Synthesis, characterization and computational modeling of cyclen substituted with dendrimeric branches. Dendrimeric and macrocyclic moieties working together in a collective fashion

Article abstract of DOI:10.1016/j.molstruc.2005.06.035

A molecular system, formed by the aza-macrocycle known as cyclen and functionalized dendrons, was synthesized in a convergent fashion. An aliphatic spacer between the two molecular units was incorporated to avoid the sterical hindrances. The N-acylation reactions to carry out the cyclen substitution, particularly those in solid state, resulted in very good yields since no further purifications were necessary. The obtained molecular system was characterized by common spectroscopic techniques. According to X-ray analysis (powder analysis), there is a random arrangement of the molecules, with an average interplanar distance (the most recurrent) of 4.8 ? between first neighbors. A theoretical study at DFT level of theory was carried out, simulating the inclusion process of Ni²⁺ into two different model molecules (A and C), with and without dendrons present. According to the calculated interaction energies in the inclusion process, there is a significant contribution in terms of energy (-60.89 Kcal/mol) due to the presence of functionalized dendrons, which enfold the metal ion, forming a molecular cage and increasing in this way the degree of pre-organization due to cooperative effects created by the dendrimeric environment. These molecular arrangements would be relevant in areas like catalysis, having influence on the selectivity of catalytic processes, by controlling the accessibility of the active site (the metal atom) via sterical hindrances. Further experimental work of the incorporation of metal ions of the first transition series will be done in a rational way, taking into account the theoretical results obtained so far.

Full text of DOI:10.1016/j.molstruc.2005.06.035

Journal of Molecular Structure 779 (2005) 1–10
www.elsevier.com/locate/molstruc
Synthesis, characterization and computational modeling of cyclen
substituted with dendrimeric branches. Dendrimeric and macrocyclic
moieties working together in a collective fashion
Jorge A. Cruz-Morales, Patricia Guadarrama
*
Instituto de Investigaciones en Materiales, Universidad Nacional Aut o´ noma de M e´ xico, Apartado Postal 70-360, CU,
Coyoac a´ n, M e´ xico DF 04510, M e´ xico
Received 10 May 2005; revised 28 June 2005; accepted 29 June 2005
Available online 26 August 2005
Abstract
A molecular system, formed by the aza-macrocycle known as cyclen and functionalized dendrons, was synthesized in a convergent
fashion. An aliphatic spacer between the two molecular units was incorporated to avoid the sterical hindrances. The N-acylation reactions to
carry out the cyclen substitution, particularly those in solid state, resulted in very good yields since no further purifications were necessary.
The obtained molecular system was characterized by common spectroscopic techniques. According to X-ray analysis (powder analysis),
˚
there is a random arrangement of the molecules, with an average interplanar distance (the most recurrent) of 4.8 A between first neighbors. A
2
C
theoretical study at DFT level of theory was carried out, simulating the inclusion process of Ni into two different model molecules (A and
C), with and without dendrons present. According to the calculated interaction energies in the inclusion process, there is a significant
contribution in terms of energy (K60.89 Kcal/mol) due to the presence of functionalized dendrons, which enfold the metal ion, forming a
molecular cage and increasing in this way the degree of pre-organization due to cooperative effects created by the dendrimeric environment.
These molecular arrangements would be relevant in areas like catalysis, having influence on the selectivity of catalytic processes, by
controlling the accessibility of the active site (the metal atom) via sterical hindrances.
Further experimental work of the incorporation of metal ions of the first transition series will be done in a rational way, taking into account
the theoretical results obtained so far.
q 2005 Elsevier B.V. All rights reserved.
Keywords: Aza-macrocycles; Dendrimers; Collective effects; Convergent synthesis; DFT Methods
1
. Introduction
Aza-macrocycles (with nitrogen as heteroatom) have
found also many applications in medical diagnosis and
therapy when they form complexes with metal ions;
particularly the complexes with paramagnetic lanthanides
like Gd (III) which are in current clinical use as contrast
agents for magnetic resonance imaging [3]. In molecular
recognition, aza-macrocycle-metal complexes have shown
abilities as binding sites [4].
Since their discover [1], the macrocyclic ligands have
been successfully used in complexation processes by means
of the possibility to have suitable cavity sizes and different
donor atoms (N, O, S, etc) as well as for the observed
extra thermodynamic stabilization due to the macrocyclic
effect [2].
Among others, some applications founded for macro-
cycles are: (i) as ion-carriers between different phases; (ii)
as ion separation agents; (iii) as ion-based sensors and (iv)
as isotopes-separating agents.
Macrocyclic compounds by themselves confer high
pre-organization conditions, which are very convenient in
terms of complexation stability and molecular recognition
selectivity. This term of pre-organization, coined by Cram
[5], refers to the degree of similitude in the arrangement
*
Corresponding author. Tel.: C52 55 56224594; fax: C52 55 56161201.
E-mail address: patriciagua@correo.unam.mx (P. Guadarrama).
of the donor atoms before and after the complexation, that
is to say, the similitude between the free ligand as a host
and the formed host-guest complex (e.g. a metal ion as
guest).
0
022-2860/$ - see front matter q 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2005.06.035

2
J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
Fig. 2. Hetero-Macrocycles. Cyclen, labeled as N12C4, is shown in red (For
interpretation of the reference to colour in this legend, the reader is referred
to the web version of this article).
Fig. 1. Pre-organization of two different macrocycles.
the macrocycle cyclen (labeled as N12C4 in red) was the best
candidate in terms of interaction energy, when complexes
with transition metals (Fe, Ni, Cu, Zn) are formed.
Furthermore, the previous calculations showed that the
attachment of some model dendrons to cyclen results in an
energetically favorable cooperative behavior (‘umbrella-
like’ arrangements).
The degree of pre-organization usually depends on the
flexibility of the aliphatic framework forming the macro-
cycle and also depends on the hardness-softness features of
the heteroatoms involved in the structures.
It was observed before [6] that there is a compromise in
energy between the pre-organization of the macrocycles and
their self-stabilization by means of internal hydrogen bonds
Therefore, in order to design and synthesize novel,
chemically collective systems and, taking into account the
preliminary theoretical results, the aza-macrocycle known
as cyclen was substituted with dendrons separately prepared
according to the procedures described by Newkome et al.
(
or another non-covalent interaction). In some cases (e.g.
macrocycle N12C4 in Fig. 1), the macrocycles are
conformationally able to interact with metal ions, whereas
in other cases (e.g. macrocycle NyS12C4), the macrocycles
become unable to interact ‘easily’ with metal ions because
they are collapsed before the complexation occurs. To
interact with metal ions, those ‘collapsed’ macrocycles must
pay the energetic price by some conformational rearrange-
ments, resulting in a lower interaction energy with metal
guests.
[7]. Two different aliphatic spacers were incorporated
between cyclen and dendrons to avoid sterical hindrances
throughout the synthesis. In order to study the inclusion
process of metal ions into the molecular systems syn-
thesized, theoretical calculations were carried out at DFT
(density functional theory) level. In terms of interaction
Considering some of the factors involved in complexa-
tion processes, it has been generally established that the
level of pre-organization expected in a variety of ligands
follows the next order:
energy, a collective effect due to the dendrimeric branches
was observed when the incorporation of metal transition
2
C
ions (e.g. Ni ) was simulated.
LIGAND
2
2
2
. Experimental section
Unidentate Chelate Macrocyclic Macrocyclic + chelating Cryptand
donor groups
.1. Materials and methods
Increasing level of pre-organization
.1.1. Materials
Tosylate chloride was crystallized from petroleum ether.
According to the above information, the combination of
macrocycles with an additional set of donor groups appears
as an attractive alternative, in terms of pre-organization
conditions. In the present study, the incorporation of such
additional donor groups in a multifunctional fashion was
carried out by the substitution of macrocycles like cyclen (1,
All other reagents were used as received from Aldrich and
Baker.
2.1.2. Measurements
FT-IR spectra: Spectrometer Nicolet (FT-IR) 510p.
1
13
4
,7,10-tetraaza-cyclododecane) with functionalized dendri-
NMR H, C: Spectrometer Bruker 400 MHz.
X-Ray analysis: Diffractometer Brucker-AXS,
model D8-Advance (radiation Cu Ka with graphite
monochromator).
meric branches (dendrons) bearing additional polar groups.
Theoretical work previously done [6] showed that, among a
series of macrocycles (like those illustrated in Fig. 2),

J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
3
1
3
2
.1.3. Dendron synthesis
Di-tertbutyl-4-nitro-4-[2-(tertbutoxycarbonyl)ethyl]-1,7-
heptanodiate (3).
CH –O–CH ) 3.69 (s, 2H, N–CH ) 5.21 (s, 1H, OH). C:
2 2 2
26.62 and 38.58 (C and C ), 59.83 and 68.52 (C and C ),
1 4 7 2
70.50 and 74.99 (C and C ) 154.50 (C ).
5 6 3
A mixture of nitromethane (6.1 g, 100 mmol) and
benzylammonium hydroxide (3 ml of 40% solution in
MeOH) was stirred at 65–70 8C in 20 ml of dioxane.
Further addition of tertbutylacrylate (39.7 g; 310 mmol)
was carried out, maintaining the temperature at 70–80 8C for
[N-tertbutoxycarbonyl-2-(aminoethoxy)] ethyltosylate
(6)
12 g (58.5 mmol) of precursor ([N-tertbutoxycarbonyl-2-
(aminoethoxy)] ethanol) and 22 ml of pyridine was mixed in
dichloromethane (100 ml) at room temperature and anhy-
dride conditions. 13 g (58.6 mmol) of tosylate chloride
newly crystallized was added and, after 1 day, the reaction
mixture was treated with diluted HCl (10 ml). The product
was extracted with CHCl (2!100 ml), washed with H 0
1
CHCl (200 ml) and washed with 50 ml of HCl (10%) and
h. The mixture was concentrated and re-dissolved in
3
with brine (saturated solution NaCl/H 0) (3!50 ml) and
2
then dried with MgSO . The solvent was removed, resulting
4
3
2
in a yellow solid. By crystallization from EtOH, the
compound (3) was obtained in 72% yield as a crystalline
(2!100 ml) and dried with MgSO . The solvent was
4
removed to obtain the product as a yellow oil. Yield: 80%.
K1
FT-IR (cm ): 3403 (–NH), 2979 (C–H aliph.), 1727
K1
white solid (m.p. 98–100 8C). FT-IR (cm ): 2975 (C–H
1
aliph.), 1724 (CaO), 1536 (–NO ). NMR (CDCl )
(
H
v(ppm)): 1.42 (s, 27H, OtBu), 2.25 (s, 12H, aliph.). C:
2
3
(CaO), 1390 (–SO ), 1159 (–O–), 1549 (CaC arom.).
2
1
3
1
NMR (CDCl ) H (v(ppm)): 1.19 (s, 9H, methyl groups),
3
2
9
6.89 and 28.53 (C and C ), 29.66 and 79.65 (C and C ),
6 9 5 8
2.13 (s, 3H, arom-CH ), 2.94 (d, 2H, NH ), 3.21 (t, 2H,O–
2
3
2.46 and 170.42 (C and C (CO)).
4
7
CH ) 3.34 (s, 1H, NH) 3.41 (t, 2H, CH –O), 3.61 (t, 2H,
2 2
Di-tertbutyl-4-amino-4-[2-(tertbutoxycarbonyl)ethyl]-1,
-heptanodiate. (4)
A solution of compound (3) (4.46 g; 10 mmol) in
CH -Tos), 7.76 (t, 4H, ortho-arom.), 7.57 (t, 4H, para-
2
1
3
arom.). C: 22.45 and 29.70 (C12 and C ), 44.36 and 62.28
7
1
(C and C ), 70.05 and 71.30 (C and C ), 72.02 and 125.54
4 7 6 2
absolute EtOH (100 ml) was prepared and Raney nickel
catalyst (4 g) was added. The mixture was shaken under
hydrogen atmosphere (40 psi) at room temperature during
(
C and C ), 131.24 and 138.54 (C and C ), 149.99 and
5 9 10 8
1
57.49 (C11 and C3).
1
ethyl)))]-1,4,7,10-tetraazacyclo dodecane (7)
,4,7,10-Tetra-[N-(tertbutoxycarbonyl-2-(aminoethoxy
2
8 h, time when no further hydrogen was consumed. The
(
catalyst was filtered through celite and washed with absolute
EtOH. The solvent was removed to give a yellow viscous
liquid which was purified by chromatography on SiO2
To a solution of cyclen (0.1 g, 0.58 mmol) in DMF
20 ml) was added K CO (42 mg) under nitrogen atmos-
(
2
3
phere at TZ60 8C. Slow addition of aliphatic spacer (6)
849 mg, 2.36 mmol) was carried out for 2 h. The reaction
column, eluting with AcOEt to obtain a white solid in 88%
K1
yield (m.p. 51–52 8C). FT-IR (cm ): 3376 (–NH ), 2979
(
2
mixture was stirred for 1 day at 60 8C, then it was cooled at
room temperature, the suspended solids were filtered and
the solvent was removed in vacuum to obtain a brown oil
(
C–H aliph.), 1724 (CaO).
1
NMR (CDCl ) H (v(ppm)): 1.43 (s, 27H, OtBu), 1.6 (t,
3
1
H, aliph.), 1.9(s,2H, NH), 2.23 (t, 6H, aliph. CO). C 28.3
3
6
which was crhomatographed on SiO , eluting with AcOEt.
2
(
C ), 173.27 (C (CO)).
C ), 30.20 and 34.61(C and C ), 52.56 and 80.53(C and
9 6 5 4
According to NMR spectra, mono-, di- and tri-substitution
could happen.
8
7
Aliphatic spacer (6)
[N-tertbutoxycarbonyl-2-(aminoethoxy)] ethanol (pre-
cursor)
K1
Yield: 10%. FT-IR (cm ): 2975 (C–H aliph.), 1590
–NH–), 1779 (CaO), 1153 (–O–).
(
1
NMR (CDCl ) H (v(ppm)): 1.54 (s, 36H, -t-Boc), 2.33
3
O
3
(s, 16H, cyclen), 3.18 (m, 8H methylene groups, neighbors
to cyclen), 3.37 (m, 8H, methylene groups, neighbors to
amine), 3.90 (m, 16H, CH –O).
1
4
7
O
2
OH
O
HN
2
5
6
2.1.4. Solid-state reaction
A solution of di-tertbutyl bicarbonate (t-BOC) (14.6 g,
6.9 mmol) in THF (50 ml) was stirred at room temperature
and 2-(2-aminoethoxy) ethanol (7 g, 66.6 mmol) was added
1,4,7,10-Tetra-[3-carboxy-propanoyl]- 1,4,7,10-tetraaza-
cyclododecane (9).
6
350 mg (3.50 mmol) of succinic anhydride and 150 mg
(0.87 mmol) of cyclen were placed in a mortar and the
reaction mixture was grinded by repeated contacts over and
over again for 2 h, until the appearance was changed to a
viscous solid. The product was obtained as a gray viscous
slowly with liberation of CO . The mixture was refluxed for
2
4 h. Afterwards the mixture was concentrated at vacuum to
obtainanoilwhichwaschromatographedonSiO elutingwith
AcOEt to achieve the final product in 75% yield (yellow oil).
2
K1
FT-IR (cm ): 3409 (OH), 2931 (C–H aliph.), 1700
K1
solid in 90% yield. FT-IR (cm ): 3200–2900 (–COOH),
1750 (CaO), 1610 (N–CaO). NMR: (D
1
O) H (v(ppm)):
(
CaO), 3409 (–NH), 1100 (–O–).
1
NMR (CDCl ) H (v(ppm)): 1.4 (s, 9H, methyl groups),
2
1.80 (d, 8H, CO–CH –), 2.24 (d, 8H, –CH –COO–), 3.46 (s,
2 2
3
1
3
2
.90 (s, 1H, NH), 3.28 (d, 2H, CH –OH), 3.52 (m, 4H,
2
16H, macrocycle CH ’s), 10.40–13.60 (H-acids). C: 27.96
2

4
J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
cooled at room temperature, filtering the precipitate
dicyclohexylurea) formed in the course of the reaction.
Unreacted DCC was eliminated by acetic acid (15 ml). The
(
crude product was washed with NaHCO K10% (3!10 ml)
3
and H 0 (1!10 ml) and dried with MgSO . The solvent was
2
4
removed and the product was re-dissolved with petroleum
ether. After evacuation of the solvent, a viscous solid
K1
yellow-colored was obtained. Yield: 45%. FT-IR (cm ):
2
975 (C–H aliph.), 1727 (CaO), 1530 (–CO–NH–). NMR
1
CDCl ) H (v(ppm)): 1.44 (s, 108H, OtBu), 1.76 (m, 24H, –
(
N–C–CH –), 1.94 (m, 24H, –CH –COOtBu), 2.70 (m, 16H,
3
2
2
1
3
–
2
3
1
COO–CH –CH –COO–), 3.21 (m, 16H, cyclen CH ‘s). C:
2 2 2
5.62 and 26.10 (C and C ), 28.03 and 29.95 (C and C4),
3
8
11
3.92 and 34.35 (C and C ), 52.29 and 80.28 (C and C ),
6
7
1
10
72.38 (C and C ), 173.03 (C ).
2
5
9
2
.1.5. Computational details
All the initial structures were equilibrated by confor-
mational searching (Force Field OPLS-AA [8]) using the
Monte Carlo statistical method [9] included in Macromodel
software; choosing the algorithm of multiple minimum
(MCMM) [10] without limits on the number of variable
torsions allowed in the search. The Geometry optimizations
were carried out at DFT level of theory, using the hybrid
functional BHandHLYP [11], which considers the exchange
term as 50% of exact Hartree-Fock (HF) exchange functional
and 50% of Slater local exchange functional, and the
correlation term as the Lee-Yang-Parr local and nonlocal
functional[12]. The used basis set LACVP (included inJaguar
5.0 program [13]); considers effective core potentials (ECP’s)
generated to replace the innermost core electron for third-row
Fig. 3. Synthesis of Dendron (4).
and 29.01 (C and C ), 44.45 (C ), 186.00 and 189.01 (C
2
3
4
1
and C5).
,4,7,10-Tetra-[1-propanoyl (N-[3 -(tertbutoxycar-
0
1
bonyl)-1 ,1 -bis[2 -(tertbutoxycarbonyl) ethyl] propyl]-1 -
0
0
00
0
amino]carbonyl])]-1,4,7,10-tetraazaciclododecane (10).
1
40 mg (0.24 mmol) of compound (9) was dissolved in
CHCl (15 ml) at pHZ5 and DCC (dicyclohexylcarbodii-
3
mide) (410 mg) was added. The reaction mixture was
refluxed and, after 2 h, dendron (4) (409 mg, 0.98 mmol)
was put in there. The reaction was followed by tlc (eluting
with AcOET:hexane (2:1)). After 20 h, the mixture was
Fig. 4. Cyclen functionalization by N-alkylation (compound (7)) and N-acylation (compound (9)).

J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
5
(
K–Cu), fourth-row (Rb–Ag) and fifth-row (Cs–Au), in order
to incorporate the relativistic effects in case of Nickel atom
many electrons). LACVP, as well as LAV3P and LACV3P
tertbutylacrylate, aliphatic dendron (3) was obtained in
good yield (72%) as a crystalline white solid. The NH focal
2
(
point, required to attach the dendrimeric moieties to cyclen,
was incorporated by catalytic hydrogenation of dendron (3)
at room temperature to achieve the dendron (4) in 88% yield
basis sets use the pseudospectral method while all other ECP
basis sets use the analytic method.
Molecular dynamics (MD) calculations were carried out at
heating rate of 1 kcal per atom per femtosecond (fs) and a
targettemperatureof300 K [14]. Successive conformationsof
each system were generated by the integration of Newton’s
laws of motions, resulting in trajectories that specify how the
positions and the velocities of the particles in the systems vary
with time, under the Beeman’s algorithm [15].
(
white solid).
An important feature of the dendron (4) is an appropriate
equilibrium between the reactivity of the focal point (–NH2)
on the one hand, and the sterical effects on the other hand.
Since a quaternary carbon is adjacent to the –NH group,
2
this reactive group posses a fairly low reactivity, enough to
permit that structural rearrangements can take place when
the dendron reacts with something else, avoiding in some
extent the sterical hindrances.
3
. Results and discussion
In terms of design, the aliphatic framework of dendron
4) was incorporated to improve the solubility properties,
(
3
.1. Synthesis
but also to allow the free-movements for encapsulation
purposes. The terminal carboxy groups are expected to
interact in a collective fashion as a part of the molecular
system.
According to Fig. 3, after successive Michael-type
additions between the carbanion of nitromethane and
Fig. 5. Synthesis of molecule (10) by DCC coupling.

6
J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
pathways, the N-acylation gave the best results in the
present study. Whereas compound (7) was obtained in a
very low yield (10%) by N-alkylation reactions of cyclen in
presence of the alkyl-tosylate (6) previously prepared (see
experimental section), the N-acylation reaction by the ring
opening of succinic anhydride resulted in a very good yield
(90%) to obtain compound (9). Further purification in case
of compound (9) was not necessary since the reaction was
carried out in solid state (in absence of any solvent).
Compound (9) was soluble in common organic solvents,
even in water. The solid–solid reaction required repeated
contacts over and over again but it was very efficient and
waste-free, which is environmentally important. The solid
state reactions has been described as thermal intracrystalline
conversions where loss of volatile fragments might be
involved [16]. Thus, reactions like complexation, hydro-
genation, aliphatic substitutions among others, can be
carried out in solid state.
By N-acylation reactions, not only higher yields were
obtained but also further reactions considered here as part of
the synthetic route, were avoided. If we compare compounds
(
7) and (9), while compound (9) is ready to react with the
dendron (4) by means of the COOH groups, in a convergent
fashion, compound (7) needs to be transform, removing the
t-BOC protecting groups for the subsequent formation of the
isocyanate groups, in presence of triphosgene, in order to
Fig. 6. Comparison of FT-IR spectra. Compound (9) is the precursor and
Compound (10) is the product.
react afterwards with the –NH focal point of the dendron (4).
2
The coupling of compounds (9) and (4) in presence of
DCC resulted in compound (10) in 45% yield (Fig. 5)
Infrared spectra gave a good insight of the coupling
reaction shown in Fig. 5. The disappearance of the broad
The functionalization of the aza-macrocycle, cyclen,
started with the incorporation of aliphatic spacers to connect
it afterwards with dendron (4). In Fig. 4 are shown the two
alternatives tested to attach aliphatic spacers to cyclen,
obtaining compounds (7) and (9).
K1
band due to –COOH group at 3200 cm (O–H vibration)
coming from the cyclen-spacer moiety, as well as the
enlargement of the N–H bond in compound (10) in
comparison with that bond present in the free dendron (4),
are both evidences of the transformation. The vibration
As it has been reported, the main synthetic routes for the
substitution of aza-macrocycles are: N-alkylation or
N-acylation reactions. Comparing these two synthetic
Fig. 7. Diffraction pattern of compound (10).

J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
7
frequency (mm) of the N–H bonds in compound (10) and free
K1
K1
dendron (4) are 6.06 (1650 cm ) and 6.66 (1500 cm ),
respectively. Considering that the frequencies are inversely
proportional to the bond distance, the slight elongation of the
N–H bond adjacent to the carbonyl group in compound (10)
is an indication of the aliphatic substitution (Fig. 6).
1
The NMR H data described above in experimental
section are in agreement with the proposed structures.
Compound (10) is a viscous solid (amorphous) randomly
oriented, according with the powder X-ray diffraction
analysis (Fig. 7). From the gaussian-like curve, generated
over 2 h approximately (step: 0.0208, step time:2.4 s) at
room temperature, it is only possible to distinguish an
˚
average distance of 4.87002 A between molecular planes, at
first neighbors, as predominant distance. This molecular
arrangement can be visualized in Fig. 8 where, arbitrarily,
six of the molecules under study were put together and a
molecular dynamics calculation was carried out (see
computational details).
Fig. 8. Cumulus modeled by molecular dynamics, with six molecules in an
approximate interplanar distance of 16/4 angstroms.
2
C
Fig. 9. Model compounds taken for the theoretical study. Molecules B and D correspond to the hypothetical inclusion of Ni . Molecules A and C correspond
to the synthesized compounds (9) and (10), respectively.

8
J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
Table 1
Total energies (Ha) calculated at BHandHLYP/LACVP level of theory
Molecule
Total energy
A
B
C
D
K2038.095299
K2205.964385
K5755.766243
K5923.732370
K167.320198
2
C
Ni
1HaZ627.51 kcal/mol.
To distinguish the limits of each molecule, the two
terminal entities are illustrated as full-filled models whereas
the other four are showed in a wire type. It can be observed
that roughly 4 of the 6 molecules are aligned along the
double arrow showed in the Fig. 8, in a nearly distance of
˚
16 A which is in agreement with the more recurrent distance
of 16/4 A observed in X-ray spectrum.
˚
3
3
.2. Theoretical study
Fig. 10. (a) Model compound D involving dendrimeric branches; (b) Model
compound B without dendrimeric branches.
.2.1. Interaction energy calculations
In order to evaluate the collective effects, presumably
due to the dendrimeric environment, theoretical calculations
were carried out considering, on the one hand, the molecule
after the geometry optimization of the four model molecules
are shown in Table 1.
(
10) with and without metal ion incorporated and, on the
Thus, the calculated interaction energies for the two
different modeled inclusion complexes (Fig. 9) are:
other hand, the precursor (9) (without dendrones) in the
same two situations described above for molecule (10). In
both cases the Ni
2
C
2C
was considered as guest, in a
(a) ACNi /B
Einteraction ZK344:4316 Kcal=mol
Einteraction ZK405:3269 Kcal=mol
2C
configuration of low spin (square planar geometry).
Interaction energies were obtained according to the
variation method [17] which indirectly estimates such
energies as the difference between the energy of the
(b) CCNi /D
(Comparative information: Binding energy calculated for
2C
Porphyrin-Ni complex: K412.1604 kcal/mol [6]).
As it can be observed, there is a significant contribution
in energy due to the dendrimeric environment (DE DKBZK
‘
complex’ and the energies of their isolated parts
ðDEZE
KðE CEmetal ionÞ.
complex
host
C2
As it was mentioned in computational details, the basis
set labeled as LACVP was used in order to incorporate the
relativistic effects caused by the many electrons of atoms
like transition metal ions [18]. Into the DFT method, the
hybrid functional chosen was BHandHLYP, in spite of its
good results in the evaluation of non-covalent interactions
60.89 Kcal/mol) in the process of inclusion of Ni . The
resulting structures (Fig. 10) showed that the metal ion
remains in both cases in the middle of the aza-macrocycle,
after the full-optimization of the geometries, maintaining a
little distorted square geometry by the assistance of both, the
cyclen and the neighbor groups. The introduction of
carbonyl groups, directly attached to the nitrogen atoms
of cyclen, undoubtedly reduces their basicity; however,
there are still some interactions with the metal. The
[
in Fig. 9. Molecules B and D correspond to the initial
e.g. Ref. 19]. The model compounds calculated are shown
2
C
structures for the hypothetical inclusion of Ni . Molecules
2
C
A and C correspond to the synthesized compounds (9) and
(
distances between Ni and N (cyclen) atoms are enlisted
in Table 2.
As a reference, the reported average distance between
10), respectively.
It is important to mention that the metal ion in structures
2
C
B and D was deliberately put inside the aza-macrocycle, as
initial input, only to start the geometry optimization
˚
Zn and deprotonated N atoms of cyclen is 2.136 A [20].
Comparing with the distances shown in Table 2, there are
similar ranges of interaction.
2
C
calculations with an initial square geometry for the Ni
.
Hence, the electronic distribution around the metal,
apparently unlikely due to the presence of amide groups
Table 2
2
C
Theoretical interaction distances (A) between Ni and N atoms of cyclen
˚
(
formed after the cyclen substitution), may possibly change
during the full-geometry optimization, as a result of the
assistance of the neighbor groups, with the possible shift of
the metal ion through the molecule and the consequent lost
of a square planar geometry. The total energies obtained
Molecule
Distance 1
Distance 2
Distance 3
Distance 4
B
D
2.62
1.94
2.45
1.87
1.807
1.85
1.802
1.87

J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
9
Fig. 11. Several aza-macrocycles (including those under study; A and C) and the visualization of their HOMOs.
It is possible to find many examples in the literature
where aza-macrocycles form complexes with metal ions, by
the deprotonation of the nitrogen atoms at basic pH (pHO
geometry is kept and the neighbor groups are assisting the
inclusion of the metal ion. An examination of the frontier
orbitals (HOMO–LUMO) of the model molecules A and C
(before de inclusion of Ni ) give an idea of the
participation of nitrogen atoms to these molecular orbitals.
The HOMO–LUMO orbitals of other aza-macrocycles were
also calculated with comparative purposes. Fig. 11 shows
the atomic contribution to the HOMOs only.
2
C
1
3) [21–23]; generating a strong interaction M–N. Since the
potential application of the molecules reported here is as
molecular cages for metal ions, where such ions would be
eventually involved in a catalytic cycle, the strong
interactions M–N might not be necessarily convenient.
According with the obtained results, there is a compromise
between geometrical and energetic environment, and the
best situation seems to be that one where the square
Molecules 1 and 2 actually form complexes with
transition metal ions [24]. The carbonyl groups in Molecule
1 are not directly attached to nitrogen atoms; however,

10
J.A. Cruz-Morales, P. Guadarrama / Journal of Molecular Structure 779 (2005) 1–10
˚
interplanar distance of 4.8 A between first neighbor
molecules, randomly accommodated. As a result of a
theoretical study at DFT level of theory, the comparative
interaction energies of the inclusion complexes formed by
compound (9) (without dendrones) and (10) (with
Urea group
O
O
Bu
Bu
Bu
Bu
N
H
N
N
N
N
N
H
2
C
dendrones) with Ni
were obtained, showing that there
is a significant contribution in terms of energy
(K60.89 Kcal/mol) due to the functionalized dendrons
which enfold the metal ion in a molecular cage. In terms of
pre-organization, the presence of flexible dendrones results
in more organized molecular arrangements, being more
efficient the encapsulation of guests like metal ions. In
areas like catalysis, the relevance of this kind of molecular
arrangements might be the possibility to influence on the
selectivity of catalytic processes by controlling the
accessibility of active sites.
H
N
H
N
O
O
Fig. 12. Koning’s aza-macrocycle.
the effect on the HOMO orbital is rather similar to that
observed in case of model molecules A and C. To overcome
this effect, in case of Molecule 1 exists the possibility to
deprotonate the nitrogen atoms, to induce strong inter-
actions with metal ions. In the case of the model molecules
under study, the assistance of neighbor groups (like oxygen
atoms) helps to hold the metal ion, but evidently in a
weakest way. The Molecule 2, without N–H’s, forms
The experimental further work of the formation of
complexes with metal ions of the first transition series will
be aimed in a rational way, taking into account the
theoretical results obtained so far.
2
C
complexes with Ni
and, judging for the atomic
contributions to the HOMO, this complexation is highly
assisted by inductive effects. Thus, there are different effects
operating when a process of inclusion occurs.
References
[
[
1] C.J. Pedersen, J. Am. Chem. Soc. 89 (1967) 2495; C.J. Pedersen,
J. Am. Chem. Soc. 89 (1967) 7017.
An interesting paper published by Koning’s group [25],
dealing with the synthesis of a series of substituted aza-
macrocycles, showed experimentally that the presence of
urea groups, formed when the nitrogen atoms of the aza-
macrocycle are substituted, (Fig. 12) results in no
complexation of metal ions (e.g. Zn (II) and Cu(II)). Thus,
there is a thin line dividing the appropriate and non-
appropriate conditions to carry out inclusion processes by
the aza-macrocycle nitrogen atoms.
2] R.M. Izatt, J.S. Bradshaw, S.A. Nielsen, J.D. Lamb, J.J. Christensen,
D. Sen, Chem. Rev., 85271; Y. Inoue, G.W. Gokel, Cation Binding by
Macrocycles, Marcel Dekker, New York, 1990.
[
[
3] S.Aime,M. Botta,M.Fasano, E.Terreno,Chem. Soc.Rev.27(1998)19.
4] B. Konig, M. Pelka, M. Klein, I. Dix, P.G. Jones, J. Lex, J. Incl.
Phenom. Macrocycl. Chem. 37 (2000) 39.
[
5] D.J. Cram, T. Kaneda, R.C. Helgeson, S.B. Brown, C.B. Knobler,
E. Maverick, K.N. Trueblood, J. Am. Chem. Soc. 107 (1985) 3645.
[6] D. Soto-Castro, P. Guadarrama, J. Comput. Chem. 25 (2004) 1215.
[
[
[
7] G.R. Newkome, C.D. Weis, B.J. Childs, Designed Monomers Polym.
(1998) 3.
8] W.L. Jorgensen, D.S. Maxwell, J. Tirado-Rives, J. Am. Chem. Soc.
18 (1996) 11225.
9] I. Kolossv a´ ry, W.C. Guida, J. Am. Chem. Soc. 118 (1996) 5011.
Returning to Fig. 10(a), it is clear that the binding process
is assisted by dendrimeric branches, when they are present,
by means of encapsulation effects, evolving the metal ion.
Once that the theoretical evaluation has demonstrated the
complexation properties of the synthesized molecular
system, the obvious further work will be the experimental
encapsulation of metal ions of the first transition series, like
Ni, keeping in mind that the incorporation of collective
effects contributes in a favorable way in these processes.
1
1
[10] G. Chang, W.C. Guida, W.C. Still, J. Am. Chem. Soc. 111 (1989)
379; M. Saunders, K.N. Houk, Y.D. Wu, W.C. Still, M. Lipton,
4
G. Chang, W.C. Guida, J. Am. Chem. Soc. 112 (1990) 1419.
11] A.D. Becke, J. Chem. Phys. 98 (1993) 1372.
[
[
12] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1998) 785; B. Miehlich,
A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 157 (1989) 200.
13] Jaguar 5.0, Schrodinger, LLC Portland, OR, 2000.
14] M.J. Dudek, J.W. Ponder, J. Comput. Chem. 16 (1995) 791.
15] D. Beeman, J. Comput. Phys. 20 (1976) 130.
[
[
[
4
. Conclusions
[16] G. Kaupp, Top. Curr. Chem. 254 (2005) 1.
[
[
[
[
[
17] K. M u¨ ller-Dethlefs, P. Hobza, Chem. Rev. 100 (2000) 143.
18] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299.
19] E. Ruiz, D.R. Salahub, A. Vela, J. Am. Chem. Soc. 117 (1995) 1141.
20] E. Kimura, T. Koike, T. Shiota, Y. Iitaka, Inorg. Chem. 29 (1990) 4621.
21] E. Kimura, T. Koike, R. Machida, R. Nagai, M. Kodama, Inorg.
Chem. 23 (1984) 4181.
Compound (10), formed by the aza-macrocycle known
as cyclen and a functionalized dendron, was synthesized in
a convergent fashion, incorporating an aliphatic spacer to
avoid, in some extent, the sterical hindrances. N-acylation
reactions to carry out the cyclen substitution, particularly
those in solid state, resulted in very good yields since no
further purifications were necessary. The molecule (10)
was characterized by common spectroscopic techniques.
According to X-ray powder analysis, there is an average
[
22] S. Aoki, H. Kawatani, T. Goto, E. Kimura, M. Shiro, J. Am. Chem.
Soc. 123 (2001) 1123.
[
[
[
23] E. Kimura, T. Goto, S. Aoki, M. Shiro, Inorg. Chem. 41 (2002) 3239.
24] M.J. D’Aniello Jr., E.K. Barefield, J. Am. Chem. Soc. 98 (1976) 1610.
25] B. Koning, M. Pelka, M. Subat, I. Dix, P.G. Jones, Eur. J. Org. Chem.
(2001) 1943.