Nanodiamonds in sugar rings: An experimental and theoretical investigation of cyclodextrin-nanodiamond inclusion complexes

Article abstract of DOI:10.1039/c2ob06915f

We report on the noncovalent interactions of nanodiamond carboxylic acids derived from adamantane, diamantane, and triamantane with β- and γ-cyclodextrins. The water solubility of the nanodiamonds was increased by attaching an aromatic dicarboxylic acid via peptide coupling. Isothermal titration calorimetry experiments were performed to determine the thermodynamic parameters (K_a, ΔH, ΔG and ΔS) for the host-guest inclusion. The stoichiometry of the complexes is invariably 1:1. It was found that K_a, ΔG and ΔH of inclusion increase for larger nanodiamonds. ΔS is generally positive, in particular for the largest nanodiamonds. β-Cyclodextrin binds all nanodiamonds, γ-cyclodextrin clearly prefers the most bulky nanodiamonds. The interaction of 9-triamantane carboxylic acid shows one of the strongest complexation constants towards γ-cyclodextrin ever reported, K_a = 5.0 × 10⁵ M^-1. In order to gain some insight into the possible structural basis of these inclusion complexes we performed density functional calculations at the B97-D3/def2-TZVPP level of theory.

Full text of DOI:10.1039/c2ob06915f

View Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Volume 10 | Number 23 | 21 June 2012 | Pages 4473–4628
ISSN 1477-0520
PAPER
Bart Jan Ravoo et al.
Nanodiamonds in sugar rings: an experimental and theoretical
investigation of cyclodextrin–nanodiamond inclusion complexes

Dynamic Article Links
Organic &
View Online
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 4524
www.rsc.org/obc
PAPER
Nanodiamonds in sugar rings: an experimental and theoretical investigation
of cyclodextrin–nanodiamond inclusion complexes†
Jens Voskuhl,a Mark Waller,*a Sateesh Bandaru,a Boryslav A. Tkachenko,b Carlo Fregonese,a
Birgit Wibbeling,a Peter R. Schreiner*b and Bart Jan Ravoo*a
Received 13th November 2011, Accepted 1st March 2012
DOI: 10.1039/c2ob06915f
We report on the noncovalent interactions of nanodiamond carboxylic acids derived from adamantane,
diamantane, and triamantane with β- and γ-cyclodextrins. The water solubility of the nanodiamonds was
increased by attaching an aromatic dicarboxylic acid via peptide coupling. Isothermal titration calorimetry
experiments were performed to determine the thermodynamic parameters (Ka, ΔH, ΔG and ΔS) for the
host–guest inclusion. The stoichiometry of the complexes is invariably 1 : 1. It was found that Ka, ΔG and
ΔH of inclusion increase for larger nanodiamonds. ΔS is generally positive, in particular for the largest
nanodiamonds. β-Cyclodextrin binds all nanodiamonds, γ-cyclodextrin clearly prefers the most bulky
nanodiamonds. The interaction of 9-triamantane carboxylic acid shows one of the strongest complexation
constants towards γ-cyclodextrin ever reported, Ka = 5.0 × 105 M−1. In order to gain some insight into the
possible structural basis of these inclusion complexes we performed density functional calculations at the
B97-D3/def2-TZVPP level of theory.
spherical shape, which nicely matches the cavity of β-CD.8
Introduction
Typical binding constants are around Ka = 104 M−1. There are
Inclusion complexes of cyclodextrins (CDs) are among the most
studied host–guest pairs in supramolecular chemistry. The first
examples of these non-covalent complexes were described by
Pringsheim and by Freudenberg.1,2 Nowadays, CDs are widely
available via a cheap enzyme-catalyzed synthesis from starch.
CDs have become increasingly important in a range of pharma-
ceutical and cosmetic applications, and they have also become
valuable components of HPLC columns to separate enantiomers
by stereoselective host–guest complexation.3 In addition, CDs
are key components in a variety of supramolecular materials,4
including surfaces,5 hydrogels,6 and vesicles.7
The stability of the host–guest complexes of CDs is primarily
determined by the hydrophobicity and the shape of the guest
molecule and driven by the release of water molecules from the
cavity of the CD into the bulk water.3 Adamantane and ferrocene
derivatives bind strongly to β-cyclodextrin (β-CD) due to their
few examples of 1 : 1 host–guest complexes of CDs that show
higher affinity. In most of these cases, secondary interactions
such as hydrogen bonding or electrostatic attraction lead to
binding constants that are about one magnitude higher than those
reported for adamantane and ferrocene derivatives.9 Alterna-
tively, large hydrophobic molecules such as steroids can be
accommodated in CD dimers. In that case, the affinity can be as
high as Ka = 106 M−1, but the host–guest stoichiometry is effec-
tively 2 : 1.10
Inclusion complexes of γ-cyclodextrin (γ-CD) have been
much less investigated. γ-CD is more flexible than β-CD. In sol-
ution, the cavity of γ-CD partially collapses and hence the ten-
dency to replace internal hydration water with a suitable
hydrophobic guest is significantly reduced. However, bulky mol-
ecules such as substituted azobenzene derivatives form strong
inclusion complexes with γ-CD (Ka > 104 M−1).11
Diamondoids are nanometer-sized hydrocarbons (i.e., nano-
diamonds (NDs)) resembling the diamond lattice terminated by
hydrogen atoms.12 While adamantane (C10H16)13 and diaman-
tane (C14H20)14 have long been synthetically available and there-
fore comprehensively studied, recent discovery of triamantane
(C18H24)15 and various higher diamondoids in oil and deep
natural gas condensates16 promoted a surge of investigations
regarding their properties and applications. Diamondoids can be
selectively functionalized at different positions, leading to
various derivatives with unique properties that already found
applications in diverse fields ranging from catalysis,17
aOrganisch Chemisches Institut, Westfälische Wilhelms-Universität
Münster, Corrensstrasse 40, 48149 Münster, Germany. E-mail:
b.j.ravoo@uni-muenster.de, m.waller@uni-muenster.de
bInstitut für Organische Chemie, Justus-Liebig Universität,
Heinrich-Buff-Ring 58, 35392 Giessen, Germany.
E-mail: prs@org.chemie.uni-giessen.de
†Electronic supplementary information (ESI) available: Synthetic pro-
cedures, NMR spectra, details of ITC measurements, coordinates of cal-
culated complexes and a selection of geometries and energies are
provided in the electronic supplementary information. CCDC
854337–854339. For ESI and crystallographic data in CIF or other elec-
tronic format see DOI: 10.1039/c2ob06915f
4524 | Org. Biomol. Chem., 2012, 10, 4524–4530
This journal is © The Royal Society of Chemistry 2012

View Online
Lipkowitz in 1998.22 The use of empirical potentials to model
the CD was primarily due to the limited computing resources at
that time. Since then, advances in computing hardware and com-
putational techniques have opened up a new vista of accuracy for
modelling CD complexes. DFT is often referred to as the “work-
horse” of quantum chemistry, and due to its accuracy and favor-
able scaling properties it has become widely used to study CD
complexes. For example, Snor et al. published a series of papers
on the structure of anhydrous CDs,23–25 including complexes
with spironolactone26 and meloxicam.27 Furthermore, Rarati
et al. investigated decyltrimethylammonium bromide (DTAB)
and tetradecyltrimethylammonium bromide (TTAB) inclusion
complexes with α-CD and β-CD using semi-empirical and DFT
computations.28 The binding affinities of resorcinol and a series
of drugs (flubiprofen, nabumetone and naproxen) with CDs
showed good agreement between classical simulations and
experimental values.29
Fig. 1 Substituted nanodiamond (left) and γ-cyclodextrin (right).
It should be noted that in the gas phase the binding of an ada-
mantane moiety will be predominately a result of dispersion type
interactions while in solution hydrophobic interactions will dom-
inate. The inability of most available functionals to describe van
der Waals type interactions is a topic of current research. Further-
more, a cursory consideration of structure of the substituted NDs
shows that the ligands may form an additional hydrogen bonding
interaction via the attached isophthalic acid. Therefore, a model-
ling method that is capable of delivering an accurate and
balanced description of dispersion bound and hydrogen bound
complexes is required. In this study, we will use B97 density
functional that is augmented with a third generation dispersion
correction, denoted D3,30 which has been shown recently to
perform well in benchmark studies.31
Scheme 1 Synthesis and overview of guest and host molecules.
Reagents and conditions: (I) Boc2O, NaOH, H2O, t-BuOH, 12 h, rt,
99%, (II) EDCI, Oxyma Pure, NMM, DMF, 5-aminoisophthalic acid–
dimethylester, 18 h, rt, 52%, (III) TFA, CH2Cl2, 4 h, rt, 99%, (IV)
EDCI, Oxyma Pure, NMM, DMF, RCO2H, 18 h, rt, 41–55%, (V) THF,
MeOH, THF, Dowex HCR 20, 4 h, rt, 88–95%.
Methods and results
pharmaceuticals,18 polymer science19 to surface modification20
and nanoelectronics.21
Synthesis
The carboxylic acids of diamantane and triamantane were syn-
thesized as described before.32 The synthesis of the substituted
nanodiamonds starts from glycine 1 which was first amine pro-
tected followed by peptide coupling using EDCI, NMM and
Oxyma Pure (ethyl 2-cyano-2-(hydroxyimino)acetate) to 5-ami-
noisophthalic acid methylester yielding 2. After removal of the
protection group using TFA and coupling of the free amine to
the diamondoid carboxylic acids 7a–10a the protected guest
molecules 7b–10b were obtained. The final step was a saponifi-
cation with sodium hydroxide followed by neutralization with
ion exchange resin (HWCR-20) yielding the desired highly
water soluble nanodiamonds 7c–10c.
The analytical data of 7b–10c are consistent with their mol-
ecular structure. Details are available as electronic supplementary
information (ESI†). It was possible to obtain X-ray structures of
compounds 7b, 8b and 10b, which confirm the diamondoid
structures (Fig. 2). It is evident that the compounds contain the
compact hydrophobic NDs linked to 3 via a peptide bond.
On the basis of their hydrophobicity and symmetry, it would
be expected that NDs are excellent guests for β-CD and γ-CD
(Fig. 1). To the best of our knowledge, the host–guest inter-
actions of CDs and NDs have not been investigated before. The
aim of this study is to characterize the inclusion complexes of
the lower NDs with β-CD and γ-CD. To this end, we determined
the thermodynamic parameters (association constant Ka,
enthalpy ΔH, free energy ΔG and entropy ΔS) for the inclusion
complexation of 1-adamantane, 1-diamantane, 4-diamantane and
9-triamantane carboxylic acids by isothermal titration calorime-
try (ITC). We also prepared a range of hydrophilic derivatives of
the NDs by peptide coupling with 5-aminoisophthalic acid
(Scheme 1). The inclusion of these NDs was also characterized
by ITC. It should be emphasized that both the ND carboxylic
acids and their isophthalic acid derivatives may show secondary
interaction with CDs through hydrogen bonding, resulting in
additional stabilization of the inclusion complex. However, we
do not expect inclusion of the hydrophilic aromatic substituent
in the CD cavity.
In order to gain some insight into the possible structural basis
of these inclusion complexes we also performed density func-
tional theory (DFT) computations. An exhaustive review on the
application of force fields for studying CDs was published by
Isothermal titration calorimetry (ITC)
With guest molecules 7b–10c in stock it was possible to investi-
gate their ability of forming inclusion complexes with β- and
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 4524–4530 | 4525

View Online
γ-cyclodextrins. With exception of 10c, the NDs were titrated to
the CDs. Table 1 presents an overview of the ITC data; details
are available as ESI.† All NDs interact with β-CD, but only the
bulkiest ones (8c, 10a, 10c) interact with γ-CD. The stoichi-
ometry is 1 : 1 and the interaction is endothermic in each case,
with increasingly negative ΔH for the larger NDs. No significant
enthalpy change was detected for the titrations of 7a, 7c, 8a, 9a
and 9c to γ-CD. Compared to the complexes of β-CD, the com-
plexes of γ-CD have a much smaller ΔH and much larger ΔS.
overview of energetic and descriptors is provided in Table 2.37
Subsequently, single point energies were computed at the B97-
D330/def2-TZVPP for binding and deformation energies. The
degree of penetration for the ND·CD complexes is presented as
the distance, Rp, defined as the distance between the centre-of-
mass (COM) of the CD and the COM of the adamantane subunit
that is the closest to the CD COM. To quantify the tilt angle of
the guest molecule with respect to the CD, we present the angle
(θ) between CDCOM–NDCOM′–Clinker in Table 2. The system
specific density functional validation results are given in the
ESI.† We have validated our functional selection from amongst
four popular gradient corrected functionals (GGA). They show a
small functional dependence across the small model systems of
our ND·x-CD systems.
Solvation effects were modelled using a 17 Å sphere of
TIP3P38 water molecules centred on the COM of the ND
(∼1800 atoms). A QM/MM strategy was employed whereby the
gas-phase optimized NDs 7b–10c and CDs were both assigned
to the QM layer and the solvation sphere was assigned to the
MM layer (Charmm C35ba39). The Chemshell v3.3.2 package40
was employed for the QM/MM energy minimization at the B97-
D3/def-SV(P):CHARMM level of theory. A root mean square
distortion (RMSD) overlay was performed within the VMD visu-
alization package.41
Computational details
Hereafter, we use the nomenclature ND·x-CD, where ND indi-
cates the guest, and x can be either β or γ. Therefore, we have
four possible ND·β-CD and four possible ND·γ-CD. We note
that the adamantane typically binds end-on,33 in contrast to the
prototypical threaded binding motif of rotaxane. Therefore we
performed a series of calculations where the adamantane entered
the larger secondary face (S), and a series where the narrower
primary face (P) is penetrated. Gas-phase structural optimizations
were undertaken with the B97-D334 method combined with the
def-SV(P)35 basis set using the TURBOMOLE 6.336 suite. An
Discussion
The diameter of the CD cavity increases as the number of α-D-
glucopyranoside units are increased from six (α-CD) to seven
(β-CD) to eight (γ-CD). When the rim of the CD is penetrated
by the adamantane, the CD can increase its interstitial diameter
to accommodate the ND ligand. We note that there is a larger
degree of distortion, i.e., higher RMSD, for the DFT optimized
γ-CDs models than for the β-CD counterparts (Table 2). This
distortion can be traced back to the increased ellipticity of the
γ-CD that is required to envelope the guest molecules, thereby
maximizing close contacts (Fig. 3).
Fig. 2 X-ray structure of 7b, 8b and 10b. Thermal ellipsoids indicate
50% probability.
Table 1 Thermodynamic parameters of the host–guest complexation between nanodiamonds (NDs) and cyclodextrins (CDs) obtained by isothermal
titration calorimetry
Nr
ND
Guest
Host
[Guest] (mM)
[Host] (mM)
ΔH (kJ mol−1
)
ΔG (kJ mol−1
)
ΔS (J kmol−1
)
Ka (M−1) 104
A
B
C
D
E
1-Adamantane
7a
7a
7c
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
10.0
10.0
10.0
10.0
1.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
1.0
1.0
1.0
1.0
1.0
0.1
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.1
0.5
10.0
10.0
−17.8
—
−24.8
—
23.6
—
2.2
—
a
a
a
a
−17.3
—
−24.6
—
24.4
—
2.0
—
a
a
a
a
7c
1-Diamantane
4-Diamantane
9-Triamantane
8a
8a
8c
−21.9
—
−29.1
—
24.1
—
12.8
—
a
a
a
a
F
G
H
I
J
K
L
M
N
Ob
Pb
−25.1
−4.9
−26.3
−26.2
−30.3
4.2
4.1
3.9
8c
71.3
25.3
9a
9a
9c
−22.8
—
20.7
—
a
a
a
a
—
—
−20.1
—
−31.1
—
37.0
—
28.7
—
a
a
a
a
9c
10a
10a
10c
10c
−34.9
−8.7
−35.7
−11.3
−29.5
−32.5
−24.8
−31.9
−18.4
80.0
−36.3
69.1
14.7
50.0
2.3
5.0
1.0
1.0
39.1
a No detectable signals were obtained during ITC measurements. b Due to micelle formation, the titration was performed in the inverse mode.
4526 | Org. Biomol. Chem., 2012, 10, 4524–4530
This journal is © The Royal Society of Chemistry 2012

View Online
Table 2 The gas-phase energetics of nanodiamond (ND)·cyclodextrin (CD) complexes at the B97-D3/def2-TZVPP level of theory, and the
structural descriptors of ND·CD complexes that were optimized at the B97-D/def-SV(P) level of theory
BE
DECD
DEND
RMSDCD RMSDND
Rp
(Å)
Pept
(ψ)
Tilt
(θ)
Nr
ND
Guest
Host
(kJ mol−1
)
(kJ mol−1
)
(kJ mol−1
)
(Å)
(Å)
A
B
C
D
E
F
G
H
I
1-Adamantane
7a
7a
7c
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
β-CD
γ-CD
−77
33
20
85
−1
41
−23
48
10
75
7
79
−1
21
−31
15
−24
6
0.52
1.67
0.65
0.91
0.56
0.83
0.52
0.89
1.49
1.51
1.29
1.15
0.48
1.18
0.46
1.42
0.69
0.19
2.88
2.72
0.18
2.62
1.59
4.35
0.24
2.65
1.65
3.58
0.51
2.59
2.92
4.39
1.2
1.2
0.8
0.3
1.8
1.2
2.6
1.5
1.7
2.1
0.9
0.9
1.5
1.2
1.4
0.6
—
—
80
27
—
—
−176
−37
—
−
68
76
—
—
69
61
124
23
−134
−135
−117
−84
38
40
17
9
13
2
56
−2
19
13
68
19
11
12
17
141
136
126
110
141
156
138
14
154
36
105
144
108
142
7c
1-Diamantane
4-Diamantane
9-Triamantane
8a
8a
8c
−142
−97
8c
−183
−32
9a
9a
9c
J
−161
−82
K
L
M
N
O
P
9c
−233
−121
−172
−125
−250
10a
10a
10c
10c
Note: only the complexes where the secondary side of the CD ring was penetrated by the ligands are shown in above. Table S1 (ESI†) contains the
corresponding data for the primary complex.
narrow P side with the entire bulky ND, with the acid group(s)
anchoring the insertion process; this obviously results in a large
structural deformation of the CD. Both of these insertion scen-
arios would be associated with a substantial kinetic barrier to
insertion. For these reasons, we only show the computations
with the ND bound to the S side of the CD in Table 2. All of the
computational results presented in Table 2 are gas phase values.
To ensure our geometries were still reasonable, given that the
ITC experiments are performed in an aqueous environment, we
performed an explicit solvation computation. A hybrid QM/MM
Fig. 3 (a) The overlaid β-CDs taken from the optimized complexes
method wherein the complex was assigned to the QM layer and
(NDs removed for clarity), of note is the circular cavity. (b) The overlaid
the solvent was placed in the MM layer. The smallest substituted
γ-CD taken from the optimized complexes (ND removed for clarity).
ND was chosen (7c) in conjunction with γ-CD, which has a
larger cavity than β-CD. Although there was no observed inter-
the 7c·γ-CD (purple).
Note the elliptical distortion resulting from the bound ND, especially for
action via ITC, we believe that looking at this particular complex
is instructive to ascertain the effects of explicit solvation. The
1-Adamantane
optimized conformation of the complex is not significantly
altered (RMSD of only 0.4 Å) by the inclusion of ∼1800 explicit
solvent water molecules (Fig. 4). The small changes with the
inclusion of solvent are in agreement with previous finding by
Miyamoto et al.43 for coenzyme Q10 complexed with γ-CD.
However, we do note that the effect of explicit solvation was
found to be critical when Jamie et al. investigated CDs with MD
simulations, and care should therefore be taken when modelling
in the gas-phase.
The association constant Ka for the interaction of 7a and 7c with
β-CD was found to be around 2.0 × 104 M−1. No significant ΔH
was determined for the titration of 7a and 7c with γ-CD.42 It is
intriguing that the isophthalic acid substituent does not cause a
decrease or increase in binding constant or have any significant
influence on other thermodynamic parameters. This similarity is
not borne out in the DFT optimized structures where a nearly
two-fold increase in binding energy is observed (BE, Table 2).
This large difference is associated with a dramatic decrease in
the distance (Rp) between the shallow bound nonsubstituted
7a·β-CD complex and the deeply bound substituted 7c·β-CD
complex. There is a rather small difference in BE between the
ND that is complexed with the S side of the CD in Table 2, com-
pared to binding at the P side (see Table S1 in the ESI†). There-
fore, although thermodynamically the preference is not that large
in the gas phase, binding via the narrow P side of the CD is far
less likely than via the wider S side. Either the ND has to push
the entire hydrophilic carboxylic acid group(s) through the
centre of the hydrophobic CD core to penetrate through to the P
side. Alternatively, one could imagine the ND penetrating the
1-Diamantane
The interaction of ND 8a with β-CD shows a remarkably high
affinity of Ka = 1.2 × 105 M−1. The oblate shape and larger
volume of the ND is somewhat complimentary to the cavity of
the β-CD, as evidenced by an extremely small RDSDND of
0.18 Å (Table 2). Furthermore, in the DFT gas-phase optimiz-
ations of 8a·β-CD, the complex is ∼7 kJ mol−1 more stable than
the respective 7a·β-CD complex and these two factors combine
to explain the enhanced binding affinity of 1-diamantane in com-
parison with the 1-adamantane carboxylic acid 7a. Interestingly
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 4524–4530 | 4527

View Online
Fig. 4 Explicitly solvated QM/MM model of 7c·γ-CD.
Fig. 5 cis–trans Isomerization of the glycine peptide bond for 8c. cis
(red) and trans (blue) Isomers are shown.
Fig. 6 Isothermal titration calorimetry of 9a and 9c with β-cyclodex-
trin. For concentrations and thermodynamic parameters see Table 1.
the binding affinity decreases significantly upon substitution of
the acid functionality of the ND (Ka = 4.1 × 104 M−1 for 8c). It
is possible that the hydrophilic and bulky isophthalic acid sup-
presses the complete insertion of the guest molecule in the
cavity. Nevertheless, the interaction of 8c and β-CD is substan-
tial. Two possible scenarios can explain this. The first possibility
is that the isophthalic acid sticks up (perpendicularly) out of the
CD cavity into solvent bulk and does not have any influence.
Alternatively, the carboxylic acid of 8a can itself form a hydro-
gen bond with the CD rim that is effectively isoenergetic with
the hydrogen bonding ability of the carboxylic acid of the iso-
phthalic acid group of 8c. The DFT optimized structures (and to
some extent energetics) for the gas-phase 8a·β-CD and 8c·β-CD
complexes certainly suggests that the second scenario occurs.
The DFT optimizations show that the NDs lie inside the bowl
shaped interior of the CDs as observed by the tilt angle (θ),
which is significantly distorted away from linearity, and this tilt
is more acute than observed for the isophthalic ND derivative
8c.
Fig. 7 Nanodiamond 9a occupying the P opening of β-cyclodextrin.
4-Diamantane
The 4-diamantane carboxylic acid 9a reveals one of the strongest
binding constants towards β-CD ever reported for a 1 : 1
complex: Ka = 2.1 × 105 M−1. Representative ITC results are
shown in Fig. 6. Based on the tubular, prolate structure of the
ND, the hydrophobic guest can penetrate the β-CD cavity and
effectively occupy this void (Fig. 7). If the ND ligand has the
hydrophilic isophthalic acid substituent (9c) the binding is even
stronger (Ka = 2.8 × 105 M−1). Based on the small differences in
enthalpy, see Table 1, we speculate that this enhanced binding
affinity is due to the 9c·ND sitting more deeply in the CD cavity.
Indeed the Rp obtained from modeling is rather low at 0.9 Å,
which indicates significant penetration, and therefore, ipso facto,
a significant number of water molecules are displaced from the
CD cavity into the bulk.
As expected NDs 9a and 9c show no significant complexation
with γ-CD due to their reduced girth compared to 1-diamantane,
i.e., they possess a cross section more comparable to 1-adaman-
tane. This is clearly reflected in the gas phase models where the
γ-CD secondary rim must dramatically distort away from its
minima (circular), to form an elliptical shaped cavity in order to
Furthermore, in spite of a substantial affinity (Ka = 3.9 × 104
M−1) for the complexation of 8c with γ-CD, the experimental
ΔH value is surprisingly small at only −4.9 kJ mol−1. A large
positive ΔS value of +71.3 J K−1 mol−1 observed with ITC
suggests that the disorder in the system increases and the com-
plexation is no longer primarily driven by enthalpy but by
entropy instead. The release of water molecules from the CD
cavity as well as a release of hydration water from the guest mol-
ecules indicates a significant penetration of the ND into the CD
cavity. Indeed the DFT gas-phase optimized geometry confirms
this (Rp = 1.5 Å) showing the long flexible linker forming a sta-
bilizing hydrogen bond to the CD rim. The DFT optimized
models of 8c suggest that there is rotation about the peptide
bond (Fig. 5). This is in stark contrast to 8a, which shows no
detectable complexation with γ-CD.
4528 | Org. Biomol. Chem., 2012, 10, 4524–4530
This journal is © The Royal Society of Chemistry 2012

View Online
form a strong complex with γ-CD and should also give a strong
complex with β-CD. The larger volume of ND 10a is in good
agreement with the space-filling model of the hydrophobic γ-CD
cavity. Fig. 9 shows ND 10a nicely occupies the γ-CD bowl-
shaped cavity, with the carboxylic acid clearly pointing towards
the secondary rim forming strong hydrogen bonds.
The ITC results reveal that the interaction of the triamantanes
with γ-CD is both enthalpically and entropically highly favour-
able. The highly positive entropy of inclusion confirms a deep
penetration of the NDs into the γ-CD cavity, with concomitant
release of hydration water. On the other hand, the interaction of
the triamantanes with β-CD is no longer entropically favourable,
which indicates that the hydrophobic part does not penetrate the
cavity completely. The substituent on the triamantane cage also
influences the complexation. The Ka value for the β-CD com-
plexation decreases by a factor of six for ND 10c in comparison
with ND 10a. However, the isophthalic acid substituent on ND
10c does not lead to a substantial decrease of the binding con-
stant (3.9 × 105 M−1) for the complex with γ-CD.
Conclusion
Herein we have investigated a new class of host–guest com-
plexes using a combined experimental and theoretical approach.
ITC revealed very strong interactions between the 4-diamantanes
and β-CD. Triamantane shows the highest affinity towards com-
plexation with γ-CD ever reported. Computational studies pro-
vided some insights into the structure of the inclusion
complexes. Proposed structures are based on a careful inspection
of the gas phase as well as explicitly solvated models of the
ND·CD complexes. The adamantane moiety of the ND occupies
the CD cavity, and upon rotation of the glycine linker, secondary
hydrogen bonding can occur between the carboxyl group and the
CD rim. Our findings consistently indicate that an optimal fit of
host and guest results in the most stable inclusion complexes.
Fig. 8 Isothermal titration calorimetry of 10a and 10c with γ-cyclodex-
trin. For concentrations and thermodynamic parameters see Table 1.
Fig. 9 Nanodiamond 10a lying in the γ-cyclodextrin cavity forming
hydrogen bonds to the cyclodextrin rim.
Acknowledgements
increase possible interactions with the ND located in its centre.
The γ-CD cavity is simply too large to bind the slim 4-diaman-
tane tube.
Cyclodextrins were kindly donated by Wacker Chemie AG. The
Deutsche Forschungsgemeinschaft (grants Ra 1732/1-1 to BJR,
Schr 597/12-1 to PRS and SFB 858 to MW) are acknowledged
for financial support.
9-Triamantane
Notes and references
It is expected that triamantanes 10a and 10c should have a sub-
stantial affinity for CDs due to their bulky hydrophobic structure.
Indeed triamantane 10a forms the strongest ever reported
inclusion complex with γ-CD for a 1 : 1 complexation (Ka = 5.0
× 105 M−1). ND 10a also binds to β-CD, but weaker than to
γ-CD (Ka = 1.5 × 105 M−1). Representative ITC results are
shown in Fig. 8. This reduced binding affinity is evidently due
to the high deformation energy required to expand the smaller
β-CD cavity to accommodate the bulky triamantane group
(RMSDCD = 0.45 Å and Rp = 1.6 Å). The hydrogen bonding
network of β-CD44 is presumably destroyed by such significant
distortions. This observed expansion is in stark contrast to the
9a·γ-CD case (see above) where the γ-CD must contract and
become elliptically distorted. The 9-triamantane structure should
1 H. Pringsheim, Ber. Dtsch. Chem. Ges., 1913, 46, 2959–2974.
2 K. Freudenberg and F. Cramer, Chem. Ber., 1950, 83, 296–304.
3 For an overview of the chemistry and applications of cyclodextrins, see:
(a) Chem. Rev., 1998, 98, 1741–2076 (special issue on cyclodextrins);
(b) Cyclodextrins and Their Complexes, ed. H. Dodziuk, Wiley-VCH,
2006.
4 (a) A. Mulder, J. Huskens and D. N. Reinhoudt, Org. Biomol. Chem.,
2004, 2, 3409–3424; (b) G. Chen and M. Jiang, Chem. Soc. Rev., 2011,
40, 2254–2266.
5 (a) M. T. Rojas, R. Koniger, J. F. Stoddart and A. E. Kaifer, J. Am.
Chem. Soc., 1995, 117, 336–343; (b) X. Y. Ling, D. N. Reinhoudt and
J. Huskens, Chem. Mater., 2008, 20, 3574–3578; (c) M. D. Yilmaz, S.
H. Hsu, D. N. Reinhoudt, A. H. Velders and J. Huskens, Angew. Chem.,
Int. Ed., 2010, 49, 5938–5941.
6 (a) I. Tomatsu, A. Hashidzume and A. Harada, Macromolecules, 2005,
38, 5223–5227; (b) A. Harada, R. Kobayashi, Y. Takashima,
A. Hashidzume and H. Yamaguchi, Nat. Chem., 2011, 3, 34–37.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 4524–4530 | 4529

View Online
7 (a) J. Voskuhl, M. C. A. Stuart and B. J. Ravoo, Chem.–Eur. J., 2010, 16,
2790–2796; (b) S. K. M. Nalluri, J. Voskuhl, J. L. Bultema, E.
J. Boekema and B. J. Ravoo, Angew. Chem., Int. Ed., 2011, 50, 9747–
9751.
26 P. Weinzinger, P. Weiss-Greiler, W. Snor, H. Viernstein and P. Wolschann,
J. Inclusion Phenom. Macrocyclic Chem., 2007, 57, 29–33.
27 W. Snor, E. Liedl, P. Weiss-Greiler, H. Viernstein and P. Wolschann,
Int. J. Pharm., 2009, 381, 146–152.
8 (a) B. Zhang and R. Breslow, J. Am. Chem. Soc., 1993, 115, 9354–9354;
(b) M. Weickenmeier and G. Wenz, Macromol. Rapid Commun., 1996,
17, 731–736; (c) L. A. Godinez, K. Saeki, S. Patel, C. M. Criss and A.
E. Kaifer, J. Phys. Chem., 1995, 99, 17449–17455.
28 A. A. Rafai, S. M. Hashemianzadeh, Z. B. Nojini and M. A. Safarpour, J.
Mol. Liq., 2007, 135, 153–157.
29 W. Chen, C. E. Chang and M. K. Gilson, Biophys. J., 2004, 87, 3035–
3049.
9 K. Kano, R. Nishiyabu and R. Doi, J. Org. Chem., 2005, 70, 3667–3673.
10 (a) M. R. de Jong, J. F. J. Engbersen, J. Huskens and D. N. Reinhoudt,
Chem.–Eur. J., 2000, 6, 4034–4040; (b) Y. Liu, Y. W. Yang, E. C. Yang
and X. D. Guan, J. Org. Chem., 2004, 69, 6590–6602.
11 A. F. D. Denamor, R. Traboulssi and D. F. V. Lewis, J. Am. Chem. Soc.,
1990, 112, 8442–8447.
12 H. Schwertfeger, A. A. Fokin and P. R. Schreiner, Angew. Chem., Int.
Ed., 2008, 47, 1022–1036.
13 P. von R. Schleyer, J. Am. Chem. Soc., 1957, 79, 3292–3292.
14 C. Cupas, P. v. R. Schleyer and D. J. Trecker, J. Am. Chem. Soc., 1965,
87, 917–918.
30 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010,
132, 154104.
31 L. Goerigk and S. Grimme, J. Chem. Theory Comput., 2011, 7, 291–309.
32 (a) T. M. Gund, M. Nomura and P. v. R. Schleyer, J. Org. Chem., 1974,
39, 2987–2994; (b) N. A. Fokina, B. A. Tkachenko, J. E. P. Dahl,
R. M. K. Carlson, A. A. Fokin and P. R. Schreiner, Synthesis, 2012, 259–
264.
33 See for example:M. Rekharsky and Y. Inoue, Chem. Rev., 1998, 98,
1875–1917.
34 S. Grimme, J. Comput. Chem., 2004, 25, 1463–1473.
35 A. Schaefer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97, 2571.
15 V. Z. Williams, P. von R. Schleyer, G. J. Gleicher and L. B. Rodewald, J.
Am. Chem. Soc., 1966, 88, 3862–3863.
16 J. E. P. Dahl, S. G. Liu and R. M. K. Carlson, Science, 2003, 299, 96–99.
17 H. Richter, H. Schwertfeger, P. R. Schreiner, R. Fröhlich and F. Glorius,
Synlett, 2009, 193–197.
18 A. A. Fokin, A. Merz, N. A. Fokina, H. Schwertfeger, S. H. Liu,
J. E. P. Dahl, R. K. M. Carlson and P. R. Schreiner, Synthesis, 2009, 909–
912.
19 C. Sinkel, S. Agarwal, N. A. Fokina and P. R. Schreiner, J. Appl. Polym.
Sci., 2009, 114, 2109–2115.
20 (a) T. M. Willey, J. D. Fabbri, J. R. I. Lee, P. R. Schreiner, A. A. Fokin,
B. A. Tkachenko, N. A. Fokina, J. E. P. Dahl, R. M. K. Carlson, A.
L. Vance, W. Yang, L. J. Terminello, T. van Buuren and N. A. Melosh, J.
Am. Chem. Soc., 2008, 130, 10536–10544; (b) T. M. Willey,
R. I. L. Jonathan, J. D. Fabbri, D. Wang, M. H. Nielsen, J. C. Randel, P.
R. Schreiner, A. A. Fokin, B. A. Tkachenko, N. A. Fokina, J. E. P. Dahl,
R. M. K. Carlson, L. J. Terminello, N. A. Melosh and T. van Buuren, J.
Electron Spectrosc. Relat. Phenom., 2009, 172, 69.
21 (a) W. L. Yang, J. D. Fabbri, T. M. Willey, J. R. I. Lee, J. E. P. Dahl,
R. M. K. Carlson, P. R. Schreiner, A. A. Fokin, B. A. Tkachenko, N.
A. Fokina, W. Meevasana, N. Mannella, K. Tanaka, X. J. Zhou, T. van
Buuren, M. A. Kelly, Z. Hussain, N. A. Melosh and Z. X. Shen, Science,
2007, 316, 1460–1462; (b) W. A. Clay, Z. Liu, W. Yang, J. D. Fabbri,
J. E. P. Dahl, R. M. K. Carlson, Y. Sun, P. R. Schreiner, A. A. Fokin, B.
A. Tkachenko, N. A. Fokina, P. A. Pianetta, N. A. Melosh and Z.
X. Shen, Nano Lett., 2009, 9, 57–61; (c) S. Roth, D. Leuenberger,
J. Osterwalder, J. E. P. Dahl, R. M. K. Carlson, B. A. Tkachenko, A.
A. Fokin, P. R. Schreiner and M. Hengsberger, Chem. Phys. Lett., 2010,
495, 102–108.
36 TURBOMOLE V6.2 2010,
a
development of University of
Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989–2007,
TURBOMOLE GmbH, since 2007; available from http://www.turbomole.
com
37 To ensure our results are not biased due to our rather small basis set,
further selected models were re-optimized at the B97-D29 GGA func-
tional def2-TZVPP35 level of theory. This level of theory was shown to
be an excellent method in a recent benchmark study on related systems.31
The RMSD for selected models is only 0.2 Å, with the only notable
difference being a slightly shorter distance between the carboxylic acid of
the substituent and the rim of the CD, this was due to slight distortion
away from planarity.
38 W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey and M.
L. Klein, J. Chem. Phys., 1983, 79, 926–935.
39 B. R. Brooks, C. L. Brooks III, A. D. Mackerell, L. Nilsson, R.
J. Petrella, B. Roux, Y. Won, G. Archontis, C. Bartels, S. Boresch,
A. Caflisch, L. Caves, Q. Cui, A. R. Dinner, M. Feig, S. Fischer, J. Gao,
M. Hodoscek, W. Im, K. Kuczera, T. Lazaridis, J. Ma, V. Ovchinnikov,
E. Paci, R. W. Pastor, C. B. Post, J. Z. Pu, M. Schaefer, B. Tidor, R.
M. Venable, H. L. Woodcock, X. Wu, W. Yang, D. M. York and
M. Karplus, J. Comput. Chem., 2009, 30, 1545–1615.
40 P. Sherwood, A. H. de Vries, M. F. Guest, G. Schreckenbach,
C. R. A. Catlow, S. A. French, A. A. Sokol, S. T. Bromley, W. Thiel, A.
J. Turner, S. Billeter, F. Terstegen, S. Thiel, J. Kendrick, S. C. Rogers,
J. Casci, M. Watson, F. King, E. Karlsen, M. Sjøvoll, A. Fahmi,
A. Schäfer and Ch. Lennartz, THEOCHEM, 2003, 632, 1–28.
41 W. Humphrey, A. Dalke and K. Schulten, J. Mol. Graphics, 1996, 14,
33–38.
42 It has been reported that 7a forms a complex with γ-CD, however the ΔH
of complexation is rather small:W. C. Cromwell, K. Bystrom and M.
R. Efting, J. Phys. Chem., 1985, 89, 326–332.
43 S. Miyamoto, A. Kawai, S. Higuchi, Y. Nishi, T. Tanimoto, Y. Uekaji,
D. Nakata, H. Fukumi and K. Terao, Chem-Bio Inf. J., 2009, 9, 1–11.
44 W. Saenger, C. B. V. Zabel, G. M. Brown, B. E. Hingerty, B. Lesyng and
S. A. Mason, Proc. Int. Symp. Biomol. Struct. Interact., Suppl. J. Biosci.,
1985, 8, 437–450.
22 K. B. Lipkowitz, Chem. Rev., 1998, 98, 1829–1873.
23 A. Karpfen, E. Liedl, W. Snor and P. Wolschann, J. Inclusion Phenom.
Macrocyclic Chem., 2007, 57, 35–38.
24 W. Snor, E. Liedl, P. Weiss-Greiler, A. Karpfen, H. Viernstein and
P. Wolschann, Chem. Phys. Lett., 2007, 441, 159–162.
25 A. Karpfen, E. Liedl, W. Snor, H. Viernstein and P. Weiss-Greiler,
Monatsh. Chem., 2008, 139, 363–371.
4530 | Org. Biomol. Chem., 2012, 10, 4524–4530
This journal is © The Royal Society of Chemistry 2012