Flavin mimetics: Synthesis and photophysical properties

Article abstract of DOI:10.1016/j.tet.2021.131925

We report the synthesis of new isoalloxazines using a microwave-assisted approach to make N-substituted-2-nitroanilines followed by one-pot reduction and condensation via Hemmerich's method. The influence of substituents on positions 7, 8, and 10 of flavin core on the optical properties is investigated. The aliphatic functionalities on N10 give rise to quantum yields of 0.7, while aromatic side-chains quench fluorescence. Relaxed geometries (DFT) of chiral and achiral derivatives have been used for TD-DFT calculations, which yielded good agreement with the experimental UV and CD data.

Full text of DOI:10.1016/j.tet.2021.131925

Tetrahedron 82 (2021) 131925
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Flavin mimetics: Synthesis and photophysical properties
a
b
c,
Dora-M. Rasadean , Takashi Machida , Kazuki Sada d, Christopher R. Pudney e,
ꢀ ꢀ
,
*
G. Dan Pantoș a
a Department of Chemistry, University of Bath, Claverton Down, Bath, BA2 7AY, UK
b International Center for Materials Nanoarchitectonics (WPI-MANA), National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba, 305-0044,
Japan
c Graduate School of Chemical Sciences and Engineering, Hokkaido University, Kita 10 Nishi 8, Kita-ku, Sapporo, Japan
d Department of Chemistry, Faculty of Science, Hokkaido University, Kita 10 Nishi 8, Kita-ku, Sapporo, Japan
e Department of Biology and Biochemistry, University of Bath, Claverton Down, Bath, BA2 7AY, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
We report the synthesis of new isoalloxazines using a microwave-assisted approach to make N-
substituted-2-nitroanilines followed by one-pot reduction and condensation via Hemmerich’s method.
The influence of substituents on positions 7, 8, and 10 of flavin core on the optical properties is inves-
tigated. The aliphatic functionalities on N10 give rise to quantum yields of 0.7, while aromatic side-chains
quench fluorescence. Relaxed geometries (DFT) of chiral and achiral derivatives have been used for TD-
DFT calculations, which yielded good agreement with the experimental UV and CD data.
© 2021 Elsevier Ltd. All rights reserved.
Received 30 November 2020
Received in revised form
29 December 2020
Accepted 4 January 2021
Available online 13 January 2021
Keywords:
Flavin
Isoalloxazine
Microwave synthesis
Chirality
Fluorescence
1. Introduction
given in Fig. 1) is responsible for the protein binding and specificity,
while the rest of the substituents influence the stability, redox and
Flavins are coenzymes and photoreceptors that are ubiquitous
in biotic systems, taking part in numerous biochemical trans-
formations [1,2]. Representative examples of flavins are riboflavin
(RF), flavin mononucleotide (FMN) and flavin adenine dinucleotide
(FAD), which are prosthetic groups in flavoenzymes [3]. All these
share a key structural feature that modulates their activity: the
isoalloxazine core, which is a benzo[g]pteridine-2,4(3H)-dione able
to accept one or two electrons (Fig. 1) [4]. The isoalloxazine is
therefore one of the electroactive units within the biological envi-
ronment and contributes to redox reactions, activation of oxygen,
cell apoptosis, halogenation of aromatic substrates, respiratory
electron transport chain, dehydrogenation of metabolites, DNA
repair and many other [2,5,6].
acid-base properties as well as solubility [3]. The literature reports a
plethora of experimental and theoretical studies on flavins and
their interactions within biological environments [1,7,8]. However,
these are highly complex systems with many interplaying com-
ponents, making difficult to assess the binding mode, reactivity and
properties of isoalloxazines. On this basis, synthetic flavin models
have been designed to get insights into the cofactor-protein binding
modes [9]. These models have opened up a new path towards
development of artificial enzymes that mimic the behaviour of
natural ones [10e12]. Most of flavin mimetics maintain the cata-
lytically active isoalloxazine core but have been structurally
modified to control their redox and photophysical properties.
The design of artificial enzymes based on flavins follows either
the functionalisation of precursors or chemical modification and
photodegradation of RF, FMN and FAD. The isoalloxazine core can
be functionalised with small electron withdrawing or donating
groups in any position [13e17] as well as with more complex
groups. Examples of the latter include flavins functionalised with
calixarenes [18], cyclophanes [19], fullerenes [20], fused
macrocycles-isoalloxazine dyads [5,21] and flavin-peptide
Isoalloxazine-based flavins interact with cellular components
through non-covalent interactions such as
bonds, electrostatic and charge-transfer interactions [5]. The side
chain on the N10 position of isoalloxazine core (atom numbering is
p p stacking, hydrogen
ꢀ
* Corresponding author.
E-mail address: g.d.pantos@bath.ac.uk (G.D. Pantoș).
https://doi.org/10.1016/j.tet.2021.131925
0040-4020/© 2021 Elsevier Ltd. All rights reserved.

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
the ring [29]. Attempts have been made to overcome this issue, and
current methods include ultrasound-assisted N-alkylation [29,30]
and reaction of 2,4-dinitroanilines with silyl carboxylates in the
presence of titanium(IV) salts as catalysts [31,32]. However, despite
their success, both methods have certain drawbacks in terms of the
specialised equipment or expensive reagents, long reaction times
and laborious preparation of initial substrates [29]. We have
developed two strategies for the synthesis of N-substituted-2-
nitroanilines using microwave-assisted protocols. Both methods
are reliable and effective, using small volumes of solvents and
giving good conversions. The first route follows an SN2 pathway, in
which 2,4-dinitroaniline is reacted with alkyl bromides in the
presence of sodium hydride in acetonitrile (Scheme 1). Reaction
progress has been monitored by TLC, with best conversion being
observed after 3 h. Derivatives 1 and 2 have a nitro group in para
position relative to the amine, while 3 has a fluoro substituent on
the meta position. The reactivity of the amine towards nucleophilic
substitutions is influenced by the electronic properties and position
of these functionalities. The strongly electronegative fluoro sub-
stituent decreases the nucleophilicity of the amine more than the
nitro group, as reflected in the lower yield of 3 compared to 1 and 2
(Scheme 1).
Fig. 1. Structures, oxidation states and IUPAC atom numbering of isoalloxazine col-
oured as in aqueous solution: oxidised (yellow), semiquinone radical (red) and reduced
forms (transparent).
nanotubes [22]. Synthetic flavins are of paramount importance in
modelling flavoenzymatic interactions that lead to a better under-
standing of how enzymes work, and the way co-factors interact
with the protein environment. They also find extensive applications
as photoreceptors and in many catalytic processes such as oxida-
tion and reduction of a large variety of substrates. The specificity/
selectivity of such reactions can be modulated either by taking
advantage of the isoalloxazine core (able to transfer one or two
electrons) or by introduction of chiral substituents. Chiral iso-
alloxazines participate in redox reactions in an asymmetric or
enantioselective manner. The first chiral synthetic isoalloxazine
was reported in 1984 by Ohno et al. and contained a chiral centre in
The other approach is based on the SNAr of 1-chloro-2,4-
dinitrobenzene with amines. Derivative 4 was obtained via the
classical method [33], but the reaction time was reduced from 15 h
to 1 h. This protocol did not work for other amines, thus we have
developed a more versatile method. It involves irradiating the re-
action mixture for as short as 10 min at 140 ꢁC in the microwave
reactor using ethanol as solvent. This has allowed the synthesis of
chiral derivatives 5e7 in good to quantitative yields.
the
a position of the N10 substituent [23]. Since then, just a limited
number of chiral flavins has been reported, including examples
with chiral substituents on the N3 [22], N5 [24] and N10 [25] po-
sitions, ethylene bridged flavinium salts [26] and atropisomers with
CeN axial (planar) chirality [27,28].
We report herein the synthesis of a series of new isoalloxazines
functionalised on the 7 or 8 positions with electron withdrawing
groups and alkyl or aryl-alkyl side chains on the N10. The synthetic
route involves two steps: alkylation of nitrobenzene-derivatives
followed by one-pot reduction and condensation to obtain the
isoalloxazines in good yields. We put forward optimised methods
for alkylation and subsequent reduction of N-substituted o-phe-
nylenediamines. All isoalloxazines were characterised in terms of
their absorption and fluorescent properties. We have also investi-
gated the chiroptical properties of the chiral derivatives in the se-
ries both in solution and in silico using TD-DFT.
2. Results and discussion
2.1. Alkylation of (di)nitroanilines
The first step in the synthesis of the isoalloxazine core is the
preparation of the N-substituted-2-nitroaniline derivatives. There
are two main mechanistic pathways towards their synthesis, each
using commercially available precursors. The most commonly used
route involves the reaction of 2,4-dinitrohalobenzenes with alkyl/
acyl amines via a nucleophilic aromatic substitution (SNAr) mech-
anism [29]. The downside of this method is that the 2,4-
dinitrohalobenzenes are generally expensive and the yields vary
significantly based on the substrates used. Another pathway is the
N-alkylation/acylation of 2,4-dinitroaniline with alkyl halides,
which undergoes via a nucleophilic substitution (SN2) mechanism.
Such reactions are generally not favourable because the amino
group is a weak nucleophile due to the electronic effects present in
Scheme 1. The two pathways (SN2 and SNAr) to synthesise N-substituted-2,4-
dinitroanilines. The yield for each derivative is given in parentheses after product
number. The general methods used in each case (A and B) are detailed in the Exper-
imental section.
2

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
2.2. One-pot synthesis of isoalloxazines
[27,40]. The current work further expands the existent knowledge
and tackles in detail the influence of substituents in positions 7, 8
and 10 on the (chiro)optical properties of synthetic isoalloxazines.
The N-alkylated (di)nitroanilines 1e7 were reduced to the cor-
responding o-phenylenediamines followed by one-pot condensa-
tion with alloxan to form the isoalloxazines 8e14. We have used the
classical azeotropic mixture of formic acid-triethylamine (FA-Et3N)
[33,34] and catalytic amounts of palladium on carbon (0.1 equiv)
that generates hydrogen in situ for reduction of compounds 1, 2 and
4e7 (Scheme 2). This is a mild reducing system, which has allowed
the chemoselective reduction of the nitro group in position 2, but
not of that in position 4. Selectivity was achieved by cooling the
system to ꢀ15 ꢁC and dropwise addition of FA. The benefit of FA-
Et3N in the reduction process over other reducing agents is that the
resulting carbon dioxide is thermodynamically stable, which makes
the reaction irreversible [35]. Reduction of derivative 3 with just
one nitro group was carried out in much stronger reducing envi-
ronments (i.e. zinc powder and acetic acid; Scheme 2) following a
protocol reported in the literature [36].
The catalysts used during reductions were filtered off and the
crude mixtures were used immediately in the flavin formation. The
isoalloxazine core was formed via Hemmerich’s method [37],
which involved the reaction of N-substituted o-phenylenediamines
with alloxan in the presence of boric acid and acetic acid as solvent
(Scheme 2). The condensation method is versatile, with shorter
reaction times when performing the reaction at high temperatures
(90 ꢁC). The yield over the two steps (reduction and condensation)
is higher if only one nitro group is present in the starting material
(14). The substituents on the N10 position also influence the reac-
tion progress, with the lowest yield over two steps being obtained
for the methyl-substituted isoalloxazine 10 (12%). A possible
explanation for this is the higher solubility of 10 in the reaction
mixture. The one-step reduction/condensation method is versatile;
however, for certain derivatives does not lead to isoalloxazine, but
quinoxaline by-products [38].
2.3. Photophysical properties of isoalloxazines 8e14
The photophysical features of isoalloxazines 8e14 were inves-
tigated by means of absorption and fluorescence spectroscopies.
The chiroptical properties of chiral derivatives 11e13 were explored
using circular dichroism (CD). All measurements were carried out
in acetonitrile.
The UVeVis region of the absorption spectra of all isoalloxazines
display similar pattern: a broad band around 425 nm flanked by
two shoulders at 455 and 405 nm (Fig. 2A). This band arises from
the lowest allowed
p / p* transitions of isoalloxazine chromo-
phore. This spectral feature is consistent with spectra of other flavin
mimics in acetonitrile reported in the literature [41e44]. The rest of
absorption profile has different characteristics depending on the
substituent on the benzene ring of the flavin core (positions 7 and
8). The substituent on the N10 position does not significantly in-
fluence the absorption behaviour (all absorption spectra shown in
Fig. S38). Therefore, all the nitro-isoalloxazines (8e13) show one
large band around 295 nm and a weak one at 230 nm. A comparison
between the 10-butyl-7-nitro (8) and 10-butyl-8-fluoro (14) de-
rivatives is given in Fig. 2A. Flavin 14 displays a hypsochromically-
shifted band compared to the nitro one: 265 vs 294 and another
Compounds 8 and 10 are reported in the literature [39], but no
thorough characterisation is provided (only melting point and IR
are reported). Literature reports ample information on flavins in
general and other nitro- [9] and fluoro-substituted isoalloxazines
Scheme 2. One-pot reduction of N-alkylated (di)nitroanilines and condensation with
alloxan to synthesise isoalloxazines 8e14. The overall yields (over two steps) corre-
sponding to each derivative are given in parentheses after product number.
Fig. 2. A: Normalised absorption spectra of isoalloxazines 8 and 14 in acetonitrile. B:
Normalised emission spectra of isoalloxazines 8, 14 and 11 in acetonitrile at 425 nm (8
and 14) and 426 nm (11) excitation wavelength.
3

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
band of similar amplitude around 220 nm (Fig. 2A). Literature data
shows that strongly electron withdrawing functionalities (such as
F3C-) in positions 7 or 8 [41] as well as substituents on the N3
position give rise to a band at 325 nm [43,44]. This explains the
presence of the weaker band at 325 nm in the spectrum of fluoro-
isoalloxazine 14, which is not observed in the spectra of nitro de-
rivatives (Fig. 2A).
The absorption profile across the UVeVis region of biological
flavins (e.g., RF) in acetonitrile is similar to that of our derivatives
[45]. In terms of strength of isoalloxazine chromophore, the syn-
thetic flavins have larger molar extinction coefficients (ε) than RF
(εRF ¼ 2100 L molꢀ1 cmꢀ1 [46]), with 10-butyl-7-nitro flavin 8 being
the strongest and the 10-methyl-7-nitro flavin 10 the weakest
(12800 vs 2670 L molꢀ1 cmꢀ1, respectively; Table 1; Fig. S42). These
values are comparable to those of other synthetic isoalloxazines
recorded in the same solvent [44]. The ε of derivatives 8e14 in
acetonitrile are given in Table 1.
analogues are also much smaller compared to these (FFL of RF in
acetonitrile is 0.47 [47]).
The FFL of derivatives 9 and 11e13 are lower compared to the
alkyl-substituted flavins (Table 1). We associate this behaviour of 9
and 11 with fluorescence quenching processes induced by the ar-
omatic units on these molecules. It is well-known that aromatics
with organic donor or acceptor functionalities quench fluorescence
in organic solvents through an electron transfer process [50]. The
electron-acceptor nature of aromatic substituent in position N10 of
compound 11 leads to a significant decrease down to 0.05, while
the phenyl group on 9 induces a drop to 0.25. The higher FFL of 9
compared to 11 can be attributed to the position of the phenyl
group relative to the isoalloxazine core (
electronic properties. Derivatives 12 and 13 have a FFL of 0.31,
suggesting that molecular crowding in the position relative to the
flavin core lowers the FFL
g vs a in 11) and different
a
.
The chiroptical properties of chiral derivatives 11e13 have been
The fluorescence emission spectra of derivatives 8e10 and
12e14 in acetonitrile show similar spectral features (Figs. S41 and
S42), with one broad band around 480 nm and a shoulder on its
right centred at 520 nm (typical fluorescence emission spectra are
shown in Fig. 2B). The emission band of nitro-derivatives 8, 9 and
10 is shifted by 5 nm and that of 12 and 13 by 8 nm to longer
wavelengths compared to the fluoro-substituted flavin 14. This
leads to slightly larger Stokes shifts for the nitro analogues. Fig. 2B
compares the emission spectra of compounds 8 and 14. Derivative
11 exhibits a different emission profile than the rest of iso-
alloxazines in the series, with a broad band shifted at longer
wavelengths (529 nm) and a shoulder on its right part (Fig. 2B).
The maximum wavelength from the emission spectra of each
derivative and the corresponding Stokes shifts are summarised in
Table 1. The emission spectral characteristics of compounds 8e14
are similar to those of other synthetic flavins reported in the
literature [44]. The emission maximum has been attributed to S01 /
S00 vibronic transition [47].
The fluorescence quantum yields (FFL) of derivatives 8e14 in
acetonitrile are provided in Table 1. All FFL were calculated using
fluorescein as standard (we have used the FFL of fluorescein in
ethanol as 0.79 [48]; details are given in the Experimental section).
The substituents on the N10 position strongly influence the FFL of
derivatives 8e14, while the functionalities attached on the benzene
ring of the flavin core do not affect it that much. Among the flavins
in the series, the 10-methyl-7-nitro isoalloxazine 10 has the highest
FFL (0.71). This is followed closely by derivatives 8 and 14, which
have a butyl side-chain on the N10 position. The nitro (8) and fluoro
(14) groups do not influence the FFL. The large FFL of these flavins
indicate their low non-radiative decay rates [49]. To the best of our
knowledge, these values are among the largest reported so far in
the literature for synthetic isoalloxazines. The FFL of biological
Table 1
Spectral characteristics of isoalloxazines 8e14 shown in Scheme 2.
Product Absorption
lmax (nm) ε (L molꢀ1 cmꢀ1
Emission
b
)
lmax (nm) Stokes shift (nm)a FFL
8
9
425
426
426
426
428
428
425
12800 0.1
4960 4.0
2670 0.3
4070 0.1
8950 0.9
2070 0.1
9860 0.6
485
486
485
529
488
488
480
60
60
59
103
60
60
55
0.70
0.25
0.71
0.05
0.31
0.31
0.70
10
11
12
13
14
a
The Stokes shifts were calculated as: lmax emission e lmax absorption (corre-
sponding to the isoalloxazine core).
Fluorescence quantum yields at excitation wavelengths as lmax in absorption
spectra were calculated using fluorescein as standard.
Fig. 3. Overlaid experimental (2 ꢂ 10ꢀ4 M; red) and predicted (black) spectra of 11. A:
absorption and B: CD. The y axes on the left correspond to the experimental data, while
those on the right to the predicted spectra.
b
4

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
explored by circular dichroism (CD). The CD spectrum of derivative
11 in acetonitrile is shown in Fig. 3 and provides valuable insights
into optoelectronic and structural properties of this chiral flavin.
The spectrum displays a negative Cotton effect centred at 425 nm,
indicating that the chirality is transferred from the N10 substituent
onto the intrinsically achiral isoalloxazine core. There are also two
large positive bands around 290 and 230 nm. We have also per-
formed variable temperature (VT) CD studies to assess if the
chirality transfer is temperature dependant (Fig. S39). No signifi-
cant modification was observed upon VT ramp (5e75 ꢁC and re-
turn) apart from minor photodecomposition. Isoalloxazines 12 and
4. Experimental section
4.1. Materials and methods
All reagents were purchased from commercial suppliers: Merck,
Fisher, TCI and Fluorochem and used without further purification.
1H and 13C NMR spectra were recorded on 500 MHz Agilent Pro-
pulse or 500 MHz Bruker Avance IIþ (1H at 500 MHz, 13C at
126 MHz) instruments, as stated. Chemical shifts (d) are reported in
parts per million (ppm). Coupling constants (J) are reported in
Hertz (Hz), and signal multiplicity is denoted as broad singlet (br s),
singlet (s), doublet (d), doublet of doublets (dd), triplet (t), quartet
(q), pentet (p), sextet (sext), doublet of doublets of doublets (ddd),
doublet of triplets (dt), triplet of doublets (td) and multiplet (m).
Samples for NMR spectroscopy were prepared with CDCl3, acetone-
d6 or DMSO-d6 as stated in each case. All spectra were acquired at
25 ꢁC and were referenced to the residual solvent peaks. The 1H and
13C NMR spectra (Figs. S1e28) contain small traces of water,
dichloromethane, acetone, ethyl acetate and petroleum ether
40e60%. Electrospray ionisation quadrupole time-of-flight (ESI-Q-
TOF) mass spectrometry was performed on an Agilent Technologies
6545 Q-TOF LC-MS instrument using a positive- or negative-ion
mode as stated in each case (Figs. S29e37). Microwave-assisted
reactions were conducted in a CEM Discover microwave reactor
using a sealed reaction vessel.
CD and absorption experiments (Figs. S38e41) were performed
in acetonitrile on an Applied Photophysics Chirascan Circular Di-
chroism Spectrophotometer equipped with a Peltier temperature
controller using a 1 cm pathlength quartz cuvette. The background
corresponding to the acetonitrile and cuvette’s absorption was
subtracted from subsequent measurements. The following param-
eters were used: wavelength range 210e550 nm, time-per-point
1 s, monochromator bandwidth 1 nm, temperature 20 ꢁC. Fluo-
rescence spectra (Figs. S42e43) were recorded in acetonitrile using
a 1 cm fluorescence quartz cuvette on an Applied Photophysics
Chirascan Circular Dichroism Spectrophotometer equipped with a
CS/SEM accessory. The following parameters were used: wave-
length range 440e650 nm (emission spectra) and 210e550 nm
(excitation spectra), time-per-point 0.5 s, monochromator band-
width 2 nm, slit width 1 mm (bandpass 4.65 nm), PMU 1000 V,
temperature 20 ꢁC. The excitation and emission wavelengths used
in each case are given in the Supplementary material. The quantum
yields of compounds 8e14 were calculated using fluorescein in
ethanol (0.79) as reference and 1.36 and 1.344 the refractive indices
of ethanol and acetonitrile, respectively.
13 also have chiral centres in the
a position relative to the core as 11.
These bear nonane-functionalised alkyl chain (12) or
a
a
cyclohexyl-based substituent (13) in position N10 instead of aro-
matic units. The CD spectra of 12 and 13 display a similar pattern to
that of 11, with the chiral information also being transferred onto
the core. Their CD profile shows positive Cotton effects around
425 nm (Fig. S40). The literature reports some CD spectra of iso-
alloxazines, but most of them are on biological flavins, their
modified analogues or flavinyl peptides. Therefore, no direct com-
parison of chiroptical properties could be done between our syn-
thetic derivatives and previously reported examples.
In order to understand the optoelectronic properties of these
synthetic flavins, we undertook (TD)DFT studies. The structures of
derivatives 11e14 have been pre-optimised using semi empirical
methods (PM7, convergence criterion 0.01 kcal/mol/Å) followed by
DFT optimization using B3LYP-D3/6311þG* theory level. The
relaxed geometries of 11e14 obtained in the previous step were
used in sTDDFT calculations for predicting the absorption (all) and
CD (11e13) spectra. We have used hybrid-GGA (B3LYP), hybrid
meta GGA (M06e2X, PW6B95) and range separated hybrid (CAM-
B3LYP,
uB97X, uB97X-D3) functionals with Pople style and Def2-
Ahlrichs basis sets. Of the various combinations of functionals
and basis sets used, M06e2X/QZVP/CH3CN gave the best results
(compound 11, Fig. 3A and B) with similarity factors of 0.96
(absorption, ꢀ1 nm shift) and 0.96 (CD, ꢀ7 nm shift) calculated
using SpecDis [51] (for details see section 6 and Table S1 in the
Supplementary material). These results allow us to lay the foun-
dation for future pre-synthetic electronic/optical properties pre-
diction of synthetic flavins.
3. Conclusion
We report herein an optimised method to synthesise a series of
seven new isoalloxazines starting from commercially available (di)
nitroanilines or dinitrohalobenzenes. The first step uses microwave
irradiation to generate N-substituted-2-nitroanilines in good to
quantitative yields via either SN2 or SNAr mechanisms. The next
step is the one-pot reduction of nitro groups to amines followed by
condensation with alloxan to yield the isoalloxazines 8e14. The
optical properties (absorption/emission features, molar extinction
coefficients, quantum yields) of all synthetic flavins in acetonitrile
are discussed in detail. Three derivatives (8, 10, 14) in the series
have large quantum yields of 0.7, while 9 and 11e13 undergo
fluorescence quenching induced by aromatic units or bulky sub-
Molecular modelling studies were carried out using PM7
(MOPAC2016) [52] for pre-optimizing the structures followed by
DFT geometry optimization using B3LYP-D3/6311þG*. The TD-DFT
studies were performed at various functionals with 6311þþG** or
QZPT basis sets using, where appropriate, D3 dispersion correction
and an acetonitrile CPCM solvent model. All DFT calculation were
performed using ORCA v. 4.2.1 [53] with Gabedit v. 2.5.0 [54] as GUI
(further information is provided in the Supplementary material,
Table S1, Fig. S44). The predicted UVevis and CD spectra were
constructed in SpecDis by fitting Gaussian curves over the oscillator
strength (UVevis) and rotational strength (CD) computed with the
sTDDFT (M06e2X/QZVP/CH3CN). The exponential half-width for
the Gaussian fit ranged between 0.25 and 0.30 eV as obtained from
similarity factor calculations implemented SpecDis.
stituents in the
a position relative to the flavin core. The CD spectra
of 11e13 show that the chiral information is efficiently transferred
from the chiral substituent onto the isoalloxazine core. Predicted
spectra using (TD)DFT calculations match well the experimental
data. This work puts forward an optimised and versatile strategy to
making synthetic flavins, which are important in model enzymatic
studies as well as catalysis and as photoreceptors. The catalytic
potential of all flavins reported in this work will be reported in due
course.
4.2. General procedure A to synthesise compounds 1e3
A 10-mL microwave tube was charged with 2,4-dinitroaniline
(100 mg, 0.546 mmol, 1 equiv) for derivatives 1 and 2 or 5-
fluoro-2-nitroaniline (100 mg, 0.640 mmol, 1 equiv) for derivative
5

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
3 and acetonitrile (5 mL). The microwave tube was placed on a
stirrer plate and sodium hydride 60% suspension in mineral oil (2
equiv) was added. The yellow solution turned deep red upon so-
dium hydride addition. The reaction mixture was stirred 5 min on
bench and the corresponding alkylating reagent was added (2
equiv). The microwave tube was placed in a microwave reactor and
the reaction mixture was heated for 3 h at 120 ꢁC under microwave
irradiation (at 150 W). The solvent was removed under reduced
pressure to yield a red-brown residue, which was re-dissolved in
water (150 mL). The pH of the solution was adjusted to 2 using 1 M
HClaq. The solution was extracted with ethyl acetate (2 ꢂ 100 mL).
The collected organic fractions were washed with brine
(3 ꢂ 100 mL), dried over anhydrous magnesium sulfate and solvent
removed under reduced pressure to obtain a yellow solid. The
product was purified by flash column chromatography using silica
as stationary phase (60 Å particle size). The solvents were removed
under reduced pressure and the solid dried in vacuo to yield the
corresponding alkylated amine.
4.3.1. N-methyl-2,4-dinitroaniline (4)
Yellow solid (3.9 g, 19.748 mmol, quant.). 1H NMR (500 MHz,
acetone-d6)
d
9.01 (d, J ¼ 2.7 Hz,1H), 8.82 (br s,1H), 8.35 (dd, J ¼ 9.6,
2.7 Hz, 1H), 7.25 (d, J ¼ 9.6 Hz, 1H), 3.25 (d, J ¼ 5.1 Hz, 3H). The 1H
NMR data matches the one reported in the literature [33].
4.4. General procedure B to synthesise compounds 5e7
A 10-mL microwave tube was charged with 1-chloro-2,4-
dinitrobenzene (200 mg, 0.987 mmol, 1 equiv) and absolute
ethanol (5 mL). To this, the relevant primary amine (2 equiv) was
added and the reaction mixture was heated for 10 min at 140 ꢁC
under microwave irradiation (at 300 W). The solvent was removed
under reduced pressure to give an orange residue. Product 5 was
purified by flash column chromatography using silica as stationary
phase (60 Å particle size). The crude residue obtained after solvent
evaporation in the case of derivatives 6 and 7 was re-dissolved in
dichloromethane (50 mL) and extracted with saturated NaHCO3
(3 ꢂ 50 mL) and 1 x 100 mL saturated NaCl. The collected organic
fractions were dried over anhydrous magnesium sulfate and the
solvent removed under reduced pressure. The excess of unreacted
amine was removed under high vacuum to yield the corresponding
alkylated amines.
4.2.1. N-butyl-2,4-dinitroaniline (1)
Started with 1-bromobutane (149.6 mg, 118 mL, 1.092 mmol) and
sodium hydride (26.2 mg, 1.092 mmol). Purified by column chro-
matography (CH2Cl2:petroleum ether 1:1, v/v ratio), product
collected as second fraction. Yellow solid, 94.2 mg, 0.394 mmol,
72%. 1H NMR (500 MHz, CDCl3)
d
9.14 (d, J ¼ 2.7 Hz, 1H), 8.55 (br s,
1H), 8.27 (dd, J ¼ 9.5, 2.7 Hz, 1H), 6.92 (d, J ¼ 9.5 Hz, 1H), 3.41 (td,
J ¼ 7.1, 5.3 Hz, 2H), 1.77 (p, J ¼ 7.3 Hz, 2H), 1.50 (dt, J ¼ 14.8, 7.4 Hz,
2H), 1.01 (t, J ¼ 7.4 Hz, 3H). The 1H NMR data matches the one re-
ported in the literature [29].
4.4.1. (S)eN-(1-(4-chlorophenyl)ethyl)-2,4-dinitroaniline (5)
Started with (S)-4-chloro-a-methylbenzylamine (307.3 mg,
1.975 mmol) Purified by column chromatography (CH2Cl2:petro-
leum ether 1:1, v/v ratio), product collected as third fraction. Yellow
solid (206 mg, 0.641 mmol, 65%). 1H NMR (500 MHz, CDCl3)
d 9.11
4.2.2. 2,4-dinitro-N-(3-phenylpropyl)aniline (2)
(d, J ¼ 2.7 Hz, 1H), 8.86 (d, J ¼ 5.9 Hz, 1H), 8.12 (dd, J ¼ 9.5, 2.7 Hz,
Started with 1-bromo-3-phenylpropane (218 mg, 1.092 mmol)
and sodium hydride (26.2 mg, 1.092 mmol). Purified by column
chromatography (CH2Cl2:petroleum ether 1:1, v/v ratio), product
collected as second fraction. Yellow solid, 132 mg, 0.438 mmol, 80%.
1H), 7.36 (d, J ¼ 8.3 Hz, 2H), 7.29 (d, J ¼ 8.3 Hz, 2H), 6.72 (d,
J ¼ 9.5 Hz, 1H), 4.80 (p, J ¼ 6.6 Hz, 1H), 1.73 (d, J ¼ 6.6 Hz, 3H). 13
C
NMR (126 MHz, CDCl3)
d 147.3, 140.5, 136.5, 133.8, 130.8, 130.1,
129.5, 126.9, 124.0, 115.2, 53.3, 24.6. HRMS m/z calculated for
1H NMR (500 MHz, CDCl3)
d
9.11 (d, J ¼ 2.6 Hz, 1H), 8.59 (t,
C
14H12ClN3O4: [M ꢀ H]- 320.0444, found: 320.0445.
J ¼ 5.6 Hz, 1H), 8.24 (dd, J ¼ 9.5, 2.6 Hz, 1H), 7.38e7.29 (m, 2H),
7.27e7.17 (m, 3H), 6.86 (d, J ¼ 9.5 Hz, 1H), 3.45 (q, J ¼ 5.6 Hz, 2H),
2.83 (t, J ¼ 7.4 Hz, 2H), 2.15 (p, J ¼ 7.3 Hz, 2H).
4.4.2. (R)-2,4-dinitro-N-(nonan-2-yl)aniline (6)
Started with (R)-2-aminononane (283.0 mg, 362
mL,
4.2.3. N-butyl-5-fluoro-2-nitroaniline (3)
1.975 mmol). Purified by extraction with NaHCO3(sat) and NaCl(sat)
followed by the excess amine removal under reduced pressure.
Orange viscous oil, 305.1 mg, 0.987 mmol, quantitative. 1H NMR
Started with 1-bromobutane (175.5 mg, 138 mL,1.281 mmol) and
sodium hydride (30.7 mg, 1.281 mmol). Purified by column chro-
matography (CH2Cl2:petroleum ether 1:1, v/v ratio), product
collected as first fraction. Yellow solid, 55.1 mg, 0.259 mmol, 41%. 1H
(500 MHz, CDCl3)
d
9.15 (d, J ¼ 2.7 Hz, 1H), 8.53 (d, J ¼ 7.8 Hz, 1H),
8.25 (dd, J ¼ 9.6, 2.7 Hz, 1H), 6.91 (d, J ¼ 9.6 Hz, 1H), 3.78 (p,
J ¼ 6.6 Hz, 1H), 1.76e1.58 (m, 2H), 1.46e1.36 (m, 2H), 1.34 (d,
J ¼ 6.4 Hz, 4H), 1.33e1.24 (m, 6H), 0.88 (t, J ¼ 6.8 Hz, 4H). 13C NMR
NMR (500 MHz, CDCl3)
d
8.20 (dd, J ¼ 9.5, 6.1 Hz, 1H), 8.16 (br s, 1H),
6.48 (dd, J ¼ 11.5, 2.6 Hz, 1H), 6.34 (ddd, J ¼ 9.5, 7.2, 2.6 Hz, 1H), 3.25
(td, J ¼ 7.1, 5.1 Hz, 2H), 1.72 (p, J ¼ 7.3 Hz, 2H), 1.49 (sext, J ¼ 7.4 Hz,
(126 MHz, CDCl3)
d 147.8, 130.3, 124.7, 114.0, 77.3, 77.0, 76.7, 49.3,
2H), 1.00 (t, J ¼ 7.4 Hz, 3H). 13C NMR (126 MHz, CDCl3)
d 168.59,
36.6, 31.7, 29.4, 29.1, 25.9, 22.6, 20.4, 14.1. HRMS m/z calculated for
166.55, 147.58 (d, J ¼ 13.7 Hz), 129.96 (d, J ¼ 12.4 Hz), 128.71, 103.72
(d, J ¼ 24.8 Hz), 99.20 (d, J ¼ 26.9 Hz), 30.74, 20.19, 13.71.
C
15H23N3O4: [M ꢀ H]- 308.1616, found: 308.1615.
4.3. Synthesis of compound 4
4.4.3. (R)-N-(1-cyclohexylethyl)-2,4-dinitroaniline (7)
Started with (R)-1-cyclohexylethylamine (251.1 mg, 290
mL,
The reaction protocol was modified from Ref. [33]. A 250-mL
round-bottomed flask was charged with 1-chloro-2,4-
dinitrobenzene (4 g, 19.748 mmol, 1 equiv) and ethanol (150 mL)
was added. The solution was stirred 15 min at room temperature,
then methylamine (40% w/w solution in water; 1.23 g, 1.75 mL,
39.496 mmol, 2 equiv) was added (a precipitate started to form).
The reaction mixture was stirred at room temperature for 1 h and
the precipitate formed was filtered off, washed with a small
amount of ethanol and dried in vacuo.
1.975 mmol). Purified by extraction with NaHCO3(sat) and NaCl(sat)
followed by the excess amine removal under reduced pressure.
Orange viscous oil, 289.3 mg, 0.987 mmol, quantitative. 1H NMR
(500 MHz, CDCl3)
d
9.15 (d, J ¼ 2.7 Hz, 1H), 8.67 (s, 1H), 8.23 (ddd,
J ¼ 9.7, 2.8, 0.8 Hz, 1H), 6.92 (d, J ¼ 9.6 Hz, 1H), 3.70e3.60 (m, 1H),
1.89e1.67 (m, 7H), 1.57 (tdt, J ¼ 8.7, 5.7, 2.8 Hz, 2H), 1.28e1.19 (m,
2H), 1.19e1.03 (m, 3H). 13C NMR (126 MHz, CDCl3)
d 148.2, 130.5,
124.9, 114.2, 54.1, 43.2, 29.4, 29.0, 26.4, 26.3, 26.2, 17.5. HRMS m/z
calculated for C14H19N3O4: [M ꢀ H]- 292.1303, found: 292.1302.
6

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
4.5. General procedure to synthesise compounds 8e13
1H NMR (500 MHz, DMSO-d6)
d
11.62 (br s, 1H), 8.84 (d, J ¼ 2.7 Hz,
1H), 8.59 (dd, J ¼ 9.4, 2.7 Hz, 1H), 8.11 (d, J ¼ 9.4 Hz, 1H), 7.32e7.20
The two-step synthesis of isoalloxazines 8e13 from precursors
1, 2, 4e7 was performed in one-pot: reduction of corresponding
alkylated amines followed by immediate coupling. The reduction
protocol was modified from Ref. [33], and the coupling method was
modified from Ref. [37]. Both steps are sensitive to light, therefore
the reaction vessels were covered in aluminium foil and exposure
to light was minimised during work-up.
(m, 4H), 7.21e7.16 (m, 1H), 4.66 (t, J ¼ 7.7 Hz, 2H), 2.84e2.78 (m,
2H), 2.05 (t, J ¼ 7.8 Hz, 2H). 13C NMR (126 MHz, DMSO-d6)
d 171.2,
159.6, 156.0, 151.6, 144.3, 141.4, 137.2, 134.0, 128.8, 128.6, 128.3,
127.2, 126.4, 44.9, 32.5, 28.4. HRMS m/z calculated for C19H15N5O4:
[MþH]þ 378.1197, found: 378.1202.
4.5.3. 10-methyl-7-nitrobenzo[g]pteridine-2,4(3H,10H)-dione (10)
Started with compound 4 (500 mg, 2.536 mmol), Pd/C (27.6 mg,
A 50-mL two-necked round-bottomed flask was charged with
the corresponding alkylated aniline (1 equiv) suspended in dry
acetonitrile (1.35 mL/mmol alkylated amine). The flask was placed
under nitrogen atmosphere and dry triethylamine (Et3N; 4.7 equiv)
was added. Palladium on carbon (Pd/C, 10 wt%; 0.1 equiv) was
added. The reaction mixture was cooled to ꢀ15 ꢁC using a mixture
of sodium chloride:ice in 1:3 w/w ratio. While in the cooling bath, a
freshly prepared solution of formic acid (FA; 4.8 equiv) in dry
acetonitrile (1.35 mL/mmol alkylated aniline) was added dropwise
to the reaction mixture. The cooling bath was removed and the
reaction mixture was refluxed (at 80e85 ꢁC) for 3 h under nitrogen
atmosphere. The suspension colour changed during the reaction
from yellow to orange, green and lastly to red. After 3 h, the reac-
tion mixture was cooled down and the Pd/C was quickly filtered off
through a pad of Celite® using a sintered funnel and suction. For
compounds 8, 9, 10, 12 and 13: the Celite® was washed with glacial
acetic acid (5 mL). To the red-coloured solution, a mixture of
alloxan monohydrate (1.5 equiv) and boric acid (1.5 equiv) dis-
solved in hot glacial acetic acid (10 mL) was added and the reaction
mixture was stirred overnight at room temperature under nitrogen
atmosphere. For compound 11: the Celite® was washed with glacial
acetic acid (15 mL) and alloxan monohydrate (1.5 equiv) and boric
acid (1.5 equiv) were added as solids. The reaction mixture was
heated for 1 h at 90 ꢁC under nitrogen atmosphere. In all cases: the
solvent was removed under reduced pressure to yield a dark red
viscous oil. The product was purified by column chromatography
(except compound 10) using silica as stationary phase (60 Å particle
size). The solvents were removed under reduced pressure and the
solid dried in vacuo to yield the corresponding isoalloxazine.
0.259 mmol), Et3N (1650
mL, 11.838 mmol) and FA (460 mL,
12.192 mmol). For second step: alloxan monohydrate (302.5 mg,
1.890 mmol) and boric acid (116.8 mg, 1.890 mmol). The precipitate
formed over time was filtered off, washed with diethyl ether
(100 mL) and dried in vacuo. Orange solid, 78 mg, 0.284 mmol, 12%
over two steps. 1H NMR (500 MHz, DMSO-d6)
d
11.61 (s, 1H), 8.85 (d,
J ¼ 2.7 Hz, 1H), 8.64 (dd, J ¼ 9.4, 2.7 Hz, 1H), 8.11 (d, J ¼ 9.4 Hz, 1H),
3.99 (s, 3H). 13C NMR (126 MHz, DMSO-d6)
159.6, 155.8, 151.9,
d
144.5, 141.6, 138.0, 133.7, 128.4, 127.0, 118.4, 32.8. HRMS m/z calcu-
lated for C11H7N5O4: [MþH]þ 274.0571, found: 274.0570.
4.5.4. (S)-10-(1-(4-chlorophenyl)ethyl)-7-nitrobenzo[g]pteridine-
2,4(3H,10H)-dione (11)
Started with compound 5 (100 mg, 0.311 mmol), Pd/C (3.3 mg,
0.031 mmol), Et3N (204 mL,1.464 mmol) and FA (57 mL,1.495 mmol).
For second step: alloxan monohydrate (74.8 mg, 0.467 mmol) and
boric acid (28.9 mg, 0.467 mmol). Purified by column chromatog-
raphy (CH2Cl2:acetone 9:1, v/v ratio) as fourth fraction. Dark
orange-red solid, 50 mg, 0.126 mmol, 41% over two steps. 1H NMR
(500 MHz, CDCl3)
d
9.13 (d, J ¼ 2.7 Hz, 1H), 8.58 (br s, 1H), 8.35 (dd,
J ¼ 9.5, 2.7 Hz, 1H), 7.69 (d, J ¼ 7.4 Hz, 1H), 7.46 (d, J ¼ 9.5 Hz, 1H),
7.39 (d, J ¼ 7.4 Hz, 2H), 7.22e7.17 (m, 2H), 2.06 (d, J ¼ 7.2 Hz, 3H). 13
C
NMR (126 MHz, acetone-d6) d 158.5, 152.5, 144.1, 141.5, 137.4, 135.0,
133.1, 129.0, 128.0, 127.4, 127.0, 119.2, 53.4, 15.0.
4.5.5. (R)-7-nitro-10-(nonan-2-yl)benzo[g]pteridine-2,4(3H,10H)-
dione (12)
Started with compound 6 (100 mg, 0.323 mmol), Pd/C (3.4 mg,
0.032 mmol), Et3N (212 L, 1.520 mmol) and FA (58.6 L,
m
m
4.5.1. 10-butyl-7-nitrobenzo[g]pteridine-2,4(3H,10H)-dione (8)
Started with compound 1 (294 mg, 1.406 mmol), Pd/C (15.6 mg,
1.552 mmol). For second step: alloxan monohydrate (77.7 mg,
0.485 mmol) and boric acid (29.9 mg, 0.485 mmol). Purified by
column chromatography (CH2Cl2:acetone 9:1, v/v ratio) as fourth
fraction. A second purification was performed on this fraction using
(CH2Cl2:petroleum ether 9:1, v/v ratio). The dark orange-red solid
was collected as the fourth fraction, 50 mg, 0.126 mmol, 41% over
0.146 mmol), Et3N (959
mL, 6.880 mmol) and FA (265 mL,
7.027 mmol). For second step: alloxan monohydrate (337.6 mg,
2.109 mmol) and boric acid (130.4 mg, 2.109 mmol). Purified by
column chromatography (CH2Cl2:acetone 9:1, v/v ratio) as second
fraction. Yellow solid, 119 mg, 0.376 mmol, 27% over two steps. 1H
two steps. 1H NMR (500 MHz, CDCl3)
d 9.50 (s, 1H), 9.12 (d,
NMR (500 MHz, DMSO-d6)
d
11.60 (br s, 1H), 8.85 (d, J ¼ 2.7 Hz, 1H),
J ¼ 2.6 Hz, 1H), 8.61 (d, J ¼ 8.9 Hz, 1H), 8.07 (d, J ¼ 9.5 Hz, 1H), 6.53
(q, J ¼ 7.6 Hz, 1H), 2.11 (d, J ¼ 9.2 Hz, 2H), 1.74 (d, J ¼ 7.2 Hz, 3H), 1.22
(dqt, J ¼ 13.8, 10.2, 6.0 Hz, 10H), 0.83 (t, J ¼ 6.9 Hz, 3H). 13C NMR
8.61 (dd, J ¼ 9.4, 2.7 Hz, 1H), 8.14 (d, J ¼ 9.4 Hz, 1H), 4.58 (t,
J ¼ 7.4 Hz, 2H), 1.70 (p, J ¼ 7.9 Hz, 2H), 1.49 (sext, J ¼ 7.5 Hz, 2H), 0.97
(t, J ¼ 7.4 Hz, 3H). 13C NMR (126 MHz, DMSO-d6)
d 159.6, 156.0,
(126 MHz, CDCl3)
d 158.2, 154.9, 151.8, 144.5, 139.6, 136.0, 135.0,
151.5, 144.3, 141.7, 137.2, 134.0, 128.3, 127.2, 118.2, 44.93, 28.9, 19.9,
14.2. HRMS m/z calculated for C14H13N5O4: [MþNa]þ 338.0860,
found: 338.0857.
129.1, 128.0, 118.1, 54.4, 34.6, 31.6, 29.2, 28.9, 26.9, 22.5, 18.5, 14.0.
HRMS m/z calculated for C19H23N5O4: [MþH]þ 386.1823, found:
386.1824.
4.5.2. 7-nitro-10-(3-phenylpropyl)benzo[g]pteridine-2,4(3H,10H)-
dione (9)
4.5.6. (R)-10-(1-cyclohexylethyl)-7-nitrobenzo[g]pteridine-
2,4(3H,10H)-dione (13)
Started with compound 2 (119 mg, 0.395 mmol), Pd/C (4.7 mg,
Started with compound 7 (100 mg, 0.341 mmol), Pd/C (3.6 mg,
0.044 mmol), Et3N (290
mL, 2.061 mmol) and FA (79
mL,
0.034 mmol), Et3N (224 mL, 1.603 mmol) and FA (61.8 mL,
2.105 mmol). For second step: alloxan monohydrate (94.8 mg,
0.592 mmol) and boric acid (36.6 mg, 0.592 mmol). Purified by
column chromatography (CH2Cl2:acetone 9:1, v/v ratio) as second
fraction. Dark orange solid, 75 mg, 0.199 mmol, 46% over two steps.
1.637 mmol). For second step: alloxan monohydrate (81.9 mg,
0.512 mmol) and boric acid (31.6 mg, 0.512 mmol). Purified by
column chromatography (CH2Cl2:acetone 9:1, v/v ratio) as fourth
fraction. A second purification was performed on this fraction using
7

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
(CH2Cl2:petroleum ether 9:1, v/v ratio). The dark orange-red solid
References
was collected as the second fraction, 47 mg, 0.123 mmol, 36% over
two steps. 1H NMR (400 MHz, CDCl3)
d
9.08 (d, J ¼ 2.7 Hz, 1H), 9.02
[1] A.M. Edwards, in: E. Silva, A.M. Edwards (Eds.), Compr. Ser. Photochem.
Photobiol. Sci., Royal Society of Chemistry, Cambridge, 2007, pp. 1e11.
[2] J. Richtar, P. Heinrichova, D. Apaydin, V. Schmiedova, C. Yumusak,
A. Kovalenko, M. Weiter, N. Sariciftci, J. Krajcovic, Molecules 23 (2018) 2271.
(s, 1H), 8.54 (dd, J ¼ 9.6, 2.7 Hz,1H), 8.00 (d, J ¼ 9.6 Hz,1H), 6.22 (dq,
J ¼ 11.2, 7.2 Hz, 1H), 2.23e2.13 (m, 1H), 2.02 (d, J ¼ 9.6 Hz, 2H), 1.79
(d, J ¼ 6.7 Hz, 1H), 1.65 (d, J ¼ 7.2 Hz, 3H), 1.63e1.56 (m, 4H),
1.24e1.20 (m, 1H), 1.01e0.95 (m, 2H). 13C NMR (101 MHz, CDCl3)
ꢁ
ꢁ
[3] A. Kormanyos, M.S. Hossain, G. Ghadimkhani, J.J. Johnson, C. Janaky, N.R. de
Tacconi, F.W. Foss, Y. Paz, K. Rajeshwar, Chem. Eur J. 22 (2016) 9209e9217.
[4] W. Kaim, B. Schwederski, O. Heilmann, F.M. Hornung, Coord. Chem. Rev. 182
(1999) 323e342.
d
158.3, 154.8, 152.3, 144.8, 139.8, 136.3, 135.1, 129.3, 128.2, 118.3,
ꢁ
[5] A. Farran, J. Mohanraj, G.J. Clarkson, R.M. Claramunt, F. Herranz, G. Accorsi,
77.5, 77.2, 76.8, 59.1, 41.4, 30.9, 29.5, 25.6, 16.9. HRMS m/z calculated
for C18H19N5O4: [MþH]þ 370.1510, found: 370.1510.
Photochem. Photobiol. Sci. 12 (2013) 813.
[6] P. Menova, V. Eigner, J. Cejka, H. Dvorakova, M. Sanda, R. Cibulka, J. Mol. Struct.
1004 (2011) 178e187.
ꢂ
ꢂ
ꢁ
ꢁ
ꢂꢁ
ꢁ
[7] S. Weber, E. Schleicher (Eds.), Flavins and Flavoproteins, Springer, New York,
New York, NY, 2014.
4.6. Synthesis of compound 14
[8] S.K. Chapman, G.A. Reid (Eds.), Flavoprotein Protocols, Humana Press, Totowa,
N.J, 1999.
[9] S.T. Caldwell, L.J. Farrugia, S.G. Hewage, N. Kryvokhyzha, V.M. Rotello,
G. Cooke, Chem. Commun. (2009) 1350.
[10] Y. Murakami, J. Kikuchi, Y. Hisaeda, O. Hayashida, Chem. Rev. 96 (1996)
721e758.
[11] J.T. Slama, C. Radziejewski, S. Oruganti, E.T. Kaiser, J. Am. Chem. Soc. 106
(1984) 6778e6785.
The reaction protocol was modified from Refs. [36,55]. A 50-mL
round-bottomed flask was charged with compound 14 (55 mg,
0.259 mmol, 1 equiv) and glacial acetic acid (2.5 mL) and the flask
was placed under nitrogen atmosphere. Zinc dust (170 mg,
2.598 mmol, 10 equiv) was added in portions over 3 min. The
temperature of the reaction was maintained below 40 ꢁC using a
cold water bath. The reaction mixture was stirred 30 min at room
temperature, after which the catalyst was filtered off through a pad
of Celite® using a sintered funnel and suction. The Celite® was
washed with glacial acetic acid (5 mL). Alloxan monohydrate
(48.7 mg, 0.304 mmol, 1.2 equiv) and boric acid (24.4 mg,
0.395 mmol, 1.5 equiv) were added to the obtained diamine solu-
tion. The reaction mixture was heated for 1 h at 90 ꢁC under ni-
trogen atmosphere. The solvent was removed under reduced
pressure to obtain a dark red viscous oil. The flask was covered in
aluminium foil during the reactions.
[12] S. Ghisla, V. Massey, Biochem. J. 239 (1986) 1e12.
[13] B.J. Marsh, D.R. Carbery, Tetrahedron Lett. 51 (2010) 2362e2365.
[14] S. Chen, M.S. Hossain, F.W. Foss, Org. Lett. 14 (2012) 2806e2809.
ꢁ
[15] A.T. Murray, J.D. Challinor, C.E. Gulacsy, C. Lujan, L.E. Hatcher, C.R. Pudney,
P.R. Raithby, M.P. John, D.R. Carbery, Org. Biomol. Chem. 14 (2016)
3787e3792.
[16] A.T. Murray, P. Matton, N.W.G. Fairhurst, M.P. John, D.R. Carbery, Org. Lett. 14
(2012) 3656e3659.
[17] Y.-M. Legrand, M. Gray, G. Cooke, V.M. Rotello, J. Am. Chem. Soc. 125 (2003)
15789e15795.
[18] S. Sayin, G. Uysal Akkus¸ , R. Cibulka, I. Stibor, M. Yilmaz, Helv. Chim. Acta 94
(2011) 481e486.
[19] E.M. Seward, R.B. Hopkins, W. Sauerer, S.W. Tam, F. Diederich, J. Am. Chem.
Soc. 112 (1990) 1783e1790.
ꢁ
[20] M.
Mollahosseini,
E.
Karunaratne,
G.N.
Gibson,
J.A.
Gascon,
F. Papadimitrakopoulos, J. Am. Chem. Soc. 138 (2016) 5904e5915.
[21] H.A. Staab, M.F. Zipplies, T. Müller, M. Storch, C. Krieger, Chem. Ber. 127 (1994)
1667e1680.
[22] Y. Kamano, Y. Tabata, H. Uji, S. Kimura, RSC Adv. 9 (2019) 3618e3624.
[23] S. Shinkai, H. Nakao, T. Tsuno, O. Manabe, A. Ohno, J. Chem. Soc., Chem.
Commun. (1984) 849e850.
[24] S. Shinkai, T. Yamaguchi, O. Manabe, F. Toda, J. Chem. Soc., Chem. Commun.
(1988) 1399e1401.
4.6.1. 10-butyl-8-fluorobenzo[g]pteridine-2,4(3H,10H)-dione (14)
Purified by column chromatography (CH2Cl2:acetone 9:1, v/v
ratio) as second fraction. Yellow solid, 44 mg, 0.153 mmol, 59%. 1H
NMR (500 MHz, Acetone-d6)
d
8.22 (dd, J ¼ 8.6, 6.1 Hz, 1H), 7.79 (dd,
J ¼ 10.7, 2.5 Hz, 1H), 7.50 (td, J ¼ 8.6, 2.5 Hz, 1H), 4.70 (t, J ¼ 8.0 Hz,
2H), 1.87 (p, J ¼ 7.7 Hz, 2H), 1.57 (dt, J ¼ 15.0, 7.5 Hz, 2H), 1.02 (t,
J ¼ 7.4 Hz, 3H). 13C NMR (126 MHz, DMSO-d6)
d 160.1, 156.1, 151.0,
[25] H. Nakade, B.J. Jordan, H. Xu, G. Han, S. Srivastava, R.R. Arvizo, G. Cooke,
V.M. Rotello, J. Am. Chem. Soc. 128 (2006) 14924e14929.
138.4, 135.1 (d, J ¼ 11.2 Hz), 134.7, 132.5, 115.2 (d, J ¼ 24.5 Hz), 103.4
ꢂ
ꢂ
ꢁ
ꢂꢁ
ꢁ
[26] J. Zurek, E. Svobodova, J. Sturala, H. Dvorakova, J. Svoboda, R. Cibulka, Tetra-
hedron Asymmetry 28 (2017) 1780e1791.
(d, J ¼ 28.6 Hz), 44.9, 28.8, 19.9, 14.2. HRMS m/z calculated for
C
14H13FN4O2: [MþH]þ 289.1095, found: 289.1097.
[27] S. Shinkai, H. Nakao, I. Kuwahara, M. Miyamoto, T. Yamaguchi, O. Manabe,
J. Chem. Soc. Perkin 1 (1988) 313.
[28] S. Shinkai, T. Yamaguchi, A. Kawase, A. Kitamura, O. Manabe, J. Chem. Soc.
Chem. Commun. (1987) 1506.
[29] A. Khalaj, A. Doroudi, N. Adibpour, G.M. Araghi, Asian J. Chem. 21 (2009)
997e1001.
[30] A. Khalaj, K. Abdi, S. Nasser, M.R. Khoshayand, H.A. Nedaie, Chem. Biol. Drug
Des. 83 (2014) 183e190.
[31] M. Miyashita, I. Shiina, T. Mukaiyama, Bull. Chem. Soc. Jpn. 67 (1994)
210e215.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper
[32] E.C. Butler, A.P. Davis, J. Photochem. Photobiol. Chem. 70 (1993) 273e283.
[33] G. Piersanti, L. Giorgi, F. Bartoccini, G. Tarzia, P. Minetti, G. Gallo, F. Giorgi,
M. Castorina, O. Ghirardi, P. Carminati, Org. Biomol. Chem. 5 (2007) 2567.
[34] M. Orlandi, D. Brenna, R. Harms, S. Jost, M. Benaglia, Org. Process Res. Dev. 22
(2018) 430e445.
[35] X. Zhou, X. Wu, B. Yang, J. Xiao, J. Mol. Catal. Chem. 357 (2012) 133e140.
[36] M. Bejugam, S. Sewitz, P.S. Shirude, R. Rodriguez, R. Shahid,
S. Balasubramanian, J. Am. Chem. Soc. 129 (2007) 12926e12927.
[37] P. Hemmerich, S. Fallab, H. Erlenmeyer, Helv. Chim. Acta 39 (1956)
1242e1252.
Acknowledgments
This paper is a contribution to special issue in memory of Prof.
Jon Williams. An excellent colleague and chemist, he was a strong
presence in our department who is sorely missed.
The authors gratefully acknowledge an EPSRC DTP (to D.M.R.)
and the University of Bath for a visiting PhD researcher position
(T.M.) which was also supported by Japan Student Services Orga-
nization (JASSO).
[38] When attempted to reduce an N10 Valine-substituted derivative with Pd/C
and FA-Et3N, it was observed that the newly formed amine undergoes an
intramolecular
b
ꢀelimination to form a quinoxaline and dihydroquinoxaline.
This fast and favourable intramolecular cyclisation is supported by mass
spectrometry data from the reaction mixture.
Appendix A. Supplementary data
[39] K. Mallesh, T. Kavitha, V. Malla Reddy, Acta Cienc. Indica - Phys. 15 (1989)
67e74.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tet.2021.131925.
€
[40] P. Kirsch, A. Schonleben-Janas, R.H. Schirmer, Liebigs Ann. 1995 (1995)
1275e1281.
8

ꢀ ꢀ
D.-M. Rasadean, T. Machida, K. Sada et al.
Tetrahedron 82 (2021) 131925
[48] M.M. Martin, Chem. Phys. Lett. 35 (1975) 105e111.
ꢁ
[41] A.H. Tolba, F. Vavra, J. Chudoba, R. Cibulka, Eur. J. Org Chem. 2020 (2020)
1579e1585.
[49] E. Sikorska, I.V. Khmelinskii, W. Prukała, S.L. Williams, M. Patel, D.R. Worrall,
J.L. Bourdelande, J. Koput, M. Sikorski, J. Phys. Chem. 108 (2004) 1501e1508.
[50] H. Masuhara, H. Shioyama, T. Saito, K. Hamada, S. Yasoshima, N. Mataga,
J. Phys. Chem. 88 (1984) 5868e5873.
ꢁ
[42] J. Shirdel, A. Penzkofer, R. Prochazka, J. Daub, E. Hochmuth, R. Deutzmann,
Chem. Phys. 326 (2006) 489e499.
[43] E. Sikorska, I.V. Khmelinskii, D.R. Worrall, J. Koput, M. Sikorski, J. Fluoresc. 14
(2004) 57e64.
€
[51] T. Bruhn, A. Schaumloffel, Y. Hemberger, G. Bringmann, Chirality 25 (2013)
€
[44] B. Attenberger, H. Schmaderer, B. Konig, Synthesis 2008 (2008) 1767e1774.
243e249.
[45] I. Daidone, A. Amadei, M. Aschi, L. Zanetti-Polzi, Spectrochim. Acta. A. Mol.
[52] J.J.P. Stewart, J. Mol. Model. 19 (2013) 1e32.
[53] F. Neese, WIREs Comput. Mol. Sci. 8 (2018), e1327, https://doi.org/10.1002/
wcms.1327.
Biomol. Spectrosc. 192 (2018) 451e457.
€
[46] M. Spexard, D. Immeln, C. Thoing, T. Kottke, Vib. Spectrosc. 57 (2011)
282e287.
[54] A.-R. Allouche, J. Comput. Chem. 32 (2011) 174e182.
[55] F. Kipnis, N. Weiner, P.E. Spoerri, J. Am. Chem. Soc. 69 (1947) 799e800.
[47] P.F. Heelis, Chem. Soc. Rev. 11 (1982) 15e39.
9