Modular Synthesis of pH-Sensitive Fluorescent Diaza[4]helicenes

Article abstract of DOI:10.1002/ejoc.201402863

Configurationally stable diaza[4]helicenes have been prepared in two steps by using a particularly facile N-N bond-cleavage reaction. The synthetic procedure uses hydrazine (NH₂NH₂) for the introduction of a single nitrogen atom. The strategy is general, modular and highly tolerant to functional groups. A mechanistic rationale is proposed for the spontaneous N-N bond-cleavage reaction. The resulting helical quinacridines are dyes that present absorption and emission properties that can be modulated as a function of pH; the pink quinacridine and green protonated forms (pK _a ≈ 9.0) display distinct optical features in the near-IR region. Single enantiomers were obtained by chiral stationary phase HPLC resolution. The absolute configurations were assigned by comparison of the ECD spectra of the conjugated acids with those of known dialkylquinacridinium derivatives. A rather high racemization barrier was measured by means of variable-temperature ECD experiments (ΔG ^? = 30.7 ± 4.0 kcal mol^-1 at 140 C). Configurationally stable diaza[4]helicenes have been prepared by a particularly facile N-N bond-cleavage reaction using hydrazine for the introduction of a single nitrogen atom. The helical quinacridines are dyes showing absorption and emission properties in the near-IR region that depend on pH. A high racemization barrier of 30.7 ± 4.0 kcal/mol was determined for the single enantiomers.

Full text of DOI:10.1002/ejoc.201402863

FULL PAPER
DOI: 10.1002/ejoc.201402863
Modular Synthesis of pH-Sensitive Fluorescent Diaza[4]helicenes
Antoine Wallabregue,[a] Petr Sherin,[b] Joyram Guin,[a] Céline Besnard,[c] Eric Vauthey,*[b]
and Jérôme Lacour*[a]
Keywords: Fluorescence / pH Sensing / Helicenes / Chiral resolution / Reaction mechanisms
Configurationally stable diaza[4]helicenes have been pre- pink quinacridine and green protonated forms (pKa ≈ 9.0)
pared in two steps by using a particularly facile N–N bond-
cleavage reaction. The synthetic procedure uses hydrazine
(NH2NH2) for the introduction of a single nitrogen atom. The
strategy is general, modular and highly tolerant to functional
display distinct optical features in the near-IR region. Single
enantiomers were obtained by chiral stationary phase HPLC
resolution. The absolute configurations were assigned by
comparison of the ECD spectra of the conjugated acids
groups. A mechanistic rationale is proposed for the spontane- with those of known dialkylquinacridinium derivatives.
ous N–N bond-cleavage reaction. The resulting helical quin-
acridines are dyes that present absorption and emission
properties that can be modulated as a function of pH; the
A rather high racemization barrier was measured by means
of variable-temperature ECD experiments (ΔG‡
30.7Ϯ4.0 kcalmol–1 at 140 °C).
=
phores. The neutral 1 and protonated 1·H+ conjugates (pKa
≈ 9.0) present well-separated absorption and fluorescent
signatures allowing facile distinction of the entities.
Furthermore, the P and M enantiomers were separated and
they present a rather large barrier to racemization (ΔG‡ Ͼ
30.7Ϯ4.0 kcalmol–1 at 140 °C).
Introduction
Azahelicenes are helical derivatives made of ortho-fused
aromatic rings containing at least one nitrogen atom. Like
their all-carbon analogues,[1] they possess distinctive twisted
conformations that result from steric repulsions between
the terminal aromatic rings or their substituents. Syntheses
are mostly based on photochemical,[2] transition-metal-
catalysed trimerization[3] and cross-coupling strategies.[2b,4]
Azahelicenes are utilized for a variety of applications due
to the presence of both an inherently chiral geometry and
nitrogen donor atom(s). In fact, studies of azahelicenes have
been reported in many different fields,[2j,5] including solid-
state self-assembly,[2g,6] asymmetric catalysis,[4i,7] chemosen-
sors[5h,8] and enantioselective biomolecular recognition. In
Figure 1. Colourful 1,13-dimethoxyquinacridine 1 (pink) and its
terms of optical properties, these moieties usually absorb conjugated acid 1·H+ (green). The M enantiomers are shown arbi-
trarily (R = alkyl, aryl, benzyl, allyl, NH2).
light in the UV spectral domain and, to the best of our
knowledge, there are few reports of fluorescent aza-
helicenes.[4c,9] Herein, in a new development made possible
by the use of hydrazine as nucleophile and a subsequent
surprisingly facile N–N bond-cleavage reaction, we report
Results and Discussion
Permanently charged cationic dialkyldiaza[4]helicenes of
type 2 have previously been studied[10] (Scheme 1) for their
interesting stereochemical, biological, supramolecular and
photophysical properties.[11] Compounds of type 2 are pre-
pared by treatment of the tris(2,6-dimethoxyphenyl)methyl
cation (3) with primary amines. The products of ring clo-
sures are obtained by sequences of nucleophilic aromatic
substitutions (SNAr) of MeO substituents by nitrogen nu-
cleophiles. Although acridinium cations 4 are usually af-
forded at 20 °C, elevated temperatures (90–120 °C) are nec-
essary to form quinacridinium ions 2. Recently, Takui, Mo-
rita and co-workers relied on an analogous three-step pro-
tocol to form cation 5 containing unsubstituted nitrogen
the general synthesis of configurationally stable [4]helical
quinacridines 1 (Figure 1; R = alkyl, aryl, benzyl, allyl,
NH2). These derivatives are pH-sensitive dyes and fluoro-
[a] Department of Organic Chemistry, University of Geneva,
Quai Ernest Ansermet 30, 1211 Genève 4, Switzerland
E-mail: jerome.lacour@unige.ch
http://www.unige.ch/sciences/chiorg/lacour/home
[b] Department of Physical Chemistry, University of Geneva,
Quai Ernest Ansermet 30, 1211 Genève 4, Switzerland
E-mail: eric.vauthey@unige.ch
http://www.unige.ch/sciences/chifi/Vauthey/
[c] Laboratory of Crystallography, University of Geneva,
Quai Ernest Ansermet 24, 1211 Genève 4, Switzerland
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201402863.
Eur. J. Org. Chem. 2014, 6431–6438
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6431

E. Vauthey, J. Lacour et al.
FULL PAPER
atoms.[10a] First, compound 6 was prepared by using N,N- Table 1. Optimization of the reaction conditions for the direct syn-
thesis of quinacridine 1a.
dimethylethylenediamine as nucleophile and a precursor
cation similar to 3. Then the aminoalkyl side-chains were
removed through
a Hoffmann elimination sequence
(Scheme 2).
Entry Solvent T [°C] NH2NH2 [equiv.] Time [h] Yield [%][a]
1
2
3
4
NMP
NMP
DMSO 110
DMF
DMF
DMF
DMF
110
110
25
25
25
25
25
25
10
8
14
14
14
14
8
50
85
80
85
83
92
70
Scheme 1. Synthesis of permanently charged quinacridinium salts
2 from acridinium derivatives 4. Reagents and conditions: a) R-
NH2 (2.5 equiv.), N-methylpyrrolidone (NMP), 20–50 °C, 30 min;
b) RЈ-NH2 (25 equiv.), 110 °C, NMP, 1.5 h.
110
90
90
5
6[b]
7
90
8
[a] Isolated yields. [c] Reaction performed in red-tainted glassware.
tolerated either as substituents on the side-chains (entries 6–
8) or bound directly to the acridine nitrogen atom (en-
try 9).[16] Only in the case of 4k containing an allyl moiety
was the reaction unsuccessful, the allyl group being system-
atically reduced to a propyl chain during the reaction to
form 1a in good yield (90%) instead of 1k.[17] This result
will be discussed later.
Scheme 2. Reagents and conditions: a) 1. MeI, MeOH, 110 °C;
2. tBuOK, DMF, 20 °C; 3. aq. HBF4, 20 °C, 100%; b) 1. aq. HBF4,
MeOH, 70 °C; 2. Aq. Na2CO3, CH2Cl2, 20 °C, 94%.
Despite the efficiency of this method, an alternative one-
step protocol was sought for the introduction of “naked”
nitrogen atoms. Ammonia and sulfamic acid were tested
without success by using cation 4a (R = nPr) as sub-
strate.[12] Hydrazine (NH2NH2) was then considered as a
substitute due to its availability, reasonable cost, lower vola-
tility and higher nucleophilicity.[13] The often-difficult cleav-
age of the N–N bond[14] was a problem not to be underesti-
mated, but it was decided to consider it only later. The reac-
tion between acridinium 4a and NH2NH2 (64–65% solu-
tion in water, 25 equiv. NMP, 110 °C) was thus attempted.
To our satisfaction, the reaction occurred as planned and
the desired quinacridine 1a was obtained directly (Table 1,
entry 1, 50% yield), the cleavage of the N–N bond occur-
ring seemingly during the formation of the helical adduct.
Care was taken to ensure that this facile N–N bond
cleavage was a general observation and the reaction condi-
tions were optimized accordingly. The results are summa-
rized in Table 1. With NMP as solvent, a longer reaction
time was beneficial (85% yield, entry 2). DMSO and DMF
performed equally well, however, as the reactions were eas-
ier to purify in DMF, this solvent was further studied.[15]
The yields were improved by performing the reaction at
90 °C and in red-tinted glassware to limit exposure to light
(92%, entry 6). A decrease in the amount of hydrazine,
however, was detrimental (Table 1, entry 7).
Table 2. Scope of the reaction (first approach).
Entry Substrate
R
Product Yield [%][a]
1
2
3
4
5
6
7
8
9
4b
4c
4d
4e
4f
4g
4h
4i
methyl
hexyl
hexadecyl
phenyl
benzyl
CH2CH2NMe2
CH2CH2OH
CH2CH2OCH2CH2OH
NH2
1b
1c
1d
1e
1f
1g
1h
1i
97
82
80
94
78
89
85
88
85
0[b]
4j
4k
1j
1k
10
allyl
[a] Isolated yields. [b] Instead of the desired product 1k, adduct 1a
(R = nPr) was obtained in good yield (90%).
Product 1e was found to be moderately soluble in a 4:1
mixture of hexane and dichloromethane. Single crystals
With the optimized reaction conditions in hand (Table 1, were obtained that were analysed by X-ray diffraction. Cat-
entry 6), a series of acridinium ions 4 were then tested as ion 1e adopts a twisted helical conformation typical of this
substrates. The results are summarized in Table 2. In short, type of [4]helicene.[11e,11f] No particular structural modifica-
the reaction is general and other alkyl, aryl and benzyl side- tion due to the presence of the “naked” sp2 N atom (N2,
chains are compatible (entries 1–5). Heteroatoms are also Figure 2) was noticed.
6432
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 6431–6438

Modular Synthesis of pH-Sensitive Fluorescent Diaza[4]helicenes
duction of the side-chain was found. Functional group tol-
erance was further confirmed by the reaction of [4j][BF4]
with regular hydrazine (entry 6). Amino-substituted quin-
acridine 1j (85%) was obtained without traces of a second
deamination reaction. Ethanolamine also reacted with
[4j][BF4] to afford the expected product 1h (Table 3, en-
try 5).
Table 3. Scope of the reaction (second approach).
Figure 2. Ortep view of the crystal structure of 1e (one molecule
of CH2Cl2 has been omitted). The thermal ellipsoids are drawn at
the 50% probability level.
Despite this positive result, a second (one-step) approach
was considered to shorten the synthetic strategy, obtain
some mechanistic information and to be able to work with
unsaturated side-chains (e.g., allyl). This approach used salt
[4j][BF4] (Table 2, entry 9) as a single precursor for all deriv-
atives of type 1. As outlined in Scheme 3, it was considered
that 4j ought to react with primary amines to afford quinac-
ridinium salt 7 or, ideally, quinacridines 1 as final products
if the spontaneous N–N bond cleavage was to occur again.
Entry
R
Product
Yield [%][a]
1
2
3
4
5
6
7
propyl
methyl
benzyl
CH2CH2NMe2
CH2CH2OH
NH2
1a
1b
1f
90
91
88
84
83
85
60
1g
1h
1j
allyl
1k
[a] Isolated yields.
This reactivity is explained if the existence of quin-
acridinium salts as intermediates is considered
7
(Scheme 4). In fact, the addition of hydrazine to “regular”
acridinium salts (first approach) or primary amines to 4j
(second approach) generates in both instances derivatives 7.
The difference in the two approaches is only the nature of
the nucleophilic reactant in excess in the crude mixtures.
Herein it is proposed that the remaining nucleophiles attack
Scheme 3. Second convergent synthetic approach to quinacridines.
To our satisfaction, the salt [4j][BF4] reacted with n-prop- the exocyclic amino group of 7 and generate the corre-
yl- and methylamine (25 equiv.) at 90 °C to afford directly sponding quinacridine derivatives. Different byproducts are
quinacridine 1a and 1b, respectively, in excellent yields (90 then generated. In the first approach,
a
triazane
and 91%, Table 3, entries 1 and 2). The reactions with (NH2NHNH2) is obtained that decomposes to generate di-
benzyl-, 2-(N,NЈ-dimethylamino)ethyl- and allylamines were imide and ammoniac,[19] whereas in the second strategy alk-
then attempted knowing that competing oxidation, elimi- ylhydrazines are produced that do not evolve further. If an
nation and hydrogenation reactions might occur with such unsaturated side-chain is present in the molecule, then the
amines.[18] Satisfactorily, adducts 1f, 1g and 1k were ob- second approach is better as it avoids the undesired re-
tained in yields of 88, 84 and 60%, respectively. This time, duction of the double (triple) bond by the in situ generated
in the reaction with allylamine, no evidence for the re- HN=NH.
Scheme 4. Mechanistic rationale for the synthesis of quinacridines 1.
Eur. J. Org. Chem. 2014, 6431–6438
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6433

E. Vauthey, J. Lacour et al.
FULL PAPER
To establish some evidence for this mechanism, a com- IB® column correspond to (+)-(P)- and (–)-(M)-1f, respec-
pound of type 7 was prepared and isolated by treatment of tively. Interestingly, strong differences are observed in the
acridinium [4j][BF4] with only 5 equiv. of n-propylamine, ECD spectra of enantiopure 1f and [1f·H+][BF4]; the dex-
the reaction being monitored by ESI-MS (Scheme 5). Com- tro- and levorotatory series are compared in Figure S2 in
pound 7a was isolated in 85% yield and then treated with the Supporting Information.
an excess (20 equiv.) of both hydrazine and propylamine to
give in both cases the expected product 1a (90 and 80%
yields, respectively).
Scheme 5. Reagents and conditions: a) n-propylamine (5 equiv.),
DMF, 90 °C, Ar, 8 h, without light, 85%; b) H2NNH2·H2O or n-
propylamine (20 equiv.), DMF, 90 °C, Ar, 3 h, without light, 100%
conversion, 90 and 80% yields, respectively.
With these results in hand, the resolution of one of the
helical quinacridines was pursued. In a first set of experi-
ments, racemic 1a (R = Pr) was treated with common
enantiopure acids (e.g., tartaric acid, dibenzoyltartaric acid,
mandelic acid).[3d] However, conditions that would induce
a selective precipitation of a diastereomeric salt were not
found. A semi-preparative chromatographic resolution of
Figure 3. Top: ECD spectra of (P)-[2a][BF4] (red solid line) and
rac-1f (R = Benzyl) on a chiral stationary phase was then
(M)-[2a][BF4] (blue dashed line). Bottom: ECD spectra of (+)-
attempted.[20] Successful conditions were found by using a
Chiralpak IB® column and a mixture of n-hexane and 2-
propanol (60:40) as eluent with 0.1% of ethanolamine as
additive. Starting with 25 mg of rac-1f and after several
runs (see the Supporting Information), 10.0 mg (ee Ͼ99%,
[1f·H+][BF4] (red solid line) and (–)-[1f·H+] (blue dashed line) de-
rived from (+)- and (–)-[1f], respectively. Solutions in CH3CN
(20 °C, 5ϫ10–6 m).
Finally, an attempt was made to determine the racemiza-
40% yield) and 8.0 mg (ee 99%, 32% yield) were afforded tion barrier by ECD, monitoring a single wavelength every
in the first and second eluted fractions, which correspond 5 s (357 nm, see the Supporting Information). It was neces-
to the dextro- and levorotatory enantiomers of 1f, respec- sary to heat dibutyl sulfoxide solutions of (+)-(P)-1f at
tively ([α]365 = +16500 and –16000, respectively, CH3CN, 130 °C to start observing a decrease in the Cotton effect.
10–4 m). The electronic circular dichroism (ECD) spectra After 1000 s, a loss of only 20% enantiomeric purity was
(250–700 nm) of the two fractions are displayed in Fig- observed at this temperature. Samples were then heated at
ure S1 in the Supporting Information).
140 and 150 °C and analysed for the same period of time.
To establish the absolute configurations of these sepa- At 160 °C, the start of the decomposition was observed.
rated enantiomers, (+)-1f and (–)-1f were transformed into Measurements were thus not taken at this or higher tem-
their conjugated acids, namely (+)-[1f·H+][BF4] and (–)- peratures. Kinetic constants were calculated and activation
[1f·H+][BF4].[21] Their ECD spectra were then compared parameters determined (Ea, A, ΔH‡, ΔS‡ and ΔG‡, see the
with that of classical enantiopure cations of type 2. In fact, Supporting Information). The racemization of 1f (ΔG‡ =
it was considered that the protonation of quinacridines 1 30.7Ϯ4.0 kcalmol–1 at 140 °C) occurs more quickly than
ought to generate species with chiroptical properties similar that of [2a][BF4], for which a rather high barrier was mea-
to those of permanently charged quinacridiniums 2, as was sured (ΔG‡ = 41.3 kcalmol–1 at 200 °C).[10f] Clearly, the ab-
indeed the case, as seen in Figure 3. The ECD spectra of sence of an alkyl substituent on a nitrogen atom gives a
[(P)-2a][BF4] and [(M)-2a][BF4] (R = RЈ = Pr) are presented large degree of flexibility. The very high positive value for
in the top panel, in which the configurations were unambig- the entropy tends to indicate that this transformation in-
uously assigned by VCD spectroscopy and X-ray crystal- volves a strong involvement of solvent and/or an ion-pair-
lography in the presence of a chiral auxiliary of known con- ing reorganization. The racemization barrier for 1f remains,
figuration.[10e,10f] These spectra are virtually superposable however, slightly higher than that reported for analogous
on those of (+)-[1f·H+][BF4] and (–)-[1f·H+][BF4] displayed derivative 5 (ΔG‡ = 28.4 kcalmol–1, 28 °C).[10a]
in the bottom panel.[22] As such, it is safe to consider that
the first and second fractions eluted from the Chiralpak and characterize the optical properties.
With these compounds in hand, care was taken to study
6434
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 6431–6438

Modular Synthesis of pH-Sensitive Fluorescent Diaza[4]helicenes
Photophysical Properties
in aqueous solution.[11d] Moreover, the quinacridine form
that predominates at high pH has a larger aromaticity than
quinacridinium and is neutral. Consequently, it can be ex-
pected to be less hydrophilic and more prone to the forma-
tion of large aggregates.
Absorption, pH titration and fluorescence measurements
were performed on quinacridines 1b and 1j (R = Me and
NH2, respectively) in aqueous solution containing 0.1%
DMSO. Protonation strongly affects the electronic structure
of the quinacridines and, thus, leads to significant changes
in their absorption spectra, as illustrated in Figure 4. For
both 1b and 1j, the lowest-energy absorption band shifts
from 560 to 625 nm and a new band appears at 420 nm.
The presence of the hydrazino substituent in 1j has very
little effect on the absorption spectrum of either its neutral
or cationic forms, which points to a minor involvement of
the exocyclic NH2 group in the lowest optical transitions.
Consequently, a possible acid/base reaction of this group
should have only a minor impact on the electronic absorp-
tion spectrum.
Figure 5. Absorption and emission spectra of (A) 1b and (B) 1j in
aqueous solutions at various pH.
The quinacridines 1b and 1j fluoresce weakly above
600 nm with quantum yields, Φf, of around 5ϫ10–3
(Table 4). Upon protonation, their emission bands shift to
longer wavelengths with peaks at around 700 nm. However,
protonation has a negligible effect on the fluorescence
quantum yield, as illustrated in Table 4. The fluorescence
spectrum and quantum yield of 1j with the protonated NH2
substituent are essentially the same as those of the unpro-
tonated form, as could be expected from the quasi-identical
absorption spectra.
Figure 4. Electronic absorption spectra and titration curves for
aqueous solutions of 1b (A, C) and 1j (B, D). The solid lines in C
and D are the best fits of Equations (S1) and (S2), respectively.
Table 4. Fluorescence properties of 1b and 1j.
Product
pH
Φf [10–3][a]
τf1 [ns][b]
τf2 [ns][b]
The pH-dependence of the absorbance at 620 nm (1b)
and 625 nm (1j) is depicted in parts C and D of Figure 4
(1j). These titration curves were analysed assuming one and
two acid/base equilibria for 1b and 1j, respectively [see
Equations (S1) and (S2) in the Supporting Information],
and the best fits were obtained with pK1 = 8.9 for 1b·H+
and pK1 = 6.9 for 1j·H2+ and pK2 = 9.1 for 1j·H+. The
similarity of pK2 for 1j·H+ and pK1 for 1b·H+ points to an
average pKa value of about 9.0 for the equilibrium between
1b
4.4
12
4.4
7.2
12
6
4
4
3
4
0.45 (0.81)
0.48
0.92 (0.19)
–
–
–
–
1j
0.35
0.35
0.35
[a] Fluorescence quantum yield (error: Ϯ2ϫ10–3). [b] An ad-
ditional decay component with a time constant of Ͼ4 ns and a
relative amplitude of Ͻ0.01 has been omitted.
The fluorescence decays of the various forms of 1b and
quinacridinium and quinacridine conjugates. On the other 1j recorded at the maximum emission wavelength by time-
hand, the pK1 value found for 1j can be associated with the correlated single photon counting (TCSPC) upon excitation
NH2 substituent and, as anticipated above, this equilibrium at 395 nm[23] are depicted in Figure 6. Although marked
has a minor effect on the absorption spectrum.
differences can be seen for the acid and basic forms of 1b
The steady-state absorption and emission spectra of the (Figure 6, A), the three forms of 1j exhibit very similar
quinacridine and quinacridinium forms of 1b and 1j are fluorescence dynamics. These time profiles could be well re-
shown in Figure 5. The weak feature above 700 nm in the produced by the sum of two (1j) or three exponential func-
absorption spectrum of 1b at high pH is a result of the tions (1b) convolved with the instrument response function;
scattering from small particles in suspension, most probably the resulting time constants are listed in Table 4. A compo-
aggregates. Indeed, [4]heliceniums of type 2 have been nent with a time constant greater than 4 ns and a relative
shown to form dimeric aggregates at modest concentrations amplitude smaller than 0.01 was attributed to impurities
Eur. J. Org. Chem. 2014, 6431–6438
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6435

E. Vauthey, J. Lacour et al.
FULL PAPER
present in very low concentrations (Ͻ0.1%) and will not be
further considered.
Conclusions
A series of pH-sensitive [4]helical quinacridines 1 have
been synthesized in two steps from a classical carbenium
precursor using hydrazine (NH2NH2) for the introduction
of a single nitrogen atom. This synthetic procedure is gene-
ral, modular and highly tolerant to functional groups. A
mechanistic rationale has been proposed for the facile N–
N bond cleavage that seemingly occurs during the forma-
tion of the product.
Single enantiomers of the [4]helicenes were obtained by
chiral stationary phase HPLC resolution and assignment of
the absolute configuration was realized through a perfect
comparison of the ECD spectra of the conjugated acids
with those of known quinacridinium derivatives. Moreover,
a high racemization barrier (ΔG‡ = 30.7Ϯ4.0 kcalmol–1 at
140 °C) was measured. This racemization barrier for 1f is
lower than that of [2a][BF4], for which a rather high barrier
(ΔG‡ = 41.3 kcalmol–1 at 200 °C) was measured. However,
it remains slightly higher than that reported for the analo-
gous derivative 5 (ΔG‡ = 28.4 kcalmol–1, 28 °C).
These helical quinacridines are effective dyes that present
interesting absorption and emission properties that can be
modulated as a function of pH. Despite a weak fluores-
cence and relatively modest changes in the emissive proper-
ties upon protonation, the large difference in the absorption
spectrum could be advantageously used to sense hydro-
phobic domains of proteins and other biological systems.
Figure 6. Fluorescence time profiles measured for (A) 1b and (B) 1j
in water at various pH, instrument response function (IRF) and
best fits (black).
The structures of the quinacridiniums 1·H+ are very sim-
ilar to those of [4]heliceniums 2 investigated previously,[11d]
which, because of the presence of an alkyl substituent on
both N atoms, are not pH sensitive. Therefore a similar in-
terpretation of the biphasic nature of the fluorescence decay
of 1b can be proposed, that is, the faster decay component
is attributed to 1b and the other is assigned to an aggregate.
This explanation is supported by further TCSPC measure-
ments on a solution of 1b diluted by a factor two: the rela-
tive amplitude of the slower component was found to de-
crease from 0.19 to 0.11.
In the case of 1j, the fluorescence of all three forms de-
cays exponentially with the same time constant. No aggre-
gate formation could be detected under either neutral or
basic conditions, which can be explained by the presence of
the polar and hydrophilic amino group, which effectively
counteracts hydrophobic effects.
Experimental Section
General: Experimental procedures for the synthesis of the targeted
[4]helical quinacridines 1a–k are reported below. Characterization
data and further protocols can be found in the Supporting Infor-
mation.
General Procedures for the Synthesis of Cationic Quinacridines
1a–k
Approach 1: Hydrazine monohydrate (7.3 mL, 150 mmol, 25 equiv.,
64–65% in water) was added to a solution of the corresponding
acridinium salt 4 (6 mmol) in anhydrous DMF (30 mL) in red-
tinted glassware. The reaction mixture was degassed with argon
and then heated at 90 °C without light for 10 h. After complete
consumption of the starting acridinium salt 4 (monitored by ESI-
MS analysis), the purple reaction mixture was cooled to 20 °C. The
crude product precipitated upon addition of a solution of 50%
HBF4 in water (ca. 80 mL). The solid was filtered, washed several
times with water and then dissolved in CH2Cl2. Selective precipi-
tation by addition of diethyl ether afforded the title green com-
pounds.
Compared with the non-pH-sensitive [4]helicenes 2,[11d]
the fluorescence quantum yields and lifetimes of 1b and 1j
are smaller by a factor of around 4. The excited-state dy-
namics of 2 were found to be dominated by a non-radiative
decay to the ground state, the efficiency of which increased
with the hydrogen-bond-donating property of the sol-
vent.[11d] Such hydrogen-bond-assisted non-radiative deacti-
vation has been shown to be an efficient quenching mecha-
nism for a large number of organic molecules in aqueous
solutions.[24] The faster non-radiative decay found here for
1 could be explained by the presence of the unprotected
nitrogen atom that most likely forms hydrogen bonds with
water molecules. Comparing 1b and 1j (Table 4), the some-
what lower values of Φf and τf1 for 1j speak in favour of a
higher efficiency of this non-radiative process because of
the presence of the hydrophilic amino group.
Approach 2: The corresponding primary amine (15.68 mmol,
25 equiv.) was added to a solution of acridinium salt 4j (300 mg,
627 μmol) in anhydrous DMF (3 mL) in red-tinted glassware. The
reaction mixture was degassed with argon and then heated at 90 °C
without light for 14 h. After complete consumption of starting ac-
ridinium salt 4j (monitored by ESI-MS analysis), the purple reac-
tion mixture was cooled to 20 °C. The crude product precipitated
upon addition of a solution of 50% HBF4 in water (ca. 10 mL).
The solid was filtered, washed several times with water and then
dissolved in CH2Cl2. Selective precipitation by addition of diethyl
ether afforded the title green compounds.
6436
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 6431–6438

Modular Synthesis of pH-Sensitive Fluorescent Diaza[4]helicenes
[4] a) M. Weimar, R. Correa da Costa, F.-H. Lee, M. J. Fuchter,
Org. Lett. 2013, 15, 1706–1709; b) W. Lin, G.-L. Dou, M.-H.
Hu, C.-P. Cao, Z.-B. Huang, D.-Q. Shi, Org. Lett. 2013, 15,
1238–1241; c) E. Kaneko, Y. Matsumoto, K. Kamikawa, Chem.
Eur. J. 2013, 19, 11837–11841; d) D. Waghray, J. Zhang, J. Ja-
cobs, W. Nulens, N. Basaric´, L. V. Meervelt, W. Dehaen, J. Org.
Chem. 2012, 77, 10176–10183; e) H. R. Talele, S. Sahoo, A. V.
Bedekar, Org. Lett. 2012, 14, 3166–3169; f) H. Kelgtermans, L.
Dobrzanska, L. Van Meervelt, W. Dehaen, Org. Lett. 2012, 14,
1500–1503; g) O. Crespo, B. Eguillor, M. A. Esteruelas, I. Fer-
nandez, J. Garcia-Raboso, M. Gomez-Gallego, M. Martin-Or-
tiz, M. Olivan, M. A. Sierra, Chem. Commun. 2012, 48, 5328–
5330; h) M. R. Crittall, H. S. Rzepa, D. R. Carbery, Org. Lett.
2011, 13, 1250–1253; i) N. Takenaka, R. S. Sarangthem, B.
Captain, Angew. Chem. Int. Ed. 2008, 47, 9708–9710; Angew.
Chem. 2008, 120, 9854.
Optical Resolution of Neutral [4]Helical Quinacridine 1f: The two
enantiomers of racemic azahelicene 1f differed significantly in their
HPLC retention times on a Chiralcel IB column. Separation was
therefore possible under semi-preparative conditions using
a
1ϫ25 cm column. Racemic diazahelicene 1f (25 mg) was resolved
by repeated HPLC separation using an Agilent 1100 Series instru-
ment (hexane/2-propanol 6:4 + 0.1% ethanolamine, 0.6 mL/min
flow rate, and 1.5 mg injections of the racemate in 900 μL of elu-
ent). Evaporation of the solvents and trituration with diethyl ether
afforded (+)-1f (10 mg, Ͼ99% ee) and (–)-1f (8 mg, 99% ee) as
pink amorphous solids.
CCDC-1000626 (for 1e) contains the supplementary crystallo-
graphic 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.
[5] a) F. Torricelli, J. Bosson, C. Besnard, M. Chekini, T. Bürgi, J.
Lacour, Angew. Chem. Int. Ed. 2013, 52, 1796–1800; Angew.
Chem. 2013, 125, 1840; b) Y. Si, G. Yang, J. Mater. Chem. C
2013, 1, 2354–2361; c) Y. Nakai, T. Mori, K. Sato, Y. Inoue, J.
Phys. Chem. A 2013, 117, 5082–5092; d) R. Adam, R. Balles-
teros-Garrido, O. Vallcorba, B. Abarca, R. Ballesteros, F. R.
Leroux, F. Colobert, J. M. Amigó, J. Rius, Tetrahedron Lett.
2013, 54, 4316–4319; e) H. R. Talele, A. V. Bedekar, Org. Bio-
mol. Chem. 2012, 10, 8579–8582; f) K. Goto, R. Yamaguchi,
S. Hiroto, H. Ueno, T. Kawai, H. Shinokubo, Angew. Chem.
Int. Ed. 2012, 51, 10333–10336; Angew. Chem. 2012, 124,
10479; g) J. Elm, J. Lykkebo, T. J. Sørensen, B. W. Laursen,
K. V. Mikkelsen, J. Phys. Chem. A 2012, 116, 8744–8752; h) T.
Caronna, F. Castiglione, A. Famulari, F. Fontana, L. Malpezzi,
A. Mele, D. Mendola, I. N. Sora, Molecules 2012, 17, 463–479;
i) D. G. D. a. I. A. Florea Dumitrascu, ARKIVOC 2010, i, 1–
32; j) T. Caronna, F. Fontana, A. Mele, I. N. Sora, W. Panzeri,
L. Vigano, Synthesis 2008, 413–416; k) J. Roithová, D.
Schröder, J. Mísˇek, I. G. Stará, I. Starý, J. Mass Spectrom.
2007, 42, 1233–1237; l) V. Claudio, L. Benoît, M. Pierre, L.
Jérôme, Chirality 2007, 19, 601–606; m) K. Shiraishi, A. Rajca,
M. Pink, S. Rajca, J. Am. Chem. Soc. 2005, 127, 9312–9313; n)
M. Watanabe, H. Suzuki, Y. Tanaka, T. Ishida, T. Oshikawa,
A. Tori-i, J. Org. Chem. 2004, 69, 7794–7801; o) D. Bas, P.-Y.
Morgantini, J. Weber, T. A. Wesolowski, Chimia 2003, 57, 173–
174; p) S. Honzawa, H. Okubo, S. Anzai, M. Yamaguchi, K.
Tsumoto, I. Kumagai, Bioorg. Med. Chem. 2002, 10, 3213–
3218; q) K. Tanaka, Y. Kitahara, H. Suzuki, H. Osuga, Y. Ka-
wai, Tetrahedron Lett. 1996, 37, 5925–5928; r) R. H. Martin,
M. Deblecker, Tetrahedron Lett. 1969, 10, 3597–3598; s) H. A.
Staab, M. A. Zirnstein, C. Krieger, Angew. Chem. Int. Ed. Engl.
1989, 28, 86–88; Angew. Chem. 1989, 101, 73; t) H. A. Staab,
M. Diehm, C. Krieger, Tetrahedron Lett. 1994, 35, 8357–8360;
u) K. Deshayes, R. D. Broene, I. Chao, C. B. Knobler, F. Died-
erich, J. Org. Chem. 1991, 56, 6787–6795.
Supporting Information (see footnote on the first page of this arti-
cle): General remarks and analysis conditions; synthesis and data;
1H NMR, 13C NMR, 19F NMR and HRMS-ESI spectra; X-ray
data for compound 1e; resolution, enantiomeric purity determi-
nation, ECD spectra; racemization barrier determination.
Acknowledgments
The authors thank the University of Geneva, the Swiss National
Science Foundation (grant numbers 200020-146666 and 20020-
147098). This research was also supported by the NCCR Chemical
Biology, funded by the Swiss National Science Foundation. The
contributions of the Sciences Mass Spectrometry (SMS) platform
at the Faculty of Sciences, University of Geneva, are also acknowl-
edged.
[1] a) Y. Shen, C.-F. Chen, Chem. Rev. 2012, 112, 1463–1535; b)
A. Urbano, Angew. Chem. Int. Ed. 2003, 42, 3986–3989; Angew.
Chem. 2003, 115, 4116; c) R. H. Martin, Angew. Chem. Int. Ed.
Engl. 1974, 13, 649–660; Angew. Chem. 1974, 86, 727.
[2] a) F. Aloui, R. E. Abed, A. Marinetti, B. B. Hassine, Tetrahe-
dron Lett. 2008, 49, 4092–4095; b) F. Aloui, R. E. Abed, B. B.
Hassine, Tetrahedron Lett. 2008, 49, 1455–1457; c) S. Abbate,
T. Caronna, A. Longo, A. Ruggirello, V. T. Liveri, J. Phys.
Chem. B 2007, 111, 4089–4097; d) S. Abbate, C. Bazzini, T.
Caronna, F. Fontana, C. Gambarotti, F. Gangemi, G. Longhi,
A. Mele, I. N. Sora, W. Panzeri, Tetrahedron 2006, 62, 139–
148; e) C. Bazzini, S. Brovelli, T. Caronna, C. Gambarotti, M.
Giannone, P. Macchi, F. Meinardi, A. Mele, W. Panzeri, F. Re-
cupero, A. Sironi, R. Tubino, Eur. J. Org. Chem. 2005, 1247–
1257; f) K. Sato, T. Yamagishi, S. Arai, J. Heterocycl. Chem.
2000, 37, 1009–1014; g) E. Murguly, R. McDonald, N. R.
Branda, Org. Lett. 2000, 2, 3169–3172; h) J. Howarth, J. Finne-
gan, Synth. Commun. 1997, 27, 3663–3668; i) S. Arai, M. Ishi-
kura, K. Sato, T. Yamagishi, J. Heterocycl. Chem. 1995, 32,
1081–1083; j) M. Rentzea, M. Diehm, H. A. Staab, Tetrahedron
Lett. 1994, 35, 8361–8364.
T. Kaseyama, S. Furumi, X. Zhang, K. Tanaka, M. Takeu-
[6] a)
chi, Angew. Chem. Int. Ed. 2011, 50, 3684–3687; Angew. Chem.
2011, 123, 3768; b) S. Graule, M. Rudolph, W. Shen, J. A. G.
Williams, C. Lescop, J. Autschbach, J. Crassous, R. Réau,
Chem. Eur. J. 2010, 16, 5976–6005; c) I. Alkorta, F. Blanco, J.
Elguero, D. Schroeder, Tetrahedron: Asymmetry 2010, 21, 962–
968; d) M. A. Shcherbina, X.-b. Zeng, T. Tadjiev, G. Ungar,
S. H. Eichhorn, K. E. S. Phillips, T. J. Katz, Angew. Chem. Int.
Ed. 2009, 48, 7837–7840; Angew. Chem. 2009, 121, 7977; e) S.
Graule, M. Rudolph, N. Vanthuyne, J. Autschbach, C. Roussel,
J. Crassous, R. Réau, J. Am. Chem. Soc. 2009, 131, 3183–3185;
f) W. Shen, S. Graule, J. Crassous, C. Lescop, H. Gornitzka,
R. Reau, Chem. Commun. 2008, 850–852; g) C. Bazzini, T. Ca-
ronna, F. Fontana, P. Macchi, A. Mele, I. N. Sora, W. Panzeri,
A. Sironi, New J. Chem. 2008, 32, 1710–1717; h) T. J. Katz,
Angew. Chem. Int. Ed. 2000, 39, 1921–1923; Angew. Chem.
2000, 112, 1997.
[3] a) A. Jancˇarˇík, J. Rybácˇek, K. Cocq, J. Vacek Chocholousˇová,
J. Vacek, R. Pohl, L. Bednárová, P. Fiedler, I. Císarˇová, I. G.
Stará, I. Starý, Angew. Chem. Int. Ed. 2013, 52, 9970–9975; b)
ˇ
J. Storch, J. Cermák, J. Karban, I. Císarˇová, J. Sýkora, J. Org.
Chem. 2010, 75, 3137–3140; c) O. Songis, J. Mísˇek, M. B.
ˇ
Schmid, A. Kollárovicˇ, I. G. Stará, D. Saman, I. Císarˇová, I.
Starý, J. Org. Chem. 2010, 75, 6889–6899; d) J. Misek, F. Teply,
I. G. Stara, M. Tichy, D. Saman, I. Cisarova, P. Vojtisek, I.
Stary, Angew. Chem. Int. Ed. 2008, 47, 3188–3191; Angew.
Chem. 2008, 120, 3232; e) I. G. Stará, I. Starý, A. Kollárovicˇ,
[7] a) L. Kötzner, M. J. Webber, A. Martínez, C. De Fusco, B. List,
Angew. Chem. Int. Ed. 2014, 53, 5202–5205; b) J. Chen, N.
Takenaka, Chem. Eur. J. 2009, 15, 7268–7276.
ˇ
F. Teplý, D. Saman, M. Tichý, J. Org. Chem. 1998, 63, 4046–
4050.
Eur. J. Org. Chem. 2014, 6431–6438
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6437

E. Vauthey, J. Lacour et al.
FULL PAPER
[8] R. Hassey, E. J. Swain, N. I. Hammer, D. Venkataraman, [14]
M. D. Barnes, Science 2006, 314, 1437–1439.
[9] a) Y.-H. Zheng, H.-Y. Lu, M. Li, C.-F. Chen, Eur. J. Org.
Chem. 2013, 3059–3066; b) T. J. Sørensen, A. Ø. Madsen, B. W.
Laursen, Tetrahedron Lett. 2013, 54, 587–590; c) L. Shi, Z. Liu,
a) K. Umehara, S. Kuwata, T. Ikariya, J. Am. Chem. Soc. 2013,
135, 6754–6757; b) S. E. Denmark, O. Nicaise, J. P. Edwards, J.
Org. Chem. 1990, 55, 6219–6223; c) S. F. Nelsen, M. R. Willi,
J. Org. Chem. 1984, 49, 1–6; d) J. M. Mellor, N. M. Smith, J.
Chem. Soc. Perkin Trans. 1 1984, 2927–2931.
G. Dong, L. Duan, Y. Qiu, J. Jia, W. Guo, D. Zhao, D. Cui,
X. Tao, Chem. Eur. J. 2012, 18, 8092–8099; d) L. Latterini, E.
Galletti, R. Passeri, A. Barbafina, L. Urbanelli, C. Emiliani,
F. Elisei, F. Fontana, A. Mele, T. Caronna, J. Photochem. Pho-
tobiol. A: Chem. 2011, 222, 307–313; e) R. Passeri, G. G. Aloisi,
F. Elisei, L. Latterini, T. Caronna, F. Fontana, I. N. Sora, Pho-
tochem. Photobiol. Sci. 2009, 8, 1574–1582.
[15]
[16]
With DMF, addition of brine to the crude reaction mixtures
led to a selective precipitation of quinacridines 1.
Salts [4i][BF4] and [4j][BF4] were easily prepared by treatment
of [3][BF4] with NH2NH2·H2O and ethanolamine, respectively,
see the Supporting Information.
The origin of the reduction will be discussed in the mechanism
section.
With benzylamines, it has been noticed that acridinium salts
are effective photocatalysts for the transformation of these sub-
strates into imine derivatives, see: C. Nicolas, C. Herse, J. Lac-
our, Tetrahedron Lett. 2005, 46, 4605–4608. With the dimethyl-
aminoethyl chains, elimination reactions can occur, see ref.[10a]
for details.
A distant similarity can be found with the reaction of hydroxyl-
amine with hydroxy-O-sulfonic acid in basic media. The reac-
tion generates sulfonic acid and diimide by a similar reaction
sequence, see: W. Durckheimer, Justus Liebigs Ann. Chem.
1969, 721, 240–243.
[17]
[18]
[10] a) A. Ueda, H. Wasa, S. Suzuki, K. Okada, K. Sato, T. Takui,
Y. Morita, Angew. Chem. Int. Ed. 2012, 51, 6691–6695; Angew.
Chem. 2012, 124, 6795; b) D. Conreaux, N. Mehanna, C.
Herse, J. Lacour, J. Org. Chem. 2011, 76, 2716–2722; c) J. Guin,
C. Besnard, J. Lacour, Org. Lett. 2010, 12, 1748–1751; d) B.
Laleu, M. S. Machado, J. Lacour, Chem. Commun. 2006, 2786–
2788; e) B. Laleu, P. Mobian, C. Herse, B. W. Laursen, G.
Hopfgartner, G. Bernardinelli, J. Lacour, Angew. Chem. Int.
Ed. 2005, 44, 1879–1883; Angew. Chem. 2005, 117, 1913; f) C.
Herse, D. Bas, F. C. Krebs, T. Bürgi, J. Weber, T. Wesolowski,
B. W. Laursen, J. Lacour, Angew. Chem. Int. Ed. 2003, 42,
3162–3166; Angew. Chem. 2003, 115, 3270; g) B. W. Laursen,
F. C. Krebs, Angew. Chem. Int. Ed. 2000, 39, 3432–3434; An-
gew. Chem. 2000, 112, 3574.
[19]
[20] Compound 1f was selected for its higher solubility in mixtures
of hexane and 2-propanol.
[11] a) J. Bosson, J. Gouin, J. Lacour, Chem. Soc. Rev. 2014, 43,
2824–2840; b) O. Kel, A. Fürstenberg, N. Mehanna, C. Nic-
olas, B. Laleu, M. Hammarson, B. Albinsson, J. Lacour, E.
Vauthey, Chem. Eur. J. 2013, 19, 7173–7180; c) N. Mehanna,
S. Grass, J. Lacour, Chirality 2012, 24, 928–935; d) O. Kel, P.
Sherin, N. Mehanna, B. Laleu, J. Lacour, E. Vauthey, Pho-
tochem. Photobiol. Sci. 2012, 11, 623–631; e) J. Guin, C. Be-
snard, P. Pattison, J. Lacour, Chem. Sci. 2011, 2, 425–428; f) P.
Mobian, N. Banerji, G. Bernardinelli, J. Lacour, Org. Biomol.
Chem. 2006, 4, 224–231; g) C. Nicolas, C. Herse, J. Lacour,
Tetrahedron Lett. 2005, 46, 4605–4608.
[12] The treatment of 4a with ammonia, ammonium hydroxide or
sulfamic acid in sealed vessels under a variety of conditions led
only to complex mixtures of products.
[13] a) T. A. Nigst, A. Antipova, H. Mayr, J. Org. Chem. 2012, 77,
8142–8155; b) H. Mayr, A. R. Ofial, J. Phys. Org. Chem. 2008,
21, 584–595.
[21] Dextro- and levorotatory enantiomers of 1f afforded dextro-
and levorotatory salts [1f·H+][BF4]: [α]365 = +10500 and
–10000, respectively (CH3CN, 10–4 m).
[22] The only difference can be observed at around 625 nm. The
salt [(P)-2a][BF4] displays a slightly positive cotton effect
whereas (+)-[1f·H+][BF4] presents a negative effect.
[23] a) A. Fürstenberg, E. Vauthey, Photochem. Photobiol. Sci.
2005, 4, 260–267; b) P.-A. Muller, C. Högemann, X. Allonas,
P. Jacques, E. Vauthey, Chem. Phys. Lett. 2000, 326, 321–327.
[24] a) S. R. Flom, P. F. Barbara, J. Phys. Chem. 1985, 89, 4489–
4494; b) T. Yatsuhashi, H. Inoue, J. Phys. Chem. A 1997, 101,
8166–8173; c) P. S. Sherin, J. Grilj, Y. P. Tsentalovitch, E. Vau-
they, J. Phys. Chem. B 2009, 113, 4953–4962; d) P. Fita, M.
Fedoseeva, E. Vauthey, Langmuir 2011, 27, 4645–4652.
Received: July 4, 2014
Published Online: September 4, 2014
6438
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 6431–6438