This article was downloaded by: [BYU Brigham Young University]
On: 13 February 2015, At: 09:13
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Supramolecular Chemistry
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/gsch20
Binding of α,ω-alkyldiammonium ions by
cucurbit[n]urils in the gas phase
a
a
b
c
Fan Yang , Chad A. Jones , N. Selvapalam , Young Ho Ko , Kimoon Kim
bc
a
& David V. Dearden
a
Department of Chemistry and Biochemistry, Brigham Young University, C100 Benson
Science Building, Provo, UT 84602-5700, USA
b
Department of Chemistry and Division of Advanced Materials Science, Center for Smart
Supramolecules, Pohang University of Science and Technology, Pohang 790-784, Republic of
Korea
c
Click for updates
Institute for Basic Science, Center for Self-assembly and Complexity, Pohang 790-784,
Republic of Korea
Published online: 04 Jul 2014.
To cite this article: Fan Yang, Chad A. Jones, N. Selvapalam, Young Ho Ko, Kimoon Kim & David V. Dearden (2014)
Binding of α,ω-alkyldiammonium ions by cucurbit[n]urils in the gas phase, Supramolecular Chemistry, 26:9, 684-691, DOI:
10.1080/10610278.2014.930149
To link to this article: http://dx.doi.org/10.1080/10610278.2014.930149
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Supramolecular Chemistry, 2014
Vol. 26, No. 9, 684–691, http://dx.doi.org/10.1080/10610278.2014.930149
Binding of a,v-alkyldiammonium ions by cucurbit[n]urils in the gas phase
Fan Yanga1, Chad A. Jonesa, N. Selvapalamb, Young Ho Koc, Kimoon Kimb,c and David V. Deardena*
a
Department of Chemistry and Biochemistry, Brigham Young University, C100 Benson Science Building, Provo, UT 84602-5700, USA;
Department of Chemistry and Division of Advanced Materials Science, Center for Smart Supramolecules, Pohang University of Science
and Technology, Pohang 790-784, Republic of Korea; cInstitute for Basic Science, Center for Self-assembly and Complexity, Pohang
790-784, Republic of Korea
b
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
(Received 26 March 2014; accepted 27 May 2014)
We investigated gas-phase complexes of a,v-n-alkyldiammonium ions with cucurbit[5]uril (CB[5]), decamethylcucurbit[5]
uril (mc5), penta(cyclohexyl)cucurbit[5]uril (CB*[5]), hexa(cyclohexyl)cucurbit[6]uril (CB*[6]), cucurbit[7]uril (CB[7])
and cucurbit[8]uril (CB[8]) using electrospray Fourier transform ion cyclotron resonance mass spectrometry and collisioninduced dissociation techniques. The five-membered cucurbit[n]urils (CB[n]s) form singly charged 1:1 and doubly charged
2:1 diamine:CB[n] complexes. All dissociate via loss of neutral a,v-n-alkyldiamine with only weak dependence of
dissociation thresholds on chain length. For a given diamine, threshold energies are in the order CB[5] , mc5 , CB*[5].
This is consistent with guest hydrogen bonding on the portals of the CB[5]s with no threading into the host’s interior. The n
$ 6 CB[n]s form 1:1 complexes with doubly protonated a,v-n-alkyldiamines. These collisionally dissociate via four
channels: loss of singly protonated a,v-n-alkyldiammonium; fragmentation of the CB[n] cage; loss of neutral a,v-nalkyldiamine and fragmentation of the a,v-n-alkyldiamine. The dissociation threshold energies and branching ratios exhibit
strong dependence on the length of the a,v-n-alkyldiamine and the size of the CB[n]. The data suggest that the optimum a,
v-n-alkyldiamine length for binding CB*[6] is three to four methylene groups; for CB[7], four to five methylene groups and
for CB[8], five to six methylene groups, indicating an increasing tendency for the guest to span the host cavity diagonally as
the size of the CB[n] increases.
Keywords: cucurbituril; alkyldiamine; mass spectrometry; collision-induced dissociation
1.
Introduction
Host –guest interactions in supramolecules are strongly
influenced by solvent and counterion effects in condensed
phases. Gas-phase studies are ideal for investigating such
interactions because the perturbing effects of solvation are
not present and subtle, yet important differences can be
probed (1). Gas-phase experiments, primarily using mass
spectrometry and ion mobility methods, have increasingly
been used to address such issues (2 –19).
Cucurbit[n]urils (CB[n]s) (20 – 24) are pumpkinshaped macrocycles that are cyclic condensation polymers
of glycoluril and formaldehyde. The CB[n]s examined in
this study have two identical openings that are lined with
electronegative carbonyl groups that make CB[n]s ideal
for binding with positive ions. The carbonyls are also good
hydrogen bond receptors. As a result, CB[n]s are
especially good hosts for a,v-n-alkyldiammonium ions,
which, in turn, are prototypical guests.
Complexes of the hollow, pumpkin-shaped guest
cucurbit[6]uril (CB[6]) with a,v-n-alkyldiammonium ions
have previously been characterised in the gas phase as a
function of alkyl chain length using Fourier transform ion
cyclotron resonance mass spectrometry using the sustained
*Corresponding author. Email: [email protected]
q 2014 Taylor & Francis
off-resonance irradiation collision-induced dissociation
(SORI-CID) technique (3, 25). The experimental results
yielding relative dissociation energies, combined with
HF/6-31G* and B3LYP/6-31G* computational results,
indicate the diamine complexes with CB[6] are internal
pseudorotaxane complexes and four carbons is the
optimum chain length for binding of an n-alkyldiamine
by CB[6] in the gas phase. This is shorter than the
optimum chain length of six carbons measured in
aqueous formic acid solution (24, 26) and is consistent
with the fact that in solution the terminal charges of the
diammonium ion can be solvated by solvent molecules,
whereas in the gas phase they are only solvated by the CB
[n] host. A more recent study (18) examined complexes
of CB[6] and cucurbit[7]uril (CB[7]) with a series of a,
v-alkyldiammonium ions using isothermal titration
calorimetry and NMR techniques to characterise their
properties in solution, and low-energy CID, travelling
wave ion mobility spectrometry, and density functional
calculations to characterise their gas-phase behaviour.
This study verified the earlier conclusion that four
carbons are the optimum n-dialkylammonium chain
length for binding CB[6] in the gas phase.
Supramolecular Chemistry
O
N
N
O
H2
C
N
N
N
N
were purchased from Airgas (Radnor, PA, USA) to be used
as collision gases.
H2
C
R
R
N
N
O
CB[n]
C
H2
n
O
C
H2
n
CB*[n]
Figure 1. CB[n] hosts examined in this study. R ¼ CH3 for
mc5; R ¼ H otherwise.
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
685
In this paper, we use SORI techniques to examine
other CB[n] hosts with the same a,v-alkyldiammonium
(H3Nþ(CH2)xNHþ
3 , x ¼ 2 – 10) guests. The goal of this
work was to elucidate the effects of CB ring size and
peripheral substitution on the binding of alkylammonium
ions in the gas phase, because prior work has shown that
even small changes can sometimes have surprisingly large
effects. This work can be directly compared with the
results of computational studies, which are normally
carried out under gas-phase conditions, and it also
provides a benchmark for comparison with the much
more common work done in solution. Such comparisons
help reveal the role of solvation and are important because
these systems tend to behave differently in the gas phase
than in solution (18, 25). Cucurbit[5]uril (CB[5]),
decamethylcucurbit[5]uril (mc5), penta(cyclohexyl)cucurbit[5]uril (CB*[5]), hexa(cyclohexyl)cucurbit[6]uril (CB*
[6]), CB[7] and cucurbit[8]uril (CB[8]) are examined; the
relevant structures are shown in Figure 1. This affords the
opportunity to compare several hosts (the three n ¼ 5
hosts) that have portals too small to bind the guests
internally with larger hosts (the three with n . 5) that have
the potential to form inclusion complexes as does CB[6].
This set of hosts allows us to examine the interactions
between alkyldiamines and neutral CB[n]s, and specifically to examine chain length selectivity in larger CB[n]s.
2. Experimental
2.1 Materials
n-Alkyldiamines (x ¼ 2 – 10, where x is the number of
carbon atoms in the alkyl chain) and CB[5] were
purchased from Sigma-Aldrich (St Louis, MO, USA)
and used without further purification. mc5 was synthesised
by Dr Krzystof Krakowiak using published procedures
(27). CB*[5], CB*[6], CB[7] and CB[8] were synthesised
and provided by Dr Kimoon Kim (28). Stock solutions of
CB[n] were prepared at about 200 mM by dissolving the
solid samples in 88% formic acid (Mallinckrodt Baker
Inc., Phillipsburg, NJ, USA) and then diluting with
methanol:water (50:50). The electrospray solutions were
prepared at concentrations of 100 mM in CB[n] and
200 mM in amines. Argon (99.995%) and SF6 (99.8%)
2.2 Instrumentation
All experiments were carried out using a Bruker model
APEX 47e Fourier transform ion cyclotron resonance
mass spectrometer controlled by a MIDAS/PREDATOR
data system (29, 30) and equipped with an infinity cell and
a 4.7-T superconducting magnet. Ions were generated in a
microelectrospray source modified from an Analytica
(Branford, MA, USA) design, with a heated metal
capillary drying tube based on the design of Wigger
et al. (31).
2.2.1
SORI-CID experiments
The stored waveform inverse Fourier transform method
was used to isolate the ion of interest (32). SORI-CID
experiments were carried out by irradiating 1 kHz below the
resonant frequency of the target ion (33). Argon was
introduced as the collision gas using a Freiser-type pulsed
leak valve (34). SORI events included pulsing the Ar
background pressure in the trapping cell up to 1.7 £ 1025
mbar, waiting 2 s for conditions to fully stabilise and
applying the off-resonance irradiation for a variable amount
of time (which allows variation of the total energy deposited
in the ions; herein, times ranged from 10 to about 500 ms,
with amplitudes kept constant at 11.4 V and durations kept
short to minimise radiative cooling between collisions),
followed by a 5-s delay to allow the trapping cell to return to
baseline pressure (about 1029 mbar) prior to detection. Five
scans were averaged for each SORI excitation duration.
2.3
Data analysis
Transient signals were analysed using the Igor Pro
software package (version 6; Wavemetrics, Lake Oswego,
OR, USA). For SORI experiments, the Igor program was
used to extract peak amplitudes for a set of spectra that
differ in one or more experimental parameters and then
generate tables of peak intensities as a function of SORI
excitation duration. The resulting parent and product ion
peak intensities were normalised, and the relative SORI
collision energy was scaled to account for differences in
mass and excitation amplitude (although most experiments
were conducted while maintaining a constant excitation
amplitude). Relative SORI energies were calculated as
described previously (25); relative collision cross sections
for these calculations were determined using the cross
sectional areas by Fourier transform ion cyclotron
resonance (CRAFTI) technique (35). Energies obtained
from these experiments may be compared qualitatively,
but are not quantitative due to uncertainties about the
686
F. Yang et al.
absolute kinetic energies of the colliding ions as well as
uncertainties about the efficiency of kinetic-to-internal
energy conversion.
3.
Results and discussion
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
3.1 Electrospray of CB[n]s with a,v-n-alkyldiamine
complexes
Solutions of a,v-n-alkyldiammonium complexes with CB
[5], mc5 and CB*[5] were electrosprayed, and Fourier
transform ion cyclotron resonance mass spectrometry was
carried out. Doubly charged ions with 2:1 guest:host
stoichiometry ([CB[n] þ 2H2N(CH2)xNH3]2þ) and 1:1
singly charged ions ([CB[n] þ H2N(CH2)xNH3]þ) were
observed in the mass spectra for these solutions.
Solutions of the various guests with CB*[6], CB[7] and
CB[8] were also electrosprayed. Except for the mass spectra
for 1,2-ethanediamine and 1,3-propanediamine with CB[8]
(in which peaks from complexes were not consistently
observed), all the mass spectra involving these larger hosts
have one point in common: only one main peak was
observed in each spectrum, corresponding to one doubly
protonated diamine cation attached to the CB (a 1:1
complex). Considering the similarity between CB[6] and
these larger CB[n]s, we assume that these 1:1 complexes are
also pseudorotaxanes (written as [H3Nþ(CH2)xNHþ
3 ]@CB
[n]), like the CB[6] complexes studied previously (3, 25); in
fact, ion mobility data strongly suggest that both the CB[6]
and the CB[7] complexes have threaded structures (18). The
difficulty in observing complex peaks when spraying
solutions of ethane- or propanediamine with CB[8] suggests
that complexes of these shorter guests with CB[8] have low
stability as a result of the size mismatch between the bigger
cavity and wider portal of CB[8] and the shorter chain
lengths of these guests.
3.2 Characteristic dissociation channels for SORI-CID
of CB[n] complexes
To better understand the structures of these complexes,
SORI-CID experiments were carried out on the ions of the
various complexes; the average energy deposited was
varied by varying the SORI duration while holding the
amplitude constant.
3.2.1 Dissociation behaviour of singly charged 1:1 a,vn-alkyldiamine complexes of CB[n]s
Except for 1,4-butanediamine with CB[5], the dissociation
fragments observed for CB[5], mc5 and CB*[5] 1:1
complexes are all the same type, yielding protonated CB
[5] and hydronium complexes of CB[5]. An example is
given in Figure 2, where the relative signal abundance of
the parent ion and product ions are plotted against the
E
Figure 2. (Colour online) Typical SORI dissociation curve for
singly protonated 1:1 CB[5]:a,v-n-alkyldiamine complexes.
These data are for the complex of CB[5] with a,vpentanediamine.
relative SORI dissociation energies. The [CB[5] þ H2N
(CH2)4NH3]þ complex yields one additional minor
dissociation product corresponding to loss of neutral
ammonia, [CB[5] þ H2N(CH2)4NH3 – NH3]þ.
The data suggest only one primary dissociation
pathway for these 1:1 complexes, loss of the neutral
diamine with proton transfer to the CB. At first glance, the
hydronium complexes are surprising as no primary
dissociation process could produce them. Indeed, we
believe that they are not primary dissociation products, but
rather are secondary products arising from reaction of the
protonated primary fragments with background water
vapour in the mass spectrometer. Because the SORI events
are relatively long (from tens to hundreds of milliseconds,
typically), there are ample opportunities for reaction with
background water during these events and the subsequent
5-s delay period before the ion population is measured in
each experiment. As H3Oþ has been proposed as a
template ion in the formation of CB[5] (36), [CB[n]
þ H3O]þ might be expected to form easily. Similarly, the
importance of water complexes in gas-phase CB[6]
chemistry has previously been discussed (18). Interestingly, in contrast to the alkyldiamine complexes with CB
[6] (3), no cage fragmentation is observed for the 1:1
complexes of the CB[5] hosts. We conclude that these
singly charged complexes involve external binding of the
diamine guest, with the protonated ammonium group
attached to the carbonyl groups of the CB rim while the
remainder of the molecule stays outside the CB[5] cage.
It can be shown that the average energy deposited in a
SORI event is directly proportional to the duration of the
event and to the collision cross section, and also depends
on the masses of the neutral collision gas and the ion (25).
The relative energies for 50% loss of the parent ion,
ESORI,50, were determined taking these factors into
account and are shown in Figure 3. The ESORI,50 values,
which are an indicator of the overall relative stability of
E
Supramolecular Chemistry
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
x
x
Figure 3. (Colour online) Relative SORI energies for 50%
þ
survival of [H2N(CH2)xNHþ
3 ]·CB[5], [H2N(CH2)xNH3 ]·mc5 and
[H2N(CH2)xNHþ
]·CB*[5]
complexes,
x
¼
2
–
10.
Error bars
3
represent standard errors from the linear fitting procedure used
to derive the value.
each complex, are similar for all complexes with the same
CB[5] host; no obvious trends are evident as the length of
the guest alkyl chain varies. This suggests that all these 1:1
singly charged complexes have similar dissociation
energies, as expected if the guests are all bound externally,
because each would involve similar hydrogen bonding to
the five carbonyl oxygen atoms of the portal.
Comparing CB[5] complexes with the same guest
cation, the ESORI,50 values all follow the same trend: CB[5]
, mc5 # CB*[5]. Two factors likely account for this. The
first is the relative polarisabilities of the hosts. Polarisability plays an important role in determining the strength
of ion –neutral interactions (37), and the polarisabilities of
the neutral CB[n]s go up with increasing number and size
of substituents (38). For instance, the relative polarisabilities of CB[5], mc5 and CB*[5] can be estimated using
the method of Miller and Savchik (38); they are 54, 73 and
87 Å3, respectively, in the same order as the observed
ESORI,50 values. Therefore, the larger mc5 and CB*[5]
complexes have greater intrinsic dissociation energies than
the smaller corresponding CB[5] complexes.
The second factor is related to the dissociation kinetics
for the complexes. The more highly substituted hosts have
more internal degrees of freedom in which energy
deposited by collisions may be distributed prior to
dissociation. This causes the dissociation rates of the
larger complexes to be slower, meaning that more energy
may be required to facilitate dissociation at the same rates
as observed for the less-substituted hosts (39). Because the
time scale for dissociation in SORI is quite long (several
seconds), we do not expect these degree-of-freedom
effects to be large; the polarisability effects probably
dominate and the observed CB[5] , mc5 # CB*[5]
trends in ESORI,50 probably reflect real differences in
intrinsic binding strengths.
687
3.2.2 Dissociation behaviour of doubly charged 2:1 a,vn-alkyldiamine complexes of CB[5]s
As for the 1:1 complexes, SORI-CID experiments were
carried out on the 2:1 alkyldiamine complexes of CB[5],
mc5 and CB*[5] with a constant SORI excitation
amplitude. A typical plot of normalised parent and
product ion signals in the dissociation of complexes of two
singly protonated diamines with one CB[5] host is shown
in Figure 4. These 2:1 complexes all have the same three
types of fragments: singly charged 1:1 complexes,
protonated hosts and hydronium complexes of the hosts.
Therefore, we observe two dissociation pathways for the
2:1 complexes: loss of one protonated diamine cation and
loss of an additional neutral amine. Just as with the 1:1
complexes, we observe no cage fragmentation for the 2:1
complexes. This again suggests that the alkyldiammonium
is too large to thread through the portals of CB[5] or its
substituted analogues. Rather, the ammonium cations bind
with the ammonium groups hydrogen bonded to the CB
rims without including the rest of either molecule within
the CB cavity. This conclusion is consistent with the
observed small variation in binding strengths as alkyldiamine chain length varies for complexes with CB[5] and its
analogues (Figure 3): the binding strengths are determined
primarily by the hydrogen bonding interactions, which are
similar regardless of chain length. The observed
dissociation behaviour of the 2:1 complexes also provides
one more evidence for the external structure of the 1:1
complexes: if the 1:1 complexes were inclusion complexes, dissociation of the 2:1 complexes should result in
doubly charged 1:1 complexes with the loss of a neutral
diamine, not the observed loss of a protonated diamine (3).
The average SORI energies required for 50% loss of
the various parent 2:1 complex ions as the guest chain
length is varied are plotted together in Figure 5. For
complexes with the same hosts, values of the disappearance energies do not vary greatly with diammonium chain
E
Figure 4. (Colour online) Typical SORI dissociation curve for
doubly protonated 2:1 a,v-n-alkyldiamine:CB[5] complexes.
These data are for the 2:1 complex of a,v-pentanediamine with
CB[5].
688
F. Yang et al.
x
x
E
x
x
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
x
x
x
Figure 5. (Colour online) Relative SORI energies for 50%
survival of [CB[5] (H2N(CH2)xNH2)2 þ 2H]2þ, [mc5·(H2N
(CH2)xNH2)2 þ 2H]2þ and [CB*[5] (H2N(CH2)xNH2)2 þ 2H]2þ
complexes, x ¼ 2– 10. Error bars represent standard errors from
the linear fitting procedure used to derive the value.
length. For complexes with the same diammonium cation,
values of disappearance energies follow a consistent trend:
CB[5] , mc5 , CB*[5], the same as for the 1:1
complexes, with the same explanation. Comparison of
Figure 5 with Figure 3 reveals that the energies required to
dissociate the 1:1 complexes shown in Figure 3 are
consistently larger than those required to dissociate the 2:1
complexes of Figure 5. This is consistent with the idea that
Coulombic repulsion plays a significant role in lowering
the intrinsic stability of the doubly charged 2:1 complexes,
as has been seen previously for 2:1 alkali cation:CB[5]
complexes (40).
3.2.3 Dissociation behaviour of alkyldiammonium
complexes of CB*[6]
Four general fragmentation pathways were observed in the
dissociations of the [H3Nþ(CH2)xNHþ
3 ]@CB*[6] complexes. First, we note proton transfer dissociation, in which
the main products are protonated CB*[6] and singly
charged diamine (due to mass discrimination in the
instrument, the relative abundance of H2N(CH2)xNHþ
3 is
much lower than that of CB*[6]Hþ) and diamine
fragments. Second, we observe cage fragmentation; this
includes both singly and doubly charged fragments of the
CB[n] cage. Third, we observe loss of the neutral diamine,
resulting in doubly charged CB*[6]. Finally, we observe
losses that appear to originate from the neutral diamines,
resulting in doubly charged ions of the remaining part of
the diamine with intact CB*[6]. These included losses
of neutrals with the formulae CyH2yNH2 as well as losses
of neutral ammonia. The fraction of fragmentations via
each of these four channels, at average kinetic energies
sufficient to cause the complete disappearance of each
Figure 6. (Colour online) Fragmentation channels for
[H3Nþ(CH2)xNHþ
3 ]@CB*[6] ions as a function of guest alkyl
chain length.
parent complex ion, is plotted as a function of number of
carbon atoms in the n-alkyldiamine chain in Figure 6.
For x ¼ 2 – 4, cage fragmentation is the principal
dissociation channel, with other channels accounting for
less than about 20% of the dissociation products. For x ¼ 5,
the relative importance of loss of the neutral diamine
(resulting in the [CB*[6] þ 2H]2þ product) becomes
significant, and this becomes the dominant channel for
x ¼ 6 –7. For x ¼ 7 – 10, loss of neutral pieces of diamine
become more abundant, becoming the dominant channel
for x ¼ 8 and longer. It is likely that these larger diamines
are too long for optimal binding with both rims of the CB
cage. Furthermore, the longer diamines likely protrude
from the CB*[6] cavity and are therefore less protected
from collisional cleavage.
3.2.4 Dissociation behaviour of alkyldiammonium
complexes of CB[7]
The same four dissociation channels noted for CB*[6] are
observed for the CB[7] complexes: proton transfer
dissociation, cage fragmentation, loss of neutral amines
and loss of neutral fragments of the amines. The relative
abundances of the various dissociation channels are
plotted in Figure 7.
Proton transfer dissociation and cage fragmentation are
observed for every diamine complex with CB[7], and the
relative abundances of these two channels appear to mirror
each other as a function of alkyl chain length. Proton
transfer is the most abundant channel for x ¼ 2, decreases
in relative abundance for x ¼ 3 –5, then increases again for
x ¼ 6 –8 and remains the most abundant channel for
x ¼ 9 –10. Cage fragmentation increases in abundance
from x ¼ 2 to a maximum of over 80% for x ¼ 5, then falls
steadily from x ¼ 6– 10, with , 10% abundance for x ¼ 9–
10. The prominence of the cage fragmentation channel for
x ¼ 4 –5 suggests that this chain length is close to the
optimum for hydrogen bonding between the protonated
Supramolecular Chemistry
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
x
x
Figure 7. (Colour online) Fragmentation channels for
[H3Nþ(CH2)xNHþ
3 ]@CB[7] ions as a function of guest alkyl
chain length.
diamine and the two opposite rims of CB[7]. Interestingly,
this maximum in the cage fragmentation channel occurs at a
larger value of x for CB[7] (x ¼ 5) than for the smaller CB*
[6] (x ¼ 4). The other two channels, loss of neutral diamine
and loss of neutral diamine fragments, only become
important for the longer chain lengths and are much less
prominent than was observed for the CB*[6] complexes.
This likely reflects the larger size of CB[7] and a greater
tendency to include the alkyldiamine guests than is possible
for the smaller CB*[6].
3.2.5 Dissociation behaviour of alkyldiammonium
complexes of CB[8]
SORI-CID experiments similar to those done for the CB
[7] complexes were carried out on the diamine complexes
with CB[8] (Figure 8). Dissociation of these CB[8]
complexes also follows four pathways, analogous to those
observed for the CB[7] complexes. Among the four
channels, proton transfer dissociation and cage fragmenta-
689
tion are the most abundant for the CB[8] complexes, as
they are for the CB[7] complexes. As was noted for the CB
[7] complexes, the patterns in proton transfer dissociation
and cage fragmentation roughly mirror each other. The
greatest abundances for cage fragmentation are observed
for x ¼ 5– 6. As chain lengths increase beyond that, proton
transfer, loss of neutral diamine and loss of neutral
diamine fragments all grow in prominence. The relative
abundance of the diamine fragmentation channel is greater
for the CB[8] complexes than for the corresponding CB[7]
complexes; perhaps the larger portal of CB[8] leaves the
diamine guests less protected from collisional fragmentation than in the CB[7] complexes. To sum up, although
cage fragmentation is predominant for x ¼ 4– 7, based on
the fraction of proton transfer dissociation, the x ¼ 5 and 6
diammonium ions appear to have the highest binding
affinity with CB[8].
3.3 Relative dissociation energetics of
alkyldiammonium complexes of CB*[6],
CB[7] and CB[8]
As we did with the CB[5] complexes, we can measure the
average SORI energy required to dissociate 50% of each of
the complex ions formed from the larger CB[n]s. These
values are plotted in Figure 9, along with data from Ref.
(7) for complexes of unsubstituted CB[6].
The energy required to dissociate complexes of these
four CB[n]s is much greater than was needed for
dissociation of either the 1:1 singly charged or 2:1 doubly
charged complexes of the CB[5]s. This is likely due to two
factors. First, the larger CB[n]s have more internal degrees
of freedom and would therefore be expected to require
larger internal energies to dissociate on the same time
x
x
x
Figure 8. (Colour online) Fragmentation channels for
[H3Nþ(CH2)xNHþ
3 ]@CB[8] ions as a function of guest alkyl
chain length.
x
Figure 9. (Colour online) Relative SORI energies for 50%
survival of [H3Nþ(CH2)xNHþ
3 ]@CB[6] (from Ref. (7)),
þ
þ
[H3Nþ(CH2)xNHþ
3 ]@CB*[6], [H3N (CH2)xNH3 ]@CB[7] and
[H3Nþ(CH2)xNHþ
]@CB[8]
complexes,
x
¼
2–10.
Error bars
3
represent standard errors from the linear fitting procedure used to
derive the value. Relative values within a series with a given host are
accurate, but comparisons between different hosts are approximate.
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
690
F. Yang et al.
scales as the smaller CB[5]s (39). However, we believe
that the second factor is probably more important: the CB
[5]s form externally bound complexes, whereas the data
suggest that the n . 5 CB[n]s form pseudorotaxanes with
alkyldiamines. Thus, in the CB[5] complexes, only one
end of each alkyldiamine guest is hydrogen bonded to the
CB[5] rim, whereas for n . 5, both ends of each guest are
bound (as long as the guest is sufficiently long to allow
this). In fact, the dissociation energies for the shortest
guests bound to the n . 5 CB[n]s approach the energies
observed for the CB[5] complexes, which is expected if the
guest is too short to bind simultaneously to both rims.
The variation in relative dissociation energies with
alkyldiamine chain length shows how the stabilities of the
various complexes change as the chain length increases.
The data indicate that in cases where the diamine can
thread through the CB[n], the host exhibits selectivity for
certain n-alkyldiamine chain lengths. The [H3Nþ(CH2)xNHþ
3 ]@CB*[6] complexes have the highest average
dissociation energies for x ¼ 3– 4 and the dissociation
energy peaks sharply (Figure 9), whereas the maximum for
[H3Nþ(CH2)xNHþ
3 ]@CB[7] occurs for x ¼ 4 –5, peaking
less sharply, and the maximum for [H3Nþ(CH2)xNHþ
3]
@CB[8] occurs for x ¼ 5– 6, with the weakest chain
length dependence. This suggests that the optimum
alkyldiamine chain length for binding CB*[6] is 3– 4
carbon atoms (interestingly, this is shorter than the
optimum of 4 carbon atoms observed for CB[6]) (25). The
slightly shorter optimum observed for CB*[6] may reflect
greater rigidity and stronger size selectivity in this host
than in unsubstituted CB[6]. The optimum alklydiamine
chain length for CB[7] is 4– 5 carbon atoms, and for CB
[8], it is 5– 6 carbon atoms. Because all CB[n]s have the
same height, this increase in optimum chain length must
reflect increasing ability for the alkyldiamine to pass
through the CB[n] portal at greater angles to the CB[n]
molecular axis, ‘diagonally’ through the cavity, as n
increases.
It is interesting to compare the dependence of average
dissociation threshold energies on chain length (Figure 9)
with the tendency for the complex to undergo cage
fragmentation (Figure 10). In general, the highest average
dissociation threshold energies correspond to greatest
abundances in the cage fragmentation channels. This
again suggests that these complexes are strongly bound
pseudorotaxanes in the gas phase; removal of the guest
generally requires breaking covalent bonds in the CB[n]
cage, whereas for the externally bound CB[5]s, lower
energies are needed to remove the guest and the only
important dissociation channel is loss of the neutral
alkyldiamine. One interesting exception to this trend is the
1:1 CB*[6] complex with ethylenediammonium. Despite
the relatively low ESORI,50 value for this complex
(Figure 9), cage fragmentation accounts for about 75%
of its collisional dissociation products (Figures 6 and 10).
x
x
Figure 10. (Colour online) Fractional abundance of the cage
fragmentation dissociation channel for [H3Nþ(CH2)xNHþ
3 ]@CB
[n] as a function of chain length for CB*[6], CB[7] and CB[8].
The low ESORI,50 value is easily rationalised in terms of the
short chain length of ethylenediamine, too short to span
the distance between the two CB*[6] rims. The high
incidence of cage fragmentation may occur because the
other dissociation channels are not readily accessible.
Proton transfer dissociation is not abundant for any of the
CB*[6] complexes (Figure 6), suggesting that the proton
affinity of CB*[6] is relatively low compared with the
diamines. If so, this would also disfavour loss of neutral
ethylenediamine. That would leave fragmentation of the
diamine as the only other dissociation channel, but this
process is only observed for longer diamines in all the CB
[n] complexes.
4.
Conclusions
Complexation of a,v-alkyl diammonium ions with CB[5],
mc5, CB*[5], CB*[6], CB[7] and CB[8] in the gas phase
was studied via SORI-CID techniques. The similarity in
dissociation thresholds as guest chain length varies and the
preponderance of simple diamine losses suggest that the
CB[5] complexes all involve external binding of the
diamine in the gas phase. On the other hand, observation of
extensive cage fragmentation and strong chain length
dependences in both the dissociation threshold energies
and the fragmentation branching ratios indicate the
complexes of CB*[6], CB[7] and CB[8] are pseudorotaxanes in the gas phase, with the alkyl chain threading
through the CB cavity. Although all CB[n]s have the same
intrinsic height, for CB[7] and CB[8] it is possible for the
guest to span the CB[n] cavity diagonally, so these larger
CB[n]s can accommodate longer alkyldiamine guests. The
patterns in both average dissociation energies and
dissociation channel branching ratios suggest that for
CB*[6] the optimal chain length is x ¼ 3 –4; for CB[7],
x ¼ 4 –5; and for CB[8], x ¼ 5– 6, where x is the number
of carbon atoms in the n-alkyl chain.
Supramolecular Chemistry
Funding
This work was supported by the U.S. National Science
Foundation [grant number CHE-0957757] and the Institute for
Basic Science (IBS) [grant number CA1303].
Note
1.
Present address: Dalian Institute of Chemical Physics,
Chinese Academy of Sciences, Dalian 116023, China.
Downloaded by [BYU Brigham Young University] at 09:13 13 February 2015
References
(1) Steed, J.W.; Atwood, J.L. Supramolecular Chemistry; 2nd
ed. Wiley: Chichester, 2009.
(2) Kellersberger, K.A.; Anderson, J.D.; Ward, S.M.; Krakowiak, K.E.; Dearden, D.V. J. Am. Chem. Soc. 2001, 123,
11316– 11317.
(3) Zhang, H.; Paulsen, E.S.; Walker, K.A.; Krakowiak, K.E.;
Dearden, D.V. J. Am. Chem. Soc. 2003, 125, 9284– 9285.
(4) Osaka, I.; Kondou, M.; Selvapalam, N.; Samal, S.; Kim, K.;
Rekharsky, M.V.; Inoue, Y.; Arakawa, R. J. Mass
Spectrom. 2006, 41, 202– 207.
(5) Mitkina, T.; Fedin, V.P.; Llusar, R.; Sorribes, I.; Vicent, C.
J. Am. Soc. Mass Spectrom. 2007, 18, 1863– 1872.
(6) Dearden, D.V.; Ferrell, T.A.; Asplund, M.C.; Zilch, L.W.;
Julian, R.R.; Jarrold, M.F. J. Phys. Chem. A 2009, 113,
989– 997.
(7) Zhang, H.; Grabenauer, M.; Bowers, M.T.; Dearden, D.V.
J. Phys. Chem. A 2009, 113, 1508– 1517.
(8) Deroo, S.; Rauwald, U.; Robinson, C.V.; Scherman, O.A.
Chem. Commun. 2009, 644– 646.
(9) Rauwald, U.; Biedermann, F.; Deroo, S.; Robinson, C.V.;
Scherman, O.A. J. Phys. Chem. B 2010, 114, 8606– 8615.
(10) Heo, S.W.; Choi, T.S.; Park, K.M.; Ko, Y.H.; Kim, S.B.;
Kim, K.; Kim, H.I. Anal. Chem. 2011, 83, 7916– 7923.
(11) Yang, F.; Dearden, D.V. Supramol. Chem. 2011, 23,
53 –58.
(12) Yang, F.; Dearden, D.V. Isr. J. Chem. 2011, 51, 551– 558.
(13) Cziferszky, M.; Biedermann, F.; Kalberer, M.; Scherman,
O.A. Org. Biomol. Chem. 2012, 10, 2447– 2452.
(14) Černochová, J.; Branná, P.; Rouchal, M.; Kulhánek, P.;
Kuřitka, I.; Vı́cha, R. Chem. Eur. J. 2012, 18,
13633– 13637.
(15) Choi, T.S.; Ko, J.Y.; Heo, S.W.; Ko, Y.H.; Kim, K.; Kim,
H.I. J. Am. Soc. Mass Spectrom. 2012, 23, 1786– 1793.
(16) Lee, T.-C.; Kalenius, E.; Lazar, A.I.; Assaf, K.I.; Kuhnert,
N.; Grün, C.H.; Jänis, J.; Scherman, O.A.; Nau, W.M. Nat.
Chem. 2013, 5, 376– 382.
(17) Lee, J.W.; Heo, S.W.; Lee, S.J.C.; Ko, J.Y.; Kim, H.; Kim,
H.I. J. Am. Soc. Mass Spectrom. 2013, 24, 21 –29.
691
(18) Lee, S.J.C.; Lee, J.W.; Lee, H.H.; Seo, J.; Noh, D.H.; Ko, Y.
H.; Kim, K.; Kim, H.I. J. Phys. Chem. B 2013, 117,
8855 –8864.
(19) Noh, D.H.; Lee, S.J.C.; Lee, J.W.; Kim, H.I. J. Am. Soc.
Mass Spectrom. 2014, 25, 410– 421.
(20) Freeman, W.A.; Mock, W.L.; Shih, N.-Y. J. Am. Chem. Soc.
1981, 103, 7367– 7368.
(21) Lagona, J.; Mukhopadhyay, P.; Chakrabarti, S.; Isaacs, L.
Angew. Chem. Int. Ed. 2005, 44, 4844– 4870.
(22) Kim, K.; Selvapalam, N.; Ko, Y.H.; Park, K.M.; Kim, D.;
Kim, J. Chem. Soc. Rev. 2007, 36, 267– 279.
(23) Isaacs, L. Chem. Commun. 2009, 619– 629.
(24) Mock, W.L. In Comprehensive Supramolecular Chemistry;
Vögtle, F., Ed.; Elsevier: New York, 1996; Vol. 2 pp
477– 493.
(25) Zhang, H.; Ferrell, T.A.; Asplund, M.C.; Dearden, D.V. Int.
J. Mass Spectrom. 2007, 265, 187– 196.
(26) Mock, W.L.; Shih, N.-Y. J. Org. Chem. 1986, 51,
4440– 4446.
(27) Zhang, X.X.; Krakowiak, K.E.; Xue, G.; Bradshaw, J.S.;
Izatt, R.M. Ind. Eng. Chem. Res. 2000, 39, 3516– 3520.
(28) Zhao, J.; Kim, H.-J.; Oh, J.; Kim, S.-Y.; Lee, J.W.;
Sakamoto, S.; Yamaguchi, K.; Kim, K. Angew. Chem. Int.
Ed. 2001, 40, 4233– 4235.
(29) Senko, M.W.; Canterbury, J.D.; Guan, S.; Marshall, A.G.
Rapid Commun. Mass Spectrom. 1996, 10, 1839– 1844.
(30) Blakney, G.T.; Hendrickson, C.L.; Quinn, J.P.; Marshall, A.
G. New Hardware for Ultrahigh Resolution and/or DataDependent SWIFT Ion Isolation in an FT-ICR Mass
Spectrometer, Presented at the 56th ASMS Conference on
Mass Spectrometry and Allied Topics, Denver, CO, June 1 –
5 2008.
(31) Wigger, M.; Nawrocki, J.P.; Watson, C.H.; Eyler, J.R.;
Benner, S.A. Rapid Commun. Mass Spectrom. 1997, 11,
1749 –1752.
(32) Chen, L.; Wang, T.-C.L.; Ricca, T.L.; Marshall, A.G. Anal.
Chem. 1987, 59, 449– 454.
(33) Gauthier, J.W.; Trautman, T.R.; Jacobson, D.B. Anal.
Chim. Acta 1991, 246, 211– 225.
(34) Jiao, C.Q.; Ranatunga, D.R.A.; Vaughn, W.E.; Freiser, B.S.
J. Am. Soc. Mass Spectrom. 1996, 7, 118– 122.
(35) Yang, F.; Voelkel, J.; Dearden, D.V. Anal. Chem. 2012, 84,
4851 –4857.
(36) Oh, K.S.; Yoon, J.; Kim, K.S. J. Phys. Chem. B 2001, 105,
9726 –9731.
(37) Su, T.; Bowers, M.T. In Bowers Gas Phase Ion Chemistry;
Bowers, M.T., Ed.; Academic: New York, 1979; Vol. 1,
pp 83 – 118.
(38) Miller, K.J.; Savchik, J.A. J. Am. Chem. Soc. 1979, 101,
7206 –7213.
(39) Robinson, J.P.; Holbrook, K.A. Unimolecular Reactions;
Wiley: New York, 1972.
(40) Mortensen, D.N.; Dearden, D.V. Chem. Commun. 2011, 47,
6081 –6083.