A new quadruply bound heterodimer DDAD·AADA and investigations into the association process

Article abstract of DOI:10.1002/1099-0690(200212)2002:23<4054::AID-EJOC4054>3.0.CO;2-H

DDAD·AADA heterodimers based on the structures 8 and 9 have been synthesised and characterised. Solubility problems have been overcome through the introduction of a 2,6-dimethylphenyl substituent (8c), and association constants K_ass have been determined by ¹H NMR titrations. The K_ass values obtained could be confirmed by osmometry, as could the association constants for other heterodimers. The K_ass value increases when the acidity of participating N-H hydrogen atoms is enhanced. Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002.

Full text of DOI:10.1002/1099-0690(200212)2002:23<4054::AID-EJOC4054>3.0.CO;2-H

FULL PAPER
A New Quadruply Bound Heterodimer DDAD·AADA and Investigations into
the Association Process[‡]
Stefan Brammer,[a] Ulrich Lüning,*[a] and Christine Kühl[b]
Keywords: Association constant / Heterocycles / Hydrogen bonds / Molecular recognition / Osmometry
DDAD·AADA heterodimers based on the structures 8 and 9
have been synthesised and characterised. Solubility prob-
lems have been overcome through the introduction of a 2,6-
dimethylphenyl substituent (8c), and association constants
Kass have been determined by 1H NMR titrations. The Kass
values obtained could be confirmed by osmometry, as could
the association constants for other heterodimers. The Kass
value increases when the acidity of participating N−H hydro-
gen atoms is enhanced.
( Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany,
2002)
Introduction
the respective partners are mixed. The first heterodimer
with four hydrogen bonds, DAAD·ADDA (1·2), was re-
ported in 1998.[11,12] A related pair is the heterodimer
AADA·DDAD. In each half of this heterodimer, a sub-
sequence Ϫ AAD and DAA, respectively Ϫ is identical to
those in DAAD·ADDA (1·2),[11] as indicated in Scheme 1.
The synthesis of this new pair and the determination of
association constants is the goal of this work.
Multiple hydrogen bonds are responsible for many recog-
nition processes in nature, especially in nucleic acids.[1]
While each base pair uses only two or three hydrogen
bonds,[2] dimers with more than three hydrogen bonds have
been used in supramolecular chemistry.[3Ϫ7] If four hydro-
gen bond acceptors A or hydrogen bond donors D are
placed side by side, ten different arrays are possible.
Scheme 1
Results and Discussion
Syntheses
These ten patterns may form dimers: four heterodimers
and two homodimers. The homodimers (AADD)2 and The AADA Unit
(ADAD)2
have been
used for
supramolecular
AADA needs only one more hydrogen acceptor atom
than the AAD subsequence (Scheme 1). A straightforward
approach to such a pattern is the synthesis of an AAD unit
such as 6 through a Friedländer condensation[13,14] and sub-
sequent bridging of the two NH groups with a CO unit
(Scheme 2).
The requisite aminopyridinecarbaldehyde 5 was syn-
thesised by a literature procedure,[15] as were some of the
cyanoacetamides 4.[16,17] These acetamides had substituents
at the amido nitrogen atom for two reasons: (i) a substitu-
polymers.[8Ϫ10] While these dimers may exist once the mole-
cule is synthesised, a heterodimer will only be formed when
[‡]
Multiple Hydrogen Bonds, 2. Part 1: Ref.[11]
Institut für Organische Chemie, Universität Kiel,
[a]
Olshausenstr. 40, 24098 Kiel, Germany
Fax: (internat.) ϩ 49-(0)431/880-1558
E-mail: luening@oc.uni-kiel.de
[b]
New address: Bernina Biosystems GmbH,
Am Klopferspitz 19a, 82152 Martinsried, Germany
E-mail: kuehl@bernina-biosystems.de
4054
 2002 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ193X/02/1223Ϫ4054 $ 20.00ϩ.50/0 Eur. J. Org. Chem. 2002, 4054Ϫ4062

A New Quadruply Bound Heterodimer DDAD•AADA
FULL PAPER
chloroform (up to about 5 m). Higher concentrations were
accessible by warming, but recrystallisation could be ob-
served after a few minutes. When, however, approximately
1 equiv. of the complementary DDAD urea 9a was added,
warming achieved a homogeneous solution that was stable
for weeks.
In an attempt to improve the solubility of the AADA
target, compound 8b, with a branched alkyl residue, was
proposed. The required amine 3b was synthesised from tert-
butylacetyl chloride by known amidation[21] and reduction
procedures.[22] Formation of the new cyanoacetamide 4b,
Friedländer condensation and formation of the urethane 7b
were carried out analogously to the procedures described
above. Again, the condensation of 7b to 8b was effected by
heating in pyridine or in the solid state. Unfortunately, the
latter procedure resulted in lower purity and, due to the
necessary recrystallisation, in a lower yield (60%) of 8b than
in the case of 8a. The smaller yield may be caused by the
higher melting point of the starting material 7b (not below
the conversion temperature of 170 °C). In contrast to ex-
pectations, 8b had solubility no better than that of 8a.
An alternative plan to increase the solubility of the
AADA target involved the attachment of a 2,6-dimethyl-
ated aryl substituent directly onto the N-3 atom of the
tetraazaanthracene part of 8. This residue would be forced,
due to the steric hindrance of the methyl groups, to main-
tain a twisted orientation with respect to the plane of the
heterocycle, thus potentially disrupting dense packing of the
molecules and, in consequence, favouring solution energy
over crystal energies, which might be expected to result in
a better solubility. Treatment of the dimethylaniline 3c with
ethyl cyanoacetate gave 4c, then[17] known only as a by-
product, in 21% yield. Friedländer condensation and treat-
ment with ethyl chloroformate in the usual manner pro-
vided urethane 7c. In this case, condensation to 8c was
achieved satisfactorily by heating neat solid 7c, yielding
73% of product 8c after recrystallisation of the solidified
reaction cake. Compound 8c was indeed much more soluble
than its analogues 8a and 8b (about 200 m solution in
chloroform).
Scheme 2
ent on the imido nitrogen atom of the target 8 blocks the
potentially competing ADA recognition site arising be-
tween the two CϭO oxygen atoms, and (ii) many hetero-
cycles including naphthyridines are known to be poorly sol-
uble in common solvents,[18] and long and/or branched
nonpolar groups usually confer some solubility on hetero-
cyclic systems. Related tetraazaanthracenes have already
been described in the literature.[19,20] Starting from barbit-
uric acid or its doubly N-methylated derivative,
a
Friedländer condensation with 5 gives the desired heterocy-
cle. In the products of this approach, however, either both
nitrogen atoms are alkylated or both are non-alkylated,
blocking the desired AADA domain or leaving the compet-
ing ADA pattern free, respectively.
Counterpart DDAD
A first synthesis with the cyanacetamide 4a,[16] accessible
As the complementary counterpart of the AADA sys-
from commercially available amine 3a and ethyl cyanoacet- tems 8, a DDAD unit was required. DDA subsequences are
ate, and the aminopyridinecarbaldehyde 5 afforded the present in pyridyl ureas (see Scheme 1). Thus, treatment of
naphthyridine 6a. Use of trichloromethyl chloroformate (di- 2,6-diaminopyridine with 1 equiv. of alkyl isocyanates gave
phosgene) for the subsequent bridging with a CO unit gave compounds 9a and 9b, leaving the other amino group free
only inseparable mixtures, as did heating with different de- as the solitary donor in the DDAD pattern. In one-step
rivatives of formic acid. The best procedure turned out to syntheses with commercially available starting materials, the
be the addition of ethyl chloroformate at low temperature, ureas 9 could be isolated in only approximately 30% yield
to give the urethane 7a. After separation from by-products, but in sufficient quantities.[23] To study the effects of sub-
application of heat resulted in condensation and the forma- stituents’ size and donor acidity, a derivative of 9a further
tion of 8a. Heating of 7a in pyridine at reflux for 3 d, fol- acylated on the solitary donor site (9c) was synthesised in
lowed by chromatography, yielded 64% of 8a. When the 73% yield by treatment of 9a with acetic anhydride.[23] For
neat solid 7a was heated to 170 °C in vacuo (to remove the the same reasons, a thio analogue of 9a was prepared from
ethanol), 100% conversion could be achieved even within butyl isothiocyanate and 2,6-diaminopyridine to give thio-
minutes. The AADA target 8a is soluble to some extent in urea 9d in 22% yield.
Eur. J. Org. Chem. 2002, 4054Ϫ4062
4055

S. Brammer, U. Lüning, C. Kühl
FULL PAPER
molecular hydrogen bond conceals the DDAD pattern. The
energetic cost of breaking this bond and exposing the recog-
nition site is met at the expense of complex stability. Such
an intramolecular bond was already suspected to lie behind
the relatively low association constant Kass of the
1
DAAD·ADDA couple 1·2; the H NMR spectrum of the
ADDA urea 2 in fact reveals its high mobility. In forming
this bond the molecule becomes unsymmetrical, and the
signals of the pyridyl H atoms should split, and this effect
should be greatest on the 3-H signal, due to its varying dis-
tance from the carbonyl group. There are two identical in-
tramolecular bonds for such a symmetrical dipyridylurea,
and obviously there is a fast interconversion between them,
because the 3-H signals are not split, but extremely
broadened (distinguishable from the baseline only by integ-
ration).
Complex Formation Studies
1H NMR Titrations
The association of AADA units with their DDAD
complements was mainly investigated by 1H NMR titra-
tion. Association constants Kass were calculated from the
shifts of selected host NMR signals on addition of increas-
ing amounts of a guest, by the fitting of a function con-
taining Kass as a variable parameter to the experimental
data.[24] Because of the limited solubilities of the tetra-
azaanthracenes 8a and 8b, compounds 8 were usually
chosen as hosts and the ureas 9 as the guests, added in up
to about a twentyfold excess. The data are collected in
Table 1.
Figure 1. 2-Pyridylureas may form intramolecular hydrogen bonds;
such an intramolecular hydrogen bond was found in the crystal
structure for N,NЈ-bis(6-methyl-2-pyridyl)urea;[26] the new con-
formation allows the ureas to form (AD)2 dimers
In contrast, the 1H NMR spectra of the asymmetric
ureas 9a, 9b, and 9d each show one set of sharp signals
for the pyridyl moieties. The presence of the intramolecular
hydrogen bond was shown by a dilution experiment: the
NMR signal of the NH proton next to the alkyl group (the
one involved in the internal hydrogen bond) virtually does
not change its chemical shift at δ ഠ 9 ppm when the sample
is diluted by a factor of twenty, because this bond is formed
in a unimolecular (concentration-independent) process. The
other NH proton is to some extent involved in self-comple-
mentary AD recognition (Figure 1). As this is a bimolecular
process, the association decreases upon dilution, and the
NMR signal shifts upfield from δ ϭ 8.1 to 6.5 ppm.
Table 1. Association constants Kass [Ϫ1] of tetraazaanthracenes
8aϪc and ureas 9aϪd as determined by 1H NMR titration of hosts
8 with guests 9 in CDCl3 at 25°C
9a
9b[a]
9c
9d[a]
[b]
[b]
8a
8b
8c
115
100
120
85
59
590[c]
[a]
Because of the similarity of the results for the titrations of 8aϪc
[b]
with 9a, these titrations were only performed with 8b. No deter-
mination was possible, due to the limited solubility of both com-
ponents. Because of the poor solubility of 9c, this titration was
[c]
Ureas 9 are thus fixed to a certain extent[31] in the cyclic
conformation, which has to be broken for recognition of
performed inversely: the tetraazaanthracene 8c was added to a
given solution of the urea 9c.
1
the DDAD pattern. In fact, during the H NMR titration
of AADA units 8 with DDAD units 9, the 3-H signals of
All AADA systems 8 exhibited similar association con- all ureas 9 were broadened and had positions downfield
stants Kass of about 110 Ϫ1 with the butyl-substituted relative to the spectrum of pure 9 as long as 8 was still in
DDAD unit 9a. The sterically more demanding cyclohexyl excess. As the amount of 9 was increased, the 3-H signal
substituent in 9b lowered Kass slightly to 85 Ϫ1. Compar- shifted further and further upfield and sharpened, finally to
ison of these results with those known for other quadruple resemble the spectrum of neat 9. Obviously, the ‘‘stretched’’
hydrogen bonding systems such as DAAD·ADDA (1·2)[11] conformation of 9, almost absent in the pure solution, is
(Kass ϭ 2000 Ϫ1) or self complementary AADD[25,26] (Kass stabilized by hydrogen bonding to 8 because it provides the
about 106 Ϫ1) shows that the association between 8 and 9 complementary AADA pattern.[32]
is rather weak.
For further study of the effects of substituents’ size and
This could most probably be due to an intramolecular donor acidity, thiourea 9d, analogous to 9a, was titrated
hydrogen bond (see Figure 1); such a hydrogen bond has with the complementary 8b. The NH protons of a thiourea
been found in analogous 2-azaheterocycle-substituted are more acidic[33,34] (thought to be better hydrogen bond
ureas.[27Ϫ30] The new conformation produced by the intra- donors), so heteromolecular association might be stronger.
4056
Eur. J. Org. Chem. 2002, 4054Ϫ4062

A New Quadruply Bound Heterodimer DDAD•AADA
FULL PAPER
On the other hand, though, the intramolecular hydrogen
bond would be stronger as well, possibly disfavouring the
stretched form. The association constant turned out to be
59 Ϫ1, and the NH signal from the intramolecular hydro-
gen bond in a neat sample of 9d had (and kept upon dilu-
tion) a position even further downfield than that in 9a. Ob-
viously, the cyclic form is more stable for the thio com-
pound than for its oxygen analogue. As well as the in-
creased acidity of the NH hydrogen atom, the greater steric
hindrance of the bigger sulfur atom and the pyridine moiety
in the stretched conformation may also be responsible.
The opposite effect was observed with an analogue of 9a
acetylated on the solitary amino group (9c). The bigger
acetyl residue next to the recognition-pattern motif might
sterically hamper association, but acetylation, on the other
hand, acidifies the NH proton and thus could stabilize the
heterodimer. Furthermore, the amido group withdraws elec-
tron density from the pyridine ring and should thus weaken
the intramolecular hydrogen bond. In fact, the association
constant was determined to be 590 Ϫ1, so electronic effects
outweighed steric ones. In this titration, the less soluble
urea 9c was treated with increasing amounts of the highly
soluble tetraazaanthracene 8c, and the unfolding of the
urea during the titration again became obvious. The NMR
signals of the pyridine 3-H and 5-H protons in 9c started
out at δ ϭ 6.45 and 7.65 ppm, respectively. With increasing
amounts of 8c, the proportion of stretched 9c grew. In this
conformation, 3-H and 5-H have fairly similar environ-
ments, each with equal distances to a carbonyl group, and
(3)
From the concentrations used in the osmometric experi-
ments for Equation (3) and by application of Equation (4),
the degrees of association αNMR can be compared to those
derived from the osmometric experiments (αosmo) with
Equation (2).
(4)
Besides the new heterodimer 8a·9a, the known couple
DAAD·ADDA (1·2) and another, bent, pair 10·11[36] were
investigated by osmometry (see Table 2). 1H NMR titra-
tions were usually performed at 25 °C, but the vapour pres-
sure osmometer available had a fixed operating temperature
of 36.5 °C, which of course decreases association. The
couples 8a·9a and 1·2 were therefore also titrated in the
NMR at 36.5 °C to provide an idea of the magnitude of
this decrease. The third couple had rather high α values
anyway and was not titrated again. With allowance for a
relative error of about Ϯ10% for both methods, the values
obtained by NMR titrations and by VPO correlate quite
well.
1
indeed their H NMR signals finally reached δ ϭ 7.75 and
7.73 ppm, respectively.
Vapour Pressure Osmometry
Vapour Pressure Osmometry (VPO), usually applied to
pure substances, was chosen as an alternative means of in-
vestigating association. Osmotic pressure is a colligative
property and reveals the weighted average molecular mass
Mobs of all species present in the sample solution. If a 1:1
mixture of two substances is measured and a 1:1 association
model presumed, the highest possible concentration for
each species equals c0, while their actual concentrations de-
pend on the degree of association (α). Mobs is calculated
from the total mass concentration (c0·Mass), divided by the
actual molar concentration of each species. Equation (1) is
thus obtained, and can be transformed into Equation (2).
Interpretation
Table 2 provides
AADA·DDAD heterodimer (8a·9a) and the DAAD·ADDA
pair (1·2). Surprisingly, there is a large difference in Kass
a comparison between the new
,
of a factor of 20, although the association is through four
hydrogen bonds in each pair. Even when secondary hydro-
gen bond interactions[37,38] between neighbouring hydrogen
bonds are considered, the Kass values should be similar be-
cause the numbers of positive and negative secondary inter-
actions are alike in both pairs (one and two, respectively).
One of the partners is a pyridyl urea in each pair. An intra-
molecular hydrogen bond therefore has to be cleaved in
both cases before the quadruple hydrogen bond can be
formed. Again though, this cannot cause the twentyfold dif-
ference in the association constants Kass between 1·2 and
any of the three heterodimers 8aϪc·9a (see Table 1).
(1)
(2)
1
On the other hand, the Kass values derived from the H
When the association between 8c and 9c is compared to
NMR titrations are related to the degree of dissociation (δ) that between 8c and 9a, however (see Table 1), an increase
by Equation (3), with c0 being the initial concentration of of Kass by a factor of five is observed. The only difference
the presumed complex on hypothetical total association.[35] between 9a and 9c is the fact that the lateral NH2 group in
Eur. J. Org. Chem. 2002, 4054Ϫ4062
4057

S. Brammer, U. Lüning, C. Kühl
FULL PAPER
Table 2. Degrees of association α for different heterodimers (at 1:1 stoichiometry), determined at 36.5 °C by osmometry by application
of Equation (2) (αosmo) and by 1H NMR titration by application of Equations (3) and (4) (αNMR) at the temperatures given; c0 in Equation
(3) was taken from the osmometric experiments (2·10Ϫ2 mol·LϪ1
)
Couple
Kass [Ϫ1
]
αNMR
αosmo
(36.5 °C)
Mobs
(25 °C)
(36.5 °C)
(25 °C)
(36.5 °C)
[g·molϪ1
]
AADA·DDAD (8a·9a)
DAAD·ADDA (1·2)
10·11[35]
115
2000
60000
76
0.53
0.85
0.97
0.45
0.49
0.73
0.90
392
448
753
580
0.74
[a]
[a]
[a]
These values were not determined at 36.5 °C because of the large analytical effort required and the α values being rather high anyway.
acetate (150 mL) and was filtered hot. This extraction was repeated
twice. After the combined solutions had been kept for 3 d at Ϫ18
°C, a colourless solid (6.11 g) had formed, and this was filtered off.
On concentration of the filtrate, further colourless crystals (2.99 g)
precipitated. Total yield: 13.8 g (70%; ref.[21] 82%), m.p. 126Ϫ129
°C (all batches; ref.[21] 132 °C). IR (KBr): ν˜ ϭ 3396, 3197 cmϪ1
(NϪH), 2955, 2869 (aliph. CϪH), 1661, 1628 (CϭO). 1H NMR
(200 MHz, CDCl3): δ ϭ 1.05 ppm [s, 9 H, C(CH3)3], 2.09 (s, 2 H,
CH2), 5.50 (br. s, 1 H, cis-NH), 5.95 (br. s, 1 H, trans-NH). MS
(EI, 70 eV): m/z (%) ϭ 115 (40) [M]ϩ, 100 (100) [M Ϫ NH]ϩ. MS
9a is acylated in 9c. The acidity of the amide NH group is
thus increased relative to that of an amine, suggesting that
this enhanced polarisation is the cause of the larger binding
constants. In the DAAD·ADDA pair (1·2), each NH group
is adjacent to a carbonyl group, either in an amide or in a
urea substructure. In the AADA·DDAD heterodimers
8·9aϪb, however, only three NH groups are next to a car-
bonyl group, one NH group being part of a less acidic NH2
group. Acidification of this NH group by acylation (9c in-
stead of 9a or 9b) increases the binding constant by a factor (CI, isobutane): m/z (%) ϭ 116 (100) [M ϩ 1]ϩ.
of five. In this case, the Kass values for the DAAD·ADDA
3,3-Dimethylbut-1-ylamine (3b): 3,3-Dimethylbutanamide (13.8 g,
and the AADA·DDAD pairs deviate only by a factor of
three.
120 mmol), dissolved in dry THF (300 mL), was added dropwise
under nitrogen to a suspension of lithium aluminium hydride (5.46
g, 144 mmol) in dry THF (80 mL) over a period of 2 h. The mixture
was heated to 50Ϫ55 °C for 17 h. It was cooled with ice and hydro-
lysed with water until gas evolution had ceased. The precipitate
was dissolved with 2  sulfuric acid. The organic layer was separ-
ated and concentrated in vacuo. Water was added and the mixture
was extracted four times with diethyl ether (80 mL). Solid sodium
hydroxide was added to the combined water layer until the pH was
basic. After extraction with diethyl ether (five times 50 mL), the
organic layer was dried with sodium sulfate, the solvent was re-
moved in vacuo, and the residue was purified by distillation to yield
a colourless oil (6.2 g, 51%), b.p. 115 °C (ref.[22] 112.8Ϫ112.9 °C).
IR (KBr): ν˜ ϭ 3283 cmϪ1 (NϪH), 2954, 2867 (aliph. CϪH). 1H
NMR (200 MHz, CDCl3): δ ϭ 0.89 ppm [s, 9 H, C(CH3)3], 1.28
(s, 2 H, NH2), 1.37 (mc, 2 H, NH2CH2CH2), 2.69 (mc, 2 H,
NH2CH2). MS (EI, 70 eV): m/z (%) ϭ 101 (18) [M]ϩ, 71 (100). MS
(CI, isobutane): m/z (%) ϭ 102 (100) [M ϩ 1]ϩ.
Tables 1 and 2 again show that prediction of an associ-
ation constant Kass for binding by several hydrogen bonds
cannot be based on the number of hydrogen bonds alone,[5]
and that the acidity and basicity of hydrogen bond donors
and acceptors may have a strong influence on complex
stability.[39,40]
Experimental Section
General Remarks: Commercially available starting materials were
used as received. Melting points were measured with a Büchi (Ͻ
250 °C) or with an Electrothermal Melting Point Apparatus (Ͻ 380
°C) and are not corrected. 1H NMR spectra were obtained with
Bruker AC 200, AM 300, or DRX 500 machines, IR spectra were
run with a PerkinϪElmer Paragon 1000, mass spectra with a Finni-
gan MAT 8230 or MAT 8200. Elemental analyses were obtained
with a PerkinϪElmer Elemental Analyzer 240 or an Elementarana-
N-Dodecylcyanoacetamide (4a): n-Dodecylamine (3a, 6.13 g,
33.1 mmol) was dissolved in ethyl cyanoacetate (23.0 mL,
lysen GmbH VarioEL machine. GC analysis was performed with 216 mmol) by slight warming. After 1 h at room temp., a precipit-
a Varian 3400, equipped with an FID. 1H NMR titrations were
performed in CDCl3 with a Bruker DRX 500, starting with approx-
ate started to form. After 17 h, this was filtered off, washed with
ethanol and diethyl ether and recrystallised from ethanol to yield
imately 5·10Ϫ3  of one component. For Kass ϭ 400, the choice of a colourless solid (4.40 g). Concentration of the mother liquor and
this concentration gives 50% complexation at 1:1 stoichiometry.
The partner was added in small aliquots in such a way that several
data points could be measured in the vicinity of a 1:1 stoichiometry.
recrystallisation of the residue from ethanol gave further product
(1.41 g). Total yield: 5.81 g (69%, ref.[16] 95%), m.p. 82Ϫ83 °C (both
batches, ref.[16] 90.5 °C). IR (KBr): ν˜ ϭ 3298 cmϪ1 (NϪH), 2916,
Vapour pressure osmometry (concentration: 2·10Ϫ2 mol·LϪ1) was 2849 (aliph. CϪH), 2261 (CϵN), 1652, 1560 (CϭO). 1H NMR
performed with a Mechrolab 301 A vapour pressure osmometer at
36.5 °C with benzil as a standard.
(300 MHz, CDCl3): δ ϭ 0.88 ppm (t, J ϭ 6.9 Hz, 3 H, CH3),
1.2Ϫ1.4 [m, 18 H, (CH2)9CH3], 1.54 (mc, 2 H, NHCH2CH2), 3.30
(dt, J1 ϭ 7.2 Hz, J2 ϭ 5.7 Hz, 2 H, NHCH2), 3.37 (s, 2 H, CNCH2),
6.15 (br. s, 1 H, NH). MS (EI, 70 eV): m/z (%) ϭ 252 (5) [M]ϩ, 212
(100) [M Ϫ CNCH2]ϩ. MS (CI, isobutane): m/z (%) ϭ 253 (100)
[M ϩ 1]ϩ.
3,3-Dimethylbutanamide: In a three-necked flask with thermometer
and mechanical stirrer, tert-butylacetyl chloride (25.0 mL,
179 mmol) was added dropwise to concd. ammonia (250 mL), with
the temperature being maintained below 12 °C by use of an ice/salt
bath. After the mixture had been stirred for a further 90 min, the
N-(3,3-Dimethylbutyl)cyanoacetamide (4b): 3,3-Dimethylbut-1-yl-
precipitated solid (4.71 g) was filtered off and the filtrate was con- amine (3b, 6.2 g, 61 mmol) was added with ice-cooling to ethyl cy-
centrated to dryness. The residue was heated for 10 min with ethyl
4058
anoacetate (10 mL, 94 mmol). After 17 h at room temp., colourless
Eur. J. Org. Chem. 2002, 4054Ϫ4062

A New Quadruply Bound Heterodimer DDAD•AADA
FULL PAPER
crystals (4.1 g) had formed, and these were filtered off. After the
filtrate had stood in the open flask for several days, further product
(2.4 g) had precipitated from the filtrate. The batches were dried in
vacuo over silica gel, each for 5 d. Because of its excellent solubility
in any solvent tested, the substance was not washed or recrys-
CϪH), 2920, 2847 (aliph. CϪH), 1642, 1557, 1517 (CϭO). 1H
NMR (500 MHz, CDCl3): δ ϭ 0.88 ppm (t, J ϭ 7.0 Hz, 3 H, CH3),
1.2Ϫ1.4 (m, 18 H, [CH2]9CH3), 1.6Ϫ1.7 (m, 2 H, NHCH2CH2),
3.46 (td, Jt ϭ 7.3 Hz, Jd ϭ 5.9 Hz, 2 H, NHCH2), 6.56 (br. s, 1 H,
NH), 6.65 (br. s, 2 H, NH2), 7.14 (dd, J1 ϭ 7.9 Hz, J2 ϭ 4.4 Hz, 1
tallised; GC analysis showed high purity anyway. Total yield: 6.57 g H, Ar-6-H), 7.89 (dd, J1 ϭ 7.9 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-5-H), 8.08
(s, 1 H, Ar-4-H), 8.84 (dd, J1 ϭ 4.4 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-7-
(64%), m.p. 92Ϫ93 °C (both batches). IR (KBr): ν˜ ϭ 3290, 3109
cmϪ1 (NϪH), 2960, 2865 (aliph. CϪH), 2255 (CϵN), 1659, 1571 H). MS (EI, 70 eV): m/z (%) ϭ 356 (27) [M]ϩ, 172 (100) [M Ϫ
(CϭO). 1H NMR (200 MHz, CDCl3): δ ϭ 0.94 ppm [s, 9 H, NHC12H25]ϩ, 145 (93) [M Ϫ NHC12H25 Ϫ HCN]ϩ. MS (CI, isobu-
C(CH3)3], 1.46 (mc, 2 H, NHCH2CH2), 3.32 (mc, 2 H, NHCH2), tane): m/z (%) ϭ 357 (100) [M ϩ 1]ϩ. HR-MS: calcd. for
3.37 (s, 2 H, CNCH2), 6.17 (br. s, 1 H, NH). MS (EI, 70 eV): m/z C21H32N4O 356.25760, found 356.25740 (diff.: Ϫ0.6 ppm); calcd.
(%) ϭ 168 (9) [M]ϩ, 153 (12) [M Ϫ CH3]ϩ, 128 (14) [M Ϫ C2013CH32N4O for 357.26096, found 357.26100 (diff.: 0.9 ppm).
CNCH2]ϩ, 97 (30) [M Ϫ CH2C(CH3)3]ϩ, 69 (100) [CNCH2CHO]ϩ.
2-Amino-3-{[(3,3-dimethylbutyl)amino]carbonyl}-1,8-naphthyridine
(6b): 2-Aminopyridine-3-carbaldehyde (5, 3.50 g, 28.7 mmol), N-
(3,3-dimethylbutyl)cyanoacetamide (4b, 4.78 g, 28.4 mmol) and pi-
peridine (10.0 mL, 101 mmol) were heated at reflux in dry ethanol
MS (CI, isobutane): m/z (%) ϭ 169 (100) [M ϩ 1]ϩ. HR-MS: calcd.
for C9H16N2O 168.12627, found 168.12610 (diff.: 1.0 ppm); calcd.
for C813CH16N2O 169.12962, found 169.12950 (diff.: 0.7 ppm). GC
(OPTIMA 1, 25 m, 5 min at 60, 10°C/min until 250°C, 15 min at
250°C): tR ϭ 13.95 min, 96%.
(80 mL) under nitrogen for 55 h. On cooling to 8 °C for 3 d, a
yellow precipitate formed and was filtered off, washed with ethanol
1
N-(2,6-Dimethylphenyl)cyanoacetamide (4c): 2,6-Dimethylaniline
(3c, 17.4 mL, 140 mmol) and ethyl cyanoacetate (20.0 mL,
18.7 mmol) were stirred under nitrogen at room temp. for 16 h. The
precipitate was filtered off and washed with ethanol. The solid was
heated under reflux in ethanol (100 mL) for 6 h. After cooling, it
was filtered off and dried in vacuo to provide slightly purple crys-
and dried in vacuo. The H NMR spectrum showed no impurities.
Yield: 5.60 g (72%), m.p. 202Ϫ205 °C. IR (KBr): ν˜ ϭ 3364 cmϪ1
(NϪH), 3054 (arom. CϪH), 2959, 2849 (aliph. CϪH), 1648, 1557,
1
1528 (CϭO). H NMR (300 MHz, [D6]DMSO): δ ϭ 0.94 ppm [s,
9 H, C(CH3)3], 1.48 (mc, 2 H, NHCH2CH2), 3.33 (mc, 2 H,
NHCH2), 7.22 (dd, J1 ϭ 7.9 Hz, J2 ϭ 4.4 Hz, 1 H, Ar-6-H), 7.47
tals (5.5 g, 21%, ref.[17] 3%), m.p. 203Ϫ209 °C (ref.[17] 180 °C). IR (br. s, 2 H, NH2), 8.12 (dd, J1 ϭ 7.9 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-5-
(KBr): ν˜ ϭ 3252 cmϪ1 (NϪH), 3045 (arom. CϪH), 2985, 2927 H), 8.40 (s, 1 H, Ar-4-H), 8.73 and 8.75 (2 br. s, 1 H, cis/trans-
1
(aliph. CϪH), 2258 (CϵN), 1661 (CϭO), 1543, 1472 (arom.). H NH), 8.78 (dd, J1 ϭ 4.4 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-7-H). MS (EI,
NMR (300 MHz, [D6]DMSO): δ ϭ 2.14 ppm (s, 6 H, CH3), 3.92
70 eV): m/z (%) ϭ 272 (34) [M]ϩ, 215 (19) [M Ϫ C(CH3)3]ϩ, 172
(s, 2 H, CH2), 7.04Ϫ7.14 (m, 3 H, ArЈ-H), 9.66 (br. s, 1 H, NH). (100) [M
Ϫ
NH(CH2)2C(CH3)3]ϩ, 145 (36) [M
Ϫ
MS (EI, 70 eV): m/z (%) ϭ 188 (64) [M]ϩ, 148 (100) [M Ϫ
H2CCN]ϩ, 120 (36) [M Ϫ COH2CCN]ϩ. MS (CI, isobutane): m/z
(%) ϭ 189 (100) [M ϩ 1]ϩ.
NH(CH2)2C(CH3)3 Ϫ HCN]ϩ. MS (CI, isobutane): m/z (%) ϭ 273
(100) [M ϩ 1]ϩ. HR-MS: calcd. for C15H20N4O 272.16370, found
272.16280 (diff.: Ϫ3.0 ppm); calcd. for C1413CH20N4O 273.16705,
found 273.16670 (diff.: Ϫ1.3 ppm).
2-Aminopyridine-3-carbaldehyde (5): A mixture of ammonium sul-
famate (52.0 g, 450 mmol) and nicotinamide (36.5 g, 300 mmol)
2-Amino-3-{[(2,6-dimethylphenyl)amino]carbonyl}-1,8-naphthyridine
was heated at 150 °C (in a molten state) under nitrogen. Over 8 h, (6c): 2-Aminopyridine-3-carbaldehyde (5, 1.84 g, 15.1 mmol), N-
the temperature was raised to 200 °C and maintained for another (2,6-dimethylphenyl)cyanoacetamide (4c, 2.88 g, 15.3 mmol) and
hour. After cooling, the solid was suspended in water (80 mL).
Concd. ammonia was added until the suspension was basic. After
the mixture had been stirred for 1 h, the remaining solid was fil-
piperidine (2.0 mL, 20 mmol) were heated at reflux in dry ethanol
(100 mL) under nitrogen for 22 h. On cooling to Ϫ18 °C for 18 h,
a yellow precipitate formed and was filtered off, washed with eth-
tered off and washed with diethyl ether. It was dissolved in 2  anol and dried in vacuo. The 1H NMR spectrum showed no impur-
hydrochloric acid and heated under reflux under nitrogen for 4 h.
On neutralisation of the cooled solution with solid sodium hydro-
ities. Yield: 2.39 g (54%), m.p. 280Ϫ290 °C (gas evolution, de-
comp.). IR (KBr): ν˜ ϭ 3416, 3243, 3146 cmϪ1 (NϪH), 1654, 1557,
gencarbonate, a precipitate formed. The mixture was extracted 1516 (CϭO), 1472 (arom.). 1H NMR (300 MHz, [D6]DMSO): δ ϭ
eight times with diethyl ether (80 mL). The combined organic layers
were dried with sodium sulfate and the solvent was evaporated to
2.24 ppm (s, 6 H, CH3), 7.15 (s, 3 H, ArЈ-H), 7.28 (dd, J1 ϭ 7.9 Hz,
J2 ϭ 4.4 Hz, 1 H, Ar-6-H), 7.39 (br. s, 2 H, NH2), 8.23 (dd, J1 ϭ
yield a yellow solid (4.0 g, 11%; ref.: 58%[15] or 25%[15a]) sufficiently 8.0 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-5-H), 8.69 (s, 1 H, Ar-4-H), 8.82 (dd,
pure for further synthesis, m.p. 94Ϫ96 °C (ref.:[15] 99 °C). IR (KBr):
ν˜ ϭ 3413, 3132 cmϪ1 (NϪH), 2747 (CHO), 1644 (CϭO), 1553, MS (EI, 70 eV): m/z (%) ϭ 292 (39) [M]ϩ, 172 (100) [M Ϫ
1450 (arom.). 1H NMR (300 MHz, [D6]DMSO): δ ϭ 6.74 ppm (dd, NHC6H3(CH3)2]ϩ, 144 (30) [M Ϫ CONHC6H3(CH3)2]ϩ, 117 (26)
J1 ϭ 7.6 Hz, J2 ϭ 4.8 Hz, 1 H, 5-H), 7.55 (br. s, 2 H, NH2), 8.00 [M Ϫ CONHC6H3(CH3)2 Ϫ HCN]ϩ. MS (CI, isobutane): m/z
(dd, J1 ϭ 7.6 Hz, J2 ϭ 2 Hz, 1 H, 4-H), 8.24 (dd, J1 ϭ 4.8 Hz, J2 ϭ
(%) ϭ 293 (100) [M ϩ 1]ϩ. HR-MS: calcd. for C17H16N4O
J1 ϭ 4.4 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-7-H), 10.12 (br. s, 0.5 H, NH).
2 Hz, 1 H, 6-H), 9.85 (s, 1 H, CHO). MS (EI, 70 eV): m/z (%) ϭ 292.13242, found 292.13230 (diff.: 0.4 ppm); calcd. for
122 (100) [M]ϩ, 94 (70) [M Ϫ CO]ϩ, 67 (22) [M Ϫ CO Ϫ HCN]ϩ. C1613CH16N4O 293.13577, found 293.13570 (diff.: 0.1 ppm).
MS (CI, isobutane): m/z (%) ϭ 123 (100) [M ϩ 1]ϩ.
3-[(Dodecylamino)carbonyl]-2-[(ethoxycarbonyl)amino]-1,8-naph-
2-Amino-3-[(dodecylamino)carbonyl]-1,8-naphthyridine (6a): 2-Ami-
thyridine (7a): Ethyl chloroformate (900 µl, 9.49 mmol) was added
nopyridine-3-carbaldehyde (5, 1.00 g, 8.19 mmol), N-dodecylcy- slowly by syringe, at 0 °C and under nitrogen, to a solution of 2-
anoacetamide (4a, 2.08 g, 8.24 mmol) and piperidine (3.2 mL, amino-3-[(dodecylamino)carbonyl]-1,8-naphthyridine (6a, 158 mg,
33 mmol) were heated at reflux in dry ethanol (50 mL) under nitro-
gen for 2 d. On cooling to 8 °C for 1 d, a yellow precipitate formed
and was filtered off, washed with ethanol and dried in vacuo. The
1H NMR spectrum showed no impurities. Yield: 2.43 g (83%), m.p.
141Ϫ143 °C. IR (KBr): ν˜ ϭ 3393 cmϪ1 (NϪH), 3094 (arom.
443 µmol) in dry pyridine (13 mL). The mixture was allowed to
warm to room temp. while stirring for 17 h. The volatile compon-
ents were distilled off in vacuo. The solid residue was dissolved in
chloroform and the same volume of water was added. The water
layer was extracted three times with chloroform (10 mL). The com-
Eur. J. Org. Chem. 2002, 4054Ϫ4062
4059

S. Brammer, U. Lüning, C. Kühl
FULL PAPER
bined organic layers were dried with sodium sulfate and concen-
trated to dryness. Chromatography (silica gel, dichloromethane/
1 H, Ar-5-H), 8.76 (dd, J1 ϭ 4.7 Hz, J2 ϭ 1.9 Hz, 1 H, Ar-7-H),
9.22 (s, 1 H, Ar-4-H), 12.66 (br. s, 1 H, ArЈ-NH), 14.24 (br. s, 1 H,
methanol, 6:1, Rf ϭ 0.64) yielded a pale yellow solid, which was ArNH). MS: Like 7a, 7c partially reacts to give 8c upon vaporiza-
recrystallised from n-hexane and a little methanol to give an almost
colourless powder (115 mg, 61%), m.p. 104Ϫ106 °C (the sample
solidifies at 160 °C and remelts at 244Ϫ246 °C, corresponding to
the m.p. of the next desired product 8a). IR (KBr): ν˜ ϭ 3454 cmϪ1
tion in the mass spectrometer; peaks of both species are detected
simultaneously. MS (EI, 70 eV): m/z (%) ϭ 364 (14) [M7c]ϩ, 318
(100) [M8c]ϩ, 301 (34) [M8c
Ϫ Ϫ
NH3]ϩ, 198 (65) [M7c
(CH3)2C6H3NH Ϫ H5C2OH]ϩ. MS (CI, isobutane): m/z (%) ϭ 365
(NϪH), 2915, 2850 (aliph. CϪH), 1675, 1629, 1594, 1540 (CϭO),
(8) [M7c ϩ 1]ϩ, 319 (100) [M8c ϩ 1]ϩ. HR-MS: calcd. for
1
1272 (CϪO). H NMR (300 MHz, CDCl3): δ ϭ 0.88 ppm [t, J ϭ C20H20N4O3 364.15353, found 364.15320 (diff.: 0.9 ppm); calcd. for
6.9 Hz, 3 H, (CH2)11CH3], 1.2Ϫ1.5 [m, 18 H, (CH2)9CH3], 1.38 (t, C1913CH20N4O3 365.15689, found 365.15680 (diff.: 0.2 ppm).
J ϭ 7.1 Hz, 3 H, OCH2CH3), 1.67 (mc, 2 H, NHCH2CH2), 3.50
3-Dodecylpyrimido[4,5-b]-1,8-naphthyridine-2,4(1H,3H)-dione (8a).
Method A: A solution of 3-[(dodecylamino)carbonyl]-2-[(ethoxy-
carbonyl)amino]-1,8-naphthyridine (7a, 235 mg, 548 µmol) in dry
pyridine (10 mL) was heated under reflux under argon for 2 d. The
volatile components were distilled off in vacuo. The solid residue
was dissolved in chloroform (5 mL) and the same volume of water
was added. The water layer was extracted three times with chloro-
form (5 mL each). The combined organic layers were dried with
sodium sulfate and concentrated to dryness. Chromatography (sil-
ica gel, dichloromethane/methanol, 6:1, Rf ϭ 0.74) yielded a yellow
solid (134 mg, 64%) that was pure by 1H NMR spectroscopy. For
(td, Jt ϭ 6.8 Hz, Jd ϭ 5.4 Hz, 2 H, NHCH2), 4.28 (q, J ϭ 7.1 Hz,
2 H, OCH2), 7.38 (dd, J1 ϭ 7.9 Hz, J2 ϭ 4.7 Hz, 1 H, Ar-6-H),
8.12 (ddd, J1 ϭ 7.9 Hz, J2 ϭ 1.8 Hz, J3 ϭ 0.7 Hz, 1 H, Ar-5-H),
8.73 (dd, J1 ϭ 4.7 Hz, J2 ϭ 1.8 Hz, 1 H, Ar-7-H), 9.12 (s, 1 H, Ar-
4-H), 11.03 (br. t, J ϭ 5.5 Hz, 1 H, NHCH2), 14.21 (br. s, 1 H,
ArNH). MS: The substance obviously reacts to give the next de-
sired product 8a upon vaporization in the mass spectrometer, and
so no HR-MS could be obtained. MS (EI, 70 eV): m/z (%) ϭ 382
(92) [M8a]ϩ, 215 (100) [M8a Ϫ C12H23]ϩ. MS (CI, isobutane): m/z
(%) ϭ 383 (100) [M8a ϩ 1]ϩ.
3-{[(3,3-Dimethylbutyl)amino]carbonyl}-2-[(ethoxycarbonyl)amino]- elemental analysis, a sample was recrystallised from n-hexane and
1,8-naphthyridine (7b): Ethyl chloroformate (900 µl, 9.49 mmol)
a little methanol, m.p. 243Ϫ246 °C (decomp.). Method B: 3-[(Dode-
was treated with 2-amino-3-{[(3,3-dimethylbutyl)amino]carbonyl}-
cylamino)carbonyl]-2-[(ethoxycarbonyl)amino]-1,8-naphthyridine
1,8-naphthyridine (6b, 159 mg, 587 µmol). Synthesis and workup (7a, 58.9 mg, 137 µmol) was heated to 170 °C in vacuo (approxim-
were carried out analogously to those used in the preparation of 3-
[(dodecylamino)carbonyl]-2-[(ethoxycarbonyl)amino]-1,8-naph-
thyridine (7a). Chromatography (Rf ϭ 0.57) and crystallisation
yielded a yellow solid (169 mg, 84%), m.p. Ͼ 160 °C, the sample
sinters and turns brown, m.p. 325Ϫ334 °C, corresponding to the
m.p. of the next desired product 8b. IR (KBr): ν˜ ϭ 3498, 3354 cmϪ1
ately 15 mbar), kept at that temperature for 5 min and cooled, also
in vacuo. The yellow solid residue (52.6 mg, 100%) was identical to
that obtained by Method A, m.p. 241Ϫ244 °C (decomp.). IR
(KBr): ν˜ ϭ 2919, 2849 cmϪ1 (aliph. CϪH), 1729, 1665, 1634 (Cϭ
O). H NMR (300 MHz, CDCl3): δ ϭ 0.88 ppm (t, J ϭ 6.9 Hz, 3
H, CH3), 1.2Ϫ1.4 (m, 18 H, [CH2]9CH3), 1.7Ϫ1.75 (m, 2 H,
1
(NϪH), 3067 (arom. CϪH), 2956, 2866 (aliph. CϪH), 1743, 1654, NCH2CH2), 4.09 (mc, 2 H, NCH2), 7.53 (dd, J1 ϭ 8.2 Hz, J2
1611, 1515 (CϭO), 1247, 1206 (CϪO). 1H NMR (500 MHz, 4.3 Hz, 1 H, Ar-7-H), 8.34 (dd, J1 ϭ 8.2 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-
CDCl3): δ ϭ 1.00 ppm [s, 9 H, C(CH3)3], 1.39 (t, J ϭ 7.4 Hz, 3 H, 6-H), 9.05 (s, 1 H, Ar-5-H, 9.18 (br. s, 0.6 H, NH), 9.21 (dd, J1 ϭ
ϭ
OCH2CH3), 1.61 (mc, 2 H, NHCH2CH2), 3.51 (mc, 2 H, NHCH2), 4.3 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-8-H). MS (EI, 70 eV): m/z (%) ϭ 382
4.28 (q, J ϭ 7.4 Hz, 2 H, OCH2), 7.37 (dd, J1 ϭ 7.9 Hz, J2
ϭ
(12) [M]ϩ, 215 (29) [M Ϫ C12H23]ϩ, 110 (100). MS (CI, isobutane):
m/z (%) ϭ 383 (100) [M ϩ 1]ϩ. HR-MS: calcd. for C22H30N4O2
382.23688, found 382.23660 (diff.: Ϫ0.7 ppm); calcd. for
4.7 Hz, 1 H, Ar-6-H), 8.12 (dd, J1 ϭ 7.9 Hz, J2 ϭ 1.9 Hz, 1 H, Ar-
5-H), 8.72 (dd, J1 ϭ 4.7 Hz, J2 ϭ 1.9 Hz, 1 H, Ar-7-H), 9.11 (s, 1
H, Ar-4-H), 10.90 (br. t, J ϭ 4.8 Hz, 1 H, NHCH2), 14.16 (br. s, C2113CH30N4O2 383.24023, found 383.24010 (diff.: Ϫ0.3 ppm).
1 H, ArNH). MS: Like 7a, 7b partially reacts to form 8b upon
vaporization in the mass spectrometer; peaks of both species are
detected simultaneously. MS (EI, 70 eV): m/z (%) ϭ 344 (12)
[M7b]ϩ, 329 (10) [M7b Ϫ CH3]ϩ, 298 (39) [M8b]ϩ, 287 (18) [M7b
C22H30N4O2 (382.5): calcd. C 69.08, H 7.81, N 14.65; found C
69.02, H 7.70, N 14.44.
3-(3,3-Dimethylbutyl)pyrimido[4,5-b]-1,8-naphthyridine-2,4(1H,3H)-
dione (8b). Method A: 3-{[(3,3-Dimethylbutyl)amino]carbonyl}-2-
[(ethoxycarbonyl)amino]-1,8-naphthyridine (7b, 243 mg, 706 µmol)
was heated at reflux in dry pyridine (8 mL) under argon for 3 d.
The workup was performed analogously to that used for the pre-
Ϫ C(CH3)3]ϩ, 273 (33) [M7b Ϫ CH2C(CH3)3]ϩ, 215 (100) [M8b
C(CH3)3 Ϫ CN]ϩ. MS (CI, isobutane): m/z (%) ϭ 345 (49) [M7b
Ϫ
ϩ
1]ϩ, 299 (100) [M8b ϩ 1]ϩ. HR-MS: calcd. for C18H24N4O3
344.18484, found 344.18300 (diff.: Ϫ5.3 ppm); calcd. for
C1713CH24N4O3 345.18820, found 345.18730 (diff.: Ϫ2.6 ppm).
paration
2,4(1H,3H)-dione (8a, Method A). After chromatography (Rf ϭ
3-{[(2,6-Dimethylphenyl)amino]carbonyl}-2-[(ethoxycarbonyl)- 0.66) and crystallisation, a yellow solid (81.7 mg, 39%) was ob-
of
3-Dodecylpyrimido[4,5-b]-1,8-naphthyridine-
amino]-1,8-naphthyridine (7c): Ethyl chloroformate (1.20 mL,
12.7 mmol) was treated with 2-amino-3-{[(2,6-
tained, m.p. 335Ϫ343 °C (decomp.). Method B: 3-{[(3,3-
Dimethylbutyl)amino]carbonyl}-2-[(ethoxycarbonyl)amino]-1,8-
dimethylphenyl)amino]carbonyl}-1,8-naphthyridine (6c, 199 mg, naphthyridine (7b, 77.6 mg, 225 µmol) was heated to 170 °C in
682 µmol). Synthesis and workup were carried out analogously to
those used for the preparation of 3-[(dodecylamino)carbonyl]-2-
[(ethoxycarbonyl)amino]-1,8-naphthyridine (7a). Chromatography
(Rf ϭ 0.70) yielded a yellow solid (179 mg, 72%), m.p. 158Ϫ162 °C
vacuo (approximately 15 mbar), kept at that temperature for 5 min
and cooled, also in vacuo. The remaining brown solid was recrys-
tallised from hexanes/methanol/chloroform (approximately 1:6:6),
yielding a rose-coloured powder (64.5 mg, 60%) identical to that
(gas evolution). IR (KBr): ν˜ ϭ 3448 cmϪ1 (NϪH), 3044 (arom. obtained from Method A, m.p. 340Ϫ345 °C (decomp.). IR (KBr):
CϪH), 2992 (aliph. CϪH), 1741, 1682, 1624 (CϭO), 1535, 1474 ν˜ ϭ 3434 cmϪ1 (NϪH), 2956, 2853 (aliph. CϪH), 1728, 1662, 1628
1
1
(arom.), 1220 (CϪO). H NMR (300 MHz, CDCl3): δ ϭ 1.34 ppm (CϭO), 1244 (tert-butyl). H NMR (300 MHz, CDCl3): δ ϭ 1.04
(t, J ϭ 7.1 Hz, 3 H, OCH2CH3), 2.33 (s, 6 H, CH3), 4.26 (q, J ϭ ppm [s, 9 H, C(CH3)3], 1.61 (mc, 2 H, NCH2CH2), 4.12 (mc, 2 H,
7.1 Hz, 2 H, OCH2), 7.14 (s, 3 H, ArЈ-H), 7.40 (dd, J1 ϭ 7.9 Hz,
NCH2), 7.53 (dd, J1 ϭ 8.3 Hz, J2 ϭ 4.2 Hz, 1 H, Ar-7-H), 8.34
J2 ϭ 4.7 Hz, 1 H, Ar-6-H), 8.14 (dd, J1 ϭ 7.9 Hz, J2 ϭ 1.9 Hz, (dd, J1 ϭ 8.3 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-6-H), 8.90 (br. s, 0.6 H,
4060
Eur. J. Org. Chem. 2002, 4054Ϫ4062

A New Quadruply Bound Heterodimer DDAD•AADA
FULL PAPER
NH), 9.03 (s, 1 H, Ar-5-H), 9.21 (dd, J1 ϭ 4.2 Hz, J2 ϭ 2.0 Hz, 1 (%) ϭ 209 (58) [M ϩ 1]ϩ, 173 (100). HR-MS: calcd. for C10H16N4O
H, Ar-8-H). 13C NMR (125 MHz, CDCl3): δ ϭ 29.2 [q, C(CH3)3],
29.7 [s, C(CH3)3], 38.5 (t, NCH2CH2), 40.9 (t, NCH2), 111.5 (s, Ar-
208.13242, found 208.13230 (diff.: Ϫ0.6 ppm); calcd. for
C913CH16N4O 209.13577, found 209.13570 (diff.: Ϫ0.3 ppm).
5a-C), 119.9 (s, Ar-9a-C), 121.7 (d, Ar-7-C), 138.7 (d, Ar-6-C), C10H16N4O (208.3): calcd. C 57.67, H 7.74, N 26.90; found C
141.8 (d, Ar-5-C), 149.8 (s, Ar-4a-C), 157.2 (d, Ar-8-C), 160.8 (s,
Ar-10a-C), the Ar-2-C and Ar-4-C signals were not detected due
to the small amount sampled. MS (EI, 70 eV): m/z (%) ϭ 298 (42)
[M]ϩ, 241 (34) [M Ϫ C(CH3)3]ϩ, 215 (100) [M Ϫ C(CH3)3 Ϫ CN]ϩ.
MS (CI, isobutane): m/z (%) ϭ 299 (100) [M ϩ 1]ϩ. HR-MS: calcd.
for C16H18N4O2 298.14297, found 298.14338 (diff.: 1.4 ppm); calcd.
for C1513CH18N4O2 299.14633, found 299.14685 (diff.: 1.7 ppm).
C16H18N4O2 (298.3): calcd. C 64.41, H 6.08, N 18.78; found C
64.24, H 6.08, N 18.54.
57.79, H 7.66, N 26.71.
N-(6-Amino-2-pyridyl)-NЈ-cyclohexylurea (9b): 2,6-Diaminopyrid-
ine (4.40 g, 40.4 mmol) was dissolved in dry toluene (25 mL) under
nitrogen. Freshly distilled cyclohexyl isocyanate (1.30 g, 10.5 mmol)
was added and the mixture was heated to reflux for 2 h. After cool-
ing to room temperature, excess 2,6-diaminopyridine was filtered
off and the solvent was evaporated in vacuo. The residue was dis-
solved in dichloromethane and filtered through silica gel (1 ϫ
3 cm), Rf ϭ 0.72. Evaporation of the solvent and recrystallisation
from toluene gave colourless needles (630 mg, 27%), m.p. 164Ϫ165
°C. IR (KBr): ν˜ ϭ 3441, 3339, 3254 cmϪ1 (NϪH), 2934, 2852 (al-
iph. CϪH), 1663 (CϭO), 1606, 1583, 1500 (arom.), 1537 (NϪH),
3-(2,6-Dimethylphenyl)pyrimido[4,5-b]-1,8-naphthyridine-2,4(1H,3H)-
dione (8c): 3-{[(2,6-Dimethylphenyl)amino]carbonyl}-2-[(ethoxy-
carbonyl)amino]-1,8-naphthyridine (7c, 61.3 mg, 168 µmol) was
heated to 170 °C in vacuo (approximately 15 mbar), kept at that
temperature until gas evolution from the molten material had
ceased (approximately 15 min) and cooled, also in vacuo. The re-
maining brown solid was recrystallised from hexanes/methanol/
chloroform (approximately 1:2:1), yielding pale brown crystals
(39.2 mg, 73%), m.p. 150Ϫ157 °C (gas evolution). IR (KBr): ν˜ ϭ
3392 cmϪ1 (NϪH), 3035 (arom. CϪH), 2812 (aliph. CϪH), 1730,
1674, 1624 (CϭO), 1578, 1499 (arom.). 1H NMR (300 MHz,
CDCl3): δ ϭ 2.19 ppm (s, 6 H, CH3), 7.20Ϫ7.35 (m, 3 H, ArЈ-H),
7.55 (dd, J1 ϭ 8.1 Hz, J2 ϭ 4.3 Hz, 1 H, Ar-7-H), 8.38 (dd, J1 ϭ
8.2 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-6-H), 9.14 (s, 1 H, Ar-5-H), 9.31 (dd,
J1 ϭ 4.3 Hz, J2 ϭ 2.0 Hz, 1 H, Ar-8-H), 10.80 (br. s, 0.8 H, NH).
13C NMR (125 MHz, CDCl3): δ ϭ 17.7 ppm (q, CH3), 111.6 (s,
Ar-5a-C), 120.0 (s, Ar-9a-C), 121.7 (d, Ar-7-C), 128.7 (d, ArЈ-3,5-
C), 129.3 (d, ArЈ-4-C), 132.7 (s, ArЈ-2,6-C), 135.8 (s, ArЈ-1-C),
138.8 (d, Ar-6-C), 142.2 (d, Ar-5-C), 149.6 (s, Ar-4-C), 151.2 (s,
Ar-4a-C), 157.3 (d, Ar-8-C), 157.6 (s, Ar-10a-C), 160.5 (s, Ar-2-C).
MS (EI, 70 eV): m/z (%) ϭ 318 (100) [M]ϩ, 301 (38) [M Ϫ NH3]ϩ,
172 (36) [M Ϫ (CH3)2C6H3 Ϫ N Ϫ HCN]ϩ. MS (CI, isobutane):
m/z (%) ϭ 319 (100) [M ϩ 1]ϩ. HR-MS: calcd. for C18H14N4O2
318.11166, found 318.11189 (diff.: Ϫ0.7 ppm); calcd. for
C1713CH14N4O2 319.11502, found 319.11497 (diff.: 0.2 ppm).
C18H14N4O2 (318.3): calcd. C 67.92, H 4.43, N 17.60; found C
67.87, H 4.53, N 16.97. Purity (HPLC, RP-18, CH2Cl2/MeOH
5:1): 99.2%.
1
1458 (CϪH). H NMR (200 MHz, CDCl3): δ ϭ 1.2Ϫ2.1 ppm (m,
10 H, CH2), 3.7Ϫ3.9 (m, 1 H, CH), 4.30 (br. s, 2 H, NH2), 6.06 (d,
J ϭ 8.0 Hz, 1 H, ArH), 6.13 (d, J ϭ 8.0 Hz, 1 H, ArH), 7.32 (t,
J ϭ 8.0 Hz, 1 H, ArH), 7.89 (br. s, 1 H, NH), 9.15 (d, J ϭ 7.6 Hz,
1 H, NH). MS (EI, 70 eV): m/z (%) ϭ 234 (8) [M]ϩ, 109 (100)
[(NH2)2C5H3N]ϩ. MS (CI/isobutane): m/z (%) ϭ 235 (100) [M ϩ
1]ϩ. C12H18N4O (234.3): calcd. C 61.52, H 7.74, N 23.91; found C
61.47, H 7.76, N 23.89.
N-(6-Acetamido-2-pyridyl)-NЈ-(n-butyl)urea (9c): Triethylamine
(50.5 mg, 500 µmol) was added under nitrogen to a solution of N-
(6-amino-2-pyridyl)-NЈ-(n-butyl)urea (9a, 100 mg, 481 µmol) in dry
dichloromethane (5 mL). Dry acetic anhydride (49.1 mg, 481 µmol)
was then added at 0 °C. After stirring for 1 h at 0 °C and standing
for 15 h, the mixture was hydrolysed with ice/water (20 mL). The
water layer was extracted with dichloromethane (four times 20 mL).
The combined organic layers were dried with sodium sulfate, and
the solvent was evaporated in vacuo. Recrystallisation first from
toluene/ethanol (10:1) and then from toluene gave a colourless solid
(87.6 mg, 73%), m.p. 158Ϫ160 °C. IR (KBr): ν˜ ϭ 3250, 3173, 3076
cmϪ1 (NϪH), 2957, 2928, 2870 (aliph. CϪH), 1689, 1670 (CϭO),
1594 (arom.), 1555 (NϪH), 1449 (CϪH). 1H NMR (200 MHz,
[D6]DMSO): δ ϭ 0.89 ppm (t, J ϭ 7.1 Hz, 3 H, CH3), 1.2Ϫ1.5 (m,
4 H, CH2), 2.10 (s, 3 H, CH3), 3.19 (mc, 2 H, CH2), 6.76 (d, J ϭ
7.9 Hz, 1 H, ArH), 7.40 (d, J ϭ 7.9 Hz, 1 H, ArH), 7.58 (t, J ϭ
8.0 Hz, 1 H, ArH), 8.68 (mc, 1 H, NH), 9.09 (br. s, 1 H, NH), 10.23
(br. s, 1 H, NH). MS (EI, 70 eV): m/z (%) ϭ 250 (70) [M]ϩ, 109
(100) [(NH2)2C5H3N]ϩ. MS (CI/isobutane): m/z (%) ϭ 251 (100)
[M ϩ 1]ϩ. C12H18N4O2 (250.3): calcd. C 57.58, H 7.25, N 22.38;
found C 57.34, H 7.19, N 22.20.
N-(6-Amino-2-pyridyl)-NЈ-(n-butyl)urea (9a): A suspension of 2,6-
diaminopyridine (1.00 g, 9.16 mmol) in dry toluene (50 mL) was
stirred under nitrogen at 50 °C for 1 h, after which n-butyl isocyan-
ate (1.0 mL, 9.2 mmol) was added and the mixture was heated at
reflux for 2 h. The clear solution was concentrated to dryness. The
residue was dissolved in dichloromethane/methanol (9:1) and puri-
fied by chromatography (silica gel with the same eluent). The frac-
tion eluting with Rf ϭ 0.5 was concentrated to dryness and crys-
N-(6-Amino-2-pyridyl)-NЈ-(n-butyl)thiourea (9d): A suspension of
2,6-diaminopyridine (1.00 g, 9.16 mmol) and n-butyl thioisocyan-
ate (1.0 mL, 9.2 mmol) in dry toluene (50 mL) was stirred under
nitrogen at 50 °C for 30 min and then heated under reflux for 3 h.
tallised from cyclohexane/ethanol (2:1). On vacuum filtration of the The clear solution was concentrated to dryness. The residue was
colourless solid, further product precipitated from the mother li-
quor and was also filtered off. Total yield: 559 mg (29%), m.p.
dissolved in chloroform and water (20 mL each) and the phases
were separated. The organic layer was dried with sodium sulfate,
132Ϫ135 °C. IR (KBr): ν˜ ϭ 3484, 3328, 3205 cmϪ1 (NϪH), 2956, concentrated and purified by chromatography (silica gel, dichloro-
2863 (aliph. CϪH), 1678, 1624, 1597, 1564 (CϭO). 1H NMR methane/methanol, 10:1). The fraction eluting with Rf ϭ 0.5 was
(300 MHz, CDCl3): δ ϭ 0.96 ppm (t, J ϭ 7.2 Hz, 3 H, CH3),
concentrated to dryness and crystallised from hexanes/ethanol to
1.35Ϫ1.5 (m, 2 H, CH2CH3), 1.5Ϫ1.65 (m, 2 H, CH2CH2CH3), give large, colourless needles (451 mg, 22%), m.p. 129Ϫ131 °C. IR
3.37 (td, Jt ϭ 6.9 Hz, Jd ϭ 5.6 Hz, 2 H, NHCH2), 4.33 (br. s, 2 H,
NH2), 6.06 (dd, J1 ϭ 8.0 Hz, J2 ϭ 0.7 Hz, 1 H, Ar-3-H), 6.15 (dd, CϪH), 1618, 1587, 1562. H NMR (300 MHz, CDCl3): δ ϭ 0.99
J1 ϭ 7.9 Hz, J2 ϭ 0.7 Hz, 1 H, Ar-5-H), 7.32 (t, J ϭ 7.9 Hz, 1 H, ppm (t, J ϭ 7.3 Hz, 3 H, CH3), 1.4Ϫ1.55 (m, 2 H, CH2CH3),
(KBr): ν˜ ϭ 3458, 3331, 3207 cmϪ1 (NϪH), 2959, 2863 (aliph.
1
Ar-4-H), 8.13 (br. s, 1 H, Ar-NH), 9.13 (br. t, lateral maxima only
visible as shoulders, 1 H, NHCH2). MS (EI, 70 eV): m/z (%) ϭ 208
1.65Ϫ1.75 (m, 2 H, CH2CH2CH3), 3.37 (dt, J1 ϭ 7.1 Hz, J2
ϭ
5.3 Hz, 2 H, NHCH2), 4.33 (br. s, 2 H, NH2), 4.37 (br. s, 2 H,
(14) [M]ϩ, 109 (100) [M Ϫ H9C4NCO]ϩ. MS (CI, isobutane): m/z NH2), 6.06 (dd, J1 ϭ 7.9 Hz, J2 ϭ 0.7 Hz, 1 H, Ar-3-H), 6.15 (dd,
Eur. J. Org. Chem. 2002, 4054Ϫ4062
4061

S. Brammer, U. Lüning, C. Kühl
FULL PAPER
[18]
E. C. Taylor, J. Heterocycl. Chem. 1990, 27, 1Ϫ12. Citation:
‘‘… remarkable solubility (words little known to workers in this
field of heterocyclic chemistry) …’’
J. M. Quintela, R. M. Acras, C. Veiga, C. Peinador, V. Vilas,
V. Ojea, Heterocycles 1996, 36, 2513Ϫ2522.
G. Argiropoulos, L. W. Deady, M. Dorkos, Aust. J. Chem.
1991, 44, 481Ϫ486.
F. C. Whitmore, A. H. Homeyer, V. H. Wallingford, J. Am.
Chem. Soc. 1933, 55, 4209Ϫ4214.
A. P. Howe, H. B. Hass, Ind. Eng. Chem. 1946, 38, 251Ϫ253.
NMR data for 9aϪc have been published, see ref.[11]
H.-J. Schneider, A. Yatsimirsky, Principles and Methods in Sup-
ramolecular Chemistry, J. Wiley & Sons Ltd., Chichester, 2000,
chapter D3.
J1 ϭ 8.0 Hz, J2 ϭ 0.7 Hz, 1 H, Ar-5-H), 7.32 (t, J ϭ 8.0 Hz, 1 H,
Ar-4-H), 8.32 (br. s, 1 H, Ar-NH), 11.39 (br. t, lateral maxima only
visible as shoulders, 1 H, NHCH2). MS (EI, 70 eV): m/z (%) ϭ 224
(60) [M]ϩ, 152 (35) [M Ϫ H9C4NH]ϩ, 109 (100) [M Ϫ H9C4NCS]ϩ.
MS (CI, isobutane): m/z (%) ϭ 225 (100) [M ϩ 1]ϩ. HR-MS: calcd.
for C10H16N4S 224.10957, found 224.11050 (diff.: Ϫ4.1 ppm);
calcd. for C913CH16N4S 225.11293, found 225.11340 (diff.:
2.1 ppm). C10H16N4S (224.3)[41]: calcd. C 53.54, H 7.19, N 24.97,
S 14.29; found C 54.15, H 8.01, N 24.39, S 13.31.
[19]
[20]
[21]
[22]
[23]
[24]
[1]
See for example: L. Stryer, Biochemistry, W. H. Freeman and
Company, New York, 1995, part V.
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[2]
P. S. Corbin, S. C. Zimmerman, J. Am. Chem. Soc. 1998, 120,
J. J. Voegel, S. A. Benner, J. Am. Chem. Soc. 1994, 116,
9710Ϫ9711.
6929Ϫ6930, and refs. cited.
S.-K. Chang, D. Van Engen, E. Fan, A. D. Hamilton, J. Am.
[3]
F. H. Beijer, R. P. Sijbesma, H. Kooijman, A. L. Spek, E. W.
Meijer, J. Am. Chem. Soc. 1998, 120, 6761Ϫ6769.
M. Bolte, C. Kühl, U. Lüning, Acta Crystallogr., Sect. E 2001,
57, 502Ϫ504.
L. V. Sudha, D. N. Sathyanarayana, J. Chem. Soc., Perkin
Trans. 2 1986, 1647Ϫ1650.
P. S. Corbin, S. C. Zimmerman, J. Am. Chem. Soc. 2000, 122,
Chem. Soc. 1991, 113, 7640Ϫ7645.
A. D. Burrows, C.-W. Chan, M. M. Chowdry, J. E. McGrady,
[4]
D. M. P. Mingos, Chem. Soc. Rev. 1995, 24, 329Ϫ339.
G. M. Whitesides, E. E. Simanek, J. P. Mathias, C. T. Seto, D.
[5]
C. Chin, M. Mammen, D. M. Gordon, Acc. Chem. Res. 1995,
28, 37Ϫ44.
[6]
3779Ϫ3780.
A. G. Bielejewska, C. E. Marjo, L. J. Prins, P. Timmerman,
U. Lüning, C. Kühl, M. Bolte, Acta Crystallogr., Sect. C 2001,
57, 989Ϫ990.
For an investigation on the strength of the intramolecular hy-
drogen bond see: S. H. M. Söntjens, J. T. Meijer, H. Kooijman,
A. L. Spek, M. H. P. van Genderen, R. P. Sijbesma, E. W.
Meijer, Org. Lett. 2001, 3, 3887Ϫ3889.
P. S. Corbin, S. C. Zimmerman, P. A. Thiessen, N. A. Hawry-
luk, T. J. Murray, J. Am. Chem. Soc. 2001, 123, 10475Ϫ10488.
F. G. Bordwell, D. J. Algrim, J. A. Harrelson, Jr., J. Am. Chem.
Soc. 1988, 110, 5903Ϫ5904.
E. Fan, S. A. Van Arman, S. Kincaid, A. D. Hamilton, J. Am.
Chem. Soc. 1993, 115, 369Ϫ370.
G. Jander, E. Blasius, Lehrbuch der analytischen und präpara-
tiven anorganischen Chemie, 13th ed., S. Hirzel, Stuttgart, 1989,
p. 51.
An alkylated barbituric acid (11) was commercially available;
a suitable receptor (10) was synthesised in analogy to literature
procedures: S. Brammer, Ph.D. thesis, Universität Kiel, 2001;
according to ref.[3]
F. de Jong, D. N. Reinhoudt, J. Am. Chem. Soc. 2001, 123,
7518Ϫ7533, and ref. cited.
Recent review: S. C. Zimmerman, P. S. Corbin, Struct. Bonding
[7]
(Berlin, Ger.) 2000, 96, 63Ϫ94.
R. P. Sijbesma, F. H. Beijer, L. Brunsveld, B. J. B. Folmer, J.
[8]
H. K. K. Hirschberg, R. F. M. Lange, J. K. L. Lowe, E. W.
[32]
[33]
[34]
[35]
Meijer, Science 1997, 278, 1601Ϫ1604.
B. J. B. Folmer, R. P. Sijbesma, E. W. Meijer, J. Am. Chem.
[9]
Soc. 2001, 123, 2093Ϫ2094, and ref. cited.
[10]
C. Schmuck, W. Wienand, Angew. Chem. 2001, 113,
4493Ϫ4499; Angew. Chem. Int. Ed. 2001, 40, 4363Ϫ4369.
U. Lüning, C. Kühl, Tetrahedron Lett. 1998, 5735Ϫ5738. Paral-
[11]
lel with our work, Corbin and Zimmerman were working on a
very similar system. After rejection by Angew. Chem., our pa-
per was accepted by Tetrahedron Lett. on 15 May 1998, while
the Corbin/Zimmerman paper was accepted by J. Am. Chem.
Soc. on 1 June 1998.
P. S. Corbin, S. C. Zimmerman, J. Am. Chem. Soc. 1998, 120,
9710Ϫ9711.
D. K. J. Gorecki, E. M. Hawes, J. Med. Chem. 1977, 20,
124Ϫ128.
E. M. Hawes, D. K. J. Gorecki, R. G. Gedir, J. Med. Chem.
[36]
[12]
[37]
[38]
[39]
[40]
[41]
[13]
W. L. Jorgensen, J. Pranata, J. Am. Chem. Soc. 1990, 112,
2008Ϫ2010.
[14]
T. J. Murray, S. C. Zimmerman, J. Am. Chem. Soc. 1992, 114,
4010Ϫ4011.
H. P. Hopkins, C. J. Alexander, S. Z. Ali, J. Phys. Chem. 1978,
1977, 20, 838Ϫ841.
I. Cardineau, A. Gueiffier, J.-C. Debonzy, J.-C. Milhavet, J.-P.
[15]
[15a]
82, 1268Ϫ1272.
Chapat, Heterocycles 1993, 36, 2513Ϫ2522.
reports a yield of ‘‘9.4 g (58%)’’, but 9.4 g in the run cited is
The literature
Y. Kyogoku, R. C. Lord, A. Rich, Proc. Natl. Acad. Sci. USA
1967, 57, 250Ϫ257.
The elemental analysis was poorly reproducible because 9d is
calculated to be 25%.
H. Migeon, A. Fradet, P.-J. Madec, E. Marechal, Bull. Soc.
Chim. Fr. 1995, 132, 967Ϫ974.
A. Gazit, N. Osherov, I. Posner, P. Yaish, E. Paradosu, J. Med.
Chem. 1991, 34, 1896Ϫ1907.
[16]
[17]
´
quite volatile.
Received April 24, 2002
[O02224]
4062
Eur. J. Org. Chem. 2002, 4054Ϫ4062