Evaluating the effects of carbon nanoreactor diameter and internal structure on the pathways of the catalytic hydrosilylation reaction

Article abstract of DOI:10.1002/smll.201302732

Three different types of carbon nanoreactors, double-walled nanotubes (DWNT), multi-walled nanotubes (MWNT) and graphitised carbon nanofibers (GNF) have been appraised for the first time as containers for the reactions of phenylacetylene hydrosilylation catalysed by a confined molecular catalyst [Rh₄(CO)₁₂]. Interactions of [Rh₄(CO) ₁₂] with carbon nanoreactors determining the ratio of β-addition products are unchanged for all nanoreactors and are virtually unaffected by the confinement of [Rh₄(CO)₁₂] inside carbon nanostructures. Conversely, the relative concentrations of reactants affecting the ratio of addition and dehydrogenative silylation products is very sensitive to nanoscale confinement, with all nanoreactors demonstrating significant effects on the distribution of reaction products as compared to control experiments with the catalyst in bulk solution or adsorbed on the outer surface of nanoreactors. Surprisingly, the widest nanoreactors (GNF) change the reaction pathway most significantly, which is attributed to the graphitic step-edges inside GNF providing effective anchoring points for the catalyst and creating local environments with greatly altered concentrations of reactants as compared to bulk solution. Possessing diameters significantly wider than molecules, GNF impose no restrictions on the transfer of reactants while providing the strongest confinement effects for the reaction. Furthermore, GNF facilitate the effective recyclability of the catalyst and thus represents a superior nanoreactor system to carbon nanotubes.

Full text of DOI:10.1002/smll.201302732

full papers
Nanoreactors
Evaluating the Effects of Carbon Nanoreactor Diameter
and Internal Structure on the Pathways of the Catalytic
Hydrosilylation Reaction
William A. Solomonsz, Graham A. Rance,* and Andrei N. Khlobystov*
T hree different types of carbon nanoreactors, double-walled nanotubes (DWNT),
multi-walled nanotubes (MWNT) and graphitised carbon nanofibers (GNF) have
been appraised for the first time as containers for the reactions of phenylacetylene
hydrosilylation catalysed by a confined molecular catalyst [Rh4(CO)12]. Interactions
of [Rh4(CO)12] with carbon nanoreactors determining the ratio of β-addition
products are unchanged for all nanoreactors and are virtually unaffected by the
confinement of [Rh4(CO)12] inside carbon nanostructures. Conversely, the relative
concentrations of reactants affecting the ratio of addition and dehydrogenative
silylation products is very sensitive to nanoscale confinement, with all nanoreactors
demonstrating significant effects on the distribution of reaction products as compared
to control experiments with the catalyst in bulk solution or adsorbed on the outer
surface of nanoreactors. Surprisingly, the widest nanoreactors (GNF) change the
reaction pathway most significantly, which is attributed to the graphitic step-edges
inside GNF providing effective anchoring points for the catalyst and creating local
environments with greatly altered concentrations of reactants as compared to bulk
solution. Possessing diameters significantly wider than molecules, GNF impose no
restrictions on the transfer of reactants while providing the strongest confinement
effects for the reaction. Furthermore, GNF facilitate the effective recyclability of the
catalyst and thus represents a superior nanoreactor system to carbon nanotubes.
confinement to control the pathways of chemical reac-
1. Introduction
tions.[1–4] Possessing superior mechanical, chemical and
Carbon nanoreactors represent the ultimate class of hollow
thermal stabilities relative to zeolites, nanoporous solids
nanostructured materials to utilise nanoscale spatial
and molecular containers, carbon nanoreactors are able to
encapsulate the most expansive array of guest molecules
driven into the internal cavity via ubiquitous van der Waals
forces[5,6] and as such have been successfully employed as
nanoscale reaction vessels to examine the effects of con-
finement on both single-molecule and preparative chem-
ical transformations.[7–17] For example, it has been shown
W. A. Solomonsz, Dr. G. A. Rance,
Prof. A. N. Khlobystov
School of Chemistry
University of Nottingham
University Park
Nottingham, NG7 2RD, U.K
E-mail: Graham.Rance@nottingham.ac.uk;
Andrei.Khlobystov@nottingham.ac.uk
that carbon nanoreactors facilitate the formation of unique
molecular products, such as linear oligomers of fullerene
epoxide[8] and dynamic dimers of [60]fullerene[9] inside
single-walled carbon nanotubes (SWNT), which are pre-
cluded by bulk synthetic approaches. Further studies con-
ducted using MWNT, which possess a wider internal channel
and thus the opportunity to selectively incarcerate transition
metal nanoparticle (NP) catalysts,[18] have additionally shown
Prof. A. N. Khlobystov
Nottingham Nanoscience & Nanotechnology Centre
University of Nottingham
University Park
Nottingham, NG7 2RD, U.K
DOI: 10.1002/smll.201302732
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1866 wileyonlinelibrary.com
small 2014, 10, No. 9, 1866–1872

Effects on Pathways of Catalytic Hydrosilylation Reaction
that the enhancement in activity and selectivity of the con- DWNT, MWNT and GNF and the regioselectivity of the
fined catalysts in the PtNP-catalysed asymmetric hydrogena- hydrosilylation reaction was studied at different degrees of
tion of α-ketoesters,[12] the RhPtNP-catalysed hydrogenation confinement and compared to the reactions of unconfined
of cinnamaldehyde[14] and the RhNP-catalysed conversion catalysts. Our innovative study demonstrates that the path-
of syngas to ethanol[15] is related to confinement in carbon ways of preparative chemical reactions can be effectively
nanoreactors. These effects relate to the pairwise interactions controlled by the diameter and internal structure of the
between the host nanoreactor and the components of the carbon nanoreactor.
confined reaction (reactants and catalyst) and result in drasti-
cally altered concentrations, pressures and alignment of reac-
tant molecules as compared to the bulk solution or gas phase
and are understood to become increasingly important as the
2. Results and Discussion
dimensions of the host container approach commensuration
with the size of the guest molecules.[1–4]
2.1. Preparation of Nanoreactors
Surprisingly, larger nanocontainers with diameters The hydrosilylation of phenylacetylene by hydrosilanes is
exceeding the size of reactant molecules by a factor of 50 or generally catalysed by Pt group metals, in particular rho-
more have also recently been found to fundamentally alter dium,[20,25] platinum[26] ruthenium,[27] and iridium[28] and
the mechanisms of preparative chemical reactions. Hollow yields commercially valuable products[29] in a specific dis-
graphitised carbon nanofibers (GNF) are significantly wider tribution depending on the nature of the catalyst and the
than MWNT (internal diameters typically above 50 nm) and experimental conditions.[30] Our study requires a catalyst
thus facilitate effective transport of molecules through the that possesses both high versatility and activity at low load-
tubular structure.[19–23] Furthermore, their distinctive internal ings[31] and exemplary stability to the conditions required for
surface, consisting of a succession of 3–4 nm high steps encapsulation in a wide range of host carbon nanoreactors
formed by rolled-up sheets of graphene, provides effective (see S1 and S2 in the Supporting Information);[32] hence, we
adsorption loci for catalytic nanoparticles[24] and therefore selected the molecular catalyst tetrarhodium dodecacarbonyl
localised nanoscale reaction environments, different to the as it both fulfilled these requirements and exhibited the most
bulk phase, which mimic those observed in much narrower suitable catalytic properties in the hydrosilylation reaction
carbon nanostructures. Our studies have shown that the (see S3 and S4 in the Supporting Information). The chosen
selectivity,[19,20] activity[21] and recyclability[21] of the catalysts [Rh4(CO)12] catalyst was vaporised at reduced pressure in the
in preparative synthesis can be tuned in GNF while Fickian presence of open and dry carbon nanocontainers (DWNT,
diffusion of reactants/products to/from such nanoreactors MWNT and GNF) in order to facilitate the transport and
remains unrestricted. Recently, we reported the first observa- subsequent encapsulation of gaseous [Rh4(CO)12] mole-
tion of regioselectivity switching due to spatial confinement cules inside the internal cavities of the hollow carbon struc-
of catalytic centres in GNF using the hydrosilylation of ter- tures. Once the composite material was cooled to ambient
minal alkynes as a model reaction. Systematic comparison temperature and pressure, the molecules solidified inside
of the catalytic properties of Rh and RhPt nanoparticles nanoreactors to form [Rh4(CO)12]@DWNT, [Rh4(CO)12]@
embedded in carbon nanoreactors with free-standing and MWNT and [Rh4(CO)12]@GNF nanoreactors respectively.
surface-adsorbed nanoparticles showed that the directions The loading of molecular catalyst was optimised within each
of reactions inside GNF are largely controlled by the spe- of these containers (see Table S1 in the Supporting Informa-
cific non-covalent π–π interactions between aromatic reac- tion), such that the metal content was maximised with the
tant molecules and the nanofiber channel increasing the local majority (>90%) of the catalyst molecules residing inside the
concentration of the reactant in GNF.[20]
nanoreactors. Increasing the loading beyond these optimal
Changing the pathways of chemical transformations values resulted in the inherent inability to control the loca-
inside carbon nanoreactors, a key emerging branch of chem- tion of the catalytic centres.
ical nanosciences, has often been accounted for by the struc-
High resolution transmission electron microscopy (HR-
tural characteristics of the nanoreactor and thus the specific TEM, Figure 1 and Figure S1 in the Supporting Information)
interactions between nanoreactor-catalysts and nanoreactor- was used as the primary characterisation technique for these
reactants.[1–4] While many important examples of chemical structures,[32–35] clearly showing that the catalyst is evenly dis-
reactions inside nanotubes have been reported, all of them tributed along the length of the internal channel of DWNT
remain sporadic as no attempts to compare different types and MWNT in spite of their narrow diameter. The molecular
of nanoreactors for the same transformation and the same catalysts are stabilised by van der Waals interactions with
type of catalyst have been made to date. Since systematic the concave side of the CNT (which are greater as compared
comparison of nanoreactors with different diameters and to the convex side),[1] whereas in GNF the catalyst mole-
morphology is essential for understanding the fundamental cules are immobilised at the internal graphitic step-edges
aspects of nanoscale confinement and developing real prac- (Figures 1f–g and 3). In a control experiment, [Rh4(CO)12]
tical applications of carbon nanoreactors, in this novel study was deposited selectively onto the exterior of DWNT filled
we investigate the hydrosilylation reaction of phenylacety- with fullerene C60, blocking all of the internal space in nano-
lene and triethylsilane in the presence of a [Rh4(CO)12] cata- tubes, prior to the addition of [Rh4(CO)12] ([Rh4(CO)12]/
lyst in three different types of nanoreactors. The molecular DWNT, Figure S1 in the Supporting Information). System-
catalyst [Rh4(CO)12] was inserted for the first time into atic comparison of [Rh4(CO)12]@DWNT and [Rh4(CO)12]/
www.small-journal.com
1867
small 2014, 10, No. 9, 1866–1872
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

W. A. Solomonsz et al.
full papers
a comparison between the confinement
effects in carbon nanoreactors of different
sizes and morphologies (Table 1) .
2.2. Hydrosilylation Reaction
The hydrosilylation of triethylsilane
across phenylacetylene is highly suitable
for testing the properties of different
nanoreactors as the reaction does not
proceed in the absence of rhodium and
therefore controlled positioning of the
molecular catalytic centres (i.e., inside
for [Rh4(CO)12]@DWNT and outside for
[Rh4(CO)12]/DWNT) ensures that the
reaction locus is well defined. Further-
more, rhodium chemistry dictates that
the hydrosilylation reaction proceeds via
different pathways to yield all five pos-
sible products (Scheme 1) [20,27,36] each
1
of which may be quantified by H NMR
spectroscopy (see S5 in the Supporting
Information) and thus provides a compre-
hensive chemical probe for the effects of
nanoreactors on the pathways of catalytic
reactions.
The relative ratio of β addition prod-
ucts provides an important insight into
the reaction pathway. Comparison of the
β-(Z):β-(E) ratios of the “free” molecular
[Rh4(CO)12] catalyst compared to the sup-
ported ([Rh4(CO)12]/DWNT) and con-
fined ([Rh4(CO)12]@DWNT) catalysts
shows that the environment of the catalyst
alters this ratio. When the catalyst is sup-
ported on the surface or confined inside
the DWNT, there is a 2-fold promotion
of the trans product of β addition (β-(E)
product) compared to the bulk (Table 2,
entries 1–3 and Figure 2a ) .
This shows that the interactions
between carbon nanotubes and the cata-
lyst molecules, irrespective of their loca-
tion (in or on nanoreactors), sufficiently
changes the nature of the catalyst and
results in the stabilisation of the inter-
Figure 1. HR-TEM images of (a)-(b) [Rh4(CO)12]@DWNT, (c)-(d) [Rh4(CO)12]@MWNT and (f)-(g)
[Rh4(CO)12]@GNF nanoreactors. The internal diameters of DWNT, MWNT and GNF are 1.3
0.5 nm, 5.3 3.3 nm and 52.7 16.2 nm respectively. The embedded catalyst molecules
(inset in panel a) appear as dark clusters after decomposition and coalescence induced by
electron beam radiation: (a–d) anchored to the nanotube sidewall (highlighted by white
arrows) and (f,g) residing along the graphitic step-edges (highlighted by white boxes). A
schematic representation of the GNF structure (e) highlights the unique internal structure
of this hybrid nanomaterial, comprising graphitic step-edges suitable for catalyst deposition
(highlighted by black arrows). Scale bars are 5 nm (a–d) and 20 nm (f–g).
DWNT enables discrimination between the effects of cata- mediate A in preference to intermediate B (Scheme 2).
lyst support and confinement in nanoreactors. Most impor- This leads to the preferential formation of the more ther-
tantly, appraisal of [Rh4(CO)12]@DWNT, [Rh4(CO)12]@ modynamically stable β-(E) isomer so as to remove the
MWNT and [Rh4(CO)12]@GNF will enable, for the first time, destabilising steric repulsion between adjacent Ph and SiEt3
Table 1. The unique properties of DWNT, MWNT and GNF as carbon nanoreactors operating under the Fickian diffusion regime.
DWNT
MWNT
GNF
Catalyst environment
Accessibility of interior
Stabilised by van der Waals interactions with Stabilised by van der Waals interactions with
Immobilised internally at the graphitic
step-edges
the concave interior
the concave interior
Transport resistance due to extreme spatial
confinement
Low diffusion barrier due to the wide inner Low diffusion barrier due to the extremely wide
channel
inner channel
www.small-journal.com
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1868
small 2014, 10, No. 9, 1866–1872

Effects on Pathways of Catalytic Hydrosilylation Reaction
a consequence of catalyst-nanotube interactions is a general
phenomenon applicable to variety of carbon nanoreactors
and consequently can be harnessed for the efficient promo-
tion of the β-(E) product (Figure 2).
However, although the relative stabilities of intermediates
A and B are unaffected by confinement of the [Rh4(CO)12]
catalyst as compared to the supported species, the overall
product distribution is fundamentally altered upon confine-
ment. The β-(Z):DS ratio is a useful diagnostic parameter to
assess the fate of intermediate B and hence probe the effects
of confinement as this ratio is related to the concentrations of
triethylsilane and phenylacetylene. Dehydrogenative silyla-
tion (DS) products are formed in this reaction due to the
β-H elimination process (Scheme 2), the viability of which
can be influenced by changes in the relative concentrations
of reactant molecules,[20] present in equimolar quantities in
the feedstock solution (bulk phase). Our measurements show
that confinement of the reaction inside [Rh4(CO)12]@DWNT
nanoreactors leads to a decrease in the β-(Z):DS ratio, i.e.,
Scheme 1. Reaction scheme for the hydrosilylation of phenylacetylene
and triethylsilane yielding a distribution of three addition products
(α, β-(Z) and β-(E)) and two dehydrogenative silylation products (DS),
which are produced in equimolar quantities.
Table 2. The effect of catalyst environment on the comparative selec- a promotion of DS products, compared to the reaction on
tivity for the products of hydrosilylation of phenylacetylene and tri-
ethylsilane. Comparative TOF values for these catalytic systems are
(Table 2). This indicates that the concentration of aromatic
presented in S4 of the Supporting Information file. All reactions were
the surface of [Rh4(CO)12]/DWNT and “free” [Rh4(CO)12]
reactants is increased inside DWNT, changing the pathway
of this reaction. This observation is consistent with our pre-
vious studies using RhNP@GNF nanoreactors, where a 3-fold
increase of local concentration of phenylacetylene inside
nanoreactors resulted in the promotion of the β-H elimina-
tion step so as to consume the excess of aromatic alkyne.[20]
It is important to note that this effect is observed as phenyla-
cetylene is the only reactant that possesses aromaticity and is
thus favourably encapsulated at the expense of the aliphatic
silane inside the DWNT nanoreactors, which are known to
have a special affinity for aromatic species.[22,23,37–39]
performed at a normalised catalyst loading of 2.7 mmol% [Rh4(CO)12].
Catalyst
Container
Regioselectivity
β-(Z):β-(E)
β-(Z):DS
9.3:1
3.5:1
2.2:1
2.0:1
1.1:1
[Rh4(CO)12
[Rh4(CO)12
[Rh4(CO)12
[Rh4(CO)12
[Rh4(CO)12
]
]
]
]
]
“free”
/DWNT
@DWNT
@MWNT
@GNF
1.1:1
0.6:1
0.5:1
0.5:1
0.5:1
groups in intermediate B. This effect appears to be some-
what universal and independent of the internal diameter and
internal structure of the carbon container the [Rh4(CO)12]
is anchored to, with a similar change in β-(Z):β-(E) also
observed for reactions confined in MWNT and GNF (Table 2,
entries 4–5). Although the intermediate stability appears to
be unaffected by confinement, consistent with our previous
studies regarding the nanoparticle-catalysed hydrosilylation
of phenylacetylene by triethylsilane in carbon nanoreac-
tors,[20] manipulation of the ratio of β addition products as
Such an effect had only previously been observed inside
GNF nanoreactors, which are also much more accessible
than DWNT for preparative reactions due to their wider and
thus less restrictive inner channel. Therefore, building on the
discovery that all sizes and shapes of nanocontainer can be
used as preparative vessels upon encapsulation of a suitable
catalyst, the hydrosilylation reaction studied here yielded a
unique opportunity to assess selectivity switching in different
types of hollow carbon nanocontainers and thus to examine
the extent of confinement as a function of nanoreactor
Figure 2. Graphical distributions of shifts in regioselectivity upon supporting the catalyst, where a promotion of the β-(E) product is observed
(a) and confining the catalyst within carbon nanoreactors of varying diameter and morphologies, where GNF are found to show the largest shift in
regioselectivity compared to the bulk, observed via promotion of the DS products (b).
www.small-journal.com
1869
small 2014, 10, No. 9, 1866–1872
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

W. A. Solomonsz et al.
full papers
Scheme 2. The β-(Z):β-(E) ratio is an indication of the relative stability of intermediates A and B (light boxes) whilst the β-(Z):DS ratio is related to
the favourability of the β-H elimination step (dark boxes), so as to deplete the excess of phenylacetylene.
diameter (comparison of DWNT and MWNT) and internal internal structure of the GNF nanoreactors, where the com-
structure (comparison of CNT and GNF). This was achieved bination of high mass transfer and high energy of step-edge
by evaluating the β-(Z):DS ratio for [Rh4(CO)12] in DWNT, encapsulation (Ee′) creates local reaction environments that
MWNT and GNF nanoreactors (Table 1), where it was star- concentrates an even greater excess of phenylacetylene than
tling to discover that not only are the pathways of the hydros- can be seen in CNT which possess an atomically smooth inte-
ilylation reaction altered (as compared to the bulk) inside all rior (Figure 3).
of the nanoreactors, but that the most extreme effect was
The increased concentration of phenylacetylene and the
observed inside [Rh4(CO)12]@GNF nanoreactors (Table 2 subsequent preferential β-H elimination step in nanoreac-
and Figure 2b) despite the fact that they are the widest nano- tors is related to the specific π–π interactions[40] between the
reactors in this study. Our measurements have shown that phenyl ring and the graphitic surface of the interior of the
there is a 2-fold decrease in the β-(Z):DS ratio inside GNF carbon nanocontainer, which is not available for fully ali-
compared to both DWNT and MWNT, in addition to the phatic triethylsilane molecules. Our results suggest that the
4-fold decrease these nanoreactors show
compared to the bulk.
As the internal diameter of DWNT
is close to the size of small organic mol-
ecules (the critical van der Waals diame-
ters of phenylacetylene, triethylsilane and
tetrarhodium dodecacarbonyl are 0.42,
0.53 and 0.93 nm respectively), the energy
of encapsulation of reactants (Ee) is sig-
nificantly greater than the energy of their
adsorption on DWNT surface (Ea),[2] and
thus the greatest enhancement in reac-
tant concentration was expected to be in
DWNT leading to a greater proportion
of DS products as compared to MWNT
and other wider nanoreactors. However,
our results show that there is effectively
no difference in the β-(Z):DS ratio for
reactions in DWNT and MWNT. This
indicates that local concentration effects
are independent of carbon nanotube
diameter, where the effect of higher Ee
in narrow diameter nanostructures such
as DWNT (relative to MWNT) is offset
by a corresponding reduction in the rate
of mass transfer of reactants (kc) which
will be higher in wider MWNT (relative
to DWNT). Consequently, as the greater
than 8-fold selectivity switch in GNF as
compared to the bulk cannot be explained
by nanoreactor diameter, this effect must
be related to the unique step-edged
Figure 3. Schematic representation of the contrasting internal environments of [Rh4(CO)12]@
DWNT, [Rh4(CO)12]@MWNT and [Rh4(CO)12]@GNF (from left to right, respectively) and the
relative local concentration effects induced inside these nanoreactors. The observed effects
are the result of a balance between the energy of encapsulation in CNT (Ee) and step-edge
encapsulation in GNF (Ee′) with the mass transfer rate (kc) of reactants (compared relative
to DWNT). The most extreme effects are observed inside the internally corrugated GNF,
whereas the net effects in CNT are minimised, because the step-edge encapsulation provides
a similarly constrained environment as in DWNT, but the wide diameters of GNF have the
additional ability to readily facilitate the rapid Fickian diffusion of reagents.
www.small-journal.com
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1870
small 2014, 10, No. 9, 1866–1872

Effects on Pathways of Catalytic Hydrosilylation Reaction
most preferential interactions occur within GNF (Figure 3), and graphitised carbon nanofibers (CVD, Pyrograf PR-19, Applied
causing significant mechanistic deviations (Scheme 2), whilst Science, USA) were obtained from commercial sources. All glass-
these nanoreactors also boast a more accessible internal ware was cleaned with a mixture of hydrochloric and nitric acid
cavity than CNT. Furthermore, as a consequence of the (3:1 v/v, ‘aqua regia’) and rinsed thoroughly with deionised water,
stability of the nanoreactor-catalyst interface inside GNF, cleaned with potassium hydroxide in isopropyl alcohol and finally
the catalyst in GNF can be recyclable (see S6 in the Sup- rinsed thoroughly with deionised water. 1H NMR spectra were
porting Information). TEM analysis of the [Rh4(CO)12]@ obtained using a Bruker DPX-300 (300.13 MHz) spectrometer at
GNF catalyst after one reaction cycle provides a snapshot 298K using CDCl3 as the solvent. Thermogravimetric analysis (TGA)
of the reaction, frozen in time, where the catalytic mate- was performed using a TA Instruments SDT Q600 under a flow
rial persists within the inner channel after catalysis due to of air at a rate of 100 mL min−1 at a heating rate of 10 °C min−1
anchoring at the graphitic step-edges and the reaction prod- from room temperature to 900 °C. Transmission electron micros-
ucts are observed as amorphous material within the GNF copy (TEM) was performed using a JEOL 2100F TEM (field emis-
nanoreactor interior containing C and Si, as confirmed by sion gun source, information limit <0.19 nm) operated at 100 kV
energy dispersive X-ray spectroscopy. Therefore, GNF accelerating voltage at room temperature. Energy dispersive X-ray
represent the ultimate container for preparative chemical (EDX) analysis was performed using an Oxford Instruments INCA
transformations.
560 X-ray microanalysis system. Samples were prepared via drop-
drying methanolic solutions onto a copper grid mounted “lacey”
carbon films.
Synthesis of Nanoreactor Catalysts: To the DWNT45% (10.0 mg,
see S1.1.1 in the Supporting Information), MWNT40% (6.7 mg, see
S1.2.1 in the Supporting Information) or GNF (7.13 mg, annealed
for 1 hr at 450 °C in air) was immediately added tetrarhodium
dodecacarbonyl (1.0 mg, 0.30 mg or 0.50 mg respectively),
sealed under vacuum (10−6 mbar) in a Pyrex ampoule and heated
at 140 °C for 72 hours to ensure complete vaporisation and pen-
etration of the tetrarhodium dodecacarbonyl into the hollow inte-
rior of the carbon nanostructures. The samples were cooled and
then stirred in tetrahydrofuran (15 mL) at room temperature for
1 hour in order to remove any metal carbonyl from the exterior
of the carbon nanostructures. The products were collected by
vacuum filtration (0.45 µm, PTFE), washed with tetrahydrofuran
(75 mL) and dried in vacuo to yield dark solids (10.6 mg, 5.5 mg
and 7.5 mg respectively).
The Hydrosilylation Reaction: In a typical experiment, to an
argon-flushed Schlenk tube was added the catalyst (2.7 mmol%
[Rh4(CO)12]), triethylsilane (0.72 mL, 4.5 mmol, 1 eq.) and to this
was added dropwise phenylacetylene (0.50 mL, 4. 5 mmol, 1 eq.).
The mixture was homogenised by bath sonication at room temper-
ature and then stirred at 90 °C. The progress of the reaction was
monitored by 1H NMR spectroscopy and product distributions were
generated by integrating the one-proton doublets of each product,
which have unique shifts which were found to match known litera-
ture values.[29]
3. Conclusion
Carrying out reactions inside carbon nanoreactors offers an
elegant tool to alter the pathways of conventional organic
transformations to yield products different to the bulk. In this
study, we probed nanoreactor-catalyst and nanoreactor-reac-
tant interactions via synthesis of novel [Rh4(CO)12]-based
catalytic systems allowing hydrosilylation reactions to be
performed either inside or outside the DWNT. We demon-
strate that preparative, molecular-catalysed hydrosilylation
reactions on the exterior and interior of DWNT follow a dif-
ferent reaction pathway to the bulk, as observed by a change
in the β-(Z):β-(E) products ratio. This ratio is unaffected by
confining the reaction inside DWNT showing similar value to
other, wider nanoreactors MWNT and GNF, implying that
nanoreactor-catalyst interactions determine the ratio of these
products regardless of the location of catalyst (inside or out-
side nanoreactor) or size of nanoreactor. In contrast, ratio of
β-(Z):DS products, controlled by local concentration effects,
shows that the most extreme switch in product selectivity
occurs inside GNF. We demonstrate that confinement effects
invoked by divergent local concentrations are universally
observed inside all carbon nanocontainers, although GNF
appear to have the potential to be superior nanoreactors as
compared to DWNT and MWNT. Controlling the pathways
of catalytic chemical transformations is the pinnacle goal of
the chemist and typically involves time and skill intensive
functionalisation of molecules; our results show that non-
covalent interactions of reagents with carbon nanoreactors
– essential for controlled assembly at the nanoscale[41,42]
provides a superior alternative approach and may ultimately
facilitate the formation of novel molecular structures inacces-
sible by other means.
–
Supporting Information
Supporting Information is available from the Wiley Online Library
or from the author.
4. Experimental Section
Acknowledgements
General: All reagents were purchased from Sigma-Aldrich, UK
and used without further purification. Double-walled carbon nano-
tubes (CVD, Times Nano, Chengdu Organic Chemicals, China),
This work was supported by the European Research Council (ERC),
Engineering and Physical Science Research Council (EPSRC), Royal
multi-walled carbon nanotubes (CVD, PD30L520, NanoLab, USA) Society and Nottingham Nanoscience & Nanotechnology Centre
www.small-journal.com
1871
small 2014, 10, No. 9, 1866–1872
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

W. A. Solomonsz et al.
full papers
[21] G. A. Rance, W. A. Solomonsz, A. N. Khlobystov, Chem. Commun.
2013, 1067.
[22] E. Castillejos, P.-J. Debouttiere, L. Roiban, A. Solhy, V. Martinez,
Y. Kihn, O. Ersen, K. Philippot, B. Chaudret, P. Serp, Angew. Chem.
Int. Ed. 2009, 48, 2529.
(NNNC). The authors would also like to thank Mr Scott Miners for
assistance with manuscript preparation.
[1] X. Pan, X. Bao, Acc. Chem. Res. 2011, 44, 553.
[23] E. Castillejos, R. Chico, R. Bacsa, S. Coco, P. Espinet,
M. Perez-Cadenas, A. Guerrero-Ruiz, I. Rodriguez-Ramos, P. Serp,
Eur. J. Inorg. Chem. 2010, 5096.
[24] A. La Torre, M. Fay, G. A. Rance, M. C. Gimenez-Lopez,
W. A. Solomonsz, P. D. Brown, A. N. Khlobystov, Small 2012, 8,
1222.
[2] D. A. Britz, A. N. Khlobystov, Chem. Soc. Rev. 2006, 35, 637.
[3] P. Serp, E. Castillejos, ChemCatChem 2010, 2, 41.
[4] A. N. Khlobystov, ACS Nano 2011, 5, 9306.
[5] G. Borsato, J. Rebek Jr., A. Scarso, in Selective Nanocatalysts and
Nanoscience, (Eds: A. Zecchina, S. Bordiga, E. Groppo), WILEY-
VCH, Weinheim, Germany 2011, pp. 105–108.
[25] Y. Goldberg, H. Alper, Tet. Lett. 1995, 36, 369.
[6] Molecular Encapsulation: Organic Reactions in Constrained [26] L. Lewis, K. Sy, G. Bryant, P. Donahue, Organometallics 1991, 10,
Systems, (Eds: U. H. Brinker, J.-L. Mieusset), Wiley, New York
2010.
[7] D. A. Britz, A. N. Khlobystov, J. Wang, A. S. O'Neil, M. Poliakoff,
A. Ardavan, G. A. D. Briggs, Chem. Commun. 2004, 176.
3750.
[27] M. Jiménez, J. Pérez-Torrente, M. Bartolomé, V. Gierz, F. Lahoz,
L. Oro, Organometallics 2008, 27, 224.
[28] R. S. Tanke, R. H. Crabtree, J. Am. Chem. Soc. 1990, 112, 7894.
[8] D. A. Britz, A. N. Khlobystov, K. Porfyrakis, A. Ardavan, [29] I. Fleming, J. Dunogues, R. Smithers, Org. React. 1989, 37, 57.
G. A. D. Briggs, Chem. Commun. 2005, 37. [30] R. Takeuchi, N. Tanouchi, J. Chem. Soc. Perk. Trans. 1 1994, 2909.
[9] E. Nakamura, M. Koshino, T. Saito, Y. Niimi, K. Suenaga, [31] I. Ojima, R. J. Donovan, N. Clos, Organometallics 1991, 10, 2606.
Y. Matsuo, J. Am. Chem. Soc. 2011, 133, 14151.
[10] M. Koshino, Y. Niimi, E. Nakamura, H. Kataura, T. Okazaki,
K. Suenaga, S. Iijima, Nature Chem. 2010, 2, 117.
[11] S. Miners, G. A. Rance, A. N. Khlobystov, Chem. Commun. 2013,
49, 5586.
[32] T. W. Chamberlain, T. Zoberbier, J. Biskupek, A. Botos, U. Kaiser,
A. N. Khlobystov, Chem. Sci. 2012, 3, 1919.
[33] D. V. Kazachkin, Y. Nishimura, H. A. Witek, S. Irle, E. Borguet, J.
Am. Chem. Soc. 2011, 133, 8191.
[34] T. Zoberbier, T. W. Chamberlain, J. Biskupek, N. Kuganathan,
S. Eyhusen, E. Bichoutskaia, U. Kaiser, A. N. Khlobystov, J. Am.
Chem. Soc. 2012, 134, 3073.
[12] H. Zhang, X. Pan, X. Han, X. Liu, X. Wang, W. Shen, X. Bao, Chem.
Sci. 2013, 4, 1075.
[13] Z. Chen, Z. Guan, M. Li, Q.Yang, C. Li, Angew. Chem. Int. Ed. 2011, [35] Z. Xu, H. Li, K. Fujisawa, Y. A. Kim, M. Endo, F. Ding, Nanoscale
50, 4913. 2012, 4, 130.
[14] J. Teddy, A. Falqui, A. Corrias, D. Carta, P. Lecante, I. Gerber, [36] C.-H. Jun, R. Crabtree, J. Organomet. Chem. 1993, 447, 177.
P. Serp, J. Catal. 2011, 278, 59.
[15] X. Pan, Z. Fan, W. Chen, Y. Ding, H. Luo, X. Bao, Nature Mater.
2007, 6, 507.
[16] S. Guo, X. Pan, H. Gao, Z. Yang, J. Zhao, X. Bao, Chem. Eur. J.
2010, 16, 5379.
[17] H. Yang, S. Song, R. Rao, X. Wang, Q. Yu, A. Zhang, J. Mol. Catal.:
A. 2010, 323, 33.
[37] R. J. Chen, Y. G. Zhang, D. W. Wang, H. J. Dai, J. Am. Chem. Soc.
2001, 123, 3838.
[38] Y. Gao, G. Hu, W. Zhang, D. Ma, X. Bao, Dalton Trans. 2011, 40,
4542.
[39] C. Vriamont, M. Devillers, O. Riant, S. Hermans, Chem. Eur. J.
2013, 19, 12009.
[40] C. R. Martinez, B. L. Iverson, Chem. Sci. 2012, 3, 2191.
[18] A. La Torre, G. A. Rance, J. El Harfi, J. Li, D. J. Irvine, P. D. Drown, [41] L. Tang, X. Yang, J. Phys. Chem. C 2012, 116, 11783.
A. N. Khlobystov, Nanoscale 2010, 2, 1006.
[42] Z. Xu, X. Yang, Z. Yang, Nano Lett. 2010, 10, 985.
[19] A. La Torre, M. C. Gimenez-Lopez, M. Fay, G. A. Rance,
W. A. Solomonsz, T. W. Chamberlain, P. D. Brown, A. N. Khlobystov,
ACS Nano 2012, 6, 2000.
[20] W. A. Solomonsz, G. A. Rance, M. Suyetin, A. La Torre,
E. Bichoutskaia, A. N. Khlobystov, Chem. Eur. J. 2012, 18, 13180.
Received: August 20, 2013
Revised: November 29, 2013
Published online: January 25, 2014
www.small-journal.com
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1872
small 2014, 10, No. 9, 1866–1872