Regioselectivity in sequential nucleophilic substitution of tetrabromonaphthalene diimides

Article abstract of DOI:10.1021/jo400320a

Nucleophilic substitution of tetrabromosubstituted naphthalene diimides (NDIs) with aniline was studied in detail to explore the regioselectivity as three different diamino-substituted regioisomeric products can be formed. We found that the regioselectivity of the nucleophilic disubstitution of 2,3,6,7-tetrabromonaphthalene diimide with aniline is dependent on reaction solvents and additives. In dichloromethane and chloroform without an additive the 2,7-diamino-3,6-dibromo-NDI isomer was formed regioselectively, while in DMF under similar reaction conditions the 2,3-diamino-6,7-dibromo isomer was observed as the major regioisomer. The third possible regioisomer 2,6-diamino-3,7-dibromo product was formed, if at all, in an insignificant amount. Tetrabutylammonium fluoride (TBAF) additive exerts a dramatic effect on the regioselectivity of this reaction, as in dichloromethane without TBAF the 2,7-diamino isomer was formed regioselectively, while without TBAF the 2,3-diamino isomer was formed exclusively. This remarkable effect of TBAF can be rationalized in terms of a deprotonation of the monoamino-tribromo-NDI generated in the first step of this sequential reaction as an intermediate by fluoride ions leading to an anionic species as indicated by UV-vis and NMR experiments whose electronic properties direct the regioselective attack of the second amine molecule. Our efforts led to the exclusively regioselective synthesis of 2,7-diamino-3,6-dibromo- and 2,3-diamino-6,7-dibromo-NDIs for the first time.

Full text of DOI:10.1021/jo400320a

Article
pubs.acs.org/joc
Regioselectivity in Sequential Nucleophilic Substitution of
Tetrabromonaphthalene Diimides
Sabin-Lucian Suraru and Frank Wurthner*
̈
Institut fur Organische Chemie and Center for Nanosystems Chemistry, Universitat Wurzburg, Am Hubland, 97074 Wurzburg,
̈
̈
̈
̈
Germany
S
* Supporting Information
ABSTRACT: Nucleophilic substitution of tetrabromosubsti-
tuted naphthalene diimides (NDIs) with aniline was studied in
detail to explore the regioselectivity as three different diamino-
substituted regioisomeric products can be formed. We found
that the regioselectivity of the nucleophilic disubstitution of
2,3,6,7-tetrabromonaphthalene diimide with aniline is depend-
ent on reaction solvents and additives. In dichloromethane and
chloroform without an additive the 2,7-diamino-3,6-dibromo-
NDI isomer was formed regioselectively, while in DMF under
similar reaction conditions the 2,3-diamino-6,7-dibromo
isomer was observed as the major regioisomer. The third
possible regioisomer 2,6-diamino-3,7-dibromo product was
formed, if at all, in an insignificant amount. Tetrabutylammo-
nium fluoride (TBAF) additive exerts a dramatic effect on the regioselectivity of this reaction, as in dichloromethane without
TBAF the 2,7-diamino isomer was formed regioselectively, while without TBAF the 2,3-diamino isomer was formed exclusively.
This remarkable effect of TBAF can be rationalized in terms of a deprotonation of the monoamino-tribromo-NDI generated in
the first step of this sequential reaction as an intermediate by fluoride ions leading to an anionic species as indicated by UV−vis
and NMR experiments whose electronic properties direct the regioselective attack of the second amine molecule. Our efforts led
to the exclusively regioselective synthesis of 2,7-diamino-3,6-dibromo- and 2,3-diamino-6,7-dibromo-NDIs for the first time.
chromophores, n/p heterojunction zipper assemblies9 and
INTRODUCTION
■
photofunctional multicomponent architectures were cre-
ated.10,11 Moreover, distinctly electron-poor 2,6-core-disubsti-
tuted NDIs bearing electron-withdrawing substituents such as
chlorine or cyano groups are ambient-stable high performance
n-type semiconductors.12,13 The expansion of NDI core leads
to excellent n-type performing14,15 and ambipolar materials.16
At the core mono- and 2,6-disubstituted NDIs were also used
as building blocks for oligomers17,18 and copolymers19,20 that
exhibit electron-transporting properties.
For most of the previously reported core-disubstituted NDIs,
the 2,6-substitution pattern is prevalent since those NDIs were
derived from the respective 2,6-dichloro-4,21 or 2,6-dibromo-
naphthalene dianhydrides,5 which are accessible with high
regioisomeric purity. In contrast, the disubstitution of 2,3,6,7-
tetrabromo-NDIs (Br4-NDIs)22,23 with nucleophiles can lead to
different regioisomers. However, the regioselectivity of such
disubstitution reactions has not been elucidated so far. The
tetrabromo precursor was used to substitute all four bromine
atoms by heteroatom23,24 and carbon nucleophiles25−27
uniformly, or even to realize a lateral core enlargement of the
parent NDI core, where the regioselectivity is obviously not an
Core-substituted 1,4,5,8-naphthalene diimides (NDIs) have
been a subject of very intense research activities during the past
years.1−3 In fact, core-disubstituted NDIs have been known for
many decades,4 and those very first examples were obtained
from dichloro-1,4,5,8-naphthalene dianhydride, which was
prepared in a tedious four-step synthesis starting with the
chlorination of pyrene using chlorine gas.4,5 Therefore, it was
the direct bromination of commercially available naphthalene
dianhydride5 that led to a strong push in research activities with
core-substituted NDIs. The importance of halogen-substituted
NDIs arises from an extensive tunability of the optical and
electrochemical properties of NDIs by substitution of the
halogen atoms at the naphthalene core with many nucleophiles,
making this highly popular class of dyes available in different
colors and with easily adjustable HOMO/LUMO levels for
numerous applications. For example, the substitution of
dihalogenated NDIs with heteroatom nucleophiles at the 2-
and 6-positions leads to chromophores with excellent
absorption properties, which were used for efficient light
harvesting in supramolecular antennae systems based on zinc
chlorine-NDI conjugates,6,7 and as light-absorbing charge
transport units in self-assembled systems driving a trans-
membrane proton gradient caused by a photoinitiated electron-
transfer process.8 On the basis of such core-disubstituted NDI
Received: February 11, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 1. Reactions of NDI 1 with Aniline under Different Conditions and Derivatization of NDI 3b
issue.14,28−31 Nucleophilic disubstitution of Br4-NDIs was
previously shown with 1,2-phenylenediamines, where the
ortho-constitution of the bidentate diamine enforces the 2,3-
substitution pattern on the NDI core providing a desymmet-
rization of the compound.32 However, the situation gets more
complicated when monodentate nucleophiles are involved. In
this context, it was reported that the attempted metal-catalyzed
percyanation of a Br4-NDI proceeds via the 2,6-dibromo-3,7-
dicyano isomer. However, the regioisomeric pattern of the
latter compound was proposed, but not confirmed, on the basis
of the crystal structure of another dicyano-di(tert-butylphe-
nylthio)-NDI derivative.33 Very recently a nucleophilic
disubstitution of Br4-NDI 1 (for structure, see Scheme 1)
with n-octylamine was published by the group of Zhao, and
they claimed the formation of the respective 2,6-dibromo-3,7-
di(n-octyl)amino-NDI, which is a unique triplet photo-
sensitizer.34 However, with regard to the product character-
ization, particularly the assignment of substitution pattern,
special care has to be taken in this case since the formation of
three different regioisomeric products I−III (Figure 1) with
different symmetry elements is conceivable, and hence their
identification only by NMR spectroscopy is not necessarily
straightforward. Our studies indeed indicate that the above-
mentioned regioisomeric assignment needs to be amended
(vide infra).
Our extensive literature search did not reveal any systematic
study on the regioselectivity of nucleophilic disubstitution of
tetrabromo-NDIs. Therefore, we have approached this problem
and report herein our detailed studies on the regioselectivity of
nucleophilic disubstitution of Br4-NDI 1 with aniline as
nucleophile. We found that the regioselectivity of this reaction
is dependent on the conditions applied, and more interestingly,
the regioselectivity can be controlled by additives, in particular
fluoride anions, which influence the electronic properties of the
monoamino-substituted intermediate to regioselectively direct
the second bromine substitution by amine. Our studies afforded
simple protocols for the highly regioselective synthesis of
diamino-dibromo-substituted NDIs from tetrabrominated NDI
with monodentate aniline nucleophile.
RESULTS AND DISCUSSION
■
We were interested in the 2-fold nucleophilic substitution of
bromine atoms of 2,3,6,7-tetrabromo NDI 122 with a
monodentate amine to explore the regioselectivity of this
reaction. Therefore, Br4-NDI 1 was reacted with 2 equiv of
aniline in DMF for 1 h at 135 °C, and three main products
were isolated after column chromatography in different yields
(Scheme 1).
Among the isolated products, the monoamino-substituted
compound 2 was obtained with 29% yield, indicating that this
compound acts as an intermediate in this nucleophilic
disubstitution reaction. Furthermore, two disubstitution prod-
ucts were isolated in 28 and 38% yields, respectively, and their
constitution was confirmed by HRMS and elemental analysis.
The correct assignment of the substitution pattern for the
1
diamino product with 28% yield was achieved by H and 13C
NMR spectroscopy, and it was identified as the 2,7-diamino-
3,6-dibromo isomer 3a. The 2,7-diamino substitution pattern of
3a can be easily deduced from NMR spectra since the imide
substituents of this isomer are chemically not equivalent (see
structure I in Figure 1), and thus they should not exhibit
1
isochronic signals in the H and 13C NMR spectra, which is
indeed the case as shown in Figures 2, S16 and S17
1
(Supporting Information). The distinct feature in H NMR
Figure 1. The possible regioisomers of diamino-dibromo-substituted
NDIs with their molecular symmetries.
spectrum of 3a (Figure 2) is that, in comparison to other two
B
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 2. Sections of the 1H NMR spectra (CDCl3, 400 MHz) of the three regioisomers 3a−c showing the regions for amino NH proton, aromatic,
and imide NCH2 protons.
isomers, the signals for imide NCH2 protons are split into two
separate signals at 4.18 and 4.13 ppm. On the other hand, the
identification of the third main product with 38% yield is rather
demanding. The substitution of two bromine atoms was proven
by high-resolution mass spectrometry, elemental analysis, and
stirring at room temperature for two days, only a new spot on
the TLC was detected, and no spot for the starting compound
Br4-NDI 1 was visible, indicating its total conversion. After
purification of the crude mixture the disubstitution product 3a
was obtained with a yield of 96% (Scheme 1). More
interestingly, we found that the regioselectivity of this reaction
can be influenced decisively by using additives. Particularly, the
addition of an excess of fluoride ions in form of TBAF (>10
equiv) changes the regioselectivity completely to 2,3-isomer
(3b) compared to 2,7-isomer (3a) formed without this
additive. Thus we could establish an easy protocol for the
synthesis of 2,3-diamino isomer 3b in exclusive regioselectivity
with an isolated yield of 80% by simply stirring a mixture of Br4-
NDI 1 and aniline (90 equiv) in dichloromethane in the
presence of TBAF. It is noteworthy that the presence of
fluoride ions not only has a remarkable effect on the
regioselectivity of the reaction, but also accelerates the reaction
significantly. Thus, in the presence of TBAF the 2,3-isomer 3b
was formed exclusively already after ca. 100 min, and neither
the starting compound Br4-NDI 1 nor monobrominated
intermediate 2 could be detected by TLC, while in the absence
of TBAF intermediate 2 was the predominant component in
the reaction mixture after ca. 100 min, and the latter could still
be observed after 18 h (Figure S1, Supporting Information).
In order to get more insight into the regioselectivity of the
diamination of Br4-NDI 1 with aniline, we have performed
further experiments in several common aprotic solvents such as
chloroform, dichloromethane, toluene, THF, acetonitrile and
DMF, and the regioisomeric ratios of the disubstitution
products 3a−c formed in these solvents were determined by
1H NMR analyses of the crude products directly after the
reaction. This NMR investigation has the advantage that it
avoids unbalanced losses of compounds, as it is obviously the
case after chromatographic purification on silica, and hence
1
the ratio of aromatic and imide protons in the H NMR
spectrum, but these data cannot differentiate between the two
regioisomers 3b and 3c. The usefulness of 1H NMR
spectroscopy to distinguish between the three regioisomers
remains limited to the 2,7-isomer 3a because, in contrast to the
latter, the 2,3- and 2,6-diamino isomers 3b and 3c show only
one set of signals in 1H (Figure 2) and 13C spectra (Figure S21,
Supporting Information), since the imide substituents as well as
the amino substituents on the core are chemically equivalent
because of the 2-fold rotational C2 axes in these isomers (Figure
2, II and III in Figure 1). In order to assign the product with
38% yield to one of the possible isomers (2,3- or the 2,6-
diamino), the isolated NDI derivative was reacted with 1,2-
benzenedithiol, since only the 2,3-isomer 3b should be able to
react with 1 equiv of 1,2-benzenedithiol to substitute both
adjacent bromine atoms. Indeed, the annulated benzo[b]-
thianthrene NDI derivative 4 was obtained with 91% yield
(Scheme 1). This compound was appropriately characterized
1
by H and 13C NMR and HRMS. This derivatization reaction
confirms unequivocally the identity of the 2,3-isomer 3b in this
reaction. We also found a fourth product in trace amounts
(yield <1%) after column chromatography, which was
characterized as an additional diamino-dibromo-NDI by
HRMS and 1H NMR spectroscopy (Figure 2). Since the
other two regioisomers are clearly identified, the regioisomeric
assignment of this minor product as 2,6-diamino-3,7-dibromo
isomer 3c is obvious.
We have further performed this reaction in dichloromethane,
instead of DMF, with high excess of aniline (90 equiv). After
C
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 3. Sections of the 1H NMR spectra (CDCl3, 400 MHz) of the crude reaction products of Br4-NDI 1 with aniline (90 equiv) in the different
solvents indicated above the respective spectrum (entries 4, 6, 7, 12 in Table 1). The ratios of the formed isomers 3a−c were determined after
removal of excess aniline by integration of the signals of amine NH protons 12.34 (3b), 12.00 (3a), and 11.36 ppm (3c).
Figure 4. Sections of the 1H NMR spectra (CDCl3, 400 MHz) of the crude reaction products of Br4-NDI 1 with aniline (90 equiv) in the presence
of TBAF in different solvents indicated above the respective spectrum. The ratios of the formed isomers 3a and 3b were determined after removal of
excess aniline by integration of the NH protons signals.
minimizes systematic errors in the determination of regioiso-
meric ratios. As shown in Figure 2, the chemical shifts of the
amine NH signals in the range of 11.0−12.5 ppm of the three
isomers 3a−c are distinctly different. Therefore, the identity of
these isomers and their ratios in crude product mixtures could
be easily determined by 1H NMR spectroscopy from the
integrals of the NH proton signals at 12.34 (3b), 12.00 (3a)
and 11.36 ppm (3c) (Figures 3 and 4, Figure S2, Supporting
Information). The regioisomeric ratios obtained under different
reaction conditions are summarized in Table 1.
The data collected in Table 1 reveal that, in most of the
solvents used, the regioselectivity of the disubstitution of Br4-
D
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 1. Nucleophilic Disubstitution of Br4-NDI 1 with Aniline (90 equiv) in Different Solvents under the Given Reaction
Conditions and the Ratios of the Regioisomeric Diamino-dibromo-NDIs 3a (2,7-Diamino), 3b (2,3-Diamino), and 3c (2,6-
1
Diamino) Determined by Integration of the Amine NH Proton Integrals in H NMR Spectra Recorded in CDCl3
reaction conditions
temperature (°C)
ratio of regioisomers
a
entry
solvent
time
16 h
TBAF (equiv)
3a
3b
3c
1
CHCl3
CHCl3
CHCl3
CH2Cl2
CH2Cl2
toluene
THF
rt
−
1.5
10
−
1
0.02
0.14
0.02
2
rt
14.5 h
7 h
1
b
0.03
b
3
rt
1
b
4
rt
66 h
18 h
16 h
16 h
1.5 h
7 d
1
b
0.03
b
5
rt
16
1
b
6
rt
−
−
−
−
1
1
1
0.03
7
rt
0.1
0.03
b
8
THF
65
rt
0.02
c
c
b
b
b
9
CH3CN
CH3CN
CH3CN
DMF
1
0.3
b
10
11
12
13
rt
3 h
18
1
65
rt
1 h
−
−
−
1
1
1
0.2
2.7
2.6
14.5 h
2 h
0.01
b
DMF
65
a
The reaction times were not optimized. The reactions were monitored by TLC and terminated when monoamino-substituted NDI 2 intermediate
b
c
was nearly consumed. Not observed. Because of the low solubility of Br4-NDI 1 and intermediate 2 in acetonitrile at rt, the reaction was slow and
incomplete, and thus a superposition of the NH proton signals of 3a and residual 2 obstructs their proper integration.
Scheme 2. Reaction of Br4-NDI 1 with 2 equiv of n-Octylamine in DMF under the Same Conditions Reported in Ref 34 and
a
Subsequent Transformation of the Major Product 5 to NDI 7 with 1,2-Benzenedithiol
a
The reaction of Br4-NDI 1 with 1,2-benzenedithiol to NDI 8 is shown for comparison.
NDI 1 with aniline without TBAF additive is very high with
2,7-isomer (3a) as the major product (entries 1, 4, 6−8, 9 and
11), except for DMF in which 2,3-isomer (3b) dominates
(entries 12, 13). The regioselectivity is very similar in
chloroform, dichloromethane and toluene in the absence of
TBAF with the nearly exclusive formation of 2,7-isomer
(entries 1, 4 and 6). In THF and acetonitrile the selectivity is
not that pronounced as in above-mentioned solvents since 2,3-
isomer (3b) was formed in appreciable amounts in these
solvents. However, the 2,7-isomer (3a) is still the major
product with a 3a/3b ratio of 10:1 for THF (entry 7) and 10:3
for acetonitrile (entry 9). Interestingly, an appreciable temper-
ature effect on the regioselectivity was observed in THF, as 3a
was formed nearly exclusively at 65 °C (entry 8).
It is remarkable that even with 90 equiv of aniline, no further
substitution of diamino-dibromo-NDIs to triamino product was
observed in all these solvents, except in DMF, after longer
reaction time at room temperature. In DMF further
trisubstitution was observed. Furthermore, in DMF the
regioisomeric product distribution is changed in favor of 2,3-
isomer (3b), while in all other solvents 2,7-isomer (3a) prevails
in the absence of TBAF. In DMF, a 3a/3b ratio of about 1:2.7
was observed, independent of whether the reaction was
conducted at room temperature (entry 12) or at 65 °C
(entry 13).
E
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
We have further studied the effect of TBAF as an additive on
the regioselectivity of the nucleophilic disubstitution of Br4-
NDI 1 with aniline. The addition of a small excess of fluoride
ions (1.5 equiv of TBAF) to the reaction mixture of Br4-NDI 1
and aniline (90 equiv) in chloroform significantly increased (7-
fold) the formation of 2,3-isomer, compared to the reactions
without TBAF (entry 1), leading to a 3a/3b ratio of 10:1.4
(entry 2), whereas a large excess of fluoride ions (>10 equiv)
completely reversed the regioselectivity in favor of 3b, and the
latter isomer was formed exclusively (entry 3). Under the
reaction conditions of entry 3, not only in chloroform but also
in dichloromethane (entry 5) and acetonitrile (entry 10),
exclusive formation of 2,3-isomer (3b) was observed as
1
revealed by H NMR spectra (Figure 4).
As already mentioned in the Introduction, very recently a
similar reaction of Br4-NDI 1 with n-octylamine as a
nucleophile was reported.34 Therein the authors describe that
the 2,6-diamino-3,7-dibromo-NDI (identical with the 2,6-
dibromo-3,7-diamino-NDI as the authors named this com-
pound) was obtained with 44% yield with 2 equiv of n-
octylamine in DMF at 135 °C. However, this is not in
agreement with our observations that we have made with
aniline as a nucleophile under similar reaction conditions as
described before. Therefore, we have repeated the reaction of
Br4-NDI 1 with n-octylamine, instead of aniline, under the same
conditions described in this publication,34 and found that two
main products were formed: the 2,3-diamino-substituted
product 5, as mentioned in the publication, with a yield of
55% and a 3-fold amino-substituted and monodehalogenated
NDI 6 with a yield of 20% (Scheme 2). A comparison of the 1H
NMR spectrum of 5 (Figure S25, Supporting Information) with
that published by the authors suggests that we have isolated the
same diamino-substituted compound as reported in the earlier
publication.
Figure 5. UV−vis absorption spectra of NDI derivatives 2 (red), 3a
(blue), 3b (blue dash-dotted), 4 (violet dash-dotted), 8 (black dash-
dotted), and for comparison, Br4-NDI 1 (black) in dichloromethane
(10−5 M) at 25 °C.
higher absorption coefficient. The two diamino-substituted
regioisomers 3a and 3b show similar, but not identical
absorption bands with maxima at 564 and 563 nm, respectively.
The dithiolate annulated compound 4 displays a further
redshift with an absorption maximum at 608 nm. A similar
trend was observed for the NDIs 5−7 (Figure S3, Supporting
Information). The absorption maximum of 5 appears at 530 nm
and that of the dithiolated NDI 7 at 569 nm, while the
tetrathiolated NDI 8 exhibits a broad and less intense
absorption band with a maximum at 588 nm. The octylamino
substituents in 5 and 7 provoke blueshifts of 33 and 39 nm
compared to the absorption maximum of phenylamino
derivatives 3b and 4, respectively.
Next, we have approached the question what are the reasons
for the different regioselectivities of the nucleophilic disub-
stitution of Br4-NDI 1 with aniline in different solvents and in
the absence or presence of TBAF additive. As mentioned
before, the monoaminated NDI 2 is clearly involved as an
intermediate in this reaction (Scheme 1), and thus this
intermediate should play a crucial role in regioselectivity of
the second nucleophilic bromine substitution. Therefore,
isolated NDI 2 was subjected to further UV−vis spectroscopic
investigations to get insights into the reaction processes under
various conditions. First, we have investigated the effect of
TBAF addition to the solution of NDI 2 in dichloromethane
and acetonitrile by UV−vis spectroscopy. Figure 6 shows the
absorption behavior of NDI 2 after successive addition of up to
10 equiv of TBAF to a 4.9 × 10−5 M solution of 2 in
dichloromethane. As the spectra in Figure 6a show, the initial
absorption band at 536 nm diminishes gradually with increasing
amounts of TBAF with concomitant emergence of a new
absorption band at 674 nm, and a successive color change of
the solution from violet to green was observed. A clear
isosbestic point can be seen at 594 nm during the titration
experiment, indicating the involvement of two species at an
equilibrium. In acetonitrile a similar effect was observed since
addition of TBAF generated a green solution with a new red-
shifted absorption band with a maximum at 665 nm and an
isosbestic point at 586 nm (Figure 6b). The same effect with an
identical isosbestic point and absorption maximum was
observed by adding the non-nucleophilic base DBU, and
subsequent addition of trifluoroacetic acid (TFA) reversed this
process (Figure 6c).
1
However, as stated before, on the basis of only H NMR
analysis it is not possible to distinguish between the 2,3- and
2,6-diamino regioisomers. Therefore, in the absence of a crystal
structure, for the regioisomeric assignment of the obtained
diamino product independent method needs to be applied.
Thus, we have reacted the diamino-dibromo-NDI 5 with 1,2-
benzenedithiol and obtained the dithiolated product 7, which
was characterized by NMR spectroscopy and HRMS. This
chemical derivatization clearly indicates that the diamino
product obtained from the reaction of Br4-NDI 1 with n-
octylamine, under the identical conditions reported previously,
must be the 2,3-diamino isomer 5 with two adjacent bromine
atoms and not the claimed 2,6-diamino-isomer with the two
diagonally placed bromine atoms,34 because the latter simply
cannot form an annulated product with 1,2-benzenedithiol. For
comparison, we have reacted Br4-NDI 1 with 1,2-benzenedi-
thiol leading to the symmetrically annulated tetrathiolated
compound 8 that constitutes a derivative of the NDI recently
reported by Wang and co-workers.31
We next studied the optical properties of the isolated NDIs.
The absorption spectra of the NDIs 2, 3a and 3b in
dichloromethane are displayed in Figure 5. As reported
previously, the introduction of amino groups to the NDI core
evokes a redshift of the absorption maxima compared to that of
halogenated NDIs.23 The substitution of one bromine atom in
Br4-NDI 1 by a phenylamino group results in a broad
absorption band with a maximum at 536 nm for the
monoaminated product NDI 2. The incorporation of a second
phenylamino group induces a further bathochromic shift with a
F
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
1 (see Figure S6 and Table S1, Supporting Information). This
corresponds to a LUMO energy of −3.8 eV for NDI 2, when
4.80 eV is used as the ionization potential for ferrocene. Since
the LUMO energy for NDI 2 (−3.8 eV) is higher than the
HOMO energy of fluoride (ca. −4.3 eV),37 the reduction of
NDI 2 by fluoride and hence the formation of radical species is
not very likely. Additionally, spectroelectrochemistry experi-
ments were performed with NDI 2 in dichloromethane (Figure
S7, Supporting Information). The recorded spectrum shows a
different spectral response upon electrochemical reduction
compared to that of the UV−vis titration experiments of NDI 2
in the presence of TBAF or DBU shown in Figure 6. These
results are also not supportive of radical anion formation.
Moreover, for NDI 2 we could not observe any EPR signal in
the presence of TBAF in dichloromethane, which likewise
excludes the possibility of radical species formation.
An alternative scenario might be the formation of a σ
complex through covalent bonding of fluoride to the NDI core.
However, in such a situation the addition of DBU to NDI 2
should not give the same UV−vis absorption spectrum as in the
presence of TBAF. A more likely explanation for the formation
of the new species would be an interaction between the amino
group of NDI 2 with the fluoride anion. Because fluoride is a
strong base in aprotic solvents (HF has a pKa value of 15 in
DMSO),38 it should be able to deprotonate the phenylamino
group at the NDI core, as DBU (pKa value of 12) may also do.
Since the UV−vis titration experiments (Figure 6a) indicated
that 7 equiv of TBAF are sufficient for the full transformation of
NDI 2 to the new species, we also performed NMR
experiments under similar conditions to get more information
on structural features of the new species. Figure 7 displays the
1
changes of the aromatic region of the H NMR spectra in
CD2Cl2 after addition of 2 and 7 equiv of TBAF to NDI 2 (see
Figure S9, Supporting Information, for the complete spectrum
after addition of 7 equiv of TBAF). The disappearance of the
NH proton signal after addition of 7 equiv of TBAF (Figure S9,
Supporting Information) indicates that the amine proton is
involved in the interaction with the fluoride ions, suggesting a
deprotonation of the secondary amino group at the NDI core.
Moreover, the addition of 7 equiv of fluoride ions results in a
high-field shift of the aromatic phenyl protons displaying sharp
signals. Addition of only 2 equiv of TBAF causes moderately
high-field-shifted and broadened signals, presumably due to an
equilibrium between 2 and the new species. These observations
are in accordance with those of the UV−vis experiments. It is
remarkable that the high-field shift observed for the protons of
the phenylamino group in the presence of excess TBAF is more
pronounced for the para (0.45 ppm) and ortho protons (0.32
ppm) than that for meta protons (0.27 ppm). This difference in
high-field shift is in accordance with a higher shielding effect
induced by the higher charge density after deprotonation at the
para and ortho proton as illustrated in Figure 7 inset. Moreover,
the nitrogen-bound methylene groups at imide positions
experience also a high-field shift of ca. 0.1 and 0.45 ppm in
the presence of 2 and 7 equiv of TBAF, respectively (Figure S8,
Supporting Information). This effect is also in accordance with
an anionic species, resulting in a lower electron-withdrawing
character of the imide group. In 13C NMR spectra, particularly
the phenyl 13C signals show a similar effect as observed for the
phenyl protons in the 1H NMR spectra, i.e., a high-field shift for
the carbon atoms at ortho and para positions, and merely no
effect on the meta carbons (Figures S10−S12, Supporting
Information). A pronounced low-field shift of ca. 13 ppm was
Figure 6. Titration of NDI 2 monitored by UV−vis absorption
spectroscopy at 25 °C (a) in dichloromethane (4.9 × 10−5 M) with
TBAF × 3H2O, (b) in acetonitrile (5.4 × 10−5 M) with TBAF ×
3H2O, and (c) in acetonitrile (7.0 × 10−5 M) with DBU and
subsequent regeneration of the initial absorption band upon addition
of TFA.
On the basis of the results of these UV−vis titration
experiments, we can conclude that the new species emerged in
the presence of TBAF or DBU should be crucial for the
observed 2,3-regioselectivity in the second substitution step,
and therefore, the elucidation of this species is of significant
importance for the rationalization of the observed regioselec-
tivity. The interaction of NDI 2 with a strong Lewis-basic anion
like fluoride might, for instance, suggest the formation of radical
anions by a thermal electron transfer as proposed by Saha and
co-workers for core-unsubstituted NDIs and electron-poor 2,6-
dicyano-NDI derivative,35,36 and by Ballester et al. for the
1,4,5,8,9,12-hexaazatriphenylene (HAT) derivative HAT-
(CN)6.37 To verify this possibility, we have investigated the
redox behavior of the monoamino-substituted NDI 2 by cyclic
voltammetry (CV) with ferrocene as internal standard in
dichloromethane and estimated the LUMO energy. NDI 2
reveals a reversible reduction with a half wave potential at
−0.99 V, which is similar to the one observed for the parent
NDI (−1.10 V)5 and about 0.19 V lower than that for Br4-NDI
G
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 7. Aromatic region of 1H NMR (CD2Cl2, 600 MHz) spectrum of NDI 2 before and after addition of 2 and 7 equiv of TBAF × 3 H2O. Inset:
Resonance structures of negatively charged donor-substituted benzene illustrating high electron density at ortho and para positions.
observed for the nitrogen-bound carbon atom of phenylamino
group after addition of 7 equiv of TBAF to NDI 2 (Figure S11,
Supporting Information). Similar low-field shift in carbon
resonances was reported for diphenylamide anions.39 The
results of our NMR investigations (Figures S8−S14, Supporting
Information) corroborate the existence of an anion species [2
− H]−.
the titration experiment of diamino-substituted NDI 3a with
TBAF a color change of the solution and emergence of a new
absorption band were observed (Figure 9). The occurrence of
Moreover, 13C NMR experiments also rule out the possibility
of a covalent bond between fluorine and the naphthalene core
since no 19F−13C coupling was observed. Furthermore, the
treatment of Br4-NDI 1 and 2,7-diamino-3,6-dibromo-NDI 3a
with TBAF, respectively, provided indications for the
involvement of phenylamino group in the generation of new
species from NDI 2 in the presence of TBAF. After addition of
up to 15 equiv of TBAF to Br4-NDI 1, no changes in the UV−
vis spectra were observed, and only very high excess of fluoride
ions (>60 equiv) entails a new absorption band (Figure 8).
Since this new band could not be reversed by addition of TFA
and Br4-NDI 1 lacks an amino group at the core, a different
process than the generation of an anion species should be
responsible for the emergence of this band. In contrast, during
Figure 9. Titration of NDI 3a in dichloromethane (8.9 × 10−5 M)
with TBAF × 3H2O monitored by UV−vis absorption spectroscopy at
25 °C.
an isosbestic point suggests the existence of only one new
species formed upon addition of TBAF. Even a very large
excess of 165 equiv of TBAF does not indicate the formation of
1
a second new species. H NMR experiments with 3a in the
presence of 14 equiv of TBAF finally showed that, indeed, only
one amino group is involved in this process because the new
species formed still exhibits one NH proton signal (Figure S15,
Supporting Information). Consequently, chemically equivalent
phenyl groups become nonequivalent after reaction with TBAF
giving rise to two sets of proton signals with different high-field
shifts. This observation could be explained in terms of a single
deprotonation that obviously hampers the removal of the
second proton.
Our further UV−vis spectroscopic investigations revealed
that the absorption spectrum of the monoamino-substituted
NDI 2 in DMF is markedly different than those in chloroform,
dichloromethane and acetonitrile. As shown in Figure S4,
Figure 8. Titration of Br4-NDI 1 in dichloromethane (2.2 × 10−5 M)
with TBAF × 3H2O monitored by UV−vis absorption spectroscopy at
25 °C.
H
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 3. Proposed Mechanisms for the Regioselective Formation of 2,7-Diamino and 2,3-Diamino Isomers in the Reaction of
Br4-NDI 1 with Aniline in the Absence and Presence of TBAF × 3 H2O, Respectively
Supporting Information, in DMF an additional bathochromi-
cally shifted broad absorption maximum at 671 nm was
observed. When comparing this new absorption to that of the
aforementioned fluoride anion induced spectral changes in
other solvents, it appears to be reasonable that the same species
is generated in DMF without the additive. Since NDI 2 is
poorly soluble in DMF, we have repeated these experiments
with a better soluble analogous derivative 2′ bearing 2,6-
diisopropylphenyl groups, instead of n-octyl group as in NDI 2
in the imide positions, which shows very similar absorption
properties to those of NDI 2. A freshly prepared solution of
NDI 2′ in DMF shows an absorption maximum at 535 nm and
a red-shifted shoulder as well (Figure S5, Supporting
Information). To corroborate the hypothesis that DMF and
fluoride ions induce the formation of the same new species, we
added TBAF to the DMF solution of 2′. Indeed, this causes a
decrease of the hypsochromic band and a simultaneous
appearance of a new band with an absorption maximum at
666 nm. After addition of TFA to the former solution, the
bathochromic band disappears completely, regenerating the
initial absorption spectrum of pure NDI 2′. During this
experiment an isosbestic point at 594 nm was observed as well
(Figure S5, Supporting Information).
On the basis of what we discussed before, we propose that
two different mechanisms are possibly involved in the reaction
of Br4-NDI 1 with aniline depending on the reaction conditions
employed (Scheme 3). The 2,7-selectivity observed in pure
solvents such as chloroform and dichloromethane is in good
agreement with the simple consideration that a nucleophilic
aromatic substitution (SNAr) mechanism with a Meisenheimer-
type intermediate is involved. The regioselectivity can be
rationalized in terms of the zwitterionic Meisenheimer-type
complex generated after addition of an aniline molecule to the
monoamino-substituted intermediate NDI 2. From the three
possible regioisomeric complexes, one with the amino groups at
the 2,7-positions is favored, while the remaining two with the
amino groups at 2,6- and 2,3-positions, respectively, are
disfavored because of the repulsion of nitrogen lone pair with
the adjacent negatively charged carbon (see Scheme 3).
Therefore, the 2,7-diamino regioisomer is formed in the
absence of a base. The change of regioselectivity from the 2,7-
to the 2,3-disubstitution in the presence of TBAF additive can
be explained by the generation of an anionic species [2 − H]−
from the intermediate 2 (Scheme 3, upper part). Since a
nucleophilic attack of a second aniline molecule on this anionic
species is rather unlikely, we suggest that the reaction runs via a
highly reactive, short-lived intermediate possessing an 1H-
azirine ring,40,41 generated by intramolecular nucleophilic
substitution of the adjacent bromine atom. This is in agreement
with the exclusive 2,3-selectivity in the presence of TBAF and
also with the qualitatively observed enhanced reaction rate.
As mentioned before, the regioselectivity switch of the
nucleophilic disubstitution of Br4-NDI 1 was changed by the
solvent from 2,7-diamino isomer in chloroform, dichloro-
methane, toluene and THF to 2,3-diamino isomer in DMF.
This regioselectivity switch cannot be explained on the basis of
high polarity of DMF (εr = 37.8) because such a change was
not observed for acetonitrile, which is of similar polarity (εr =
35.9). Since the UV−vis spectral behavior of monoamino-NDI
2 intermediate is similar to that observed in other solvents
(dichloromethane, acetonitrile) in the presence of fluoride ions,
it is suggestive that a deprotonation of 2 by dimethylamine,
which is a common decomposition product of DMF present in
trace amounts, provokes the observed change in regioselectivity
in this solvent. Thus, the observed regioselectivity in DMF is
apparently not due to the inherent properties of this solvent.
CONCLUSION
■
A remarkable regioselectivity has been revealed for the
nucleophilic disubstitution of 2,3,6,7-tetrabromo NDI 1 with
aniline by varying reaction conditions. The reaction in
dichloromethane, as well as in chloroform, without any additive
leads to the regioselective formation of 2,7-diamino-3,6-
dibromo-NDI, while in the presence of TBAF as an additive
the regioselectivity is completely changed to 2,3-diamino-6,7-
dibromo-NDI as the latter is formed exclusively. Thus, simple
I
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
N,N′-Di-(n-octyl)-2,3,6-tribromo-7-phenylamino-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (2). Red solid (24.5
mg, 29%); mp 160−162 °C. 1H NMR (400 MHz, CD2Cl2): 11.90 (s,
1H), 7.40−7.28 (m, 2H), 7.23−7.15 (m, 1H), 7.07−6.97 (m, 2H),
4.22−4.10 (m, 4H), 1.78−1.66 (m, 4H), 1.49−1.21 (m, 20H), 0.93-
0.83 (m, 6H). 13C NMR (151 MHz, CD2Cl2): 164.5, 160.7, 160.6,
160.3, 150.9, 141.9, 135.0, 129.9, 129.6, 128.4, 127.3, 126.1, 125.7,
125.3, 124.2, 123.5, 122.0, 107.9, 42.7, 42.2, 32.1, 29.61, 29.59, 29.57,
28.2, 28.0, 27.47, 27.46, 23.00, 22.99, 14.22, 14.21. HRMS (ESI,
acetonitrile, neg. mode): m/z 814.0485 (M − H)−, calculated for
C36H39Br3N3O4: 814.0496. Anal. Calcd. for C36H40Br3N3O4: C, 52.83;
H, 4.93; N, 5.13. Found: C, 52.36; H, 4.96; N, 5.14. UV−vis
(CH2Cl2): λmax/nm (ε/M−1 cm−1) = 536 (12100).
protocols for the complete regioselective synthesis of these two
tetrasubstituted NDIs have been developed. The third possible
regioisomer 2,6-diamino-3,7-dibromo-NDI barely plays a role
in these reactions under the applied conditions. The clear
assignment of the regioisomers of diamino-dibromo-NDIs
brought to light that the previously reported diamino-dibromo-
substituted NDI34 should be the 2,3-diamino-6,7-dibromo
rather than the 2,6-diamino-3,7-dibromo regioisomer. The
observed compelling effect of fluoride ions (in the form of
TBAF) on the regioselectivity of nucleophilic disubstitution of
tetrabrominated NDI can be attributed to a deprotonation of
the NH proton in the monoamino-tribromo-NDI, which acts as
an intermediate in this reaction, leading to the respective
anionic species whose electronic properties supposed to change
the course of regioselectivity compared to the reaction without
TBAF. Our unique results may lead to the regioselective
synthesis of further new unsymmetrical core-substituted NDI
compounds with desired functional properties.
N,N′-Di-(n-octyl)-2,7-dibromo-3,6-diphenylamino-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (3a). Dark blue solid
1
(23.5 mg, 28%); mp 178−179 °C. H NMR (400 MHz, CDCl3):
11.98 (s, 2H, NH) 7.36−7.28 (m, 4H), 7.18−7.10 (m, 2H), 7.05−6.97
(m, 4H), 4.18 (t, 3J = 7.7 Hz, 2H), 4.13 (t, 3J = 7.7 Hz, 2H), 1.82−1.64
(m, 4H), 1.46−1.17 (m, 20H), 0.90−0.82 (m, 6H). 13C NMR (101
MHz, CD2Cl2): 165.0, 161.2, 151.2, 142.5, 129.5, 128.9, 127,4, 124.8,
121.7, 119.7, 109.0, 42.5, 41.6, 32.23, 32.21, 29.71, 29.68, 29.66, 29.6,
28.3, 28.1, 24.6, 23.1, 23.0, 14.3, 14.2. HRMS (ESI, acetonitrile/
chloroform 1:1, pos. mode): m/z 829.1962 (M + H+), calculated for
C42H47Br2N4O4: 829.1959. Anal. Calcd. for C42H46Br2N4O4: C, 60.73;
H, 5.58; N, 6.74. Found: C, 61.01; H, 5.73; N, 6.81. UV−vis
(CH2Cl2): λmax/nm = (ε/M−1 cm−1) 564 (23000).
EXPERIMENTAL SECTION
■
Materials and Methods. N,N′-Di-(n-octyl)-2,3,6,7-tetrabromo-
naphthalenetetracarboxylic acid diimide (1) and N,N′-bis-(2′,6′-
diisopropylphenyl)-2,3,6,7-tetrabromo-naphthalenetetracarboxylic
acid diimide were synthesized according to the literature proce-
dures.22,23 Aniline was distilled under argon and stored under
exclusion of light. TBAF × 3 H2O was purchased from Aldrich and
stored under exclusion of moisture. DMF was dried over P2O5 and
stored under argon and exclusion of light. THF was freshly distilled
over sodium. All other reagents and solvents were obtained from
commercial suppliers and purified and dried according to standard
procedures.42 All reactions were performed without exclusion of
oxygen or humidity. Column chromatography was performed on silica
gel (particle size 0.040−0.063 mm). A preparative recycling GPC with
2H + 2.5H columns was used. Solvents for spectroscopic studies were
N,N′-Di-(n-octyl)-2,3-dibromo-6,7-diphenylamino-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (3b). Dark violet solid
1
(32.4 mg, 38%); mp 212−214 °C. H NMR (400 MHz, CD2Cl2):
3
12.32 (s, 2H), 7.17−6.92 (m, 6H), 6.49−6.46 (m, 4H), 4.20 (t, J =
7.6 Hz, 4H), 1.80−1.71 (m, 4H), 1.49−1.24 (m, 20H), 0.88 (t, 3J = 6.7
Hz, 6H). 13C NMR (101 MHz, CD2Cl2): 165.8, 161.1, 143.3, 138.4,
128.5, 128.2, 125.2, 124.8, 123.5, 121.5, 105.2, 41.9, 32.2, 29.8, 29.7,
28.3, 27.6, 23.1, 14.3. HRMS (ESI, acetonitrile/chloroform 1:1, pos.
mode): m/z 829.1960 (M + H+), calculated for C42H47Br2N4O4:
829.1959. Anal. Calcd. for C42H46Br2N4O4: C, 60.73; H, 5.58; N, 6.74.
Found: C, 60.91; H, 5.66; N, 6.92. UV−vis (CH2Cl2): λmax/nm (ε/
M−1 cm−1) = 563 (24200).
of spectroscopic grade and used as received. H and 13C spectra were
1
recorded in CD2Cl2 or CDCl3 on a 400 or 600 MHz spectrometer.
Residual undeuterated solvent was used as internal standard (CD2Cl2:
N,N′-Di-(n-octyl)-2,6-dibromo-3,7-diphenylamino-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (3c). After further
purification by HPLC (dichloromethane/hexane 1:1), trace amounts
1
1
5.32 ppm for H, 53.80 ppm for 13C; CDCl3: 7.26 ppm for H, 77.23
for 13C). High-resolution ESI-TOF mass spectrometry was carried out
on a microTOF focus instrument. UV−vis measurements were
performed in a conventional quartz cell (light pass 10 mm) on a
standard, commercial spectrometer. For cyclic voltammetry measure-
ments, a standard commercial electrochemical analyzer with a three
electrode single-compartment cell was used. Dichloromethane (HPLC
grade) was dried over calcium hydride under argon and degassed
before using. The supporting electrolyte tetrabutylammonium
hexafluorophosphate (TBAHFP) was prepared according to the
literature43 and recrystallized from ethanol/water. The measurements
were carried out in dichloromethane at a concentration of about 10−4
M with ferrocene (Fc) as an internal standard for the calibration of the
potential. A Ag/AgCl reference electrode was used. A Pt disc and a Pt
wire were applied as working and auxiliary electrodes, respectively.
Spectroelectrochemistry measurements were performed in a specially
designed sample compartment consisting of a cylindrical quartz cell, a
platinum disc (ø = 6 mm), a gold-covered metal (V2A) plate as the
auxiliary electrode, and a Ag/AgCl pseudoreference electrode. All
spectra were recorded in reflection mode, and the optical path length
was varied by adjusting the vertical position of the working electrode
with a micrometer screw. A standard, commercial spectrometer was
used for recording UV/vis/NIR absorption spectra.
1
of this compound were obtained (<1 mg, <1%), sufficient for a H
1
NMR and HRMS analysis. H NMR (400 MHz, CD2Cl2): 11.25 (s,
2H, NH), 7.36−7.27 (m, 4H), 7.15−7.09 (m, 2H), 7.00−6.94 (m,
3
4H), 4.15 (t, J = 7.7 Hz, 4H), 1.76−1.64 (m, 4H), 1.46−1.17 (m,
20H), 0.90−0.82 (m, 6H). HRMS (ESI, acetonitrile/chloroform 1:1,
pos. mode): m/z 829.1958 (M + H+), calculated for C42H47Br2N4O4:
829.1959.
N,N′-Di-(n-octyl)-8,9-diphenylamino-6,7,10,11-benzo[b]-
thianthrenetetracarboxylic Acid Diimide (4). NDI 3b (12.0 mg,
14.4 μmol) and potassium carbonate (9.0 mg, 65 μmol) were placed in
a mixture of chloroform (5 mL) and acetone (3 mL). After addition of
1,2-benzenedithiol (10 μL, 84 μmol) the mixture was refluxed for 14 h.
The solvent was removed under reduced pressure, and the residue was
purified by column chromatography (dichloromethane/pentane 1:1)
yielding a dark blue solid (10.6 mg, 91%); mp 178−181 °C. 1H NMR
(400 MHz, CDCl3): 12.05 (s, 2H), 7.49−7.46 (m, 2H), 7.29−7.26 (m,
3
2H), 7.03−7.00 (m, 6H), 6.45−6.43 (m, 4H), 4.22 (t, J = 7.7 Hz,
3
4H), 1.82−1.73 (m, 4H), 1.49−1.22 (m, 20H), 0.89 (t, J = 6.7 Hz,
6H). 13C NMR (101 MHz, CDCl3): 165.8, 162.6, 142.3, 140.4, 138.3,
135.4, 128.8, 128.5, 128.1, 124.2, 123.0, 121.1, 121.0, 105.7, 41.5, 32.0,
29.5, 29.4, 28.2, 27.5, 22.8, 14.2. HRMS (ESI, acetonitrile/chloroform
1:1, pos. mode): m/z 811.3341 (M + H+), calculated for
C48H51N4O4S2: 811.3346. UV−vis (CH2Cl2): λmax/nm (ε/M−1
cm−1) = 608 (28900).
Reaction of N,N′-Di-(n-octyl)-2,3,6,7-tetrabromo-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (1) with Aniline.
Br4-NDI 1 (84.0 mg, 0.104 mmol) was dissolved in DMF (3 mL), and
aniline (0.20 mL of a solution in DMF; c = 97.0 g/mL; 0.21 mmol)
was added. The resulting mixture was stirred at 135 °C for 1 h. After
cooling to room temperature the solvent was removed under reduced
pressure, and the residue was purified by column chromatography
(dichloromethane/pentane 1:1) yielding the following compounds.
Reaction of N,N′-Di-(n-octyl)-2,3,6,7-tetrabromo-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (1) with n-Octyl-
amine. Br4-NDI 1 (50.0 mg, 62.0 μmol) was placed in DMF (1.5
mL). After addition of n-octylamine (100 μL of a solution in DMF,
0.16 g/mL) the mixture was stirred at 135 °C for 6 h. The DMF was
J
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
distilled off, and the residue was purified by column chromatography
(dichloromethane/pentane 2:3) yielding the following compounds.
N,N′-Di-(n-octyl)-2,3-dibromo-6,7-di(n-octyl)amino-1,4,5,8-
naphthalenetetracarboxylic Acid Diimide (5). Orange solid (30.7
Hz, 2H), 1.17−1.05 (m, 24 H). 13C NMR (101 MHz, CD2Cl2): 165.0.
161.2, 160.7, 151.4, 146.8, 146.1, 141.4, 135.5, 131.1, 130.4, 130.2,
130.1, 129.6, 129.3, 128.6, 126.9, 125.8, 125.5, 124.7, 124.6, 123.8,
122.4, 107.3, 29.9, 29.7, 24.2, 24.1, 24.02, 23.99. HRMS (ESI,
acetonitrile/chloroform, pos. mode): m/z 912.0639 (M + H+),
mg, 55%); mp 93−97 °C (Lit.34 92−95 °C). H NMR (400 MHz,
1
CDCl3): 10.48 (s, 2H), 4.17 (t, 3J = 8.0 Hz, 4H), 3.57 (bs, 4H), 1.80−
calculated for C44H41Br3N3O4: 912.0642. UV−vis (CH2Cl2): λmax
/
3
nm (ε/M−1 cm−1) = 540 (5000).
1.67 (m, 4H), 1.64−1.51 (m, 4H), 1.47−1.13 (m, 40H), 0.88 (t, J =
3
6.8 Hz, 6H), 0.84 (t, J = 7.1 Hz, 6H). HRMS (ESI, acetonitrile/
Procedures for Regioselective Synthesis of 3a and 3b in
Dichloromethane. 3a: N,N′-Di-(n-octyl)-2,3,6,7-tetrabromonaph-
thalenetetracarboxylic acid diimide (31.4 mg, 38.9 μmol) was placed
in dichloromethane (4 mL). After addition of aniline (320 μL, 3.58
mmol) the mixture was stirred at room temperature for 2 days. The
mixture was directly purified by column chromatography (dichloro-
methane/pentane 1:1) yielding 3a as a dark blue solid (31.0 mg, 96%).
3b: N,N′-Di-(n-octyl)-2,3,6,7-tetrabromo-naphthalenetetracarbox-
ylic acid diimide (32.0 mg, 39.7 μmol) and TBAF × 3 H2O (85.0
mg, 0.269 mmol) were placed in dichloromethane (4 mL). After
addition of aniline (326 μL, 3.58 mmol) the mixture was stirred at
room temperature for 2 h. The mixture was directly purified by
column chromatography (dichloromethane/pentane 1:1) yielding 3b
as a dark violet solid (26.4 mg, 80%).
Reactions of Br4-NDI 1 with Aniline for NMR Analysis. Br4-
NDI 1 was dissolved in the respective solvent, and aniline was added.
The reaction was conducted under the conditions stated in Table 1.
Volatile solvents such as CHCl3, CH2Cl2, toluene, THF and
acetonitrile were removed after the reaction under reduced pressure,
whereas in the case of DMF the reaction mixture was dissolved in
diethyl ether and washed with water. The organic phases were
collected, and diethyl ether was removed under a vacuum. The crude
mixture was subjected to GPC purification prior to NMR analysis, to
remove aniline without changing the ratio of the regioisomeric
products. If TBAF × 3 H2O was used as an additive, the fluoride salt
was washed out with water prior to GPC purification.
Entry 1: Br4-NDI 1 (89.0 mg, 0.110 mmol) and aniline (0.9 mL, 1
mmol) were stirred in chloroform (10 mL) at room temperature for
16 h.
Entry 2: Br4-NDI 1 (12.0 mg, 14.9 μmol), TBAF × 3 H2O (7.0 mg,
22.2 μmol) and aniline (122 μL 1.34 mmol) were stirred in
chloroform (4 mL) at room temperature for 14.5 h.
Entry 3: Br4-NDI 1 (8.60 mg, 10.7 μmol), TBAF × 3 H2O (35.0
mg, 111 μmol) and aniline (90 μL 0.96 mmol) were stirred in
chloroform (2 mL) at room temperature for 7 h.
Entry 4: Br4-NDI 1 (12.0 mg, 14.9 μmol) and aniline (122 μL, 1.34
mmol) were stirred in dichloromethane (4 mL) at room temperature
for 66 h
chloroform, pos. mode): m/z 901.3828 (M + H+), calculated for
C46H71Br2N4O4: 901.3827. UV−vis (CH2Cl2): λmax/nm (ε/M−1 cm−1)
= 530 (25500).
N,N′-Di-(n-octyl)-2,3,6-tri-(n-octyl)-amino-1,4,5,8-naphthale-
netetracarboxylic Acid Diimide (6). Violet solid (10.8 mg, 20%);
1
3
mp 78−80 °C. H NMR (400 MHz, CD2Cl2): 10.29 (t, J = 6.0 Hz,
1H), 9.59 (t, 3J = 5.4 Hz,1H), 9.14 (t, J = 6.0 Hz, 1H), 7.75 (s, 1H),
3
4.21−4.07 (m, 4H), 3.65−3.56 (m, 2H), 3.48−3.40 (m, 2H), 3.40−
3.32 (m, 2H), 1.83−1.73 (m, 2H), 1.74−1.62 (m, 4H), 1.60−1.30 (m,
60H), 1.93−1.78 (m, 15H). 13C NMR (101 MHz, CD2Cl2): 167.0,
166.3, 166.0, 163.7, 152.5, 150.6, 146.8, 126.6, 125.6, 122.9, 100.3,
108.8, 105.5, 102.0, 46.0, 45.7, 43.5, 40.7, 40.3, 32.20, 32.19, 32.18,
32.09, 32.08, 31.8, 31.6, 29.8, 29.72, 29.67, 29.64, 29.60, 29.59, 29.58,
29.54, 29.52, 29.50, 29.49, 28.33, 28.28, 27.6, 27.53, 27.51, 27.3, 27.2,
23.02, 23.01, 23.00, 22.96, 22.95, 14.22, 14.21, 14.18. HRMS (ESI,
acetonitrile/chloroform, pos. mode): m/z 872.6983 (M + H+),
calculated for C54H90N5O4: 872.6987. UV−vis (CH2Cl2): λmax/nm
(ε/M−1 cm−1) = 596 (23680).
N,N′-Di-(n-octyl)-8,9-di-(n-octyl)-amino-6,7,10,11-benzo[b]-
thianthrenetetracarboxylic Acid Diimide (7). NDI 5 (20.6 mg,
22.8 μmol) and potassium carbonate (14.0 mg, 101 μmol) were placed
in a mixture of chloroform (5 mL) and acetone (3 mL). After addition
of 1,2-benzenedithiol (0.02 mL, 0.17 mmol) the mixture was refluxed
for 19 h. The solvent was removed under reduced pressure, and the
residue was purified by column chromatography (dichloromethane/
1
pentane 1:1) yielding a violet wax-like substance (12.6 mg, 62%). H
3
NMR (CDCl3, 400 MHz): 10.21 (t, J = 6.0 Hz, 2H), 7.50−7.43 (m,
3
3
2H), 7.24−7.19 (m, 2H), 4.22 (t, J = 8.0 Hz, 4H), 3.55 (dt, J = 6.8
3
Hz, J = 6.6 Hz, 4H), 1.85−1.67 Hz (m, 4H), 1.60−1.10 (m, 44H),
3
3
0.89 (t, J = 7.1 Hz, 6H), 0.82 (t, J = 7.1 Hz, 6H). 13C NMR (101
MHz, CDCl3): 165.8, 162.8, 151.0, 138.6, 136.1,128.7, 128.2, 130.1,
120.6, 104.1, 46.5, 41.2, 32.0, 31.8, 31.7, 29.54, 29.45, 29.2, 28.2, 27.4,
27.0, 22.8, 22.7, 14.24, 14.17. HRMS (ESI, acetonitrile/chloroform,
pos. mode): m/z 883.5234 (M + H+), calculated for C52H75N4O4S2:
883.5224. UV−vis (CH2Cl2): λmax/nm (ε/M−1 cm−1) = 569 (23600).
N,N′-Di-(n-octyl)-6,7,14,15-tetrathianthreno[2,3-b]-
thianthrenetetracarboxylic Acid Diimide (8). Br4-NDI 1 (41.0
mg, 50.7 μmol) and 1,2-benzenedithiol (104 mg, 0.731 mmol) in
chloroform (5 mL) were refluxed for 61 h. The solvent was removed
under reduced pressure, and the residue was purified by column
chromatography (dichloromethane/pentane 1:1) affording a dark blue
solid (23 mg, 59%); mp 274 °C. 1H NMR (400 MHz, CD2Cl2): 7.44−
7.38 (m, 4H), 7.29−7.22 (m, 4H), 4.22 (t, 3J = 7.6 Hz, 4H) 1.79 (t, 3J
Entry 5: Br4-NDI 1 (11.8 mg, 14.6 μmol), TBAF × 3 H2O (76.2
mg, 242 μmol) and aniline (120 μL 1.32 mmol) were stirred in
dichloromethane (4 mL) at room temperature for 18 h.
Entry 6: Br4-NDI 1 (8.80 mg, 10.9 μmol) and aniline (90 μL, 0.99
mmol) were stirred in toluene (2 mL) at room temperature for 16 h.
Entry 7: Br4-NDI 1 (9.10 mg, 11.3 μmol) and aniline (93 μL, 1.02
mmol) were stirred in dry THF (2 mL) at room temperature for 16 h.
Entry 8: Br4-NDI 1 (9.40 mg, 11.7 μmol) and aniline (860 μL of a
1.22 M aniline solution in dry THF, 1.05 mmol) were stirred in dry
THF (2 mL) at room temperature for 1.5 h.
3
= 7.6 Hz, 4H), 1.49−1.23 (m, 20 H), 0.90 (t, J = 7.6 Hz, 6H). 13C
NMR (101 MHz, CDCl3): 165.8, 162.8, 151.0, 138.6, 136.1,128.7,
128.2, 130.1, 120.6, 104.1, 46.5,41.2, 32.0, 31.8, 31.7, 29.54, 29.45,
29.2, 28.2, 27.4, 27.0, 22.8, 22.7, 14.24, 14.17. HRMS (ESI,
acetonitrile/chloroform, neg. mode): m/z 766.2033 (M−) calculated
for C42H42N2O4S4: 766.2033. UV−vis (CH2Cl2): λmax/nm (ε/M−1
cm−1) = 583 (17000).
Entry 9: Br4-NDI 1 (10.6 mg, 13.1 μmol) and aniline (108 μL, 1.18
mmol) were stirred in acetonitrile (20 mL) at room temperature for 7
days.
Entry 10: Br4-NDI 1 (11.1 mg, 13.8 μmol), TBAF × 3 H2O (79.7
mg, 253 μmol) and aniline (113 μL, 1.23 mmol) were stirred in
acetonitrile (4 mL) at room temperature for 3 h.
Entry 11: Br4-NDI 1 (9.2 mg, 11.4 μmol) and aniline (1.3 mL of a
0.752 M aniline solution in acetonitrile, 1.34 mmol) were stirred in
acetonitrile (2 mL) at 65 °C for 1 h.
Entry 12: Br4-NDI 1 (55.5 mg, 68.8 μmol) and aniline (565 μL,
0.712 mmol) were stirred in dry DMF (5 mL) at room temperature
for 14.5 h.
Entry 13: Br4-NDI 1 (14.5 mg, 14.28 μmol) and aniline (1.2 mL of
a 1.07 M aniline solution in dry DMF) were stirred in dry DMF (2
mL) at 65 °C for 2 h.
N,N′-Bis-(2′,6′-diisopropylphenyl)-2,3,6-tribromo-7-phenyl-
amino-1,4,5,8-naphthalenetetracarboxylic Acid Diimide (2′).
N,N′-Bis-(2′,6′-diisopropylphenyl)-2,3,6,7-tetrabromo-naphthalenete-
tracarboxylic acid diimide (38.5 mg, 42.7 μmol) was placed in
chloroform (6 mL), and aniline (1.5 mL of a solution in chloroform; c
= 2.65 g/mL; 42.7 μmol) was added. The resulting mixture was stirred
at room temperature for 16 h. The solvent was removed under
reduced pressure, and the residue was purified by column
chromatography (dichloromethane/pentane 1:1) yielding a red solid
(20.3 mg, 52%); mp 321−324 °C. 1H NMR (400 MHz, CDCl3): 12.0
(s, 1H), 7.47−7.38 (m, 2H), 7.31−7.22 (m, 6H), 7.14−7.08 (m, 1H),
3
3
6.98−6.92 (m, 2H), 2.65 (sept, J = 6.8 Hz, 2H), 2.58 (sept, J = 6.8
K
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(21) Wurthner, F.; Ahmed, S.; Thalacker, C.; Debaerdemaeker, T.
̈
ASSOCIATED CONTENT
* Supporting Information
Figures S1−S15 and Table S1 mentioned in the text, and
additional NMR spectra (Figures S16−S35) and ESI-HRMS
spectra (Figures S36−S45). This material is available free of
charge via the Internet at http://pubs.acs.org.
■
Chem.Eur. J. 2002, 8, 4742−4750.
S
(22) Gao, X.; Qiu, W.; Yang, X.; Liu, Y.; Wang, Y.; Zhang, H.; Qi, T.;
Liu, Y.; Lu, K.; Du, C.; Shuai, Z.; Yu, G.; Zhu, D. Org. Lett. 2007, 9,
3917−3920.
(23) Roger, C.; Wurthner, F. J. Org. Chem. 2007, 72, 8070−8075.
̈
̈
(24) Fin, A.; Petkova, I.; Doval, D. A.; Sakai, N.; Vauthey, E.; Matile,
S. Org. Biomol. Chem. 2011, 9, 8246−8252.
(25) Suraru, S.-L.; Wurthner, F. Synthesis 2009, 11, 1841−1845.
̈
AUTHOR INFORMATION
Corresponding Author
*Fax: +49 (0)931 31 84756. Tel: +49 (0)931 31 85340. E-mail:
wuerthner@chemie.uni-wuerzburg.de.
■
(26) Ye, Q.; Chang, J.; Huang, K.-W.; Chi, C. Org. Lett. 2011, 13,
5960−5963.
(27) Kruger, H.; Janietz, S.; Sainova, D.; Dobreva, D.; Koch, N.;
̈
Vollmer, A. Adv. Funct. Mater. 2007, 17, 3715−3723.
(28) Hu, Y.; Gao, X.; Di, C.; Yang, X.; Zhang, F.; Liu, Y.; Li, H.; Zhu,
D. Chem. Mater. 2011, 23, 1204−1215.
Notes
The authors declare no competing financial interest.
(29) Yue, W.; Gao, J.; Li, Y.; Jiang, W.; Di Motta, S.; Negri, F.; Wang,
Z. J. Am. Chem. Soc. 2011, 133, 18054−18057.
ACKNOWLEDGMENTS
■
(30) Cai, K.; Yan, Q.; Zhao, D. Chem. Sci. 2012, 3, 3175−3182.
(31) Li, C.; Xiao, C.; Li, Y.; Wang, Z. Org. Lett. 2013, 15, 682−685.
(32) Bhosale, S. V.; Bhosale, S. V.; Kalyankar, M. B.; Langford, S. J.
Org. Lett. 2009, 11, 5418−5421.
Generous financial support by the Bavarian State Ministry of
Science, Research, and the Arts for the Center for Nanosystems
Chemistry in the framework of the Collaborative Research
Network “Solar Technologies Go Hybrid” is gratefully
acknowledged.
(33) Chang, J.; Ye, Q.; Huang, K.-W.; Zhang, J.; Chen, Z.-K.; Wu, J.;
Chi, C. Org. Lett. 2012, 14, 2964−2967.
(34) Guo, S.; Wu, W.; Guo, H.; Zhao, J. J. Org. Chem. 2012, 77,
3933−3943.
REFERENCES
■
(35) Guha, S.; Saha, S. J. Am. Chem. Soc. 2010, 132, 17674−17677.
(36) Guha, S.; Goodson, F. S.; Corson, L. J.; Saha, S. J. Am. Chem.
Soc. 2012, 134, 13679−13691.
(1) Bhosale, S. V.; Jani, C. H.; Langford, S. J. Chem. Soc. Rev. 2008,
37, 331−342.
(2) Sakai, N.; Mareda, J.; Vauthey, E.; Matile, S. Chem. Commun.
2010, 46, 4225−4237.
(37) Aragay, G.; Frontera, A.; Lloveras, V.; Vidal-Gancedo, J.;
Ballester, P. J. Am. Chem. Soc. 2013, 135, 2620−2627.
(38) Taft, R. W.; Bordwell, F. G. Acc. Chem. Res. 1988, 21, 463−469.
(39) Kennedy, A. R.; Klett, J.; O’Hara, C. T.; Mulvey, R. E.;
Robertson, G. M. Eur. J. Inorg. Chem. 2009, 2009, 5029−5035.
(40) Gilchrist, T. L.; Gymer, G. E.; Rees, C. W. J. Chem. Soc., Chem.
Commun. 1973, 835−836.
(3) Bhosale, S. V.; Bhosale, S. V.; Bhargava, S. K. Org. Biomol. Chem.
2012, 10, 6455−6468.
(4) Vollmann, H.; Becker, H.; Corell, M.; Streeck, H. Liebigs Ann.
1937, 531, 1−159.
(5) Thalacker, C.; Roger, C.; Wurthner, F. J. Org. Chem. 2006, 71,
̈
̈
8098−8105.
(41) Gilchrist, T. L.; Gymer, G. E.; Rees, C. W. J. Chem. Soc., Perkin
Trans. 1975, 1, 1−8.
(42) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals, 2nd ed.; Pergamon Press, Ltd.: Oxford, 1980.
(43) Fry, A. J. In Laboratory Techniques in Electroanalytical Chemistry,
2nd ed.; Kissinger, P., Heineman, W. R., Eds.; Marcel Dekker, Ltd.:
New York, 1996.
(6) Roger, C.; Muller, M. G.; Lysetska, M.; Miloslavina, Y.;
Holzwarth, A. R.; Wurthner, F. J. Am. Chem. Soc. 2006, 128, 6542−
̈
̈
̈
6543.
(7) Roger, C.; Miloslavina, Y.; Brunner, D.; Holzwarth, A. R.;
̈
Wurthner, F. J. Am. Chem. Soc. 2008, 130, 5929−5939.
̈
(8) Bhosale, S.; Sisson, A. L.; Talukdar, P.; Furstenberg, A.; Banerji,
̈
N.; Vauthey, E.; Bollot, G.; Mareda, J.; Roger, C.; Wurthner, F.; Sakai,
̈
̈
N.; Matile, S. Science 2006, 313, 84−86.
(9) Sakai, N.; Bhosale, R.; Emery, D.; Mareda, J.; Matile, S. J. Am.
Chem. Soc. 2010, 132, 6923−6925.
(10) Sakai, N.; Lista, M.; Kel, O.; Sakurai, S.-i.; Emery, D.; Mareda, J.;
Vauthey, E.; Matile, S. J. Am. Chem. Soc. 2011, 133, 15224−15227.
(11) Lista, M.; Areephong, J.; Sakai, N.; Matile, S. J. Am. Chem. Soc.
2011, 133, 15228−15231.
(12) Wurthner, F.; Stolte, M. Chem. Commun. 2011, 47, 5109−5115.
̈
(13) Zhan, X.; Facchetti, A.; Barlow, S.; Marks, T. J.; Ratner, M. A.;
Wasielewski, M. R.; Marder, S. R. Adv. Mater. 2011, 23, 268−284.
(14) Gao, X.; Di, C.; Hu, Y.; Yang, X.; Fan, H.; Zhang, F.; Liu, Y.; Li,
H.; Zhu, D. J. Am. Chem. Soc. 2010, 132, 3697−3699.
(15) Zhao, Y.; Di, C.; Gao, X.; Hu, Y.; Guo, Y.; Zhang, L.; Liu, Y.;
Wang, J.; Hu, W.; Zhu, D. Adv. Mater. 2011, 23, 2448−2453.
(16) Suraru, S.-L.; Zschieschang, U.; Klauk, H.; Wurthner, F. Chem.
̈
Commun. 2011, 47, 11504−11506.
(17) Polander, L. E.; Romanov, A. S.; Barlow, S.; Hwang, D. K.;
Kippelen, B.; Timofeeva, T. V.; Marder, S. R. Org. Lett. 2012, 14, 918−
921.
(18) Polander, L. E.; Tiwari, S. P.; Pandey, L.; Seifried, B. M.; Zhang,
Q.; Barlow, S.; Risko, C.; Bredas, J.-L.; Kippelen, B.; Marder, S. R.
́
Chem. Mater. 2011, 23, 3408−3410.
(19) Yan, H.; Chen, Z.; Zheng, Y.; Newman, C.; Quinn, J. R.; Dotz,
F.; Kastler, M.; Facchetti, A. Nature 2009, 457, 679−686.
(20) Schubert, M.; Dolfen, D.; Frisch, J.; Roland, S.; Steyrleuthner,
R.; Stiller, B.; Chen, Z.; Scherf, U.; Koch, N.; Facchetti, A.; Neher, D.
Adv. Energy Mater. 2012, 2, 369−380.
L
dx.doi.org/10.1021/jo400320a | J. Org. Chem. XXXX, XXX, XXX−XXX