Solid-state NMR investigations on urea inclusion compounds: Order and dynamics of 1,6-dibromohexane

Article abstract of DOI:10.1016/j.molstruc.2006.07.031

The molecular properties of 1,6-dibromohexane in its urea inclusion compound are investigated by means of a multinuclear solid-state NMR spectroscopy. ¹³C CP/MAS and ¹H MAS NMR studies (line shapes, spin-lattice relaxation measurements) were performed for the first time to probe the guest dynamics and conformational order. Variable temperature ²H NMR studies comprising line shape analysis as well as spin-lattice relaxation (T_1Z, T_1Q) experiments were done on samples with guest molecules selectively deuterated at two different positions. A quantitative analysis of the experimental data proves that the guest dynamics are dominated by mutual exchange between two gauche conformers. It is shown that these guest motions unequivocally can be quantified (type and time-scale) by a comprehensive analysis of the T_1Z and T_1Q data. In addition, there is evidence that other motional contributions, such as overall molecular fluctuations and lateral motions, also contribute to spin relaxation. The molecular behaviour of 1,6-dibromohexane in urea is completely different from that reported for the long chain analogues or for n-alkanes where typically unhindered overall rotational motions of the guests in their all-trans conformation around the urea channel long axis are discussed. The differences in guest ordering and dynamics are a direct consequence of the differences of the urea lattice structures (monoclinic vs. hexagonal urea lattice).

Full text of DOI:10.1016/j.molstruc.2006.07.031

Journal of Molecular Structure 831 (2007) 75–89
www.elsevier.com/locate/molstruc
Solid-state NMR investigations on urea inclusion compounds:
Order and dynamics of 1,6-dibromohexane
Xiaorong Yang, Klaus M u¨ ller *
Institut f u¨ r Physikalische Chemie, Universit a¨ t Stuttgart, Pfaffenwaldring 55, D-70569 Stuttgart, Germany
Received 16 June 2006; received in revised form 18 July 2006; accepted 22 July 2006
Available online 12 September 2006
Abstract
The molecular properties of 1,6-dibromohexane in its urea inclusion compound are investigated by means of a multinuclear solid-state
13
1
NMR spectroscopy. C CP/MAS and H MAS NMR studies (line shapes, spin–lattice relaxation measurements) were performed for the
2
first time to probe the guest dynamics and conformational order. Variable temperature H NMR studies comprising line shape analysis
as well as spin–lattice relaxation (T1Z, T1Q) experiments were done on samples with guest molecules selectively deuterated at two different
positions. A quantitative analysis of the experimental data proves that the guest dynamics are dominated by mutual exchange between
two gauche conformers. It is shown that these guest motions unequivocally can be quantified (type and time-scale) by a comprehensive
analysis of the T1Z and T1Q data. In addition, there is evidence that other motional contributions, such as overall molecular fluctuations
and lateral motions, also contribute to spin relaxation. The molecular behaviour of 1,6-dibromohexane in urea is completely different
from that reported for the long chain analogues or for n-alkanes where typically unhindered overall rotational motions of the guests
in their all-trans conformation around the urea channel long axis are discussed. The differences in guest ordering and dynamics are a
direct consequence of the differences of the urea lattice structures (monoclinic vs. hexagonal urea lattice).
ꢀ
2006 Elsevier B.V. All rights reserved.
13
2
Keywords: Solid-state NMR; C NMR; H NMR; Urea inclusion compounds; Guest–host systems
1
. Introduction
therefore provides reliable information about the molecu-
lar interactions among the various constituents.
During recent years inclusion compounds have been
The most prominent examples for host systems are urea
or thiourea, both of which form hexagonal channel struc-
tures. Primarily they differ by the average diameter of the
host channels, resulting in a different selectivity for the
enclathrated guest species. The smaller diameter of the urea
channels thus favours the inclathration of n-alkanes and
their derivatives while bulkier species remain excluded
[3–8]. In the past, several studies about the thermal and
structural properties of the host lattice as well as the molec-
ular behaviour of the guest species have been performed
[17–21].
studied in great detail, also addressing, for instance, the
structural properties and dynamic features of the guest spe-
cies [1–13]. Inclusion compounds can be formed by various
types of organic or inorganic host components, such as
urea, thiourea, cyclophosphazenes, and cyclodectrins. They
are of great importance in pure and applied chemistry since
they possess high selectivity with respect to the guest mol-
ecules [14–16]. Inclusion compounds are relatively simple
model systems for the study of molecules under very
restricted spatial conditions. A thorough examination
It is known that the urea inclusion compound (UIC)
with 1,6-dibromohexane [22,23] differs in several aspects
from most other UICs as, for instance, UICs with long-
*
Corresponding author. Tel.: +49 711 685 64470; fax: +49 711 685
chain a,x-dibromoalkanes Br(CH ) Br (n = 7–11) [8] or
6
4467.
2 n
E-mail address: k.mueller@ipc.uni-stuttgart.de (K. M u¨ ller).
n-alkanes CH (CH ) CH (n P 8) [17–19]. Hence, UICs
3
2 n
3
0
022-2860/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2006.07.031

76
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
with long chain dibromoalkanes exhibit a solid–solid phase
transition in the range between about 140 K and 170 K,
reflecting a transition from the low-temperature ortho-
rhombic phase to the high-temperature hexagonal phase
1,6-dibromohexane/urea with guest molecules which are
deuterated either at the a- or b-position (see Fig. 1),
and present the first comprehensive relaxation study of
this system. The data analysis primarily relies on spin–
lattice relaxation data (T ,T1Q) which turn out to match
[
8,24]. Furthermore, these long-chain analogues possess
1
Z
an incommensurate relationship between the host and
guest substructures [7,22,23], and the end-groups primarily
adopt the trans-conformation with only a small fraction
the time-scale of the guest motions. It is shown that a
data analysis, comprising relaxation times as well as
relaxation time anisotropy, provides consistent and reli-
able results for the aforementioned gauche–gauche
(
ca. 5–14%) of gauche conformers [8].
1
3
1
For 1,6-dibromohexane/urea a similar solid–solid state
exchange dynamics. In addition, C CP/MAS and H
phase transition below room temperature does not exist.
Upon sample heating, the isomorphous monoclinic struc-
MAS NMR studies (variable temperature spectra, T1
and T1q relaxation measurements) were performed for
the first time on samples with non-deuterated guests in
ture (space group P2 /n), determined by X-ray diffraction
1
at 100 K, gradually transforms towards hexagonal symme-
try [23]. Moreover, the host and guest substructures of the
urea and urea-d . It is demonstrated that the derived
4
data from these latter studies are consistent with the
2
1
,6-dibromohexane UIC are thought to be commensurate.
assumptions employed during the H NMR data analy-
The 1,6-dibromohexane guests are therefore fixed in well-
defined positions within the urea substructure. As a result,
the end-groups of the incorporated 1,6-dibromohexanes
are exclusively stabilized in a gauche conformational state
even upon cooling down to 90 K [22,23].
sis. Moreover, they provide additional information about
the structural and dynamic features of these materials
which so far was not available.
2. Experimental
The study of molecular motions in condensed matter is
of fundamental importance, for instance, for the under-
standing of transport phenomena or of the respective mac-
roscopic appearance of these materials [25–27]. Among the
different techniques for studying molecular motions,
dynamic NMR spectroscopy is well established because
various pulse experiments are available which give access
to different motional time-scales. Hence, from the com-
bined data analysis of relaxation, line shape, and stimu-
lated echo experiments molecular motions can be
followed over several orders of magnitude [5,6,24,28,29].
2.1. Materials
All chemicals for the chemical synthesis were obtained
from Aldrich Chemicals, and were used without further
purification. UICs with two deuterated dibromohexanes,
1,6-dibromohexane-1,1,6,6-d4 (a-position) and 1,6-dib-
romohexane-2,2,5,5-d (b-position), were examined in the
present work. Their synthesis and preparation is briefly
described as follows:
4
2
In this context, H NMR spectroscopy with selectively deu-
2.1.1. (a) 1,6-Dibromohexane-1,1,6,6-d4
Hexanedioic acid dimethylester was reduced with
terated compounds has demonstrated its particular appli-
cability since the data analysis is simplified by the
presence of only one dominant internal magnetic interac-
tion, namely the quadrupolar interaction [28,29]. In the
LiAlD using standard procedures. The resulting hexanedi-
4
ol-1,1,6,6-d was transformed into its bromide by reaction
4
with sodium bromide in an aqueous solution of sulphuric
acid. The product 1,6-dibromohexane-1,1,6,6-d was puri-
2
past, dynamic H NMR spectroscopy was successfully
4
employed during the molecular characterization of quite
different materials, such as molecular and liquid crystals,
polymers, biomembranes, guest–host systems, etc [3,5,29–
fied by distillation [33].
2.1.2. (b) 1,6-Dibromohexane-2,2,5,5-d4
3
1].
H/D exchange in the a-position of hexanedioic acid
dimethylester was done by using sodium in methanol-d1
[34]. The resulting selectively deuterated diester was
reduced by lithium aluminium hydride to give 1,6-hexane-
2
H NMR spectroscopy was employed previously to
determine the dynamics of selectively deuterated 1,6-dib-
romohexane in urea [7]. With the information from a com-
plementary X-ray diffraction study [7,23], it was found that
the guest molecules exist in two non-degenerate gauche
conformational states (with unequal equilibrium popula-
tion) which undergo mutual exchange at elevated tempera-
diol-2,2,5,5-d . The final product, 1,6-dibromohexane-
4
2,2,5,5-d , was obtained by reaction with sodium bromide
4
in an aqueous solution of sulphuric acid, followed by puri-
fication via distillation [33].
2
tures. H NMR line shapes were used to extract the kinetic
parameters for this motional process. However, it turned
out that the overall line shape effects are rather small, since
the motional correlation times were almost in the fast
exchange limit, and are thus outside the sensitive range of
line shape alterations [4–6,32].
2.1.3. (c) Urea-d4
Urea-d was prepared by slowly cooling a warm solution
4
of urea in methanol-d (molar ratio urea to methanol-
1
d1 = 1:4), followed by filtration of the solution and drying
of the product. This procedure was repeated once to ensure
an almost complete exchange of the urea protons by
deuterons.
We have therefore undertaken a comprehensive vari-
2
able temperature H NMR study on two samples of

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
77
4 4
Fig. 1. Molecular structures for 1,6-dibromohexane-1,1,6,6-d /urea (a) and 1,6-dibromohexane-2,2,5,5-d /urea (b), and schematic drawing of the two
gauche conformers of 1,6-dibromohexane in urea and overall motions discussed in the text (c).
2
.2. Urea inclusion compounds
spin-lock pulse sequences (see Fig. 2) were used for the
1
3
1
1
determination of C and H T1q relaxation times.
H
UICs were prepared by slowly cooling warm solutions
of 1,6-dibromohexane or 1,10-dibromodecane with urea
MAS spectra were recorded with a Varian 4 mm triple res-
onance MAS probe (at 399.75 MHz) with a p/2 pulse
length of 3.5 ls. Typically, 64 scans were accumulated with
or urea-d in methanol or methanol-d , respectively. The
4
1
1
3
white layers were filtered, washed with 2,2,4-trimethylpen-
tane, and dried.
recycle delays of 4 s. C chemical shifts were determined
relative to the external standard adamantane. This value
was then expressed relative to TMS (d = 0 ppm). Likewise,
1
2
.3. Differential scanning calorimetry
H chemical shift values are given relative to TMS
(
d = 0 ppm).
2
Calorimetric studies were performed with a differential
H NMR experiments were performed at 46.07 MHz on
scanning calorimeter Netzsch DSC 204 (Selb/Germany)
under nitrogen flow at heating rates of 10 ꢁC/min.
a Bruker CXP 300 spectrometer interfaced to a Tecmag
spectrometer control system. The quadrupolar echo
sequence (p/2) –s –(p/2) –s with p/2 pulses of 2.0 ls and
x
e
y
e
2
.4. NMR studies
a pulse spacing of s = 20 ls was used to obtain the exper-
imental H NMR spectra. A modified inversion recovery
e
2
1
3
1
C CP/MAS and H MAS NMR experiments were per-
sequence, p–s –(p/2) –D–(p/2) –D, combined with the
r
x
y
formed on Bruker CXP 300 (Rheinstetten/Germany) and
Varian InfinityPlus 400 (Palo Alto/USA) spectrometers
operating at static magnetic fields of 7.04 and 9.39 T,
quadrupolar echo sequence for signal detection
(D = 20 ls), variable relaxation intervals s , was employed
r
for the determination of T1Z (relaxation time for Zeeman
order). Here, instead of the inversion p-pulse, a composite
1
3
respectively. The resonance frequencies were 75.47 ( C),
1
3
1
00.13 MHz ( H) for the CXP 300 spectrometer, and
pulse, given by [p/2] [p/2]
[p/2] , with appropriate
/± p/2 /
/
1
3
1
00.52 ( C), 399.75 MHz ( H) for the InfinityPlus 400
phase cycling (/ = 0, p/2, p and 3p/2) [36] was used.
spectrometer. A double tuned Bruker 4 mm MAS probe
Recycle delays were set to be at least five times the spin–lat-
1
3
13
13
was used for the C CP/MAS, T ( C), T ( C), and
tice relaxation time T . Relaxation times for quadrupolar
1
1q
1Z
1
T1q ( H) experiments. p/2 pulse lengths were 4 ls. Contact
times varied between 300 ls and 8 ms, as indicated in the
text, and the recycle delays between successive scans were
order, T1Q, were measured applying the broadband
Jeener-Broekaert sequence (BBJB) (90ꢁ –2s –67.5ꢁ
ꢀy
x
w
–2s –45ꢁ –s –45ꢁ –s –45ꢁ –D–90ꢁ ) [37]. Here, (90ꢁ –2s
w
y
w
y
r
x
x
x
w
6
s. A sample rotation frequency of 5 kHz was used, and
–67.5ꢁ –2s –45ꢁ –s –45ꢁ ) represents the excitation
ꢀ
y
w
y
w
y
1
3
the number of scans was between 512 and 1024. C T1
data were obtained with cross-polarization (CP) excitation
as proposed by Torchia et al. (see Fig. 2) [35]. Standard
sequence, whose phase cycle maximizes quadrupolar order
while fully suppressing Zeeman order [38]. A refocusing
pulse was used to generate an echo, resulting in partially

78
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
2.5. Data processing and simulations
Data processing of the experimental (from the CXP 300
NMR spectrometer) and theoretical NMR signals was
done on a SUN workstation using NMR1 and Sybyl/triad
software packages (Tripos, St. Louis, MO). For the data
handling from the InfinityPlus 400 NMR spectrometer
the MestReC software was used [39].
Simulation programs were employed which describe the
behaviour of a spin = 1 system during the corresponding
2
2
H NMR line shape and H relaxation experiments [6,40].
These FORTRAN programs are very general, and consider
various types of molecular motions of the systems under
investigation. The theoretical line shapes and relaxation
times were obtained by numerical diagonalization of the cor-
responding relaxation matrices [40,41] using standard soft-
ware packages.
2
For the H NMR line shape simulations, a slow-motion-
al approach was used [41,42]. This approach assumes d-
pulses, and explicitly accounts for the interval s in the
e
quadrupolar echo experiment, responsible for spin–spin
2
relaxation. The theoretical description of the H spin–lat-
tice relaxation data (T ,T1Q) is based on the Redfield
1
Z
approach [43]. Here, partially relaxed spectra were
2
obtained after multiplication of motionally averaged H
NMR spectra in the fast exchange limit with theoretical
damping factors that account for orientation-dependent
spin–lattice relaxation during the interval s of the inver-
r
sion recovery experiment [42] and BBJB experiments [37].
2
For instance, simulations of the partially relaxed H
NMR spectra from the modified inversion recovery exper-
iment are feasible by Eqs. (1)–(3). Here, the detected signal
S(t,2s ,s ) can be represented by
e
r
Sðt; 2s ; s Þ ¼ ½1 ꢀ 2 ꢁ expðꢀs =T1ZÞꢂ ꢁ Sðt; 2seÞ
ð1Þ
e
r
r
where S(t,2s ) is the FID signal from the former quadru-
e
pole echo experiment. The delay s corresponds to the
r
relaxation interval between the inversion pulse and the
quadrupole echo sequence.
The spin–lattice relaxation times T , which in the most
1
Z
general case exhibit a distinct orientational dependence,
were calculated on the basis of a dominant quadrupolar
interaction. In this case, T1Z is related to the spectral
densities by [43]
Fig. 2. Pulse sequences used in the present work: (a) quadrupolar echo;
b) inversion recovery; (c) broadband Jeener-Broekaert (BBJB); spin–
ꢀ
ꢁ
2
2
(
1
3
e qQ
1
13
¼
ꢁ ½J ðxÞ þ 4J ð2xÞꢂ
1
2
ð2Þ
lattice relaxation experiments for H (d) and C; (e) in the rotating frame,
T
1Z
16
ꢀh
13
respectively; (f) spin–lattice relaxation experiments of C.
The spectral densities, J , were derived by solving Eq. (3)
m
for a general N-site exchange: [44,45]
2
relaxed spectra with minimal baseline distortion and no
first-order phase corrections. During the H NMR studies,
X
ð2Þ
0
ð2Þ
ðh0Þ
2
J
m
ðxÞ ¼ 2 ꢁ
d
ðh Þd
0
ma
ma
0
a;a ¼ꢀ2
the number of scans varied between 512 and 1300 depend-
ing on the signal/noise ratio. The sample temperature was
controlled by a Bruker BVT 2000 unit. The samples were
initially held at each new temperature for at least 15 min
to allow full thermal equilibration. In general, the accuracy
of temperature was found to be within ± 1 K.
N
X
ð0Þ ðnÞ ð0Þ ðnÞ ð2Þ 00
ð2Þ 00
ꢁ
ꢁ
X
X
X
l
X
d
ðh Þd ðh Þ
l
j
j
0a
l
0a
j
n;l;j¼1
k
n
0
cosða/ ꢀ a / Þ
ð3Þ
l
j
2
n
2
k þ x

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
79
0
0
0
(n)
with / ¼ / ꢀ / . Here, X and k are the corresponding
13
3.1. C CP/MAS and H MAS NMR studies
1
i
i
n
eigenvectors and eigenvalues of the symmetrized rate ma-
trix K . h and / are the spherical polar angles between
the principal axis and an intermediate axis (motional axis)
system, while the angles h and / connect the intermediate
0
00
00
13
Fig. 3 depicts C CP/MAS spectra of 1,6-dibromohex-
ane/urea and 1,10-dibromodecane/urea which were record-
ed at room temperature. Peak assignments along with the
corresponding data for the guest species from solution
NMR studies are summarized in Table 1. It can be seen
0
0
ð2Þ
and the laboratory axis system. dab ðhÞ are elements of the
reduced Wigner rotation matrix. In quite the same way
the partially relaxed spectra from BBJB sequence were cal-
culated and T1Q is given by the expression
1
3
that the C NMR signals of the dibromoalkanes experi-
ence a pronounced downfield shift upon enclathration in
urea. In fact, unusual large shifts up to about 5 ppm for
ꢀ
ꢁ
2
2
2
1
9p e qQ
1
3
¼
J
1
ðx
0
Þ
ð4Þ
1,6-dibromohexane can be observed. Similar C chemical
shift alterations were reported earlier, for instance, for n-al-
kanes [46] and branched n-alkanes [47] in urea, and were
attributed to changes of the conformational state. On the
T
1Q
2
h
3
. Results and discussion
Urea inclusion compounds with 1,6-dibromohexane
were investigated by variable temperature H NMR spec-
troscopy. In addition, H and C solid-state NMR studies
basis of these observations and the results from
2
H
NMR, Raman, IR and X-ray investigations
2
[3–8,22,23,46,47], it was concluded that n-alkanes in urea
adopt a nearly all-trans conformational state while in solu-
tion a fast equilibrium between gauche and trans conform-
ers exists. The present data for the UIC with the long-chain
analogue 1,10-dibromodecane can be understood in quite
the same way. That is, 1,10-dibromodecane in urea – unlike
the situation in solution – again exists in an almost all-trans
conformation which also has been established from Raman
experiments [8]. However, in the case of 1,6-dibromohex-
ane even larger downfield shifts are registered which cannot
be explained in the same way. As mentioned earlier, these
guests exclusively exist in two states [7,22,23] which are
characterized by a gauche conformation at the chain ends.
1
13
2
were performed for the same system. H NMR studies were
conducted on two inclusion compounds with non-deuterat-
ed urea, and 1,6-dibromohexane deuterated either at car-
bons C-1/C-6 (a-positions) or at carbons C-2/C-5
b-positions). For the C and H NMR studies, samples
with 1,6-dibromohexane and urea or urea-d were prepared.
For comparison, some data will be also reported for 1,10-
dibromodecane/urea [24]. Samples were checked by differen-
tial scanning calorimetry. In agreement with earlier reports
1
3
1
(
4
[
7,23] for 1,6-dibromohexane/urea no solid–solid state phase
transitions could be observed between 130 K and the disinte-
gration point of the inclusion compound at 337 K. The ther-
mal behaviour of 1,6-dibromohexane/urea is different from
that encountered for UICs with n-alkanes or long-chain
dibromoalkanes, such as 1,10-dibromohexane, where char-
acteristic solid–solid transitions, accompanied by a change
of the host lattice structure, are registered [8,24].
1
3
That is, for 1,6-dibromohexane/urea the observed
downfield shifts must reflect a change from coexisting
gauche and trans conformers in solution towards two
C
It should be noted that studies about 1,6-dibromohex-
ane/urea were published recently which addressed the
structural properties of this system and the dynamic
features of the guest species [7,22,23]. Variable temperature
X-ray diffraction studies showed that the isomorphous
monoclinic structure (space group P2 /n) at 100 K gradual-
1
ly transforms towards hexagonal symmetry when
approaching room temperature [23]. Moreover, the guest
components were found to exist exclusively in two non-de-
generate gauche conformational states (with one predomi-
nant site), and are assumed to undergo mutual exchange
between these two states. It is important to note that the
exchange is not an internal process. Rather, it affects the
whole molecule, and restricted reorientational motion of
the molecule thus takes place. In addition, motional corre-
lation times were reported for the gauche–gauche intercon-
2
version process on the basis of variable temperature H
NMR line shapes on UICs with selectively deuterated
1
,6-dibromohexane guests [7], although the observed spec-
tral effects were rather subtle. The existence of a predomi-
nant gauche conformational state was also confirmed
independently by a Raman investigation [22].
1
3
Fig. 3. C CP/MAS NMR spectra of (a) 1,6-dibromohexane, and (b)
1,10-dibromodecane in urea.

80
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
Table 1
1
3
1
NMR parameters from C and H NMR studies on 1,6-dibromohexane and 1,10-dibromodecane in solution and UICs
c
d
13
C
[ppm]
Dm1/2 (13C) [Hz]
d
1
H
[ppm]
a
b
d
e
Solution
Urea
Urea-d
4
Urea
Urea-d
4
Solution
Urea
6
(a) BrC H12Br
C-1
C-2
C-3
C@O
33.5
32.5
27.3
38.4
34.7
30.5
38.1
34.4
30.2
164
37.3
32.0
29.4
220
65.7
49.3
47.5
137
H (C-1)
H (C-2)
H (C-3)
3.4
1.8
1.5
3.8
2.2
1.8
164
(b) BrC10H20Br
C-1
C-2
C-3,4
C-5
C@O
33.9
32.8
28.7–28.2
29.3
36.2
33.1
32.0
32.5
36.1
32.9
31.9
32.3
164
20.4
22.7
22.5
31.8
91.6
28.5
36.5
33.8
39.6
51.9
H (C-1)
H (C-2)
H (C-3)
H (C-4,5)
3.4
1.9
1.5
1.3
3.5
2.1
1.5
1.5
164
a
b
c
13
13
13
1
3
C chemical shifts of 1,6-dibromohexane, 1,10-dibromodecane in CDCl . Assignments from Refs. [57,58].
C chemical shifts of 1,6-dibromohexane and 1,10-dibromodecane in their UICs derived in this work from C CP/MAS studies.
C NMR line width.
13
d
e
3
H chemical shifts of 1,6-dibromohexane and 1,10-dibromodecane in CDCl .
1
1
H chemical shifts of 1,6-dibromohexane and 1,10-dibromodecane in their UICs derived in this work from H MAS experiments.
coexisting gauche conformers. This implies that in general
the chemical shift alterations in urea inclusion compounds
are not solely based on conformational effects. Rather they
also reflect other alterations of the electronic environment
around the guest carbon nuclei due to interactions with
the urea host lattice.
can be attributed to the loss of urea protons which are
an important source for magnetization transfer during
cross-polarization. This assumption is also confirmed by
1
the H MAS NMR data discussed below. Similar trends
1
3
for the C signal intensity during cross-polarization were
also observed recently for other urea inclusion compounds
[24].
The interaction between the urea host lattice and the
guest species also becomes evident from the comparison
1
Fig. 4 exhibits H MAS NMR spectra recorded at
1
3
of the C NMR line widths in the UICs from urea and
15 kHz for 1,6-dibromohexane and 1,10-dibromodecane
urea-d (see Table 1). Thus, for the latter samples larger
in urea and urea-d . For 1,6-dibromohexane/urea highly
resolved H NMR spectra are observed even at moderate
4
4
1
line widths of the guest species are found than for the sam-
ple made from non-deuterated urea, which is attributed to
spinning rates of 15 kHz which indicates a reduced spin dif-
fusion for this UIC. Deuteration of the urea host lattice is
accompanied by additional resolution enhancement for the
1
3
2
remaining C– H dipolar couplings between the guest spe-
1
3
cies and the host matrix. The C NMR line width of the
urea carbonyl group is found to become smaller upon urea
deuteration, which reflects interference effects [48,49]
1
H NMR signals of the guest species. Obviously, upon urea
deuteration proton spin diffusion is further suppressed, as
the large proton spin reservoir from the non-deuterated
urea molecules is missing, and only reduced intermolecular
dipolar interactions between isolated guests exist. A com-
1
between the H decoupling r.f. field and the urea reorienta-
tional motions, i.e., 180ꢁ flips around the C@O and the
1
CAN bonds [50,51]. Insufficient H decoupler power can
1
1
be ruled out. In fact, an increase of the H decoupler power
parison of the derived H chemical shift values again
1
3
is accompanied by a broadening for the carbonyl
C
reveals downfield shifts upon enclathration for both urea
samples with 1,6-dibromohexane and 1,10-dibromodecane
where for 1,6-dibromohexane/urea larger effects are
registered (shifts up to 0.4 ppm). As before, these chemical
shift alterations are traced back to conformational changes
and/or interactions between guest molecules and host
matrix.
NMR signal which is in line with the aforementioned inter-
ference effect. Similar observations were made for other
UICs [24].
It is worth mentioning that the optimum contact time
for cross-polarization is found to vary considerably with
guest species and deuteration of the urea matrix. Thus,
1
3
for 1,6-dibromohexane/urea, a maximum C CP signal
was observed at a contact times of 0.3 ms and 0.9 ms for
Further line narrowing can be achieved through sample
spinning at 34 kHz (data not shown). However, for 1,6-dib-
romohexane/urea additional resonances show up which
stem from the presence of free guest molecules. That is,
due to temperature increase upon fast MAS, guest mole-
cules diffusive out of the urea host lattice and a disintegra-
tion of the UIC sets in. Similar effects arising from partial
disintegration of the UIC are also found during the vari-
urea and urea-d , respectively, while for 1,10-dibromode-
4
cane these values are 2 and 8 ms. Two major reasons might
be responsible for these findings: (i) differences for the
motions of guest species which interfere with the cross-po-
larization process, and/or (ii) differences for the average
distances between guest and host molecules. The longer
contact times needed for the deuterated urea clathrates
2
able temperature H NMR studies presented below.

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
81
1
Fig. 4.
4
H MAS NMR spectra of (a) 1,6-dibromohexane/urea, (b) 1,6-dibromodecane/urea-d , (c) 1,10-dibromodecane/urea, and (d) 1,
1
0-dibromodecane/urea-d .
4
The dynamic properties of the guest components are
it can be concluded that the underlying motions are on
the high temperature branch of the relaxation curve. The
C T values are slightly longer than the C T1q values
1
1
3
13
1
examined by C T , C T and H T measurements.
1
q
1
1q
1
3
13
The derived data are summarized in Tables 2 and 3. For
1
3
1
,6-dibromohexane/urea it is found that the C T1q values
(about a factor of 4), they do not exhibit a distinct position
dependence, and are almost unaffected by urea deuteration.
slightly vary with chain position, as reflected by a T1q
1
3
increase towards the inner part of the chain. Urea deutera-
Moreover, C T minima are found at 305 K, while the
1
1
3
tion is characterized by an increase for both the C T1q
UIC with 1,10-dibromodecane/urea behaves completely
1
13
and H T values due to reduced effective hetero- and
different. For instance, considerable larger C T1q and
1
q
1
3
homonuclear dipolar interactions between guest and host
C T values of about 400 ms and 8 s, respectively, are
1
1
3
components. From the variable temperature C T1q data
observed at room temperature. Further details can be
found elsewhere [24].
We made an attempt for a semi-quantitative analysis of
1
3
Table 2
these C relaxation data by using the well-known expres-
sions [52] for T1q and T1 in the presence of isotropic
motions and a dominant heteronuclear dipolar interaction
T
1q(13C) and T1q(
1
H) data for the BrC
6
H
12Br in urea and urea-d
4
b
T
1q(13C) [ms]
T
1q(1H) [ms]
2
94 K
305 K
310 K
320 K
1
2
a
12Br/urea (CT : 0.3 m)
¼ A ½Jðx0;H ꢀ x0;CÞ þ 3 ꢁ Jðx0;CÞ þ 6 ꢁ Jðx0;H þ x0;CÞꢂ
BrC
C-1
C-2
C-3
6
H
H
T
T
ð CÞ
13
1
45.6
57.9
63.5
55.9
65.7
73.8
61.4
69.4
76.5
64.6
71.8
82.2
1.2
1.2
1.2
ð5Þ
2
1
A
¼
½4 ꢁ Jðx1;CÞ þ Jðx0;H ꢀ x0;CÞ þ 3
2
1
3
BrC
C-1
C-2
C-3
a
6
12Br/urea-d
54.0
4
(CT: 0.9 m)
73.7
1qð CÞ
80.7
103
122
88.0
114
146
9.0
9.3
9.4
ꢁ
Jðx0;CÞ þ 6 ꢁ Jðx0;H þ x0;CÞ þ 6 ꢁ Jðx0;HÞꢂ ð6Þ
80.6
88.8
87.3
98.7
with
ðx
Contact time.
Value at 294 K.
b
sc
J
n
i
Þ ¼ 1
;
2
c
x
i
¼ cB
i
ð7Þ
2
i
þ x s
2
Here A is a factor which contains the gyromagnetic ratios
of the interacting nuclei, and the average internuclear dis-
tance as Æ1/r æ. x and x0,H are the Larmor frequencies
of the C and H nuclei, respectively, and x is the nuta-
tion frequency of the radio frequency field B ( C).
In a first step, the experimental C T curve – which
exhibits a minimum (see above) – was fitted by means of
Table 3
T
1(13C) data for the BrC
6
H
12Br in urea and urea-d
4
3
0
,C
T
1(13C) [ms]
1
3
1
1
,C
2
94 K
305 K
310 K
320 K
13
1
a
12Br/urea (CT : 0.3 m)
BrC
C-1
C-2
C-3
6
H
13
1
216
226
228
203
193
216
205
193
223
216
194
225
Eq. (5). In this way, factor A and the correlation times sc
of the motion responsible for spin–lattice relaxation were
a
BrC
C-1
C-2
C-3
a
6
H
12Br/urea-d
4
(CT : 0.9 m)
determined. In the second step, we assumed that the same
factor A can be employed for the analysis of the experimen-
242
251
255
203
189
203
209
193
211
209
194
219
1
3
tal C T1q data. That is, Eq. (6) along with the factor
1
3
A – derived via the C T analysis – were used to obtain
Contact time.
1

82
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
2
the correlation times for the motions which dominate T
relaxation. The two sets of correlation times from the
obtained from the quantitative H T analysis discussed
1
q
1
1
3
C
below. Nevertheless, the derived kinetic data from the
C relaxation experiments can be only discussed on a qual-
1
3
T and T data analysis are summarized in Tables 4 and
1
1q
5
. It can be seen that the correlation times derived from
itative basis, as a simple isotropic motion was employed
during this part of the data analysis. The corresponding
H NMR relaxation data analysis, however, is straightfor-
1
3
the C T data analysis are in a very similar range as those
1
2
ward, as here the correct (anisotropic) motional model for
the guest dynamics was employed.
Table 4
Kinetic parameters derived from T1q(13C) values for the BrC
6
H
12Br in UICs
On the basis of these results from the semi-quantitative
1
3
13
ꢀ
9
Correlation time s
94 K 305 K
c
[10 s]
Activation energy [kJ/mole]
C T1q analysis it is concluded that (i) C T relaxation
1
2
reflects the same motion being responsible for H T1Z
2
310 K
320 K
1
3
and T1Q relaxation, and (ii) C T1q relaxation is governed
BrC
C-1
C-2
C-3
6
H
12Br/urea (CT: 0.3 m)
by another motional process which occurs on a slower
8.5
6.0
6.0
6.5
5.1
5.0
6.0
4.6
4.8
5.6
4.5
4.4
12.8 ± 2.4
9.1 ± 1.9
9.3 ± 1.1
1
3
time-scale. C T relaxation is therefore considered to be
1
caused by the aforementioned interconversion between
two gauche conformers of the guest molecules, while over-
BrC
C-1
C-2
C-3
6
H
12Br/urea-d
4
(CT: 0.9 m)
1
3
6.9
3.7
3.6
4.6
3.3
3.1
4.1
2.5
2.1
3.7
2.1
1.5
19.1 ± 3.7
17.9 ± 3.8
27.4 ± 5.9
all fluctuations of the dibromohexane guests dominate
C
T1q relaxation. A discussion of this issue will be provided
2
after the presentation of the H NMR results.
2
.2. H NMR line shapes
3
Table 5
Kinetic parameters derived from T1(13C) values for the BrC
2
6
H
12Br in UICs
Variable temperature H NMR studies were performed
ꢀ
9
Correlation time s
c
[10 s]
in a temperature range between 220 and 320 K on two urea
samples, prepared with 1,6-dibromohexane selectively deu-
terated either at carbon positions C-1/C-6 or positions C-2/
C-5. Representative variable temperature H NMR line
shapes of both samples (from quadrupolar echo experi-
294 K
305 K
310 K
320 K
BrC
C-1
C-2
C-3
6
H
12Br/urea (CT: 0.3 m)
2
2.5
3.2
2.5
1.6
1.7
1.6
1.4
1.6
1.2
1.0
1.5
1.1
ments with s = 20 ls) are displayed in Fig. 5. It can be
e
seen that for the lowest temperature typical axially sym-
metric Pake pattern (asymmetry parameter g = 0.0) are
observed. Characteristic spectral alterations occur upon
sample heating which imply an onset of the guest motions.
BrC
C-1
C-2
C-3
6
H
4
12Br/urea-d (CT: 0.9 m)
3.4
4.1
3.7
1.6
1.6
1.6
1.2
1.3
1.1
1.1
1.2
1.0
2
Fig. 5. H NMR line shapes as a function of temperature for 1,6-dibromohexane-1,1,6,6-d
4
/urea (left: experiment and simulation) and 1,6-
dibromohexane-2,2,5,5-d /urea (right: experiment and simulation).
4

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
83
data analysis, as outlined below). It can be seen that the
populations derived from the analysis of the two deuterat-
ed guest species are almost identical. Likewise, the enthalpy
differences between the two gauche conformers, calculated
from these populations with values of DH = 9.3 kJ/mol for
the a-deuterated and 8.6 kJ/mol for the b-deuterated sam-
ples, are very close to the previously reported value of
8.5 kJ/mol [7].
The major difference between the present and the former
study is the time-scale of gauche–gauche interconversion.
Unlike the former work, where the authors proposed that
the motions should occur on a time-scale compatible with
2
H NMR line shape effects, according to our analysis a
faster time-scale has to be considered which matches the
sensitive range of spin–lattice relaxation. In fact, there
are various arguments which support our assumption:
Fig. 6. Populations of major gauche site as a function of temperature
C6Br-1,6: 1,6-dibromohexane-1,1,6,6-d /urea, C6Br-2,5: 1,6-dibromohex-
ane-2,2,5,5-d /urea).
(
4
4
(i) Model simulations based on the populations and rate
constants published in the former work only exhibit
subtle effects on the theoretical H NMR spectra
which hardly can be used to extract accurate motion-
al correlation times.
2
2
The observed non-axially symmetric H NMR spectra at
higher temperatures, characterized by an averaged value
of Ægæ „ 0.0, are typical for a molecules undergoing motions
in the fast exchange limit with a symmetry lower than C . It
(ii) Motional induced spin–spin relaxation effects (i.e.
3
is further registered that the spectral alterations also
depend on the deuterated position of the guest species.
For both compounds variable temperature spin–spin
angular dependence of T ) were not found during
the quadrupolar echo experiments with variable pulse
delay se.
2
(
T ) relaxation studies were performed (data not shown),
(iii) The experimental spin–lattice relaxation data pre-
sented below display a strong temperature and pro-
nounced angular dependence (T , T anisotropy),
2
from which there was no evidence for motionally induced
spin–spin relaxation effects. These findings clearly demon-
strate that the underlying guest motions are too fast and
outside the sensitive time-scale for NMR line shape analy-
sis. Rather, as outlined below they occur on a time-scale
suitable for spin–lattice relaxation studies.
1
Z
1Q
and are therefore a very sensitive tool for the exami-
nation of the kinetic properties for the guest
molecules.
2
The present variable temperature H NMR spectra are
2
2
identical to those presented in an earlier H NMR line
3.3. H spin–lattice relaxation data (T and T1Q)
1Z
shape study on UICs with a-, b- and c-deuterated 1,6-dib-
romohexanes [7]. Along with the structural data from a
complementary X-ray analysis, the authors described the
The guest dynamics were examined in a quantitative
way by spin–lattice relaxation experiments comprising both
samples with selectively deuterated guest molecules. For an
isolated I = 1 nucleus, two linear combinations of the
eigenstate populations (zero-quantum elements) define the
Zeeman and quadrupolar order which decay with time con-
stants T1Z and T1Q, respectively. The spin–lattice relaxa-
tion rates 1/T1Z and 1/T1Q depend on the molecular
motion responsible for spin relaxation via the spectral den-
2
experimental H NMR line shapes by a non-degenerate
two-site jump model which was consistent with mutual
exchange of the guest species between two coexisting
gauche conformers (see Fig. 1). With the assumption of
ideal tetrahedral geometry [7], the jump angles were first
calculated to 56ꢁ for the a-deuterons and 109ꢁ for the b-
deuterons. A further refinement [7,23] provided a jump
angle of Du = 58ꢁ for the a-deuterons, while the jump
sities J (nx), as given earlier by Eqs. (2 and 4) [43,53].
a
n
angle for the b-deuterons was kept to Du = 109ꢁ.
Both spectral densities J (x ) and J (2x ) depend on the
b
1
0
2
0
2
The theoretical H NMR line shapes of 1,6-dibromohex-
time-scale and type of the underlying motional process
which yields a distinct angular dependence for the spectral
densities and relaxation times T1Z and T1Q. In polycrystal-
line solids, the anisotropy of T1Z and T1Q, extracted from
the corresponding partially relaxed spectra [3,6,40,41,47],
can be further exploited to examine the quality of the mod-
el assumptions during the experimental relaxation data
analysis.
ane deuterated in the a- and b-positions, shown in Fig. 5,
were obtained with the same jump angles from the earlier
2
H NMR work [7]. However, as there was no evidence
for any spin–spin relaxation effects, the motional correla-
tion times were set to the fast exchange limit with
ꢀ8
sC = 10 s. The derived populations of the major gauche
site slightly differ from the former published data, and
are given in Fig. 6, Tables 6 and 7 (please note that the giv-
The experimental T1Z data from inversion recovery
experiments are given in Fig. 7. The T1Z values are powder
2
en rate constants are derived from the H T relaxation
1

84
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
Table 6
Simulation parameters for 1,6-dibromohexane-1,1,6,6-d
4
/urea
a
a
Temperature T [K]
Quadrupolar coupling
constant e qQ/h [kHz]
Correlation
Population
major site p
Rate constant
Rate constant
Line width
2
ꢀ1
ꢀ1
1=T2 ½sꢀ1
ꢂ
0
time s
c
[s]
10
10
10
9
1
k
12 [s
]
k21 [s
]
ꢀ
ꢀ
ꢀ
ꢀ
8
8
8
8
8
7
7
7
8
8
8
7
7
7
6
6
6
6
5
320
310
300
292
280
270
260
250
240
230
220
a
157
160
162
162
162
164
164
164
165
165
165
7 · 10
8 · 10
9 · 10
5 · 10
0.68
0.71
0.74
0.78
0.81
0.84
0.86
0.88
0.90
0.93
0.95
9.7 · 10
8.9 · 10
8.2 · 10
1.6 · 10
1.2 · 10
6.5 · 10
4.8 · 10
3.1 · 10
4.6 · 10
3.6 · 10
2.9 · 10
4.4 · 10
2.9 · 10
1.2 · 10
7.8 · 10
4.3 · 10
2.2 · 10
1.2 · 10
6.3 · 10
2500
2500
3000
3000
3000
3000
3500
3500
4000
4500
5000
ꢀ9
ꢀ8
ꢀ8
ꢀ8
ꢀ8
ꢀ8
6.5 · 10
1.3 · 10
1.8 · 10
2.8 · 10
4.5 · 10
5.7 · 10
7
2 · 10
7
7
1.6 · 10
1.2 · 10
ꢀ
8
8 · 10
= 58ꢁ.
kij: rate constant for transition j fi i, jump angle Du
a
Table 7
Simulation parameters for 1,6-dibromohexane-2,2,5,5-d
4
/urea
Correlation
time s [s]
a
a
Temperature T [K]
Quadrupolar coupling
Population
major site p
Rate constant
Rate constant
Line width
2
ꢀ1
ꢀ1
1=T2 ½sꢀ1
ꢂ
0
constant e qQ/h [kHz]
c
1
k
12 [s
]
k21 [s
]
ꢀ
10
9
8
8
8
7
7
7
6
6
6
6
5
320
310
300
292
280
270
260
250
240
230
220
a
162
162
162
162
162
162
163
164
164
164
164
6.5 · 10
0.69
0.72
0.75
0.78
0.81
0.83
0.85
0.87
0.89
0.92
0.95
1.1 · 10
9 · 10
4.8 · 10
3.5 · 10
2.8 · 10
3.7 · 10
2.4 · 10
1.3 · 10
7.1 · 10
4.2 · 10
2.3 · 10
1.1 · 10
6.3 · 10
1500
1500
1500
2000
2500
2500
3000
3500
4000
4500
5000
ꢀ
ꢀ
ꢀ
ꢀ
10
10
9
8
8 · 10
9 · 10
6 · 10
8 · 10
8
8
8.3 · 10
1.3 · 10
9
8
1 · 10
ꢀ
8
8
8
8
7
1.3 · 10
2.1 · 10
3.1 · 10
4.8 · 10
6.4 · 10
ꢀ
ꢀ
ꢀ
7
4 · 10
7
7
7
7
2.8 · 10
1.9 · 10
1.3 · 10
1.2 · 10
ꢀ
ꢀ
8
7 · 10
8 · 10
8
k
ij: rate constant for transition j fi i, jump angle Du = 109ꢁ.
b
values and were derived from the analysis of FID intensi-
ties from the inversion recovery experiments. It can be seen
that the T1Z curves display a shallow minimum of 8.8 ms
for the a-deuterons and 7.9 ms for the b-deuterons at
ticular, as shown by several other applications, the T1Z
anisotropy was found to be very sensitive for type and
time-scale of the molecular motions.
It should be noted that during this part of the data anal-
ysis the same values for the quadrupolar coupling constant,
populations and jump angles were used as before during
3
10 K. Moreover, the slope above the T1Z minimum
appears to be somewhat smaller than for the low tempera-
ture range.
2
the H NMR line shape analysis (see Tables 6 and 7).
2
Experimental partially relaxed H NMR spectra,
The only adjustable variable was the motional correlation
obtained from the inversion recovery experiment at
time s , which is related to the rate constants k and k21
c
12
2
1
60 K, are given in Figs. 8a and b for the samples with
,6-dibromohexane deuterated at the a- and b-positions,
by Eq. (8).
1
k12 k21
p1
respectively. The experimental data sets are contrasted with
the theoretical counterparts, which were obtained by means
of the simulation programme described in the experimental
part, and in general exhibit a very good agreement, sup-
porting the quality of the underlying model assumptions.
The same analysis of the partially relaxed spectra was done
for the other points of the experimental T1Z curve dis-
played in Fig. 7.
At this stage we also underwent attempts to describe the
guest motions by other motional models (e.g. degenerate
and non-degenerate 2- and three-site jumps at variable
opening angles). However, none of them provided a consis-
tent description of all available experimental data. In par-
¼
¼
p2
ð8Þ
sC
Here, kij represents the pseudo first-order rate constant for the
transition from site j to site i with the populations p and p
.
i
j
The rate constants k12 and k21, derived from the simula-
tions of the partially relaxed spectra of the inversion recov-
ery experiments, are plotted in Fig. 9 as a function of the
reciprocal temperature. In the Arrhenius representations
discontinuities at around 300 K are observed which stem
from the fact that the experimental and theoretical T1Z val-
ues at the T1Z minimum do not match. That is, the values
of the theoretical T1Z minimum were somewhat lower than
those for the experimental T1Z minimum. The only way to

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
85
Fig. 7. Experimental and theoretical T1Z data as a function of temper-
ature for (a) a-deuterons and (b) b-deuterons in 1,6-dibromohexane/urea.
2
Fig. 8. Partially relaxed H NMR spectra from inversion recovery
experiments at 260 K (left: experiment, right: simulation) on (a) 1,6-
reproduce the experimental data was by shifting the corre-
lation times to smaller values which then gave rise to the
discontinuity in the Arrhenius plots.
dibromohexane-1,1,6,6-d /urea and (b) 1,6-dibromohexane-2,2,5,5-d4/
urea.
4
Inspection of Table 8 reveals that practically the same
kinetic parameters were achieved, independent of particu-
lar deuterated samples. In this context, it should be empha-
sized that a completely independent data analysis of both
deuterated samples was made. Hence, activation energies
Ea,12 of 20.0 kJ/mol (deuterated a-position) and 19.1 kJ/
mol (deuterated b-position) were found below the T -min-
Iðs Þ ¼ Ið0Þ exp½ꢀs =T1Qꢂ
ð9Þ
r
r
In this equation, I(s ) is the difference between the peak
r
intensities of the quadrupolar doublet components in
anti-phase at time s , and I(0) is the initial difference at
r
relaxation delay s = 0. From the experimental T data,
r
1Q
minimum values of 11.2 and 9.3 ms for the a- and b-deuter-
ated samples, respectively, were derived at 300 K.
1
Z
imum, while Ea,12 = 6.7 kJ/mol (deuterated a-position) and
1.2 kJ/mol (deuterated b-position) were derived above the
T -minimum. The difference DH of about 12 kJ/mol
As can be deduced from the representative spectra in
Fig. 11, in general a very satisfactory reproduction of the
experimental data can be achieved which also holds for
the reproduction of the experimental T1Q values, given in
Fig. 10. In fact, the T1Q analysis yielded the same simula-
tion parameters, given in Tables 6 and 7, as obtained from
the aforementioned T1Z data analysis.
1
1
Z
between the two activation energy values is somewhat high-
er than the value of about 9 kJ/mol derived from the tem-
perature dependence of population p (see above).
1
For completeness a variable temperature T1Q study was
done for the same temperature range. The experimental
T1Q curves are depicted in Fig. 10, while representative par-
tially relaxed spectra along with their theoretical counter-
parts are given in Fig. 11. The experimental T1Q data
were calculated from the intensities of the perpendicular
singularities, and were obtained by a two-parameter fit
using Eq. (9) [27].
3.4. Discussion
1
3
1
2
The present C, H and H NMR study provides a con-
sistent picture for the molecular properties of the 1,6-dib-
romohexane/urea inclusion compound. The observed

86
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
1,6-dibromohexane and the long-chain analogue 1,10-
dibromodecane. Primarily, they arise from structural differ-
ences of the urea lattices which in turn affect the guest
mobility and conformational properties.
1
3
There is strong evidence that C spin–lattice relaxation is
caused by the same motional contributions which are
2
responsible for H spin–lattice relaxation. This is born out
by the similar temperature dependence of the relaxation
curves, and the results from the semi-quantitative analysis
1
3
of the C T data providing similar motional correlation
1
Fig. 9. Arrhenius representation for rate constants of the gauche–gauche
interconversion in 1,6-dibromohexane/urea ((a) C6Br-1,6: 1,6-dibromo-
hexane-1,1,6,6-d /urea, (b) C6Br-2,5: 1,6-dibromohexane-2,2,5,5-d /urea).
4 4
1
3
1
strong alterations of the C and H chemical shifts for the
guest molecules upon enclathration clearly point to chang-
es in conformational order, although an influence by inter-
actions with the urea host lattice also plays some role. The
examination of samples with deuterated and non-deuterat-
ed urea clearly demonstrates that strong dipolar interac-
tions exist between the guest and host components which
1
is also expressed by an enhanced H spin diffusion (shorter
1
H T1q values) for the non-deuterated sample. Further-
Fig. 10. Experimental and theoretical T1Q data as a function of
temperature for (a) a-deuterons and (b) b-deuterons in 1,6-dibromohex-
ane/urea.
more, it is shown that there exist substantial differences
between the C and H NMR data for the UICs with
1
3
1
Table 8
Kinetic parameter for gauche–gauche interconversion of 1,6-dibromohexane-1,1,6,6-d
4
and 1,6-dibromohexane-2,2,5,5-d
4
in urea
1
,6-Dibromohexane-1,1,6,6-d
4
1,6-Dibromohexane-2,2,5,5-d
4
b
b
b
c
c
b
b
b
c
c
E
a,12
E
a,21
DH
A
12
ꢀ
A
21
ꢀ1
E
a,12
E
a,21
DH
A
12
ꢀ1
A
21
ꢀ1
1
(
kJ/mole)
(kJ/mole)
(kJ/mole)
(s
)
(s
)
(kJ/mole)
(kJ/mole)
(kJ/mole)
(s
)
(s
)
a
10
11
11
12
10
11
12
13
HT
LT
6.7 ± 0.3
20.0 ± 1.3
18.4 ± 0.7
32.0 ± 1.0
11.7
12
1.5 · 10
5.3 · 10
4.7 · 10
2.0 · 10
11.2 ± 2.5
19.1 ± 1.3
21.5 ± 2.8
31.0 ± 0.6
10.3
11.9
6.7 · 10
3.1 · 10
1.5 · 10
1.3 · 10
a
a
b
c
LT: temperature range below relaxation minimum; HT: temperature range above relaxation minimum.
a,ij: activation energy for jump from site j to i, DH: activation energy difference.
Pre-exponential factors.
E

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
87
earlier NMR work. However, the time-scale of this motional
process must lie in the sensitive range of spin–lattice relaxa-
tion, and not – as proposed earlier – within the time-scale of
line shape effects [7]. Therefore, considerably shorter correla-
tion times (by about one two orders of magnitude) are
obtained in the present work.
In summary, it is the commensurability of the 1,6-dib-
romohexane guest and the urea host substructures, result-
ing in a monoclinic crystal structure, which gives rise to
larger spatial constraints on the guest species and stabilizes
the gauche conformational states. In quite contrast, UICs
with the long-chain analogues C H Br (n = 10 and 11)
n
2n
2
or n-alkanes are characterized by hexagonal channels in
which the guests adopt a nearly all-trans conformational
state. The latter guests undergo almost unhindered rota-
tional motions around their long axes, while for the 1,6-
dibromohexane guests restricted overall reorientation
between two gauche conformations is registered along with
a large activation energy. The nearly identical activation
energies of Ea,12 = 20.0 kJ/mol and 19.1 kJ/mol for the a-
and b-deuterons, respectively, below the T1Z minimum
clearly demonstrate that the underlying motional process
affects the guest molecules as a whole. In quite contrast,
for n-hexadecane (12.4 kJ/mol, six-site rotational motions,
high temperature phase) [6] and n-nonadecane (8.4 kJ/mol,
rotational diffusion) [5] in urea, a low sterical hindrance for
the overall rotation is discussed. For alkanes with branched
end-groups [47], due to the spatial requirement of the larger
end groups, a slightly larger activation energy of 16.7 kJ/
mol is reported. For carboxylic acids [3] also larger activa-
tion energies are found which are attributed to both inter-
actions between adjacent guest molecules (hydrogen
bonding) and between guest species and urea host lattice.
The differences in sterical hindrance are also manifested
in the absolute rate constants. Thus, the rate constants for
guest motions for 1,6-dibromohexane/urea are found to be
several orders of magnitude smaller than those reported for
n-alkanes, branched alkanes or long-chain dibromoalkanes
incorporated in hexagonal urea channels [4–6,24,47].
In connection with the present T1Z and T1Q data analy-
sis it is worth mentioning that the anisotropy of these relax-
ation times is very useful for the discrimination between
different motional models. That is, although it might be
2
Fig. 11. Partially relaxed H NMR spectra from BBJB experiments at
2
1
60 K (left: experiment, right: simulation) on (a) 1,6-dibromohexane-
,1,6,6-d
4
4
/urea, and (b) 1,6-dibromohexane-2,2,5,5-d /urea.
2
times as derived from the quantitative H T data analysis
1
discussed next.
1
3
It should be noted that recently a comprehensive
NMR relaxation study for a series of n-alkane/UICs was
published [54]. It was shown that the C T values are almost
C
1
3
2
possible to reproduce the experimental fast exchange H
1
independent of the particular chain length, and vary with the
chain position. Both results were attributed to the overall
and internal chain mobility, as also verified, for instance,
NMR line shapes by different data sets, the corresponding
T1Z and T1Q anisotropies – as directly reflected by the line
shape alterations of the partially relaxed spectra as a func-
2
by variable temperature H NMR studies on n-alkane/UICs
tion of the relaxation interval s – strongly vary with the
r
[
4–6].
Reliable information about the guest mobility is obtained
underlying motional model. For instance, in the case of a
two-site jump model it is always possible to describe a fast
exchange spectrum by degenerate or non-degenerate jump
processes; the only difference would be the jump angle.
The corresponding anisotropies of the relaxation times,
however, show completely different behaviour and can be
clearly discriminated.
2
from the analysis of the variable temperature H NMR data.
It could be demonstrated that the experimental H NMR line
2
shapes, T1Z and T1Q data are reproducible by a single param-
eter set (quadrupolar coupling constants, jump angles, site
populations) which again emphasizes the quality of the
underlying motional model of mutual exchange between
two non-degenerate gauche states, as proposed during an
The discontinuities in the Arrhenius plots around the
minima of spin–lattice relaxation curves, resulting in two

88
X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
different slopes above and below the T1 minima, are
addressed next. As mentioned earlier, at first sight they
can be traced back to the fact that the theoretical minima
of the relaxation curves (based on the assumption of a
gauche–gauche interconversion process) are lower than
their experimental counterparts. Attempts were therefore
made to overcome these discrepancies. Hence, a distribu-
tion of correlation times was considered. Likewise, the
quadrupolar coupling constant was reduced, and the site
populations and jump angles were varied. None of these
attempts were successful, since they gave rise to consider-
able deviations between the experimental and theoretical
H NMR line shapes. Therefore, these options were not
considered any further.
On the other hand, the discontinuities in the Arrhenius
plots are surprising, since calorimetric studies do not show
any phase transition in this temperature range. However,
due to the continuous change of the urea lattice structure
lattice strain arising from the internal stress due to the
guest molecules, and gradually shifts from a monoclinic
towards hexagonal symmetry upon temperature increase
[7,23] along with enhanced fluctuations of the guest
molecules.
In summary, the correlation times for overall fluctua-
tions are about one order of magnitude smaller than the
values for gauche–gauche interconversion. The actual acti-
vation energies for the non-deuterated UICs are about
10 kJ/mol, and 20 kJ/mol for the deuterated UICs (see
Table 4). The stronger temperature dependence for the
guest fluctuations in the deuterated UIC can be explained
by weaker hydrogen bonds and a reduced host lattice
stability. As a consequence, a stronger temperature
dependence is found for the guest fluctuations.
2
2
The theoretical description of the H NMR line shapes
required a temperature dependent residual line width
0
2
1=T (see Tables 6 and 7). This line width reduction with
[
23], the guest molecules might have gained enough spatial
increasing temperature can be attributed to a reduction
of the heteronuclear dipolar interactions between guest
molecules and host matrix due to the increasing guest
mobility and/or the temperature dependent increase of
freedom that at about room temperature the gauche–
gauche interconversion is suddenly accelerated. This
assumption is supported by the observation that overall
molecular fluctuations, as reflected by the reduction of
the quadrupolar coupling constant and by the C T1q
data (see below), also play an important role at higher
temperature. Similar discontinuities for the guest overall
motions are also observed for other guest–host systems,
although a first-order phase transition could not be
observed [55].
1
3
the channel diameter (see also former discussion of
C
1
3
2
NMR line widths and H NMR relaxation data). Similar
temperature dependent changes of the residual NMR line
widths have been reported for other systems [6,24].
4. Conclusions
The gain in spatial freedom is also reflected by the lower
activation energy above the discontinuity. Likewise, there
is evidence for lateral diffusion of the guests at higher tem-
peratures which also requires sufficient spatial freedom.
The presence of lateral motion is indirectly visible by an
The structural and motional properties of 1,6-dibromo-
hexane in urea were explored by solid-state NMR spectros-
1
3
1
copy. The results from C CP/MAS and H MAS NMR
experiments are in agreement with the assumption that
the guests exist exclusively in two gauche conformational
states. The quantitative analysis of the variable tempera-
2
isotropic signal in the H NMR spectra at elevated temper-
2
atures which stems from free, non-clathrated 1,6-dibromo-
hexane escaped from the urea lattice. As mentioned earlier,
free non-clathrated 1,6-dibromohexane can be also
ture H NMR line shape experiments and spin–lattice
relaxation studies, performed on two samples with selec-
tively deuterated guest molecules, supports the proposed
non-degenerate two-site jump model which describes mutu-
al exchange between the two gauche conformers within the
urea matrix. It is concluded that the gauche–gauche
1
deduced from H MAS spectra of samples spun at 34 kHz.
1
3
C T1q relaxation is assumed to be caused by overall fluc-
tuations of the guest molecules. This assumption is support-
1
3
2
13
ed by the increase of C T1q towards the center of the alkyl
exchange process is the dominant source for H and
C
chains. That is, the fluctuations are more efficient for the
spin–lattice relaxation of 1,6-dibromohexane in urea. Anal-
ysis of the T1Z and T1Q data provided the correlation times
for this motional process. The activation energy Ea,12 of
about 20 kJ/mol for the low temperature range is rather
large, and reflects considerable spatial constraints by the
surrounding urea matrix. Above room temperature, a low-
er activation energy of about 9 kJ/mol along with a discon-
tinuous change for the mobility is observed which is
attributed to the gain in spatial freedom. Moreover, it is
1
3
chain ends giving rise to shorter C T1q values. Moreover,
2
the reproduction of the overall experimental H spectral
widths required smaller quadrupolar coupling constants
(
see Tables 6 and 7) at higher temperatures. This effect is
more pronounced at the chain ends than in the center of
the chains which again gives strong evidence that the reduc-
tion in overall spectral width stems from fast overall
fluctuations.
The amplitude of guest wobbling is extremely sensitive
to the channel diameter [56], and becomes larger if the
channel is allowed to deform. X-ray studies [23] have dem-
onstrated that the guest molecules coil into gauche confor-
mations and tilt away from the channel axis by about 12–
1
3
assumed that C T1q relaxation is caused by overall fluctu-
ations of the guest molecules which are also responsible for
the temperature dependent reduction of the overall widths
2
in the H NMR spectra. Finally, there is evidence to prove
that lateral motions of the guests also play a role at elevat-
ed temperatures.
1
4ꢁ. The net result is that the host channel experiences a

X. Yang, K. M u¨ ller / Journal of Molecular Structure 831 (2007) 75–89
89
Acknowledgements
[23] M.D. Hollingsworth, U. Werner-Zwanziger, M.E. Brown, J.D.
Chaney, J.C. Huffman, K.D.M. Harris, S.P. Smart, J. Am. Chem.
Soc. 121 (1999) 9732.
We thank Mrs. Gabrielle Br a¨ uning and Ms. Annette
Hirtler for their help during the chemical synthesis and
preparation of the samples. Financial support by the Deut-
sche Forschungsgemeinschaft and the Graduiertenkolleg
No. 448 ‘‘Modern Methods of Magnetic Resonance in
Materials Science’’ is gratefully acknowledged.
[
[
24] X. Yang, K. M u¨ ller, in preparation.
25] P. Vashishta, in: J.N. Mundy, G.K. Shenoy (Eds.), Fast Ion
Transport in Solids, North-Holland press, Amsterdam, 1979.
[26] J.D. Ferry, Viscoelastic Properties of Polymers, Wiley press, New
York, 1980.
[
27] R.W. Pastor, R.M. Venable, M. Karplus, A. Szabo, J. Chem. Phys.
9 (1988) 1128.
28] H.W. Spiess, H. Sillescu, J. Magn. Reson. 42 (1981) 381.
8
[
References
[29] H.W. Spiess, J. Chem. Phys. 72 (1980) 6755.
30] Y.D. Ronald, J. Magn. Reson. 66 (1986) 422.
[
[
[
[
[
1] M.D. Hollingsworth, A.R. Palmer, J. Am. Chem. Soc. 115 (1993)
881.
2] F. Imashiro, T. Maeda, T. Nakai, A. Saika, T. Terao, J. Phys. Chem.
0 (1986) 5498.
3] R.L. Vold, G.L. Hoatson, R. Subramanian, J. Chem. Phys. 108
1998) 7305.
4] H.L. Casal, D.G. Cameron, E.C. Kelusky, J. Chem. Phys. 80 (1984)
407.
[31] V. Kurze, B. Steinbauer, T. Huber, K. Beyer, Biophys. J. 78 (2000)
2441.
[32] G.L. Hoatson, T.Y. Tse, R.L. Vold, J. Magn. Reson. 98 (1992) 342.
[33] B. Westermark, J. Labelled Compd. Radiopharm. 25 (1988) 1381.
[34] E. Oldfield, M. Meadows, D. Rice, R. Jacobs, Biochemistry 14 (1978)
2727.
[35] D.A. Torchia, J. Magn. Reson. 30 (1978) 613.
[36] N.J. Heaton, R.R. Vold, R.L. Vold, J. Magn. Reson. 77 (1988) 572.
[37] S. Wimperis, J. Magn. Reson. 86 (1990) 46.
[38] G.L. Hoatson, J. Magn. Reson. 94 (1991) 152.
[39] J.C. Cobas, F.J. Sardina, Concepts Magn. Reson. Part A 19A (2003)
80.
5
9
(
1
[5] M.S. Greenfield, R.L. Vold, R.R. Vold, Mol. Phys. 66 (1989) 269.
[6] J. Schmider, K. M u¨ ller, J. Phys. Chem. A 102 (1998) 1181.
[7] U. Werner-Zwanziger, M.E. Brown, J.D. Chaney, E.J. Still, M.D.
Hollingsworth, Appl. Magn. Reson. 17 (1999) 265.
[
[
8] S.P. Smart, A.E. Baghdadi, F. Guillaume, K.D.M. Harris, J. Chem.
Soc., Faraday Trans. 90 (1994) 1313.
9] A.E. Baghdadi, E.J. Dufourc, F. Guillaume, J. Phys. Chem. 100
[40] J.A. Villanueva-Garibay, K. M u¨ ller, Lect. Notes Phys. 684 (2006) 65.
[41] R.J. Wittebort, E.T. Olejniczak, R.G. Griffin, J. Chem. Phys. 86
(1987) 5411.
(
1996) 1746.
[42] K. M u¨ ller, P. Meier, G. Kothe, Prog. Nucl. Magn. Reson. Spectrosc.
17 (1985) 211.
[
[
10] E. Meirovitch, I. Belsky, J. Phys. Chem. 88 (1984) 6407.
11] I. Kanesaka, S. Matsuzawa, T. Ishioka, Y. Kitagawa, K. Ohno,
Spectrochim. Acta A 60 (2004) 2621.
12] M.C. Menziani, P.G. de Benedetti, U. Segre, M. Brustolon, Mol.
Phys. 92 (1997) 903.
13] Y. Chatani, H. Anraku, Y. Taki, Mol. Cryst. Liq. Cryst. 48 (1978) 219.
14] J.L. Atwood, in: J.E.D. Davies, D.D. MacNicol (Eds.), Inclusion
Compounds, vols. 1–3, Academic Press, New York, 1984.
15] J.L. Atwood, in: J.E.D. Davies, D.D. MacNicol (Eds.), Inclusion
Compounds, vols. 4 and 5, Oxford University Press, Oxford, 1991.
16] J.L. Atwood, in: J.E.D. Davies, D.D. MacNicol, E. V o¨ gtle (Eds.),
Comprehensive Supramolecular Chemistry, Pergamon Press, Lon-
don, 1996.
[43] A.G. Redfield, Adv. Magn. Reson. 1 (1965) 1.
[44] R.J. Wittebort, A. Szabo, J. Chem. Phys. 69 (1978) 1722.
[45] D.A. Torchia, A. Szabo, J. Magn. Reson. 42 (1982) 107.
[46] M. Okazaki, C.A. McDowell, J. Mol. Struct. 118 (1984) 149.
[47] T. Handel, F. Lissner, T. Schleid, K. M u¨ ller, Appl. Magn. Reson. 27
(2004) 225.
[48] D.L. VanderHart, W.L. Earl, A.N. Garroway, J. Magn. Reson. 44
(1981) 361.
[49] W.P. Rothwell, J.S. Waugh, J. Chem. Phys. 74 (1981) 2721.
[50] J.C. Williams, A.E. McDermott, J. Phys. Chem. 97 (1993) 12393.
[51] N.J. Heaton, R.L. Vold, R.R. Vold, J. Am. Chem. Soc. 111 (1989)
3211.
[
[
[
[
[
[
[
17] J. Ripmeester, in: J.L. Atwood, J.E.D. Davies, D.D. MacNicol
[52] R.K. Harris, Nuclear Magnetic Resonance Spectroscopy, Pitman
press, London, 1983.
[53] J.P. Jacobsen, H.K. Bildsoe, K. Schaumburg, J. Magn. Reson. 23
(1976) 153.
[54] T. Nakaoki, H. Nagano, T. Yanagida, J. Mol. Struct. 699 (2004) 1.
[55] J.A. Villanueva-Garibay, K. M u¨ ller, J. Phys. Chem. B 108 (2004)
15057.
[56] R.L. Vold, R.R. Vold, N.J. Heaton, Adv. Magn. Reson. 13 (1989) 17.
[57] C. Francisco, G. Vicens, G. Angel, Synthesis 5 (1987) 511.
[58] F. Sato, S. Sato, H. Kodama, M. Sato, J. Organomet. Chem. 142
(1977) 71.
(Eds.), Inclusion Compounds, vol. 5, Oxford University Press,
Oxford, 1991, p. 37.
18] M.D. Hollingsworth, K.D.M. Harris, in: J.L. Atwood, J.E.D. Davies,
D.D. MacNicol, E. V o¨ gtle (Eds.), Comprehensive Supramolecular
Chemistry, vol. 6, Pergamon Press, London, 1996, p. 177.
19] K. Fukushima, J. Mol. Struct. 34 (1976) 67.
20] K. Fukushima, K. Sugiura, J. Mol. Struct. 41 (1977) 41.
21] D.R.F. Gilson, C.A. McDowell, Mol. Phys. 4 (1961) 125.
22] L. Elizabe, A.E. Baghdadi, S.P. Smart, F. Guillaume, K.D.M. Harris,
J. Chem. Soc., Faraday Trans. 92 (1996) 267.
[
[
[
[