimide (NDI) radical
anion is presented. Its properties allow a reliable comparison point
O
O
for other systems involving the NDI radical anion, such as systems
iPr
iPr
claimed to perform the oxidation of the fluoride anion. In addition to
R N
N R
O
1 : R=
reiterating the obvious thermodynamic objections to such an unlikely
oxidation, we present several observations that show that the
O
fluoride anion does not act as a one-electron reducing agent toward
the NDI investigated in this work.
Figure 1. Left: Generic naphthalene diimide (NDI). Right: N-substituent used
in this study.
Owing to their unique redox and optical properties, napthalene
diimides (NDIs, Figure 1) have been employed in numerous
-
-
[
1]
[2]
applications including anion transport and organic redox and
F + NDI
F
NDI
[
3]
photosystems, and organic electronics where electron-poor n-
-
-
[
4]
type semiconductors are desired.
Owing to their electron
F
NDI
F +
[NDI ]
deficient nature and their demonstrated ability to engage in
[
5]
anion-π interactions, NDIs have also found applications in the
[
6]
Scheme 1. Steps typically invoked to describe the oxidation of fluoride by NDI
domain of anion sensing.
An intriguing aspect in the anion-
derivatives
sensing chemistry of these species concerns the nature of the
response displayed toward the fluoride anion. Indeed, it was
proposed that the interaction of NDI with fluoride anions is not
limited to the formation of a complex but rather involves an
electron transfer process in which fluoride is oxidized to atomic
[
7]
fluorine and NDI reduced to the corresponding radical anion.
The use of fluoride as a "reducing agent" has also been
extended to various other systems ranging from electron-
[
8]
deficient organic compounds to silver cations. While the
formation of the organic radical anion has been confirmed by
[
7]
EPR and by UV-Vis spectroscopy, the possibility of an electron
transfer from the fluoride anion to the organic moiety seems out
of reach at least on thermodynamic grounds. To the best of our
[
9]
knowledge, few articles directly criticizes these claims,
although some have noted that the oxidation of fluoride is an
• -
2 2 2
Figure 2. ORTEP plot of the solid state structure of [CoCp ][1 ] CH Cl with
[
10]
unlikely mechanism.
Given that the oxidation of fluoride is a
ellipsoids drawn at the 50% level. Only one orientation of the disordered
cobaltocenium moiety is displayed. Hydrogen atoms were omitted for clarity.
highly endergonic (and endothermic) process, we have decided
to compare the ability of the fluoride anion to reduce NDI with
2
that of cobaltocene (Cp Co), a well know one-electron molecular
reductant.
To work in a solvent that can be rigorously dried, we decided
to
propylphenyl)napthalenediimide
soluble in CH Cl . The reaction of 1 with an equimolar amount
Co in CH Cl
cobaltocenium salt. Examination of the structure shows that the
synthesize
the
known
N'N'-di-(2,6-di-iso-
[
11]
(DippNDI) (1), which is
2
2
[
12]
• -
Cp
2
2
2
afforded [1 ] as a deep purple, air-sensitive
•
-
cobaltocenium and molecules of [1 ] are well separated (Figure
[
13]
[a]
Dr. G. Bélanger-Chabot, A. Ali and Prof. Dr. F. P. Gabbaï
Department of Chemistry
Texas A&M University
College Station, TX, USA
E-mail: francois@tamu.edu
2).
2 2
The UV-visible and EPR spectra (Figure 3) of CH Cl
solutions of the isolated material are essentially identical to the
ones published for similar electrochemically generated NDI
[
12]
•
radical anions.
characterized NDI radical anion salts
stabilized in a zwitterionic form using phosphonium-substituted
2
[Cp Co][1 ] is one of a few structurally
[14]
which can also be
Supporting information for this article (complete experimental,
computational and crystallographic details and further remarks) is
given via a link at the end of the document
[
15]
NDIs as demonstrated by the work of Mukhopadhyay.
•
[
Cp
2
Co][1 ] is a very sensitive compound which cannot be
dissolved and diluted without some level of decomposition. For
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201705223
COMMUNICATION
•
-
this reason, [1 ] was generated by addition of an excess of
CoCp
, allowing us to measure an extinction coefficient (ε698) of
,500 mol cm for the band at 698 nm Figure 3. This band,
which has no interference with those arising from Cp Co and
Co] , can be used as a reliable indicator of the concentration
vis and EPR spectroscopy, arguing against the correlation
•
-
1
[7a,7b]
2
between the concentration in [1 ] and H NMR signals.
It is
-1
-1
4
interesting to note that the reaction with TBAF is accompanied
by the appearance of unidentified aromatic and aliphatic
resonances, signaling a small amount of decomposition (see
Figure 4).
2
+
[
Cp
2
•
-
of [1 ] . Based on the value of ε698 determined above, we
observed that the dissolution of a 1:1 solution of 1 and Cp Co to
a concentration of 2 x 10 M produces the radical anion in 78%
yield, indicating some degree of decomposition caused by
adventitious solvent impurities. When the same experiment was
reproduced by diluting a 1:1 solution of 1 and TBAF a as a
2
1
9
.
-
4
The F NMR resonance of TBAF 3H
2 2 2
O in CD Cl never
disappeared upon reaction with one equivalent of NDI, in sharp
[
7a,7b]
3
contrast with the published results in DMSO and CD CN.
This disappearance is usually interpreted as the tell-tale sign of
the formation of paramagnetic fluorine radicals in TBAF/NDI
-
3
fluoride source to 2 x 10 M, only traces of the radical could be
detected in 4-5% yield. These results alone show that, unlike
1
systems. The fact that the H NMR NDI-core proton signal did
19
vanish while the F signal of fluoride but did not in TBAF/1
solutions shows that NMR evidence is an unreliable tool to
assess the presence of paramagnetic species in these systems.
Cp
This conclusion is supported by EPR measurements carried out
on solutions containing CoCp /1 and TBAF/1 in CH Cl . While
2
Co, fluoride does not act as a one electron reducing agent.
2
2
2
1
9
We attempted to reproduce the disappearance of the F signals
in DMSO and noticed no major change in the total concentration
the expected 13-line signal is observed in both cases, the
normalized double integrals of the respective EPR signals show
that the concentration of [1 ] in samples generated with Cp
-
•
-
of F , thereby arguing against its oxidation to a reactive fluorine
2
Co is
radical (Figure 4). The only notable effect is a change in the
1
6 times higher than those obtained with TBAF.
-
-
2
speciation of the F with respect to HF . The amount of bifluoride
increases more acutely in the case of DMSO, which possibly
results from the increased acidity of DMSO when compared to
2 2
CH Cl . This view is supported by the appearance of a triplet
20
18
16
14
12
10
8
6
4
2
0
EPR
19
-
1
+ 5.6 CoCp2
(1:1:1) in the F NMR spectrum corresponding to DF
TBAF is dissolved in DMSO, as previously explained.
when
The
2
[
16]
-
-
formation of DF
2
shows that
F
is sufficiently basic to
deprotonate DMSO or at the very least promote H/D exchange.
Field / G
Previous reports on the oxidation of fluoride by NDI have
[7d]
3
340
3350
superficially rejected
acid-base phenomena as the
explanation for the observations on the TBAF/NDI system,
despite the well-known basic character of the fluoride anion,
1 + CoCp2
[
9a,17]
ε
even with respect other NDI derivatives.
Invoking the
[18]
1 + TBAF
500
elegant work of Matile, we will go on to suggest that NDIs, due
to their electron-deficient nature, may in fact acidify the
solvent, further promoting its deprotonation by fluoride anions.
400
600
700
800
[19]
λ / nm
-
This is supported by the observation that the quantity of HF
2
increases when NDI is added to TBAF solutions, especially in
DMSO. We also note that this interpretation is consistent with
earlier work by Würthner who pointed out that the fluoride anion
is unlikely to reduce amino-substituted NDIs while being capable
Figure 3. Comparison of normalized UV-Vis absorbances for CH
solutions of CoCp /1 1:1 (0.2 mM 1) (gray hashed), 5.6:1 (0.1 mM 1)(solid
black) and TBAF/NDI 1:1 (2 mM 1) (gray dotted). Inset: EPR spectrum of
isolated [CoCp ][1] in CH Cl (0.5 mM). 50 KHz modulation frequency, 0.1 G
2 2
Cl
2
2
2
2
modulation amplitude, 0.6 mW. The signal is centered at ca g=2.004. The 13-
[9a]
of promoting the deprotonation of pendent amino groups.
line has been attributed to coupling to four equivalent protons and two
14
[12]
equivalent N atoms.
To complement these experimental results, we have used
computational methods to probe the energetic parameters that
-
govern the hypothetical reduction of NDI by F (Scheme 2).
We next decided to investigate the system TBAF/1 using
These calculations suggest that the proposed electron transfer is
endothermic by at least 76 kcal/mol. The highly endothermic
character of these reactions supports the obvious conclusion
that, fluoride, as the anion of the most electronegative element,
is not a reducing agent. We have also used computational
methods to investigate possible complexes formed between 1
.
2 2 2
NMR spectroscopy. The addition of TBAF 3H O to 1 in CH Cl ,
even in substoichiometric amounts, resulted in the
1
disappearance of the H NMR NDI-core proton signals, similarly
[
7a,7b]
to published results in DMSO and CD
3
CN.
All other
resonances undergo a distinct broadening but remain detectable.
It has been previously argued that these changes are indicative
of the formation of the NDI radical anion. However, the presence
of signals for the non-core hydrogen atoms is incompatible with
-
-
and F . On the basis of previous calculations on related NDI/ F
[
7a]
complexes, we first optimized a structure with a fluoride anion
residing above a carbonyl carbon atom in the gas phase.
However, no minimum could be identified when the SMD-PCM
solvation model was applied to the gas-phase structure, the
optimization leading the fluoride anion to drift away from the NDI
core.
2
this interpretation since the addition of Cp Co to solutions of 1 in
CH Cl leads to the complete disappearance of all identifiable
2
2
signals, as would be expected upon formation of a radical.
Moreover, the disappearance of the core hydrogen atoms signal
•
-
occurs in solutions assessed as containing traces of [1 ] by UV-
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201705223
COMMUNICATION
iPr
O
N
O
O
N
O
iPr
iPr
H1
a
H4
H5
H2
CH2
H3
H3
Me
H1
H5
H4
H2
/
/
/
b
c
8
x
/
8.6
8.2
7.8
7.4
ppm
2.6
2.2
1.8
1.4
1.0 ppm
4%
e
f
TBAF in DMSO-d6
TBAF in CD Cl2
2
87%
13%
96%
d
1
+ TBAF in DMSO-d6
1 + TBAF in CD Cl2
2
45%
55%
83%
17%
-110
-120
-130
-140 ppm
-120
-130
-140
-150 ppm
1
.
.
Figure 4.Top: H NMR spectra of 1 in CD
2
Cl
2
without (a) and with TBAF 3H
2
O (b). The signals denoted by ** belong to TBAF 3H
2
O and those denoted by * to
19
.
19
.
unidentified side-products. Bottom: F NMR spectra of TBAF 3H
2
O in wet DMSO-d
6
without (c) and with 1 (d); F NMR spectra of TBAF 3H
2
O in wet CD
2
Cl
2
without (e) and with 1 (f).
-
+
-
literature originates from similar events. Indeed, it is interesting
to note that most of these investigations are carried out in
solvents such as DMSO. DMSO is acidic and prone to
1
F
1
+ F
]
[
∆
Hrxn
=
7
6 kcal/mol
[
16]
deprotonation by bases such as fluoride anions.
There is
Scheme 2. Enthalpy change for the oxidation of fluoride by 1 calculated at the
unrestricted B3LYP/6-311+G(d,p)-SMD-PCM level (solvent = CH Cl ).
precedence in the literature for the involvement of deprotonated
2
2
[21]
DMF and DMSO in redox processes. We therefore propose
that it is the deprotonated solvent which acts as a reducing
agent to produce the NDI radicals, an interpretation consistent
From the very beginning, Occam's razor would suggest that
an extremely thermodynamically unfavorable redox event is
unlikely the correct explanation for the observed reduction of
NDIs in the presence of fluoride anions. Our results show that
fluoride anions are not one-electron reducing agents. Indeed,
with the anion-induced nature of this effect as previously
[9a,10b,17,22]
advanced by a few researchers.
We conclude this
paper by stating the obvious: the fluoride anion does not act as a
reducing agent towards NDIs.
•
-
2 2
when 1 is mixed with TBAF in CH Cl , the radical [1 ] is formed
in very limited amounts. It is in fact surprising that any radical is
detected in these conditions. However, our NMR results show
that the addition of TBAF to 1 leads to a complex system and is
accompanied by side reactions. We assign the presence of
these unidentified products to the strongly basic character of
fluoride (pKa =15 in DMSO) and propose that one or some of
these trace side-products act as a reducing agent. We suspect
that the NDI-promoted “fluoride oxidation” described in the
Acknowledgements
We gratefully acknowledge the Welch Foundation (Grant A-
1423) and Texas A&M University (Arthur E. Martell Chair of
[
20]
Chemistry) for funding this research. The authors would also
like to thank Drs Vladimir Bakhmoutov, Lisa Perez, Joseph
Reibenspies, Hamidreza Samouei, Daniel Tofan and Gregory
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201705223
COMMUNICATION
[
[
7
[
20] F. G. Bordwell, Acc. Chem. Res. 1988, 21, 456-463.
21] C. L. Øpstad, T.-B. Melø, H.-R. Sliwka, V. Partali, Tetrahedron 2009, 65,
616-7619.
22] A. Lafleur-Lambert, J.-B. Giguère, J.-F. Morin, Macromolecules 2015,
8, 8376-8381.
Wylie, for their input. We would like to thank the Texas A&M
Laboratory for Molecular Simulation.
4
Keywords: Napthalenediimide (NDI) • Anion Sensing• Fluoride
Oxidation• Paramagnetic Species • Radicals
References
[
3
4
1]
a) S. V. Bhosale, C. H. Jani, S. J. Langford, Chem. Soc. Rev. 2008, 37,
31-342; b) N. Sakai, J. Mareda, E. Vauthey, S. Matile, Chem. Commun. 2010,
6, 4225-4237; c) M. A. Kobaisi, S. V. Bhosale, K. Latham, A. M. Raynor, S. V.
Bhosale, Chem. Rev. 2016, 116, 11685-11796.
2] A. Vargas Jentzsch, A. Hennig, J. Mareda, S. Matile, Acc. Chem. Res.
013, 46, 2791-2800.
3] a) A. C. Fahrenbach, C. J. Bruns, H. Li, A. Trabolsi, A. Coskun, J. F.
[
2
[
Stoddart, Acc. Chem. Res. 2014, 47, 482-493; b) A. Das, S. Ghosh, Chem.
Commun. 2016, 52, 6860-6872; c) P. Spenst, F. Würthner, J. Photochem.
Photobiol., C 2017, 31, 114-138.
[4]
a) X. Zhan, A. Facchetti, S. Barlow, T. J. Marks, M. A. Ratner, M. R.
Wasielewski, S. R. Marder, Adv. Mater. 2011, 23, 268-284; b) S.-L. Suraru, F.
Würthner, Angew. Chem. Int. Ed. 2014, 53, 7428-7448; c) W. Jiang, Y. Li, Z.
Wang, Acc. Chem. Res. 2014, 47, 3135-3147.
[5]
a) B. L. Schottel, H. T. Chifotides, K. R. Dunbar, Chem. Soc. Rev. 2008,
37, 68-83; b) R. E. Dawson, A. Hennig, D. P. Weimann, D. Emery, V.
Ravikumar, J. Montenegro, T. Takeuchi, S. Gabutti, M. Mayor, J. Mareda, C. A.
Schalley, S. Matile, Nat Chem 2010, 2, 533-538; c) M. M. Watt, M. S. Collins,
D. W. Johnson, Acc. Chem. Res. 2013, 46, 955-966; d) P. Ballester, Acc.
Chem. Res. 2013, 46, 874-884.
[6]
a) M. R. Ajayakumar, P. Mukhopadhyay, S. Yadav, S. Ghosh, Org. Lett.
2010, 12, 2646-2649; b) M. R. Ajayakumar, D. Asthana, P. Mukhopadhyay,
Org. Lett. 2012, 14, 4822-4825; c) J. Zhao, G. Li, C. Wang, W. Chen, S. C. J.
Loo, Q. Zhang, RSC Advances 2013, 3, 9653-9657; d) A. Mitra, R. J. Clark, C.
T. Hubley, S. Saha, Supramol. Chem. 2014, 26, 296-301; e) Y. Zhao, Y.
Cotelle, N. Sakai, S. Matile, J. Am. Chem. Soc. 2016, 138, 4270-4277; f) M.
Giese, M. Albrecht, K. Rissanen, Chem. Commun. 2016, 52, 1778-1795.
[7]
a) S. Guha, S. Saha, J. Am. Chem. Soc. 2010, 132, 17674-17677; b) S.
Guha, F. S. Goodson, S. Roy, L. J. Corson, C. A. Gravenmier, S. Saha, J. Am.
Chem. Soc. 2011, 133, 15256-15259; c) S. Guha, F. S. Goodson, R. J. Clark,
S. Saha, CrystEngComm 2012, 14, 1213-1215; d) S. Guha, F. S. Goodson, L.
J. Corson, S. Saha, J. Am. Chem. Soc. 2012, 134, 13679-13691; e) Y. Li, Y.
Zhao, R. Jiang, H. Liu, Y. Li, Inorganic Chemistry Frontiers 2014, 1, 661-667;
f) A. Rananaware, M. Samanta, R. S. Bhosale, M. A. Kobaisi, B. Roy, V.
Bheemireddy, S. V. Bhosale, S. Bandyopadhyay, S. V. Bhosale, Scientific
Reports 2016, 6, 22928; g) N. A. Young, S. C. Drew, S. Maniam, S. J.
Langford, Chem. Asian J., DOI: 10.1002/asia.201700459.
[8]
a) F. S. Goodson, D. K. Panda, S. Ray, A. Mitra, S. Guha, S. Saha, Org.
Biomol. Chem. 2013, 11, 4797-4803; b) G. Aragay, A. Frontera, V. Lloveras, J.
Vidal-Gancedo, P. Ballester, J. Am. Chem. Soc. 2013, 135, 2620-2627; c) K.
Maity, D. K. Panda, E. Lochner, S. Saha, J. Am. Chem. Soc. 2015, 137, 2812-
2
[
815.
9]
Li, X. Pang, Y. Wang, Y. Che, J. Zhao, Catal. Today 2014, 224, 258-262.
10] a) M. T. Stephenson, H. J. Shine, J. Org. Chem. 1981, 46, 3139-3141;
a) S. L. Suraru, F. Würthner, J. Org. Chem. 2013, 78, 5227-5238; b) J.
[
b) Y. Hu, Z. Wang, X. Yang, Z. Zhao, W. Han, W. Yuan, H. Li, X. Gao, D. Zhu,
Tetrahedron Lett. 2013, 54, 2271-2273.
[
11] L. H. Tong, P. Pengo, W. Clegg, J. P. Lowe, P. R. Raithby, J. K. M.
Sanders, S. I. Pascu, Dalton Trans. 2011, 40, 10833-10842.
12] G. Andric, J. F. Boas, A. M. Bond, G. D. Fallon, K. P. Ghiggino, C. F.
[
Hogan, J. A. Hutchison, M. A. Lee, S. J. Langford, J. R. Pilbrow, G. J. Troup,
C. P. Woodward, Aust. J. Chem. 2004, 57, 1011-1019.
[13] CCDC 1480105 ( [Cp2Co][1]-CH2Cl2 ), contain the supplementary
crystallographic data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[
14] a) A. Mizuno, Y. Shuku, R. Suizu, M. M. Matsushita, M. Tsuchiizu, D.
Reta Mañeru, F. Illas, V. Robert, K. Awaga, J. Am. Chem. Soc. 2015, 137,
612-7615; b) Y. Wu, M. D. Krzyaniak, J. F. Stoddart, M. R. Wasielewski, J.
Am. Chem. Soc. 2017, 139, 2948-2951.
15] S. Kumar, M. R. Ajayakumar, G. Hundal, P. Mukhopadhyay, J. Am.
Chem. Soc. 2014, 136, 12004-12010.
7
[
[
[
16] H. Sun, S. G. DiMagno, J. Am. Chem. Soc. 2005, 127, 2050-2051.
17] M. R. Ajayakumar, G. Hundal, P. Mukhopadhyay, Chem. Commun.
2
013, 49, 7684-7686.
18] Y. Zhao, C. Beuchat, Y. Domoto, J. Gajewy, A. Wilson, J. Mareda, N.
Sakai, S. Matile, J. Am. Chem. Soc. 2014, 136, 2101-2111.
19] T. Lu, S. E. Wheeler, Org. Lett. 2014, 16, 3268-3271.
[
[
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201705223
COMMUNICATION
COMMUNICATION
-
No
Guillaume Bélanger-Chabot,
Ahmed Ali and François P.
Gabbaï*
Back to Occam's Razor: Can
Naphthalene Diimides (NDIs) Really
Oxidize the Fluoride Anion? A study of
the reaction of N,N'-di-(2,6-di-iso-
propylphenyl)napthalenediimide with
Cp Co and TBAF indicates that the
2
fluoride anion does not act as a
reducing agent towards NDIs.
F
F
[
a]
e-
e-
O
R N
O
O
Page No. – Page No.
N R
On the reaction of Naphthalene
Diimides with Fluoride ions:
Acid/Base vs Redox reactions
O
+
Cp Co
Cp Co
2
2
Yes
This article is protected by copyright. All rights reserved.