Interactions of Enolizable Barbiturate Dyes

Article abstract of DOI:10.1002/chem.201504932

The specific barbituric acid dyes 1-n-butyl-5-(2,4-dinitro-phenyl) barbituric acid and 1-n-butyl-5-{4-[(1,3-dioxo-1H-inden-(3 H)-ylidene)methyl]phenyl}barbituric acid were used to study complex formation with nucleobase derivatives and related model compo

Full text of DOI:10.1002/chem.201504932

DOI: 10.1002/chem.201504932
Full Paper
&
Molecular Recognition
Interactions of Enolizable Barbiturate Dyes
Alexander Schade,[a] Katja Schreiter,[a] Tobias Rüffer,[b] Heinrich Lang,[b] and Stefan Spange*[a]
Abstract: The specific barbituric acid dyes 1-n-butyl-5-(2,4-
dinitro-phenyl) barbituric acid and 1-n-butyl-5-{4-[(1,3-dioxo-
1H-inden-(3H)-ylidene)methyl]phenyl}barbituric acid were
used to study complex formation with nucleobase deriva-
tives and related model compounds. The enol form of both
compounds shows a strong bathochromic shift of the UV/
Vis absorption band compared to the rarely coloured keto
form. The keto–enol equilibria of the five studied dyes are
strongly dependent on the properties of the environment as
shown by solvatochromic studies in ionic liquids and a set
of organic solvents. Enol form development of the barbituric
acid dyes is also associated with alteration of the hydrogen
bonding pattern from the ADA to the DDA type (A=hydro-
gen bond acceptor site, D=donor site). Receptor-induced al-
tering of ADA towards DDA hydrogen bonding patterns of
the chromophores are utilised to study supramolecular com-
plex formation. As complementary receptors 9-ethyladenine,
1-n-butylcytosine, 1-n-butylthymine, 9-ethylguanidine and
2,6-diacetamidopiridine were used. The UV/Vis spectroscopic
response of acid–base reaction compared to supramolecular
1
complex formation is evaluated by H NMR titration experi-
ments and X-ray crystal structure analyses. An increased
acidity of the barbituric acid derivative promotes genuine
salt formation. In contrast, supramolecular complex forma-
tion is preferred for the weaker acidic barbituric acid.
coupled at the five position of the six-membered BA ring.[7] As
long as the BA exists in its keto form, those compounds are
rarely coloured because the p-conjugation is interrupted by
a sp3-hybridised carbon atom. In contrast, the enol form of
such BA dyes possesses a strong positive resonance effect and
a push–pull chromophore results, which features an intense
UV/Vis absorption band (Scheme 1).
Introduction
The pairing of nucleobases and of related model compounds
by multiple hydrogen bonds is a fascinating field of research
owing to its importance in nature.[1] Multiple hydrogen-bond
formation has been studied by NMR[2] and IR spectroscopy,[3] X-
ray diffraction,[4] melting points,[5] or theoretical calculations.[6]
In several previous works we have shown that base pairing
can also be observed by UV/Vis spectroscopy if one of the
complementary partners contains a chromophore, which is
characteristically influenced by hydrogen bond formation.[7]
Barbituric acid (BA) derivatives are well-established as model
receptors for studying multiple hydrogen-bond formations
with nucleobase derivatives and related compounds.[3,8] BAs
are readily available and can be chemically modified on
demand.[9] For this purpose, barbiturate dyes have been con-
structed in such a way that the hydrogen bond formation af-
fects the electron density of one of the substituent of the chro-
mophore. Conceptually, an electron-withdrawing substituent
such as the 4-nitrophenyl group or the 4-pyridinium ring is
Scheme 1. Principle of induction in the push–pull system by keto-enol tau-
tomerism of a barbiturate dye and resonance formulary of the enol form.
EWG=electron withdrawing group.
However, several external influences can trigger the occur-
rence of the enol form. Supramolecular complex formation or
common acid–base reaction can generate the enol or enolate
form. Furthermore, the nature of solvents can have a fluctuat-
ing effect on that keto–enol equilibrium and thus the associat-
ed colour. If the precipitating agent will abstract the acidic
proton of the BA, the overall HBA (hydrogen-bond-accepting)
strength of the former ADA sequence is significantly enhanced.
This effect was utilised by pyridinium-substituted barbiturate
betaine dyes of the Reichardt-type,[7b] which showed optical re-
sponse towards nucleobase derivatives via the anionically
charged barbiturate moiety.[7b]
[a] A. Schade, Dr. K. Schreiter, Prof. Dr. S. Spange
Technische Universität Chemnitz, Department of Polymer Chemistry
09107 Chemnitz (Germany)
Fax: (+49)371-531-21239
E-mail: stefan.spange@chemie.tu-chemnitz.de
[b] Dr. T. Rüffer, Prof. Dr. H. Lang
Technische Universität Chemnitz, Department of Inorganic Chemistry
09107 Chemnitz (Germany)
Supporting information for this article can be found under http://
dx.doi.org/10.1002/chem.201504932.
Chem. Eur. J. 2016, 22, 5734 – 5748
5734
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Related studies were done with the enolizable chromophore
1-n-butyl-5-(4-nitrophenyl)barbituric acid (NPBA), which
showed the complexity of this field.[7a] Unfortunately, the opti-
cally measured effects by UV/Vis spectroscopy, which are
caused by supramolecular complex formation of NPBA, were
not very significant. In this work, we will continue the study on
chromophoric BA derivatives. The aim is to locate the way to
detect optically any nucleobase by a colour change and deter-
mine crucial properties of the receptor dye, which are required
to achieve this target.
Concerning this long-term objective, there is one important
issue, which will be treated in this manuscript to complement
the conceptual approach: Does an increased Brønsted acidity
of the BA chromophore enhance the complex formation with
a nucleobase derivative or does it gather preferentially the
genuine salt formation? In this part, the anionically charged BA
component will be operating on supramolecular complex for-
mation.
Scheme 3. Keto–enol tautomerism of 1-n-butyl-5-(2,4-dinitrophenyl)barbitu-
ric acid (4b). The resulting hydrogen bonding sequence pattern (A=H-bond
acceptor, D =H-bond donor) for each tautomer is indicated.
The expected tautomeric species among the keto–enol equi-
libria of 1-n-butyl-5-(2,4-dinitrophenyl)-barbituric acid (4b) are
shown in Scheme 3.
An increased acceptor strength can be simply achieved by
insertion of an additional nitro group in ortho-position of the
phenyl ring of NPBA. Those compounds 4 are accessible by nu-
cleophilic aromatic substitution of 2,4-dinitrofluoro-benzene
with barbituric acid salts.[10] The mono N-alkyl substituted BA
4b is beneficial to obtain a better solubility in nonpolar sol-
vents.[7a,11] Additionally, the n-butyl group barricades one hy-
drogen-bonding sequence of the BA, thus solely a 1:1-complex
formation can take place.[7a,c]
Optical properties of the enolizable barbituric acid dyes 4a–
c, 7b, and 7c are measured in a set of 36 various organic sol-
vents and twelve ionic liquids to investigate the solvatochrom-
ism of the dyes as function of various solvent properties. Goal
of this specific work is to obtain quantitative information
about the effect of the environments HBD (hydrogen-bond-do-
nating), HBA (hydrogen-bond-accepting), and dipolarity/polar-
izability properties on the shift of the UV/Vis absorption band
of the dyes by using the linear solvation energy (LSE) relation-
ships of Kamlet–Taft[13] and Catalµn,[14] respectively. This solva-
tochromic method was applied for investigations of several
solvatochromic dyes of the BA type and others.[7b,15]
As a second chromophore with weaker acidity, 5-(4-((1,3-
dioxo-1H-inden-2(3H)-ylidene)methyl)phenyl)barbituric acid de-
rivatives 7b,c have been chosen (Scheme 2). Compounds of
type 7 possess a larger conjugated p-system and show
a longer wavelength UV/Vis absorption band and a weaker
acidity than 4 or NPBA.
Brønsted acid strength of the dyes has been measured by
common acid–base titration and the results have been evaluat-
ed by means of the established Henderson–Hasselbalch equa-
tion (HHE) according to our previous study.[7a]
According to our previous studies,[7a,b,11] the nucleobase de-
rivatives 1-n-butylthymine (BuTy), 9-ethyladenine (EtAd), 1-n-
butylcytosine (BuCy), and 9-ethylguanine (EtGu), and 2,6-diace-
tamidopyridine (DACP) were used as receptors (Scheme 4) to
study the base pairing effect.
Complex formations of 4b,c and 7b,c, respectively, with the
different receptors of Scheme 4 were studied by means of UV/
Vis and NMR titration experiments. Solid-state structures were
examined using X-ray structure analyses of suitable crystals of
BAs and their available adducts with nucleobase derivatives.
Scheme 2. Molecular structures of the five barbituric acid dyes 4a–c and
7b,c that were used in this study.
The N,N-dialkylated derivatives 4c and 7c (Scheme 2) were
used as reference chromophores to study the genuine salt for-
mation, because both hydrogen bonding sequences are
blocked. Therefore, the UV/Vis spectroscopic respond is prefer-
entially related to an acid–base reaction of the chromophore
with a nucleobase derivative or a related compound.[12]
Scheme 4. Receptors and their hydrogen-bonding sequences (A=H-bond
acceptor, D =H-bond donor) used in this work.
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5735
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
nitro group, and one has the enol group on the opposite side
(Scheme 3) in the asymmetric unit together with two ethanol
solvent molecules (Figure 1). The carbon atoms C8 and C18 of
solid 4a are sp2-hybridised, as revealed by the sum of their
bond angles of 359.9(3)/359.9(3)8 for C8/C18 (Figure 1). Conse-
quently, the hydrogen atoms were found as enol species at
the oxygen atoms O6 and O13, interacting as hydrogen bond
donors with ethanol molecules (Scheme 7; Supporting Infor-
mation, Figure S29). The different CÀC and CÀO bond lengths
at the C7/C9 and C17/C19 atoms of the barbituric acid
(Figure 1) validate the formation of an enol species.
Results and Discussion
Synthesis
The synthetic procedure for dyes 4a–c is shown in Scheme 5
and for dyes 7b,c in Scheme 6. A detailed procedure is given
in the Experimental Section.
Scheme 5. Synthesis route for 5-(2,4-dinitrophenyl)barbituric acids 4a–c.
Figure 1. ORTEP (ellipsoids set at 50% probability) of the asymmetric unit of
4a. Selected bond lengths [Š] and angles [8]: C8–C4 1.474(2), C8–C7
1.437(2), C8–C9 1.375(2), C7–O5 1.241(2), C9–O6 1.321(2), C18–C14 1.472(2),
C18–C17 1.432(2), C18–C19 1.372(2), C17–O12 1.244(2), C19–O13 1.324(2);
C4-C8-C7 118.9(2), C4-C8-C9 122.9(2), C7-C8-C9 118.1(2), C14-C18-C17
120.2(1), C14-C18-C19 121.8(2), C17-C18-C19 117.9(2).
Scheme 6. Synthesis of 5-(4-((1,3-dioxo-1H-inden-2(3H)-ylidene)methyl) phe-
nyl)barbituric acids (7b,c).
The barbituric acid 5-(4-((1,3-dioxo-1H-inden-2(3H)-ylidene)-
methyl)phenyl) derivatives 7b,c were synthesised utilising nu-
cleophilic aromatic substitution of the corresponding fluoro
compound (5), which is accessible by a Knoevenagel conden-
sation reaction. Mentionable for a successful reaction is the
use of CaCO3 to trap the generated fluoride anion completely.
Otherwise, the fluoride anion is able to catalyse the retro-Knoe-
venagel condensation, which has a negative impact on the
overall yield.
Scheme 7. Illustration of the two strands of 4a within the crystallographic
structure and their hydrogen bonds. Dark grey and grey squares represent
the different conformation of 4a, the grey ovals are ethanol molecules, the
lighter squares and ovals belong to the strand in the back.
Molecules of 4a form a chain along the crystallographic
c axis, connected by two hydrogen bonds. Involved into these
hydrogen bonds are the hydrogen bond donor groups N3–
H3N, N4–H4N, N7–H7N, and N8–H8N as well as the hydrogen
bond acceptors O5, O7, O12, and O14 (Figure 1). Two of such
chains interact with each other by hydrogen bonds between
two ethanol molecules. Both strands run in opposite directions
and always the different conformational molecules lay next to
each other (Scheme 7). These hydrogen bonded double-
strands are connected further by p–p interactions, forming
a 3D network (Figure S30).
In spite of the low yield in several cases, by reason of the
simple procedure the products can be easily separated from
the reactants.
Structure of 4a and 4c in the solid state
Compound 4a crystallises from an ethanol/CH2Cl2 mixture
¯
within the triclinic space group P1. Two crystallographically dif-
ferent molecules of 4a in two different conformations were
obtained; one has the enol group on the side of the ortho-
Compound 4c crystallises from a CH2Cl2/n-pentane mixture
in the monoclinic space group P21/c. Importantly, 4c forms
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5736
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. pKa values of 4a–c and 7b,c.
4a
4b
4c
7b
7c
pKa (1)
pKa (2)
0.14
11.76
0.13
12.42
À0.23
1.35
/
1.34
/
[a]
[a]
/
[a] Decolourisation at pH>6 is due to base reacting with the benzylidene
double bond.
tion spectra of the titration experiments are given in the Sup-
porting Information (Figures S16–S27).
Figure 2. ORTEP (ellipsoids set at 50% probability) of the molecular structure
of 4c. Selected bond lengths [Š] and angles [8]: C8–C4 1.508(2), C8–C7
1.516(2), C8–C9 1.517(2), C7–O5 1.209(2), C9–O6 1.209(2); C4-C8-C7 114.7(1),
C4-C8-C9 112.4(1), C7-C8-C9 113.5(1).
Solvatochromic studies
The evaluation of the measured solvatochromic data (n˜max) has
been performed by LSE (linear solvation energy) relationships
using either the Kamlet–Taft[13] [Eq. (4)] or the Catalµn equa-
tion[14] [Eq. (5)].
a ketobarbituric acid in the solid state (Figure 2) which is in
contrast to 4a. We point out that the carbon atom C8 is sp3-
hybridised with a sum of the bond angles of 340.6(2)8. Addi-
tionally, the bond lengths of the atoms C7/C9 to C8 are of
equal distance (Figure 2).
*
~
~
nmax ¼ nmax,0 þ aa þ bb þ sp
ð4Þ
ð5Þ
~
~
nmax ¼ nmax,0 þ aSA þ bSB þ dSP þ eSdP
Compound 4c in the solid state is present as dimer, owing
to an intermolecular p–p interaction between the ortho-nitro
group and the phenyl ring of two different molecules (Fig-
ure S31).
n
˜
is the measured solvent dependent UV/Vis absorption
max
maximum of a dye in a specific solvent. n
˜
is the calculated
max,0
These crystallographic studies show that depending on the
substitution pattern and used solvents either the keto or enol
tautomeric constitution occurs in the solid state.[16]
UV/Vis absorption maximum of the solvent-unaffected dye,
a is the HBD parameter, SA is the solvent acidity, which relates
to a, b is the HBA parameter, SB is the solvent basicity, which
relates to b, p* is the solvent dipolarity/polarizability parame-
ter, SdP is the solvent dipolarity, and SP the solvent polarizabili-
ty. a, b, and s as well as d and e are solvent-independent corre-
lation coefficients. The advantage of the Catalµn LSE relation-
ship is that it considers the solvent polarizability and dipolarity
as independent terms, whereas Kamlet–Tafts’s p* is a blend of
both.
Acidity measurements in aqueous solution
The pKa values for all of the compounds were determined by
the HHE [Eq. (1)][17] in a mixture of water/methanol (ratio 5:1,
v/v) at 208C by pH-dependent UV/Vis titration. The absorbance
of the UV/Vis absorption maximum of the neutral, the mono-
anionic, and the dianionic barbituric acids, respectively, were
plotted as function of the pH [Eq. (2) and (3)]. The parent bar-
bituric acid derivatives show UV/Vis absorption maxima at
about 350 nm, and the mono-anionic form at about 400 nm.
The dianions of barbituric acids 4a or 4b absorbs at about
440 nm in the respective media. The second deprotonation
step of 4a or 4b takes place at the NH group, which is verified
by the fact that the N,N-dialkylated derivative 4c did not show
this UV/Vis absorption band at 444 nm at pH>11.
Multiple square correlation analysis of n˜max of the dyes 4a–
c and 7b,c measured in 36 organic solvents and twelve ILs
with the empirical solvent parameters according to Equa-
tion (4), or (5) are performed to examine the correlation coeffi-
cients a, b, s, d, e.
Correlation coefficients of solvatochromic dyes determined
from Eq. (4) and (5) deliver predictions on preferred solvation
sites of the dye by a distinct property of the solvent. Thus, the
knowledge of the coefficients allows a semi-quantitative inter-
pretation on specific and nonspecific solvation of a solvatochro-
mic molecule.[15b,h]
pKa ¼ pHÀlog ðcAÀ =cHA
Þ
ð1Þ
ð2Þ
ð3Þ
Recently, SP and SdP parameters of ILs were firstly deter-
mined by our group.[19] Hence, we are specifically interested in
the application of these SP and SdP parameters of IL for inter-
pretation of solvatochromic data to show their reliability.
The UV/Vis spectra of compounds 4a–c and 7b,c were mea-
sured in 36 various organic solvents and compounds 4a–c ad-
ditionally in twelve ILs. The UV/Vis absorption maxima of 4a–
c and 7b,c or their deprotonated species 4aq–cq and 7bq,cq
are presented in the Supporting Information, Table S2 and S3.
The Kamlet–Taft and Catalµn empirical solvent parameters for
pKa ð1Þ ¼ pHÀlog ðABSÀ =ABSÞ
pKa ð2Þ ¼ pHÀlog ðABS2À =ABSÀ
Þ
The obtained pKa values are summarised in Table 1. Compared
to NPBA (pKa ꢀ2.0),[7a] 5-phenylbarbituric acid (pKa =2.54)[18]
and the parent BA (pKa =4.02),[18] 4a–c are more acidic (pKa
ꢀ0.0). The BA derivatives with an enlarged p-system (7b,c) are
surprisingly more acidic than NPBA (pKa ꢀ1.3), but less acidic
than 4a–c. Detailed information as well as the UV/Vis absorp-
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5737
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
these solvents were taken from previous reports and are given
in the Supporting Information, Table S1.
number of solvent molecules are involved on the molecular
level.
Representative UV/Vis spectra of 4bq measured in six sol-
vents are shown in Figure 3 and in six ILs in Figure 4.
This issue must be considered to understand the measured
effects of the solvent upon the position of the UV/Vis absorp-
tion band. If the solvent shell serves as base medium, a proton
is abstracted from the dye and the anionic form is observed.
This feature occurs when the dye is measured in solvents pos-
sessing moderate and strong hydrogen-bond-accepting (HBA)
properties. Hence, the dissolution process of 4a–c in such sol-
vents is associated with an acidochromic process. Whether the
anionic shape of the dye (4aq–cq) is released in any solvent in-
stead of 4a–c. This can be easily tested when the strong base
DBU (1,8-diazabicyclo[5.4.0]undec-7-ene) (pKa DBU-H+ =12.0) is
added in a small portion (cDBU =cdye) to the dye solution. If the
UV/Vis absorption band of the dissolved dye is not affected by
DBU traces, the anionic form is already present. Due to the po-
sition of UV/Vis absorption, bands of anionic dyes are also sol-
vent-dependent,[15a] the DBU test is necessary to ensure the be-
haviour of the dye in those solvents, which are very weak
bases. Thus, in the following classes of solvents the dye is pres-
ent in its anionic form: aliphatic alcohols, amides, ketones, car-
bonic acid ester, amines, and ionic liquids.
For all of the compounds, the largest hypsochromic shift
was measured in the most acidic solvent used, 1,1,1,3,3,3-hexa-
fluoro-2-propanol (HFIP), and the largest bathochromic shift in
the most basic solvent used, hexamethylphosphoric triamide
(HMPA). The solvatochromic range Dn˜ [Dn˜ =n˜ (HFIP)Àn˜
(HMPA)] of these compounds amounts to 5110 cmÀ1 for 4cq,
5840 cmÀ1 for 4bq, and 7600 cmÀ1 for 4aq.
The concentration of the dye is very low (c=5”
10À5 molLÀ1). Thus the solvent is existent in about 100000-fold
excess to the dye. Therefore, also a weak base with pKa <2 at
10À4 molLÀ1 becomes partly protonated, for example, ethanol
(pKa C2H5OH2+ =À2.4). This can be estimated by applying the
acid base equilibrium equation but it is not clear which
It is observable that even in the strong HBD solvent
1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), the anionic form is
formed, as shown by the fact that DBU addition does not
affect the UV/Vis absorption maximum in this solvent. Further-
more, in alcoholic solvents the dyes show an asymmetric UV/
Vis absorption band, which indicates that the solvation of the
enolate species seems manifold. Consistently the longest UV/
Vis absorption band has been used in the multiple square cor-
relation analyses for the evaluation of Equations (4) and (5).
The keto form of the dye is only present in few solvents,
such as halogenoalkanes (alkyl halides). Therefore, a colourless
or pale-yellow-coloured (lmax <300 nm) solution is observed.
Usually, for the multiple square correlation analyses of n˜max
as function of SA, SB, SP, and SdP (or a, b, and p*) these sol-
vents have to be excluded where a chemical alteration of the
dye takes place.
Figure 3. UV/Vis spectra of 4bq in six different solvents (HFIP: 1,1,1,3,3,3-hex-
afluoro-2-propanol; TFE: 1,1,1-trifluoroethanol; EtOH: ethanol; DMSO: dime-
thylsulfoxide; HMPA: hexamethylphosphoric triamide).
The results of the multiple square correlation analyses are
compiled in Table 2. Graphically depicted results were given in
Figure 5 and the Supporting Information (Figures S1–S3).
Altogether, the calculated coefficient a shows a positive al-
gebraic sign for the compounds studied, which indicates a hyp-
sochomic effect with increasing acidity of the solvent (a or
SA). This result indicates the preferred solvation of the barbitu-
rate anion in acidic solvents. This explanation is in agreement
with solvatochromic phenomena of established betaine dyes
such as Reichardt’s dye[14h] or other negatively charged dyes.[15a]
The positive solvatochromic effect of the HBA ability of the
solvents (b<0) results from the interactions of the NH groups
of the BA anion with an electron lone pair of the solvent. With
increasing the number of N-alkylation sites, the b coefficient
decreases:
Figure 4. UV/Vis spectra of 4bq in six ionic liquids. [BMIM]=1-butyl-3-meth-
ylimidazolium; [OMIM]=1-octyl-3-methylimidazolium; [BPyr]=1-butylpyridi-
nium; [TBMN]=tributylmethylammonium;
b from Kamlet–Taft: À1.397 (4aq)!À0.732 (4bq)!À0.434 (4cq)
b from Catalµn: À0.792 (4aq)!À0.493 (4bq)!Æ0.000 (4cq)
NTf2 =bis(trifluoromethylsulfonyl)imide.
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5738
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
electronic ground state. This result is well-known for chromo-
phores with a neutral ground state and a polar first exited
state, such as nitroaniline derivatives.[20] A related result was
found for anionic thiazole based chromophores with a nitro
group as acceptor moiety.[15a]
Table 2. Calculated solvent-independent correlation coefficients a, b, and
s according to the Kamlet–Taft Equation (4) and a, b, d, and e of the
Catalµn Equation (5) for the solvatochromic response of 4a–c.[a]
4aq
23.042
4bq
22.330
4cq
22.217
Kamlet–Taft Equation (4)
n˜max,0
With the help of the Catalµn Equation (5), the influence of
polarizability and dipolarity on shift of n˜max can be separately
quantified. Both coefficients are negative indicating a batho-
chromic shift with increased polarizability (d<0) as well as di-
polarity (e<0).
a
b
s
n
2.662
À1.397
À1.869
32
2.305
À0.732
À1.553
30
2.171
À0.434
À1.754
33
sd
0.310
0.323
0.307
r
0.98
0.97
0.97
The polarizability of the solvent influences the UV/Vis ab-
sorption maximum of these dyes two times more than the di-
polarity. This d/e ratio matches their contributions to the p*
parameter[14h] and agrees with the results from the Kamlet–Taft
correlations.
f
ꢁ0.0001
26.863
4.478
À0.792
À4.529
À2.574
30
ꢁ0.0001
25.508
4.266
À0.493
À3.137
À2.561
28
ꢁ0.0001
25.371
3.886
Æ0.000
À4.124
À1.949
31
Catalµn Equation (5)
n˜max,0
a
b
d
e
n
sd
r
The ILs fit well in the correlation plot among the organic sol-
vents, which is shown in Figure 5 for the solvatochromism of
4cq.
This result shows that the empirical solvent parameters SP
and SdP of ILs[19] can be readily applied to interpret solvato-
chromic effects in comparison to organic solvents.
0.315
0.97
0.294
0.97
0.200
0.98
f
ꢁ0.0001
ꢁ0.0001
ꢁ0.0001
[a] Owing to the anionic form is generated in the solvents, 4aq–cq are ob-
served in reality. Additionally, given is the correlation coefficient r, the
standard deviation sd, the number of solvents n and the significance f.
A correlation of n˜max,3b versus ET(30) shows a moderately
good correlation (Supporting Information, Figure S6), because
both dyes show a strong hypsochromic effect caused by acidic
solvents (a>0).[14h] They behave different by the effect of po-
larizability and dipolarity of solvents. The anionic barbiturate
dyes 4aq–cq show a bathochromic shift of the UV/Vis absorp-
tion maximum as function of n˜max in contrast to Reichardt’s
ET(30) dye, which shows a hypsochromic shift of the UV/Vis ab-
sorption maximum with increased dipolarity/polarizability of
the solvent.
The UV/Vis absorption spectra of the chromophoric dyes 7b
and 7c were also measured within various organic solvents
(Supporting Information, Table S3). The solvatochromic range
Dn˜ of these compounds amounts to 6420 cmÀ1 for 7b (HFIP
to HMPA) and 6700 cmÀ1 for 7c (HFIP to HMPA). The results of
the multiple square correlation analyses of n˜max with the sol-
vent parameters are given in Table 3 and in the Supporting In-
formation (Figures S4 and S5).
Both dyes show a strong hypsochromic UV/Vis shift with in-
creased solvent acidity (a>0). This effect is similar to the re-
sults of chromophores 4a–c and also indicates a deprotonation
of the chromophores during dissolution of 7b,c. The hypso-
chromic shift of the UV/Vis absorption maximum results from
the interaction of HBD solvents with the anionic barbiturate
unit. The UV/Vis absorption maxima did not change after addi-
tion of DBU, indicating that the anionic forms 7bq,cq were al-
ready present.
Figure 5. Calculated UV/Vis absorption maxima of compound 4cq by Catalµn
Equation (5) with coefficients from Table 2 versus measured UV/Vis absorp-
&
tion maxima in a set of 19 organic solvents ( ) and 12 ILs (+).
This result shows that even the anionic form is able to inter-
act with a donor solvent, which is measurable by the solvato-
chromic response and it is of importance to understand the
supramolecular complex formation of the anionic forms of bar-
biturates, which will be discussed below.
A bathochromic shift of the UV/Vis absorption maximum
was found for 7bq with increasing SB of the solvent, because
of the interaction by HBA solvents with the NH function. The
solvatochromic shift of the chromophore 7cq shows no de-
pendence on the properties of HBA solvents, which is caused
by the alkylation of both NH functionalities.
Despite the negative charge of the BA anion, which contains
NH moieties, these NH groups are still able to interact with fur-
ther HBA sites. Apparently, electrostatic repulsion is clearly
overcompensated by the NH···HBA interaction strength.
The negative algebraic sign of s indicates a stronger solva-
tion of the first exited state with increasing polarity of the sol-
vents. Thus, the first excited state is more dipolar than the
Unspecific interactions (SP, SdP, or p*) cause a bathochromic
shift of n˜max of 7bq,cq, comparable with the effect observed for
4aq–cq. An exception is the correlation of solvatochromism of
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5739
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 3. Solvent-independent correlation coefficients a, b and s according
to the Kamlet–Taft Equation (4) and a, b, d and e of the Catalµn Equa-
tion (5) for the solvatochromic response of 7b and 7c.[a]
Table 4. Comparison of the solvatochromic properties of the BA dyes
4a<gq>–cq; 7bq,cq with negative[7b] and positive solvatochromic[15d] BA
dyes.[a]
7bq
19.470
2.452
À1.467
À0.978
21
7cq
17.702
3.057
Æ0.000
Æ0.000
20
ÀBAÀEWG+
betaine
BA=EWG
merocyanine
dyes[15d]
ÀBAÀEWG
this work
Kamlet–Taft Equation (4)
n˜max,0
a
b
s
dyes[7b]
algebraic sign of s from
Eq. (4) and d, e from
Eq. (5)
+
À
À
n
sd
r
0.323
0.98
0.348
0.97
algebraic sign of a from
Eq. (4) and (5)
+
À
+
f
ꢁ0.0001
21.815
4.795
À0.836
À2.894
À1.694
20
0.226
0.99
ꢁ0.0001
ꢁ0.0001
20.927
4.997
Æ0.000
À3.171
À0.979
20
0.265
0.99
ꢁ0.0001
solvatochromic range
ca. 10000
<4500
5100–
7600
Catalµn Equation (5)
n˜max,0
Dn˜ [cmÀ1
]
a
b
d
e
n
sd
r
[a] EWG=electron withdrawing group, EDG=electron donating group.
the Reichardt-type, which contain the BA moiety anion as neg-
ative part, unambiguously show a negative solvatochromic
property because a, s, d and e coefficients from Equations (4)
and (5) altogether show a positive algebraic sign.[7b]
If the original definition would be applied, the dyes 4aq–cq,
7bq,cq would ostensibly show world records in positive solva-
tochromism despite the fact that the large solvatochromic
range is actually caused by the HBD solvation rather than by
the “polarity”.
f
[a] Since the anionic form is generated in the solvents, 7bq and 7cq are
observed in reality. Also given is the correlation coefficient r, the standard
deviation sd, the number of solvents n, and the significance f.
chromophore 7cq explained by the Kamlet–Taft Equation (4),
which apparently shows no dependence on p*.
Thus, non-critical use of original definitions for positive or
negative solvatochromism makes no sense for these types of
solvatochromic dyes (for related solvatochromism of anionic
dyes see also Ref. [15a,g]). This actuality instructive demon-
strates the usefulness of LSE correlation analyses for interpreta-
tion of solvatochromic effects.
Exemplary, the solvatochromic ranges Dn˜ =7600 cmÀ1 for
4aq and Dn˜ =6700 cmÀ1 for 7cq are pretty large compared to
those of established solvatochromic dyes.[8d] It should be dis-
tinguished whether a positive or negative solvatochromic dye
is to consider to equitably evaluate the dimension of the solva-
tochromic shift.[8c,21] Positive solvatochromism of a dye is clear-
ly defined. It means that the longest wavelength UV/Vis ab-
sorption shifts bathochromically with increasing solvent polari-
ty. That would be valid for the effect of dipolarity/polarizability
of the solvent on n˜max of 4aq–cq, 7bq,cq and also for 5-(4-nitro-
phenyl)barbituric acid derivatives in the previous work.[7a] That
is because the algebraic signs of coefficients s from Equa-
tion (4) and d and e from Equation (5) are negative.
Despite this difficulty of classification, the barbiturate dyes
4aq–cq, 7bq,cq definitively show genuine solvatochromism in
most of the solvents because all n˜max data of alcohols fit well in
the correlation equations (Figure 5). Acidochromism would
only happen if the keto form of barbiturate dyes were treated
in CH2Cl2 (or another halogenoalkane) with a base. This circum-
stance must be considered by interpretation the titration ex-
periments of the dyes with the receptor molecules.
Opposing to the effect of dipolarity/polarizability, the UV/Vis
absorption bands of 4aq–cq, 7bq,cq and for the 5-(4-nitrophe-
nyl)barbituric acid derivatives[7a] are hypsochromically shifted
with increasing HBD strength (a) or acidity of the solvent (SA).
That result would suggest negative solvatochromism with re-
spect to a or SA. Thus the dyes 4aq–cq, 7bq,cq and the 5-(4-ni-
trophenyl)barbituric acid derivatives[7a] are not absolute posi-
tive solvatochromic dyes in spite of the negative algebraic
signs for the s, [Eq. (4)] or d and e [Eq. (5)] coefficients.
Complex formation studies
The interactions of 4b and 7b, respectively, with the different
receptors were also investigated (Scheme 4). The titration ex-
periments were carried out in CH2Cl2 for the pyridine derivative
DACP and in a mixture of CH2Cl2/MeOH (ratio 25:1, v/v) for the
nucleobase derivatives. The methanol addition was necessary
because of the poor solubility of reactants and products in
CH2Cl2.
Table 4 shows a comparison of the solvatochromic proper-
ties of three classes of barbiturate dyes from previous reports
and this work with respect to the solvatochromic range and
a classification as to whether positive or negative solvato-
chromism is determined.
The competition between complexation and salt formation
was investigated by reference experiments of the receptors
with 4c and 7c. As already mentioned, these compounds are
not suited for supramolecular complex formation, because
both hydrogen bonding sequences are barricaded by methyl
substitution. Therefore, UV/Vis spectroscopic changes of 4c
and 7c with the receptors can only be an effect of genuine
salt formation. Complex formation of dye 4a was not investi-
The facts of Table 4 clearly indicate a positive solvatochrom-
ism for BA dyes of the merocyanine type[9b,15d,h] because a, s, d
and e coefficients from Equations (4) and (5) altogether show
a negative algebraic sign. Likewise, pyridinium-betaine dyes of
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5740
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
gated in this study, because of its poor solubility and the ex-
pected effects due to the present NH groups.
ance of the stock solution, [S] is the concentration of the used
chromophore, KA is the association constant, e1 is the obtained
extinction coefficient of the complex, [L] is the concentration
of the added receptor and b is the path length of the beam
through the material sample of the immersion vessel.[12]
1-n-Butylcytosine
This nucleobase derivative has been used as model compound
for study the competition between complexation and genuine
salt formation, because of its complementary hydrogen bond-
ing sequence to the enol 2 form of 4b (Scheme 8). 1-n-Butyl-
cytosine (BuCy) is also the most basic nucleobase derivative
(pKa =4.60).[22]
DA ¼ ð½SŠ KA e1 ½LŠÞ=ð1 þ KA ½LŠÞ b
ð6Þ
The UV/Vis absorption intensity of 4b at 451 nm increases
with BuCy concentration. An association constant of 7220Æ
190 LmolÀ1 was calculated from Equation (7) (Supporting Infor-
mation, Table S4). The bathochromically shifted UV/Vis absorp-
tion maximum (451 nm) hints at salt formation.
The UV/Vis spectroscopic complex titration (Figure 6) was
analysed by Equation (6), where DA is the difference of the ab-
sorbance at the UV/Vis absorption maximum and the absorb-
A second hypsochromic UV/Vis absorption maximum was
found for 4b at lower concentrations of BuCy. If the concentra-
tion of BuCy exceeds the concentration of 4b, the intensity of
the new bathochromic UV/Vis absorption band at 451 nm fur-
ther increases, while the hypsochromic band is decreasing.
Isosbestic points are observed at 310 nm and 394 nm, which
indicate that above
a BuCy concentration of 2.047”
10À4 molLÀ1 only two chromophoric species are involved
(Figure 6). At lower BuCy concentrations, the UV/Vis absorb-
ance at 348 nm increases without appearing of an isosbestic
point because of the keto–enol equilibrium. The UV/Vis ab-
sorption maximum at 348 nm shifts bathochromically with in-
creased BuCy concentration. This hints at a complexation of
the enol 2-form (Scheme 3), which is further deprotonated
with increased BuCy concentration (Scheme 8).
To achieve a secure interpretation, separation of the UV/Vis
absorption spectra into two separate Gauss functions with
Origin Pro 9.1 g was employed (Figure 7). The UV/Vis absorp-
tion maximum at 468 nm belongs to the anionic form. The
hypsochromic UV/Vis absorption maximum shift bathochromi-
cally. This indicates an interaction of the enol 2 form of 4b
with BuCy.
Scheme 8. Suggested competition between complexation and genuine salt
formation of 4b with BuCy.
Figure 6. UV/Vis titration of 4b [c=9.181·10À5 mollÀ1] with BuCy in a mixture
Figure 7. Separated UV/Vis absorption spectra of 4b with BuCy in a mixture
of CH2Cl2/MeOH (ratio 25:1, v/v).
of CH2Cl2/MeOH (ratio 25:1, v/v).
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5741
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Two association constants can be determined, the first for
the UV/Vis absorption at 419 nm with KA(2)=10750Æ
240 LmolÀ1 and the second for the UV/Vis absorption at
468 nm with KA(1)=3670Æ360 LmolÀ1. The higher association
constant KA(2) shows that the complex formation is preferred
against the genuine salt formation.
(7b)=9070 LmolÀ1; Supporting Information, Table S6), which is
caused by the decreased acidity of the enolic proton com-
pared to 4b. The found association constants for genuine salt
formation KA(7b)=1925 LmolÀ1, decreases dramatically com-
pared to KA(4b)=7220 LmolÀ1
.
It could be shown that the lower acidity of 7b compared to
4b reduces the concurrence of genuine salt formation and the
complexation becomes dominant, although the association
constant between the BA and BuCy decreases slightly.
Dye 7b absorbs at 345 nm in a mixture of CH2Cl2/MeOH
(ratio 25:1 v/v), the UV/Vis absorption maximum of which is as-
signed to the keto form. By addition of BuCy to 7b, the keto–
enol equilibrium shifts towards the enol form (431 nm). This
occurrence is optically detectable by a decrease of the UV/Vis
absorption band at 345 nm, while the new UV/Vis absorption
band at 431 nm increases. The UV/Vis absorption band at
431 nm contains a bathochromic shoulder at lower concentra-
tions of BuCy. At higher concentrations of BuCy, a second UV/
Vis absorption maximum at 513 nm occurs. This UV/Vis absorp-
tion band becomes dominant at higher concentrations of
BuCy and belongs to the enolate species (Figure 8). The ob-
tained association constants show a preferred interaction of
BuCy with the enol form of 7b (KA =9070Æ530 LmolÀ1). Genu-
ine salt formation occurs again, but because of the lower acidi-
ty of compound 7b, it is less dominant (KA =1925Æ
16 LmolÀ1). These results confirm the hypothesis that a comple-
mentary hydrogen bonding sequence does influence the keto–
enol tautomerism, which induces an optical response.
NMR titration experiments were performed to support the
1
results from UV/Vis measurements. The H NMR titration of 4b
with BuCy in CD2Cl2 also shows salt formation. The signal of
the C5 hydrogen (5 ppm) at the barbituric acid keto form dis-
appears with increasing BuCy concentration. Accordingly,
a new signal appears at 14 ppm, which can be assigned to the
protonation at BuCy. With increased BuCy concentration
(excess to 4b) this signal is upfield-shifted, because of the dis-
tribution of the positive charge over more than one BuCy mol-
ecule (Supporting Information, Figure S13). Additionally, both
NMR signals of the H atoms at the five and six position of
BuCy show the expected shift to lower field indicating a re-
duced electron density. The NMR titration experiment of BuCy
with 1,1,1-trifluoroacetic acid was used to support the interpre-
tation. The 1H NMR spectrum shows signals of BuCy with a simi-
lar chemical shift, which indicates that during both titrations
a protonation of BuCy occurs (Supporting Information, Fig-
ure S14).
Compared to 4b, the association constant between BuCy
and 7b is slightly decreased (KA(2) (4b)=10750 LmolÀ1; KA(2)
Structure of 4b/BuCy in the solid state
Deep-red crystals suitable for single-crystal X-ray diffraction
studies were obtained by crystallisation of a 1:1 mixture of 4b
with BuCy from CH2Cl2/n-pentane at ambient temperature. 4b/
¯
BuCy crystallises in the triclinic space group P1. The structure
of the asymmetric unit is given in the Supporting Information
(Figure S34) together with selected bond lengths and angles.
The deep-red colour of the crystals already indicate the forma-
tion of a barbiturate anion. Indeed, the asymmetric unit of 4b/
BuCy comprises mono-deprotonated 4b (the barbiturate anion
denoted further as 4bq), two crystallographically independent
molecules of BuCy, of which each second is mono-protonated,
and one water molecule. The presence of 4bq is indicated by
the sp2-hybridised atom C8, as the sum of its bond angles is
360.1(10). Additionally, in case of 4a this compound has been
described as an enol-barbituric acid, as the bond lengths of C8
to C7 and C9 are significantly different (Figure 1). In case of
4bq, however, the C8ÀC7 and C8ÀC9 bond lengths are equal
(Supporting Information, Figure S34) and consequently this
molecule does not correspond to an enol-barbituric acid as 4a.
Instead, the barbiturate anion is detected.
The two crystallographically independent molecules of BuCy
were taken as being each half-protonated (see above). This is
revealed by the presence of dimers of the form [(BuCy)2H]+ in
the solid state for each individual BuCy molecule (Figure 9).
For example, the BuCy molecule with the atom N9 narrows to
a further N9-carrying BuCy molecule with an N9…N9A distance
of 2.835 Š (Figure 9). Such a short distance is characteristic for
Figure 8. UV/Vis titration of 7b (c=6.640·10À5 molLÀ1) with BuCy in a mixture
of CH2Cl2/MeOH (ratio 25:1, v/v). UV/Vis absorption maximum at 345 nm
(keto form) decreases; absorption maximum at 431 nm (enol form) increases
(I) but at higher concentrations it decreases again (II); absorption maximum
(enolate form) increases at 513 nm.
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5742
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
an increasing enol-1 form was found for the complex forma-
tion of 5-(4-nitrophenyl)barbituric acid with EtAd.[7a] An explan-
ation for this behaviour has still not been ascertained.
The UV/Vis titration experiment of 4b with EtAd is shown in
the Supporting Information, Figure S8. The new UV/Vis absorp-
tion maximum is found at 441 nm. An association constant of
KA =970Æ50 LmolÀ1 was calculated from Equation (6) (Sup-
porting Information, Table S7). The lower association constant
compared to BuCy, results of the lower basicity of the nucleo-
base derivative EtAd. Additionally, stabilisation of the enol-2
form at low concentrations was not found for this nucleobase
derivative. The UV/Vis absorption maximum at 348 nm did not
increase. The UV/Vis absorption maximum of 4b/EtAd shifts
10 nm hypsochromically compared to the salt formation with
BuCy or DBU, indicating an interaction of the barbiturate anion
with EtAd or the adeninium cation.
The reference experiment of EtAd with 4c shows also a re-
sponse of the chromophore, indicating a salt formation and no
supramolecular complex formation as primary reaction (Sup-
porting Information, Figure S9). A KA of 300Æ40 LmolÀ1 was
found, which is smaller than that for 4b with EtAd (Supporting
Information, Table S8).
Figure 9. Selected part of the crystal structure of 4b/BuCy. Grey dotted lines
indicate intermolecular hydrogen bonds of type I–IX, with d giving the D···A
distance and a the DÀH···A angle. All C-bonded hydrogen atoms are omit-
ted for clarity.
A titration of the anionic form of 4b, which is generated by
DBU addition, with EtAd did not show any shift of the UV/Vis
absorption maximum. This is an indication that the adeninium
cation forms a moderately strong complex with 3b.
an intermolecular hydrogen bond and thus only one N9 atom
could be protonated. Related cytosine/cytosine-H+-dimers
have already been described.[23]
This complex formation influences the equilibrium of the
acid–base reaction, resulting in a higher association constant
by a reaction of 4b instead of 4c. This could be an explanation
for the hypsochromic shift compared to salt formation with
DBU or BuCy.
In the solid state, 4b/BuCy molecules form a 3D network.
The network contains of BA anion dimers are connected by
two hydrogen bonds, which also interact with two water mole-
cules (Figure 9). The two structurally different dimers of
[(BuCy)2H]+ were stabilised by three hydrogen bonds (V and
VI, VII and VIII; Figure 9). Additionally, they form hydrogen
bonds with the BA anion dimers (IV, IX; Figure 9). One
[(BuCy)2H]+ building block links the barbiturate dimers along
the crystallographic c axis, the other along the crystallographic
b axis (Figure S35). The described hydrogen bonds form a 2D
layer, and p–p-interactions connect these layers along the crys-
tallographic a axis.
1
A H NMR titration of 4b with EtAd was not possible, owing
to insolubility of the generated salt in CD2Cl2. The resulting
solid was isolated and investigated by NMR spectroscopy in
[D6]DMSO. The reference experiment of EtAd with 1,1,1-tri-
fluoroacetic acid in [D6]DMSO is given in the Supporting Infor-
mation (Figure S15), which shows the corresponding result for
an acid–base interaction.
Solid-state structure of 4b/EtAd
Deep-red crystals were obtained by crystallisation of a 1:1 mix-
ture of 4b with EtAd from MeOH/EtOAc suitable for single
crystal X-ray diffraction studies. Compound 4b/EtAd crystallises
9-Ethyladenine
The nucleobase derivative 9-ethyladenine EtAd (pKa =4.25) is
less basic than cytosine (pKa =4.60).[22] Therefore, a reduced
tendency for salt formation is expected. Investigations on the
complex formation between barbituric acid and nucleobase
derivatives by Kyogoku showed that the complex formation of
barbituric acid with adenine is the strongest of all nucleobase
derivatives and additionally stronger than the natural pendant,
thymine with adenine.[3]
¯
in the triclinic space group P1. The structure of the asymmetric
unit is shown in the Supporting Information (Figure S32) to-
gether with selected bond lengths and angles.
The solid state structure analysis confirms the salt formation
(Figure S32). However, in contrast to 4b/BuCy, the EtAd mole-
cule in the solid structure of 4b/EtAd is mono-protonated. 4b/
EtAd forms a 3D network by hydrogen bonds and p–p interac-
tions. The barbiturate anions form dimers by two hydrogen
bonds (III, Figure 10), which are additionally linked together by
two hydrogen bonds to the adeninium cation (I and II,
Figure 10). This structure is stabilised by p–p interactions (Sup-
porting Information, Figure S33, Scheme S7). Furthermore, the
adeninium cation forms dimers by two hydrogen bonds (Fig-
The keto form of BA presents the same hydrogen bond se-
quence as thymine, namely the ADA pattern. Therefore, a com-
plex formation of the keto form with thymine by two hydro-
gen bonds was expected. This type of complexation should
not result in an UV/Vis spectroscopic response, because a com-
plexation with the enol-2 form is not expected. Consequently,
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5743
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
therefore genuine salt formation is not expected. Thus, the
UV/Vis titration of 4b with BuTy (Supporting Information, Fig-
ure S12, Table S11) shows no significant changes in the spec-
trum above 300 nm. Thus, BuTy has no measurable influence
on tautomerism of 4b.
2,6-Diacetamidopyridine
2,6-Diaminopyridine derivatives as 2,6-diacetamidopyridine
(DACP) are often used as supramolecular complexation agents
for BA because of the hydrogen bonding sequence (DAD),
which is complementary to that of the keto form (ADA) of
4b.[7a,11] As described above, this complexation should not in-
fluence the keto–enol tautomerism of 4b. A pKa value of DACP
could not be found. Data of related compounds such as pyri-
dine (pKa =5.17) and 2-acetamidopyridine (pKa =4.09)[22] indi-
cate that the pKa of DACP is likely to be about 3, which is anal-
ogous to basicity of EtGu, and thus salt formation is expected.
The measured plot of the UV/Vis absorbance of 4b at
443 nm as function of the DACP concentration could not be
evaluated with Equation (6) (Supporting Information, Fig-
ure S11, Table S10). Surprisingly, the reference experiment
DACP with 4c shows no UV/Vis spectroscopic respond in spite
of its comparable acidity (Table 1). This result suggests that salt
formation of weak basic receptors only occurs if a suitable hy-
drogen bonding pattern is available.
Figure 10. Selected detail of the crystal structure 4b/EtAd. Grey dotted lines
indicate intermolecular hydrogen bonds of type I–III, with d giving the D···A
distance and a the DÀH···A angle. Thicker dotted lines indicate intermolecu-
lar p–p interactions between aromatic rings, with d giving in the centroid to
centroid distance and a the interplanar angle of interacting rings. All C-
bonded hydrogen atoms are omitted for clarity.
ure S33), which link those structures together to result in a 3D
network.[24]
9-Ethylguanine
This phenomenon is called cooperativity and is usually eval-
uated by plotting r/x as a function of r, the so called Scatchard
plot,[25] where r is DA [Eq. (6)] and x=[DACP]. The shapes of
the Scatchard plots are concave downward curves for positive
cooperativity, straight lines for no cooperativity, and concave
upward curves for negative cooperativity.[26] The obtained
curve for the titration of 4b with DACP shows the typical be-
haviour for a negative cooperativity (Figure 11).[27]
Titration experiments with 9-ethylguanine (EtGu) as receptor
were difficult to perform because of the poor solubility of this
nucleobase derivative, especially in CH2Cl2 and in CH2Cl2/
MeOH. Therefore, complex titration could only be carried out
in a small concentration range of EtGu (Supporting Informa-
tion, Figure S10). The UV/Vis spectroscopic complex titration of
4b with EtGu in a mixture of CH2Cl2/MeOH (ratio 25:1, v/v) re-
sults in an association constant of KA =500Æ80 LmolÀ1 (Sup-
porting Information, Table S9). This value for the genuine salt
formation is in good accordance with the lower basicity of
EtGu (pKa =3.30).[22] The UV/Vis absorption spectra of 4b with
EtGu are similar to those of 4b with BuCy. A preferred com-
plexation of the enol species with the BA derivative leads to
a bathochromic shift of the former UV/Vis absorption at
377 nm. At higher concentrations, a second UV/Vis absorption
maximum at 438 nm appears which indicates the formation of
the enolate species. No statements could be made on the con-
currence of complexation versus genuine salt formation, be-
cause of the too low concentration range. Owing to the insolu-
bility of EtGu in CH2Cl2, NMR titration experiments could not
be carried out.
Cooperativity as well as the suggested 1:2 stoichiometry of
the complex formation are results of the hydrogen bonding se-
quences of 4b (ADA) and DACP (DAD). After the proton trans-
1-n-Butylthymine
The nucleobase 1-n-butylthymine BuTy shows the same hydro-
gen bonding sequence as BA in the keto form. Therefore,
a complexation is possible when two hydrogen bonds are in-
volved. Indeed, this type of complex formation should not in-
fluence the keto–enol tautomerism. Furthermore, thymine is
that nucleobase which is not strong basic (pKa =9.9)[22] and
Figure 11. Scatchard plot of the UV/Vis titration of 4b with DACP in CH2Cl2.
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5744
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
fer, the sequence of the BA remains ADA, but DACP+ changes
to a DDD sequence. This makes the protonated DACP+ unsuit-
able for a complexation of 4b or the BA anion, while the neu-
tral DACP is suited for complex formation with 4b and the BA
anion. Thus, one DACP is protonated and a second molecule
was necessary for complex formation with the BA anion, result-
ing in the 1:2 stoichiometry (Supporting Information,
Scheme S1). Such cooperative systems are described for exam-
ple with oligo-bipyridines as metal ligands.[26]
Conclusion
The pKa vales of the BAs are in the range of 0.0 for the 5-(2,4-
dinitrophenyl)-barbituric acid derivatives 4a–c and 1.3 for 5-{4-
[(1,3-dioxo-1H-inden-2(3H)-ylidene)methyl]phenyl}
barbituric
derivatives 7b,c. Therefore, when a huge excess even of
a weak base, such as a common solvent, is used, readily the
anionic form is developed in great extent. This feature deter-
mines significantly the UV/Vis spectroscopic properties of the
BA derivatives 4a–c and 7b,c.
The evaluation of the UV/Vis absorption maxima (n˜max) as
function of solvent property shows that the present anionic
form of the BA derivatives is mostly affected by the solvents
HBD properties. n˜max shifts hypsochromically with increasing
HBD strength or acidity of solvent. HBA or basicity of solvent
properties causes additionally a positive solvatochromic effect
on n˜max as long as NH moieties of the BA derivative are not
barricaded. Increase of dipolarity/polarizability of solvents
cause also a bathochromic shift of n˜max as shown by the regres-
sion coefficients determined from the LFE relationships of
Kamlet–Taft and Catalµn.
Structure of 4b/DACP in the solid state
Deep-red crystals suitable for single crystal X-ray diffraction
studies were obtained by crystallisation of a 1:1 mixture of 4b
with DACP from CH2Cl2/n-pentane. Compound 4b/DACP crys-
tallises in the triclinic space group P21/c (Supporting Informa-
tion, Figure S36).
The sp2 C8 atom and the given bond lengths as observed
and discussed for 4b/BuCy, verify the formation of a barbiturate
anion as the sum of its bond angles is 359.7(7) (Supporting In-
formation, Figure S36).
The mono N-(n-butyl)-substituted barbituric acids 4b and
7b have been found suitable to examine the interaction with
nucleobase derivatives and structurally related receptors. The
results have shown that each of the used receptors 1-n-bu-
tylthymine, 9-ethyladenine, 1-n-butylcytosine, 9-ethylguanine,
and 2,6-diacetamidopyridine, respectively, interacts in another
way with 4b or 7b. Of course, acid–base reactions take place
as expected, which is deduced from the pKa. The formed ionic
species can interact in many ways with each other by supra-
molecular complex formation as shown in Figure 13 and sum-
marised in the Supporting Information, Table S12.
In the solid state, 4b/DACP develops a 3D network, formed
by 2D layers, which are linked together by p–p interactions.
The 2D layers are formed by dimers of barbiturate anions,
which are held together by two hydrogen bonds (IV;
Figure 12). Furthermore, DACP is stabilising a proton with two
intramolecular hydrogen bonds (I and II; Figure 12). Thereby,
the NH-hydrogen bond donor functionalities were rotated out-
wards and are able to link the barbiturate anions together in
crystallographic b and c axis. These hydrogen bonds were sup-
ported by p–p interactions within this layer (Figure 12). An ad-
ditional p–p interaction link the layers in crystallographic
a axis together to form a 3D network (Supporting Information,
Figure S37, Scheme S9).
Advantageously, the potentiality of the BA species to inter-
act with HB donor or acceptor molecules can be precisely mea-
sured by employing the coefficients derived from the Kamlet–
Taft or Catalµn LFE relationships. The results of supramolecular
complex formation and solvatochromism study match amaz-
ingly to the point that negatively charged BA moieties still can
show a HBD property. As consequence of this result, the antici-
pation of the reality of those BA/receptor pairs is very difficult
to project in a synthetic concept for a suitable probe dye. The
reason for this conclusion is that many interactions can simul-
taneously play a role as shown in the summarising Figure 13.
Too strong acidic compounds are unsuitable because genu-
ine salt formation dominates. Weaker acids do interact in
a smaller extend (lower KA), but supramolecular complex for-
mation dominates over genuine salt formation. Therefore, the
focus of further studies must lie on weaker acidic dyes (pKa >
2.0), which show a strong optical response.
The precise adjusting of the optimised pKa of BA dyes for
supramolecular complex formation is objective of further stud-
ies.
Figure 12. Detail of the crystal structure of 4b/DACP. Grey dotted lines indi-
cate intermolecular hydrogen bonds of type I–IV, with d giving the D···A dis-
tance and a the DÀH···A angle. Thicker dotted lines indicate intermolecular
p–p interactions between aromatic rings, with d giving in the centroid to
centroid distance and a the interplanar angle of interacting rings. All C-
bonded hydrogen atoms are omitted for clarity.
Experimental Section
The receptors 1-butylcytosine BuCy,[28] 1-butylthymine BuTy,[29] 9-
ethyladenine EtAd,[30] and 2,6-diacetamidopyridine DACP[31] were
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5745
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 13. Overview on possible interactions of enolizable barbiturate dyes with different receptors. A=acceptor functionality.
synthesised according to previous reports. 9-Ethylguanine EtGu
was received from Sigma Aldrich (ꢂ98%) and was used as re-
ceived.
3100, 2970, 2805, 1703, 1601, 1512, 1458, 1420, 1342, 1061, 996,
878, 837, 778, 760, 745, 722, 673, 579, 531 cmÀ1
.
1-n-Butyl-5-(2,4-dinitrophenyl)barbituric acid (4b): A solution of
1-n-butyl barbituric acid (0.863 g, 4.484 mmol) in anhydrous DMF
(20 mL) was treated with 0.120 g (5.02 mmol) of sodium hydride to
form the barbituric acid sodium salt. Then, the barbiturate solution
was given into a solution of 2,4-dinitrofluorobenzene (1.000 g,
5.37 mmol) in DMF (10 mL). The solution was stirred for 1 h at am-
bient temperature and quenched with an excess of aqueous
Ca(OH)2. The solid by-product was filtered off and the DMF solu-
tion was dropped into ice/H2SO4. The precipitating solid was fil-
tered off, washed with water, and
The solvents dichloromethane and methanol were dried and fresh-
ly distilled before use. N,N-Dimethyl formamide (99.8% spectro-
scopic grade, extra dry, Acros) was used without further purifica-
tion. The UV/Vis spectroscopic measurements were carried out
with a MCS 400 diode array spectrometer by Carl Zeiss Jena
GmbH. The given melting temperatures indicate the beginning of
melting. All solutions were freshly prepared immediately before
the titrations.
5-(2,4-Dinitrophenyl)barbituric acid (4a): A solution of barbituric
acid (0.600 g, 4.684 mmol) in anhydrous DMF (30 mL) was treated
with sodium hydride (0.112 g, 4.684 mmol) to form the barbituric
acid sodium salt. Then, the barbiturate solution was given into a so-
lution of 2,4-dinitrofluorobenzene (1.000 g, 5.37 mmol) in DMF
(10 mL). The solution was stirred for 1 h at ambient temperature
and quenched with an excess of aqueous Ca(OH)2. The solid by-
product was filtered off and the DMF solution was dropped into
ice/H2SO4. The precipitating solid was filtered off, washed with
water and dried. For purification, the raw product is re-dissolved in
an aqueous NaHCO3 solution and impurities were extracted with
CH2Cl2. After acidification of the
dried. For purification, the raw prod-
uct was re-dissolved in an aqueous
NaHCO3 solution and impurities were
extracted with CH2Cl2. After acidifica-
tion of the aqueous phase with HCl,
the water slurry was extracted with
CH2Cl2. The organic phase was dried
over MgSO4, the solvent was evapo-
rated, and the solid residue was
dried under reduced pressure.
1
Overall yield 5%, yellow solid, mp: 798C; H NMR ([D6]DMSO-enol/
aqueous phase with HCl, water was
evaporated until a solid precipitate
formed. The solid was filtered off,
dried under reduced pressure, and
3
enolate): d=0.87 (t, JHH =7 Hz, 3H, 14CH3), 1.25 (m, 2H, 13CH2), 1.44
(m, 2H, 12CH2), 3.68 (t, 3JHH =8 Hz, 2H, 11CH2), 8.07 (d, 3JHH =8 Hz,
5
3
6
2
1H, CH), 8.27 (d, JHH =9 Hz, 1H, CH), 8.47 (s, 1H, CH), 10.37 ppm
(s, 1H, 15NH); 1H NMR (CD2Cl2-keto): d=0.97 (t, 3JHH =7 Hz, 3H,
14CH3), 1.39 (m, 2H, 13CH2), 1.61 (m, 2H, 12CH2), 3.91 (m, 2H, 11CH2),
recrystallised from ethanol.
Overall yield 24%, yellow solid, mp: 1658C; 1H NMR ([D6]DMSO-
5.02 (s, 1H, CH), 7.79 (d, JHH =9 Hz, 1H, CH), 8.14 (s, 1H, 15NH),
7
3
5
3
5
3
3
3
6
3
enol/enolate): d=7.98 (d, JHH =9 Hz, 1H, CH), 8.33 (d, JHH =8 Hz,
1H, 6CH), 8.52 (s, 1H, 2CH), 10.46 ppm (s, 2H, 9NH); 13C NMR
([D6]DMSO): d=88.4 (7C), 120.1 (2CH), 126.0 (6CH), 134.8 (5CH),
8.64 (dd, JHH =9 Hz, JHH =3 Hz, 1H, CH), 9,14 ppm (d, JHH =3 Hz,
1H, 2CH), 13C NMR ([D6]DMSO): d=13.8 (14CH3), 19.6 (13CH2), 30.0
(12CH2), 38.6 (11CH2), 87.5 (7C), 119.5 (2CH), 125.1 (6CH), 134.0 (5CH),
142.6, 147.5, 150.7 (1CH, 3CH, 4CH), 150.7 (1 8C), 160.2 (8C),
3
4
137.5, 143.8, 148.4 (1C, C, C), 151.0 (1 8C), 162.7 ppm (8C); elemen-
tal analysis (%) for C10H6N4O7 (294.18 gmolÀ1): calcd: C 40.83,
H 2.06, N 19.05; found: C 39.39, H 2.68, N 18.12; IR: n˜ =3520, 3416,
162.1 ppm
(9C);
elemental
analysis (%)
for
C14H14N4O7
(329.19 gmolÀ1): calcd: C 48.01, H 4.03, N 16.00; found: C 46.30,
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5746
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
H 4.10, N 15.28; IR: n˜ =3212, 3108, 2963, 2934, 2873, 1713, 1678,
was dried over MgSO4, evaporated, and the obtained solid was
dried under reduced pressure.
1607, 1530, 1439, 1412, 1379, 1343, 1296, 1229, 1200, 1150, 1098,
911, 835, 739, 716, 567, 517 cmÀ1
.
1,3-Dimethyl-5-(2,4-dinitrophenyl)barbituric acid (4c): A solution
of 1,3-dimethyl barbituric acid (0.731 g, 4.684 mmol) in anhydrous
DMF (30 mL) was treated with sodium hydride (0.112 g, 4.68 mmol)
to form the barbituric acid sodium salt. Then, the barbiturate solu-
tion was given into a solution of 2,4-dinitrofluorobenzene (1.000 g,
5.37 mmol) in DMF (10 mL). The solution was stirred for 1 h at am-
bient temperature and quenched with an excess of aqueous
Ca(OH)2. The solid by-product was filtered off and the DMF solu-
tion was dropped into ice/H2SO4. The precipitating solid was fil-
tered off, washed with water, and dried. For purification, the raw
product is re-dissolved in an aqueous NaHCO3 solution and impuri-
ties were extracted with CH2Cl2. After
1
Yield 8%, orange solid, mp:1228C; H NMR (CD2Cl2-keto): d=0.94
(t, JHH =7 Hz, 3H, 22CH3), 1.34 (m, 2H, 21CH2), 1.57 (m, 2H, 2 8CH2),
3
3.89 (t, JHH =7 Hz, 2H, 19CH2), 4.75 (s, 1H, 15CH), 7.42 (d, JHH =8 Hz,
3
3
2H, 13CH), 7.86 (d, JHH =9 Hz, 2H, 1CH, CH), 7.86 (s, 1H, 8CH), 8.01
3
4
1
(m, 2H, CH, CH), 8.48 ppm (d, JHH =8 Hz, 2H, 12CH); IR: n˜ =2960,
2
3
3
acidification of the aqueous phase
with HCl, a solid precipitated, which
was filtered off, washed with water,
2923, 2857, 1677, 1588, 1509, 1443, 1416, 1370, 1354, 1254, 1190,
1076, 1018, 990, 795, 764, 733, 670, 538, 505, 422 cmÀ1
.
1,3-Dimethyl-5-(4-((1,3-dioxo-1H-inden-2(3H)-ylidene) methyl)-
phenyl)barbituric acid (7c): The potassium salt of 1,3-dimethyl-
barbituric acid (300 mg, 1.65 mmol) was dissolved in anhydrous
and dried under reduced pressure.
Overall yield 44%, yellow solid, mp: 1958C; 1H NMR ([D6]DMSO-
9
3
5
DMF (20 mL). CaCO3 (275 mg, 2.75 mmol) and
5 (215 mg,
enol/enolate): d=3.14 (s, 6H, CH3), 8.10 (d, JHH =5 Hz, 1H, CH),
8.23 (ps, 1H, CH), 8.45 ppm (s, 1H, CH); 13C NMR ([D6]DMSO): d=
27.5 (9CH3), 87.3 (7C), 119.5 (2CH), 124.7 (6CH), 134.0 (5CH), 140.1,
6
2
0.852 mmol) were added in a single portion and the reaction mix-
ture was heated at 1508C for 4 h. The intense violet-coloured solu-
tion was cooled down to ambient temperature, filtered and
poured onto ice/H2SO4. The precipitating solid was filtered off, dis-
solved in CH2Cl2, and washed twice with a NaHCO3 solution
(100 mL). The red-coloured aqueous solution was extracted twice
with CH2Cl2 (100 mL) to remove im-
141.9, 147.2 (1C, C, C), 151.9 (1 8C), 160.9 ppm (8C); elemental anal-
ysis (%) for C12H10N4O7 (329.19 gmolÀ1): calcd: C 44.73, H 3.13,
N 17.39; found: C 44.64, H 3.18, N 17.26; IR: n˜ =3110, 3081, 3052,
2865, 1696, 1677, 1605, 1530, 1441, 1426, 1379, 1343, 1312, 1283,
1246, 1152, 1115, 1082, 994, 924, 860, 828, 754, 725, 660, 629, 592,
3
4
538 cmÀ1
.
purities. After acidification with HCl,
the aqueous phase was extracted
twice with CH2Cl2. This organic phase
was dried over MgSO4, evaporated,
and the obtained solid was dried
2-(4-Fluorobenzylidene)-1H-indene-1,3(2H)-dione (5): 1,3-indan-
dione (2.90 g, 19.82 mmol) was heated in ethanol (100 mL) until
the solid dissolved. 4-fluorobenzaldehyde (2.46 g, 19.82 mmol) and
catalytic amounts of piperidine were added. The solution was
stirred for 3 h under reflux, cooled to
under reduced pressure.
1
Yield 75%, orange solid, mp: 2208C; H NMR (CD2Cl2): d=3.35 (s,
ambient temperature, and the crys-
tallised solid was filtered of, washed
with ethanol, and recrystallised from
ethanol.
6H, 17CH3), 4.78 (s, 1H, 15CH), 7.40 (d, JHH =8 Hz, 2H, 13CH), 7.85 (d,
3
1
4
1
2
3
3JHH =9 Hz, 2H, CH, CH), 7.85 (s, 1H, 8CH), 8.01 (m, 2H, CH, CH),
8.46 ppm (d, 3JHH =8 Hz, 2H, 12CH); 13C NMR (CD2Cl2): d=29.2
(17CH3), 55.9 (15CH), 123.6, 123.7 (1CH, 4CH), 129.2 (13CH), 134.8
1
4
(12CH), 135.8, 136.0 (3CH, CH), 130.5, 133.7, 138.5, 140.5, 142.9 (5C,
Yield 81%, yellow solid, mp: 1788C; H NMR ([D6]DMSO): d=7.41
(dd, 3JHH =9 Hz, 3JHF =9 Hz, 2H, 13CH),7.86 (s, 1H, 1 8CH), 7.96 (m,
4H, 1À-4CH), 8.62 ppm (dd, 3JHH =9 Hz, 4JHF =6 Hz, 2H, 12CH);
13C NMR ([D6]DMSO): d=116.0 (d, 2JCF =22 Hz, 13CH), 123.1, 123,2
6C, 9C, 11C, 14C), 145.2 (1 8CH), 151.7 (18CO), 166.9 (16CO), 189.2,
190.0 ppm (7CO, 8CO); elemental analysis (%) for C22H16N2O5
(388.37 gmolÀ1): calcd: C 68.04, H 4.15, N 7.21; found: C 68.26,
H 4.24, N 6.27; IR: n˜ =1702, 1680, 1615, 1601, 1588, 1509, 1453,
1420, 1374, 1352, 1318, 1289, 1250, 1190, 1075, 990, 843, 795, 768,
3
6
9
4
(2CH, CH), 128.9 (d, JCF =3 Hz, C), 129.6 (d, JCF =4 Hz, 11C), 136.0,
136.1 (1CH, 4CH), 136.9 (d, 3JCF =9 Hz, 12CH), 139.5, 141.9 (5C, 6C),
144.2 (1 8CH), 164.8 (d, JCF =254 Hz, 14C), 188.8, 189.4 ppm (7C, C);
elemental analysis (%) for C16H9FO2 (252.24 gmolÀ1): calcd: C 76.19,
H 3.60; found: C 76.18, H 3.62; IR: n˜ =3096, 3073, 3040, 1730, 1690,
1580, 1506, 1420, 1371, 1227, 1200, 1165, 1072, 992, 835, 733, 513,
1
8
735, 669, 588, 538, 505, 422 cmÀ1
.
497, 411 cmÀ1
.
X-ray crystallography
All data were collected with an Oxford Gemini S diffractometer. For
data collection, cell refinement and data reduction the software
CrysAlisPro was used.[32] All structures were solved by direct meth-
ods using SHELXS-2013 and refined by full-matrix least-squares
procedures on F2 using SHELXL-2013.[33] All non-hydrogen atoms
were refined anisotropically. All C-bonded hydrogen atoms were
refined using a riding model. The positions of N- and/or O-bonded
hydrogen atoms were taken from difference Fourier maps and re-
fined isotropically.
1-n-Butyl-5-(4-((1,3-dioxo-1H-inden-2(3H)-ylidene) methyl)phe-
nyl)barbituric acid (7b): The potassium salt of 1-n-butyl-barbituric
acid (444 mg, 2.00 mmol) was dissolved in anhydrous DMF (50 mL).
CaCO3 (300 mg, 3.00 mmol) and 5 (252 mg, 1.00 mmol) were
added in a single portion and the reaction mixture was heated at
1508C for 8 h. The intense violet-coloured solution was cooled
down to ambient temperature, filtered, and poured onto ice/
H2SO4. The precipitating solid was filtered off, dissolved in CH2Cl2
and washed twice with 100 mL of a NaHCO3 solution. The red-col-
oured aqueous solution was extracted twice with CH2Cl2 (100 mL)
to remove impurities. After acidification with HCl, the aqueous
phase was extracted twice with CH2Cl2 (100 mL). The organic phase
In case of 4b/DACP, the atoms C11–C14 and the atoms N1, O1,
and O2 were refined disordered with split occupancies of 0.25/0.75
and 0.59/0.41, respectively.
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5747
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Soc. 1977, 99, 6027–6038; d) M. J. Kamlet, T. N. Hall, J. Boykin, R. W. Taft,
J. Org. Chem. 1979, 44, 2599–2604; e) R. W. Taft, M. J. Kamlet, J. Chem.
Soc. Perkin Trans. 2 1979, 1723–1727; f) M. J. Kamlet, R. W. Taft, J. Chem.
Soc. Perkin Trans. 2 1979, 349–356; g) M. J. Kamlet, J.-L. M. Abboud,
M. H. Abraham, R. W. Taft, J. Org. Chem. 1983, 48, 2877–2887.
CCDC 1420547 (4a),
1420551 (4c),
1420548 (4b/BuCy),
1420550 (4b/EtAd), and 1420549 (4b/DACP) contain the supple-
mentary crystallographic data for this paper. These data are provid-
ed free of charge by The Cambridge Crystallographic Data Centre.
Tables of bond lengths, angles, and torsion angles are also given in
the Supporting Information.
[14] a) J. Catalµn, V. López, P. PØrez, Liebigs Ann. 1995, 793–795; b) J.
Catalµn, V. López, P. PØrez, R. Martín-Villamil, J.-G. Rodríguez, Liebigs
Ann. 1995, 241–252; c) J. Catalµn, C. Díaz, V. López, P. PØrez, J.-L. G. d.
Paz, J. G. Rodríguez, Liebigs Ann. 1996, 1785–1794; d) J. Catalµn, C.
Díaz, Liebigs Ann. 1997, 1941–1949; e) J. Catalµn, J. Palomar, C. Díaz, J.-
L. G. d. Paz, J. Phys. Chem. A 1997, 101, 5183–5189; f) J. Catalµn, C. Díaz,
Eur. J. Org. Chem. 1999, 885–891; g) J. Catalµn, H. Hopf, Eur. J. Org.
Chem. 2004, 4694–4702; h) J. Catalµn, J. Phys. Chem. B 2009, 113,
5951–5960.
[15] a) A. Schade, R. Menzel, H. Gçrls, S. Spange, R. Beckert, Asian J. Org.
Chem. 2013, 2, 498–503; b) K. Hofmann, S. Spange, J. Org. Chem. 2012,
77, 5049–5055; c) A. Oehlke, A. A. Auer, I. Jahre, B. Walfort, T. Rüffer, P.
Zoufalµ, H. Lang, S. Spange, J. Org. Chem. 2007, 72, 4328–4339; d) M.
Bauer, A. Rollberg, A. Barth, S. Spange, Eur. J. Org. Chem. 2008, 4475–
4481; e) K. Schreiter, S. Spange, J. Phys. Org. Chem. 2008, 21, 242–250;
f) K. Hofmann, K. Schreiter, A. Seifert, T. Rüffer, H. Lang, S. Spange, New
J. Chem. 2008, 32, 2180–2188; g) F. Riedel, S. Spange, J. Phys. Org.
Chem. 2012, 25, 1261–1268; h) S. Seifert, A. Seifert, G. Brunklaus, K. Hof-
mann, T. Ruffer, H. Lang, S. Spange, New J. Chem. 2012, 36, 674–684.
Acknowledgements
Financial support from the Fonds der Chemischen Industrieand
the DFG (Deutsche Forschungsgemeinschaft)SP 392/26-1/26-
3is gratefully acknowledged.
Keywords: barbituric acid
molecular recognition · nucleobases · solvatochromism · UV/
Vis spectroscopy
·
keto–enol tautomerism
·
[1] J. D. Watson, F. Crick, Nature 1953, 171, 737–738.
[2] a) J. Sartorius, H.-J. Schneider, Chem. Eur. J. 1996, 2, 1446–1452; b) A. M.
Kelly-Rowley, V. M. Lynch, E. V. Anslyn, J. Am. Chem. Soc. 1995, 117,
3438–3447; c) U. Lüning, C. Kühl, A. Uphoff, Eur. J. Org. Chem. 2002,
4063–4070.
[3] Y. Kyogoku, R. C. Lord, A. Rich, Nature 1968, 218, 69–72.
[4] G. A. Jeffrey, J. Mitra, Acta Crystallogr. 1983, B39, 469–480.
[5] a) C. Vilaivan, W. Srinarang, N. Yotapan, W. Mansawat, C. Boonlua, J. Ka-
wakami, Y. Yamaguchi, Y. Tanaka, T. Vilaivan, Org. Biomol. Chem. 2013,
11, 2310–2317; b) A. B. Eldrup, C. Christensen, G. Haaima, P. E. Nielsen,
J. Am. Chem. Soc. 2002, 124, 3254–3262; c) A. B. Eldrup, B. B. Nielsen, G.
Haaima, H. Rasmussen, J. S. Kastrup, C. Christensen, P. E. Nielsen, Eur. J.
Org. Chem. 2001, 1781–1790.
˘
˘
[16] a) S. F. Oguz, I. Dogan, Spectrosc. Lett. 2004, 37, 607–618; b) P. De
Meester, S. S. C. Chu, M. V. Jovanovic, E. R. Biehl, J. Heterocycl. Chem.
1986, 23, 337–341; c) F. Zuccarello, G. Buemi, C. Gandolfo, A. Contino,
Spectrochim. Acta Part A 2003, 59, 139–151.
[17] K. A. Connors, Binding Constats, New York, Chichester, Brisbane, Toronto,
Singapore, 1987, p. 169–171.
[18] H. Koffer, J. Chem. Soc. Perkin Trans. 2 1974, 1428–1429.
[19] A. Schade, N. Behme, S. Spange, Chem. Eur. J. 2014, 20, 2232–2243.
[20] G. Wypych, Handbook of Solvents, ChemTee Publishing, Toronto, 2014,
p. 581–621.
[21] C. Reichardt, Chem. Rev. 1994, 94, 2319–2358.
[22] http://research.chem.psu.edu/brpgroup/pKa_compilation.pdf,
22.06.2015.
[6] A. C. Cheng, W. W. Chen, C. N. Fuhrmann, A. D. Frankel, J. Mol. Biol.
2003, 327, 781–796.
[7] a) I. Bolz, S. Spange, Chem. Eur. J. 2008, 14, 9338–9346; b) I. Bolz, D.
Schaarschmidt, T. Rüffer, H. Lang, S. Spange, Angew. Chem. 2009, 121,
7576–7579; c) M. Bauer, S. Spange, Eur. J. Org. Chem. 2010, 259–264.
[8] a) Y. Kyogoku, R. C. Lord, A. Rich, J. Am. Chem. Soc. 1967, 89, 496–504;
b) F. Würthner, S. Yao, T. Debaerdemaeker, R. Wortmann, J. Am. Chem.
Soc. 2002, 124, 9431–9447; c) N. S. Nudelman, C. E. S. Alvaro, J. Phys.
Org. Chem. 2011, 24, 1067–1071; d) C. Reichardt, T. Welton, Solvents
and Solvent Effects in Organic Chemistry, Wiley-VCH, Weinheim, 2011,
p. 359–424.
[9] a) R. Puig-de-la-Bellacasa, L. GimØnez, S. Pettersson, R. Pascual, E. Gonza-
lo, J. A. EstØ, B. Clotet, J. I. Borrell, J. Teixidó, Eur. J. Med. Chem. 2012, 54,
159–174; b) I. Bolz, M. Bauer, A. Rollberg, S. Spange, Macromol. Symp.
2010, 287, 8–15.
[10] a) S. S. Joshi, I. R. Gambhir, J. Am. Chem. Soc. 1956, 78, 2222–2224;
b) I. M. Fellows, D. E. Kaelin, S. F. Martin, J. Am. Chem. Soc. 2000, 122,
10781–10787.
[11] M. Bauer, S. Spange, Angew. Chem. 2011, 123, 9901–9904.
[12] V. N. Khrustalev, K. A. Krasnov, T. V. Timofeeva, J. Mol. Struct. 2008, 878,
40–49.
ˇ
´
´
´
[23] N. Bosnjakovic-Pavlovic, A. Spasojevic-de BirØ, J. Phys. Chem. A 2010,
114, 10664–10675.
[24] B. Sridhar, K. Ravikumar, Acta Crystallogr. 2007, C63, o415–o418.
[25] a) B. Perlmutter-Hayman, Acc. Chem. Res. 1986, 19, 90–96; b) A. A. Sabo-
uri, A. A. Moosavi-Movahedi, Biochem. Educ. 1994, 22, 48–49.
[26] A. Pfeil, J.-M. Lehn, J. Chem. Soc. Chem. Commun. 1992, 838–840.
[27] A. K. Bordbar, A. A. Saboury, A. A. Moosavi-Movahedi, Biochem. Educ.
1996, 24, 172–175.
[28] D. L. Helfer, R. S. Hosmane, N. J. Leonard, J. Org. Chem. 1981, 46, 4803–
4804.
[29] A. D. Hamilton, D. v. Engen, J. Am. Chem. Soc. 1987, 109, 5035–5036.
[30] J. S. Nowick, J. S. Chen, G. Noronha, J. Am. Chem. Soc. 1993, 115, 7636–
7644.
[31] K. Kurita, R. L. Williams, J. Heterocycl. Chem. 1975, 12, 789–791.
[32] G. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[33] CyrsAlisPro, version 1.171.33.64 (release 22.03.2010 CrysAlis171.NET),
Agilent Technologies.
[13] a) T. Yokoyama, R. W. Taft, M. J. Kamlet, J. Am. Chem. Soc. 1976, 98,
3233–3237; b) R. W. Taft, M. J. Kamlet, J. Am. Chem. Soc. 1976, 98,
2886–2894; c) M. J. Kamlet, J.-L. M. Abboud, R. W. Taft, J. Am. Chem.
Received: November 11, 2015
Published online on March 4, 2016
Chem. Eur. J. 2016, 22, 5734 – 5748
www.chemeurj.org
5748
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim