Steric hindrance to the solvation of melamines and consequences for non-covalent synthesis

Article abstract of DOI:10.1039/b206811g

Steric hindrance to solvation disfavors structures like 1 synanti in which the melamine exposes to the solvent faces, such as tBu-H, for whom binding to the ring nitrogen is hindered but not blocked; steric hindrance to solvation lowers the barriers to rotation in solvents which bind the triazine nitrogens, therefore these solvents display the fastest rates for assembling/disassembling processes.

Full text of DOI:10.1039/b206811g

Steric hindrance to the solvation of melamines and consequences for
non-covalent synthesis†
Ion Ghiviriga*a and Daniela C. Oniciub
a Department of Chemistry, University of Florida, Gainesville, Florida 32611-7200, USA.
E-mail: ion@chem.ufl.edu; Fax: 1 352 846 0296; Tel: 1 352 846 3001
b Esperion Therapeutics, Inc, 3621 S. Street, Ann Arbor, Michigan 48108, USA.
E-mail: doniciu@esperion.com; Fax: 1 734 332 0516; Tel: 1 734 222 1825
Received (in Columbia, MO, USA) 11th July 2002, Accepted 3rd October 2002
First published as an Advance Article on the web 18th October 2002
Steric hindrance to solvation disfavors structures like 1 syn-
anti in which the melamine exposes to the solvent faces, such
as tBu–H, for whom binding to the ring nitrogen is hindered
but not blocked; steric hindrance to solvation lowers the
barriers to rotation in solvents which bind the triazine
nitrogens, therefore these solvents display the fastest rates
for assembling/disassembling processes.
conformer of 1 and the asymmetric conformer of 2. The
equilibrium constants given in Table 1 are corrected for
symmetry, that is they would equal unity for a statistical
distribution of isomers. The effect of self-association of
melamines was eliminated by extrapolation of the equilibrium
constants measured at concentrations 15–0.5 mM to concentra-
tion 0, as described in the ESI.†
One would expect negligible reaction entropies for the
conformational exchange in the gas phase, thus negative
entropies (and negative enthalpies) indicate that the product is
more solvated than the starting material. In the range of
temperature from 240 to 0 °C, the entropic cost of solvation is
The assembling of three melamine units and three cyanuric (or
barbituric) acid units into a rosette has been used widely in non-
covalent synthesis to connect two modules through 18 hydrogen
bonds.1 The pioneering work of Whitesides et al. identified
peripheral crowding as one of the factors favoring the rosettes
over the competing motifs, linear or crinkled tapes.2 On the
other hand, simulations of the assembly process, performed in
Timmerman’s group, showed that steric effects are much less
important in determining the rosette concentration than other
factors affecting the stability of the rosette.3 We will show here
that steric hindrance to solvation could destabilize the rosette in
solvents like chloroform unless bulky substituents on melamine
completely block the triazine nitrogen.
The stability of the melamine–imide complexes has been
known to depend critically on the solvent; Whitesides et al.
have even suggested that these complexes could be used as
probes of solvent–solute interactions.2 Reinhoudt’s group
reported that solvation affects the stability of the calix[4]arene
double rosettes4 and that it destabilizes specific isomers.5
Wurthner et al. have shown that the constants for the association
of melamines with imides depend on specific solvent–solute
interactions, which are more important than the solvent polarity
alone.6
Fig. 1 Conformers of melamines 1–3.
Table 1 Thermodynamic parameters for the conformational equilibria of
melamines 1–3 as a function of the solvent
Variable temperature proton spectra of three melamines
(N,N-dimethylamino-bis(N-tert-butylamino)-s-triazine
(1),
Solvent
tris(N-tert-butylamino)-s-triazine (2) and tris(N-methylamino)-
s-triazine (3)) were recorded in four deuterated solvents
(chloroform, methylene chloride, dimethylformamide and ace-
tone). In all solvents, at low temperature melamine 1 displayed
four non-equivalent sites corresponding to three conformers
while melamines 2 and 3 displayed four non-equivalent sites
corresponding to two conformers (Fig. 1). These conformers
arise from the restricted rotation about the amino–triazine
bond.7 Deviation from the statistical distribution (which
requires signals of equal intensity for the isochronous sites in all
of these spectra) allowed the identification of the signals for the
asymmetric and propeller conformers for melamines 2 and 3
and for the syn–anti conformer of melamine 1. The assignment
of the signals for the syn–syn and the anti–anti conformers of
melamine 1 was based on the NOEs observed in the ROESY
spectrum. To our surprise, in CDCl3 at 260 °C, the conforma-
tional equilibrium was dominated by what intuitively would be
the most sterically hindered conformers, namely the anti–anti
Reaction
CDCl3 CD2Cl2 Acetone-d6 DMF-d6
(i) 1 syn–anti " 1
syn–syn
DHa
DSb
Kc
0.74
5.38
2.99
0.44
2.92
1.74
0.63
2.75
1.04
0.43
1.93
1.10
(ii) 1 syn–anti " 1
anti–anti
DH 20.84
20.47
21.46
1.35
0.10
20.13
0.74
0.37
1.04
0.76
DS
K
21.62
2.75
(iii) 2 asymmetric " 2
propeller
DH
DS
K
2.51
6.94
0.13
1.44
4.93
0.50
20.57
-1.93
1.18
20.40
21.37
1.13
(vi) 3 asymmetric " 3
propeller
DH 20.55
20.38
20.62
1.62
20.54
20.99
1.90
20.52
20.96
1.90
DS
K
21.04
1.98
a Enthalpy of reaction in kcal mol21; standard errors are given the ESI,† a
typical value being 0.05 kcal mol21 b Entropy of reaction in cal mol21 K21
.
,
† Electronic supplementary information (ESI) available: relevant properties
of the solvents, preparation of melamines 1–3, identification of the
conformers, variable temperature proton spectra, measurement of the
equilibrium constants and of the barriers to rotation. See http://
www.rsc.org/suppdata/cc/b2/b206811g/
corrected for symmetry (does not include the entropy of symmetry);
standard errors are given in the ESI, a typical value being 0.2 cal mol21
K21 c Equilibrium constant measured at 232.1 K, corrected for symmetry,
.
i.e. K = 2[syn–syn]/[syn–anti], K = 3[propeller]/[asymmetric].
2718
CHEM. COMMUN., 2002, 2718–2719
This journal is © The Royal Society of Chemistry 2002

View Article Online
comparable to the enthalpic gain, and their balance determines
the position of the equilibrium with the more solvated
conformers being favored at lower temperatures.
Data in Table 1 display a significant difference between the
hydrogen bond donor solvents as CDCl3 and CD2Cl2 and the
hydrogen bond acceptor ones, such as acetone-d6 and DMF-d6,
particularly in the case of melamine 2. Relevant properties of
the protonated solvents are given in Table 2S of ESI.†
The polarity of the solvent is not the dominant factor: the
more polar syn–syn conformer of 1 is more favored in
chloroform than in the more polar methylene chloride. Specific
interactions of these solvents with melamines might include
weak coordination between the ring nitrogen and the chlorine of
the solvent, and/or hydrogen bonding. Carbon tetrachloride is
known to interact with pyridine derivatives, and the difference
in heat of transfer from hexane to CCL4 between pyridine
Barriers to rotation about the three non-equivalent bonds in
the syn–anti conformer of 1 and the asymmetric conformer of 2
were determined by lineshape simulation in the temperature
range of 250 to 0 °C, and are reported in Table 2. Barriers in
chloroform are consistently lower than in any other solvents,
indicative of a more favorable difference between the solvation
of the transition state and the solvation of the ground state. As
one of the alkylamino groups rotates out of plane in the
transition state, the adjacent triazine nitrogens become more
exposed to the hydrogen bond donor solvent. Of the three
transition states corresponding to rotations about bonds a–c,
transition state a displays the lowest steric hindrance to
solvation, while transition state c displays the highest, and
indeed both melamines 1 and 2 present the lowest barrier for the
rotation about bond a and the highest for rotation about bond c.
Lower barriers to rotation explain why a mixture of assemblies
composed of three calix[4]arene dimelamines and six barbitu-
rates/cyanurates equilibrates in seconds at 250 °C in CDCl3
and in hours in toluene at 25 °C.13
In conclusion, we have demonstrated that steric hindrance to
solvation disfavors structures in which melamines expose to the
hydrogen bond donor solvent a face for which solvation is
hindered but not blocked. Steric hindrance to solvation can
explain why the stability of the rosettes decreases as the size of
the substituents on the melamines and isocyanurates is re-
duced.14
Steric hindrance to solvation also lowers the barriers to
rotation in hydrogen bond donor solvents, which therefore
display the fastest rates for assembling/disassembling proc-
esses. Because these rates set a limit to the size of the regular
networks that can be grown by non-covalent synthesis,15
hydrogen bond donor solvents may be used instead of
assembling catalysts when large regular networks, suitable for
nanodevices, are desired.
(21.71 kcal mol21) and g-collidine (21.27 kcal mol21
)
represents the steric interaction with the solvent.8 Of the three
potential hydrogen bond acceptor sites in melamines: the ring
nitrogen, the amine nitrogen and the p electron cloud, the
former has the higher hydrogen bond basicity, as revealed by a
comparison of the SbH values for pyrimidine (0.52), aniline
2
(0.41) and benzene (0.14).9 The conformational preferences of
melamines in hydrogen bond donor solvents can be explained
by steric hindrance to solvation imposed by the substituents on
the same face with the ring nitrogen. The concept of steric
hindrance of solvation reflects not an enthalpic process, but an
ergonic one.10 In some cases the process is entirely entropic
(e.g. 2,6-di-tert-butylpyridine is a weaker base than the
2,4-isomer because hydration of the conjugated acid restricts
the rotation of the tert-butyl groups11), in other cases the
enthalpic term dominates.12
Writing the four reactions in Table 1 in terms of faces
exposed to solvation,
(i) H–Me + tBu–H ? H–H + tBu–Me
(ii) tBu–Me + tBu–H ? Me–H + tBu–tBu
(iii) 2 tBu–H ? H–H + tBu–tBu
(iv) 2 Me–H ? H–H + Me–Me
Notes and references
it becomes apparent that the disfavored sides of the reactions
(reactants for i–iii and products for iv) all have a face for which
solvation is hindered but not blocked, like tBu–H or Me–Me.
Restriction of the motion of the solvent upon binding to such a
face makes the entropic price of solvation higher than the
enthalpic gain.
1 (a) J. L. Atwood and J. M. Lehn, Comprehensive supramolecular
chemistry, Pergamon, New York, 1996; (b) J. W. Steed and J. L.
Atwood, Supramolecular chemistry, Wiley, New York, 2000; (c) D. C.
Sherrington and K. A. Taskinen, Chem. Soc. Rev., 2001, 30, 83.
2 G. M. Whitesides, E. E. Simanek, J. P. Mathias, C. T. Seto, D. N. Chin,
M. Mammen and D. M. Gordon, Acc. Chem. Res., 1995, 28, 37.
3 A. G. Bielejewska, C. E. Marjo, L. J. Prins, P. Timmerman, F. de Jong
and D. N. Reinhoudt, J. Am. Chem. Soc., 2001, 123, 7518.
4 P. Timmerman and L. J. Prins, Eur. J. Org. Chem., 2001, 3191.
5 L. J. Prins, K. A. Jolliffe, R. Hulst, P. Timmerman and D. N. Reinhoudt,
J. Am. Chem. Soc., 2000, 122, 3617.
6 F. Wurthner, C. Thalacker, A. Sautter, W. Schartl, W. Ibach and O.
Hollricher, Chem.-Eur. J., 2000, 6, 3871.
7 (a) A. R. Katritzky, D. C. Oniciu, I. Ghiviriga and R. A. Barcock, J.
Chem. Soc., Perkin Trans. 2, 1995, 785; (b) A. R. Katritzky, I.
Ghiviriga, P. J. Steel and D. C. Oniciu, J. Chem. Soc., Perkin Trans. 2,
1996, 443.
In hydrogen bond acceptor solvents, the NH group of
melamines acts as a hydrogen bond donor (for amides SaH
=
2
0.509). In this case, hydrogen binding to the solvent is not
significantly hampered by the substituents on melamine, the NH
being on an ‘outer’ position. For melamine 1, dipole–dipole
interactions with acetone and DMF favor the syn–syn conformer
(the most polar) and disfavor the anti–anti one.
Table 2 Barriers to rotation at 232.1 K (DG≠ in kcal mol21) in melamines
1 and 2 as a function of the solvent
8 A. D. Sherry and K. F. Purcell, J. Amer. Chem. Soc., 1970, 92, 6386.
9 M. H. Abraham, Chem. Soc. Rev., 1993, 22, 73.
Solvent
10 B. Wilson, R. Georgiadis and J. E. Bartmess, J. Am. Chem. Soc., 1991,
113, 1762.
Rotation about
CDCl3
CD2Cl2
Acetone-d6
DMF-d6
11 H. P. Hopkins, D. V. Jahagirdar, P. S. Moulik, D. H. Aue, H. M. Webb,
W. R. Davidson and M. D. Pedley, J. Am. Chem. Soc., 1984, 106,
4341.
12 A. Bagno, R. L. Boso, N. Ferrari and G. Scorrano, Eur. J. Org. Chem.,
1999, 1507 and references cited therein.
13 M. C. Calama, R. Hulst, R. Fokkens, N. M. M. Nibbering, P.
Timmerman and D. N. Reinhoudt, Chem. Commun., 1998, 1021.
14 J. P. Mathias, E. E. Simanek and G. M. Whitesides, J. Am. Chem. Soc.,
1994, 116, 4326–4340.
15 L. J. Prins, D. N. Reinhoudt and P. Timmerman, Angew. Chem., Int. Ed.,
2001, 40, 2383.
Bond a in 1a
Bond b in 1
Bond c in 1
Bond a in 2
Bond b in 2
Bond c in 2
12.66
12.30
13.87
12.06
—
13.76
13.17
13.21
14.27
13.08
13.87
15.38
13.63
14.11
14.97
13.57
13.89
14.90
14.15
14.22
14.91
14.15
14.14
15.89
b
a The bonds are identified in Fig. 1. b This barrier could not be reliably
measured because in CDCl3 the concentration of the propeller conformer
was too small.
CHEM. COMMUN., 2002, 2718–2719
2719