Immobilized Grubbs catalysts on mesoporous silica materials: Insight into support characteristics and their impact on catalytic activity and product selectivity

View Article Online

View Journal

Catalysis

Science &

Technology

Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: A. Dewaele, B.

Van Berlo, J. Dijkmans, P. A. Jacobs and B. F. Sels, Catal. Sci. Technol., 2015, DOI: 10.1039/C5CY01897H.

This is an Accepted Manuscript, which has been through the

Royal Society of Chemistry peer review process and has been

accepted for publication.

Accepted Manuscripts are published online shortly after

acceptance, before technical editing, formatting and proof reading.

Using this free service, authors can make their results available

to the community, in citable form, before we publish the edited

article. We will replace this Accepted Manuscript with the edited

and formatted Advance Article as soon as it is available.

You can find more information about Accepted Manuscripts in the

Information for Authors .

Please note that technical editing may introduce minor changes

to the text and/or graphics, which may alter content. The journal’s

standard T erms & Conditions and the Ethical guidelines still

apply. In no event shall the Royal Society of Chemistry be held

responsible for any errors or omissions in this Accepted Manuscript

or any consequences arising from the use of any information it

contains.

www.rsc.org/catalysis

Please do not adjust margins

Catalysis Science & Technology

Page 1 of 18

View Article Online

DOI: 10.1039/C5CY01897H

Catalysis Science and Technology

ARTICLE

Immobilized Grubbs Catalysts on Mesoporous Silica Materials:

Insight into Support Characteristics and Their Impact on Catalytic

Activity and Product Selectivity

A. Dewaele^a, B. Van Berlo^a, J. Dijkmans^a, P. Jacobs^a, B. Sels^a

Received 00th January 20xx,

Accepted 00th January 20xx

DOI: 10.1039/x0xx00000x

Silica materials show a high ability to physisorb the 2^ndgeneration Hoveyda-Grubbs catalyst (HG2) in organic solvents. The

interaction with the complex, likely proceeding through hydrogen bonding, is particularly strong with surfaces rich in

silanols, geminal silanols showing the highest affinity, and therefore mesoporous silica are the supports of choice. As long

as the silica material is sufficiently pure and free of cages, in which high HG2 concentrations can accumulate, the

immobilisation of HG2 occurs very stable. Despite the complex stability, exploration of HG2 loaded mesoporous silica

supports in metathesis of cis-cyclooctene indicated significant diffusional and confinement effects, and therefore control

of pore size, pore architecture and morphology in balance with the intrinsic catalytic activity, is essential in the catalyst

design. As metathesis of cis-cyclooctene apparently proceeds through initial formation of linear polymers, followed by

backbiting forming the cyclic oligomers, potential interference of mass transport and space restriction issues is not

surprising. This study shows that the catalyst requirements are best met with the TUD-1 silica support (1.24 wt% HG2).

Under such conditions, the heterogeneous catalyst performes as good as the homogeneous one, presenting the

thermodynamic distribution of cyclic oligomers. The latter catalyst also showed a high catalyst stability in a continuous

fixed bed reactor, corresponding to 18000 catalytic turnover numbers. Catalytic rates and catalyst stability are lower when

operating in diffusional regime, therefore long reaction times are required to reach the thermodynamic product

distribution. Water removal from the catalyst is also important, not because of HG2 stability reasons, but lower reactions

rates were measured for hydrated samples, likely due to inhibition of cis-cyclooctene uptake in the pores. Mild removal of

physisorbed water before immobilization is therefore advised, for instance by thermal treatments, but care has to be

taken to keep the silanol density high for firm HG2 immobilization and also to avoid formation of reactive siloxanes, which

chemically react with and destroy HG2. Surprisingly, reactive siloxane formation conditions strongly depend on the silica

type, TUD-1 being fairly sensitive to their formation. Finally, the best HG2 loaded TUD-1 catalyst is used successfully in a

broad set of other metathesis reactions.

www.rsc.org/

typically symbolized by a difficult catalyst recovery, leading to

Ru-contamination of the product. This is disastrous in the

production of fine chemicals like pharmaceuticals,^8-10but it

Introduction

Olefin metathesis exchanges alkyl substituents between

olefins and is considered as a very important reaction for a

wide range of applications in organic chemistry.^1-3The reaction

also provides nice opportunities to produce unsaturated and

therefore modifiable polymers.⁴Well-defined commercially

available Grubbs complexes are established as one of the most

promising catalysts; they exhibit a high catalytic activity at

ambient conditions and they show great tolerance towards

functional groups, air and moisture.^{5, 6}Despite of the industrial

demonstrations,⁷their use has not been fully exploited due to

high cost and environmental reasons. The latter issue is

also implies a loss of precious metal.

Catalyst immobilization on a solid support offers an

attractive solution to Ru recovery. The catalyst is not only

more practical to separate from the product mixture,

immobilization also allows protection against catalyst

deactivation, caused by reported bimolecular decomposition

pathways,^11-13and even may exert beneficial confinement

effects on conversion rate and product selectivity. The first

heterogeneous metathesis catalysts were based on metal

oxides or organometallic complexes on silica/alumina. Pioneer

work of the surface organometallic chemistry was delivered by

Basset and Copéret, et al.^{14, 15}, who characterized the active

sites with NMR spectroscopy.^16-18Several classical strategies

have been developed to anchor Grubbs-like complexes to the

surface through covalent linkage via: (a) halide ligands,^19-25(b)

phosphine ligands or N-heterocyclic carbenes (NHC) ^{11, 26-36}and

(c) alkylidene ligands,^37-50or through encapsulation in cage-like

pore structures followed by post-reducing the window size

^a.Center for Surface Chemistry and Catalysis, KU Leuven, Celestijnenlaan 200F,

3001 Leuven, Belgium.

† Electronic Supplementary Information (ESI) available: Stirring experiments of

HG2/MCM-41, correlation plots of ordered mesoporous silica, conversion plots of

thermally treated TUD-1 and MCM-41 are provided. (Near)-infrared spectra and

nitrogen physisorption measurements of thermally treated TUD-1 and MCM-41,

information about ring-chain equilibria and additional ring selectivity experiments

are given, See DOI: 10.1039/x0xx00000x

This journal is © The Royal Society of Chemistry 20xx

J. Name., 2013, 00, 1-3 | 1

Please do not adjust margins

Catalysis Science & Technology

Page 2 of 18

ARTICLE

Journal Name

52

through silylation.^51,

Recent publications with regard to were compared for their activity in metathesis of cis-c_Vy_ic_el_wo_Ao_rc_tit_ce_len_Oe_n,_lia_nes

these strategies are summarized in excellent reviews.^53-57this substrate is generally used to study the catalytic activity of

Whereas these immobilization approaches often require a porous catalysts.^{26, 59, 60, 62, 64, 72}Moreover, it is an interesting

sophisticated and laborious synthesis, with alterations of both probe to investigate the impact of textural properties on

support and catalytic complex, an elegant and more practical selectivity, as both cyclic and linear oligomers/polymers are

immobilization strategy of Hoveyda-Grubbs (HG) complexes on formed. Table 1 compares the catalytic performance along with

silica supports was reported by Van Berlo et al.⁵⁸This novel the material’s characteristics. The activity is defined by the

DOI: 10.1039/C5CY01897H

immobilization strategy is based on

a subtle physical initial turnover frequency (TOF_i), i.e. the number of moles of

interaction between the 2^ndgeneration Hoveyda-Grubbs (HG2) cyclooctene converted per mol Ru per second, in the initial

and the silica surface. Its synthesis avoids complicated phase of the reaction.

modifications of either catalyst partner, therefore being

industrially very attractive. Though the soft adsorption concept

Three zeolites were tested: Silicalite-1, AlPO₄-5 and Si-VPI-5

Table 1 , entry 2-4). The pores of Silicalite-1 and AlPO₄-5 are

(

has been studied briefly and its application expanded obviously too small for adsorption of HG2 (1.18 x 1.07 nm) and

successfully by others to new ordered mesoporous silica therefore contain very little Ru. Si-VPI-5 has larger pores (1.1

(OMS) like MCM-41 and SBA-15, there is no real consensus on nm) and does adsorb HG2, but the density of hydroxyl groups

the requirements of the support material of choice.^59-65

on its surface, affording anchor points for the complex,⁵⁸is

Next to the high activity of the physically sorbed very low resulting in low Ru loading. None of the zeolites

complexes, shifts in product selectivity between showed a true heterogeneous activity. The few complexes

polymerization (ROMP) and ring-opening ring-closing present on the material leached into the solution, as was

metathesis (RO-RCM) were reported, especially in metathesis proven by a split-test. In such test, after 15 minutes reaction,

of cyclooctene.^{59, 60, 66}The products of RO-RCM of cyclooctene, part of the reaction suspension was removed and separated by

viz. macrocyclic structures, exhibit unique properties filtration and transferred into a new vial. The reaction progress

compared to their linear counterparts due to the lack of chain of the filtered sample was compared to that of the remaining

ends,^{67, 68}making them potential intermediates as lubricants reaction mixture at different time intervals. The result of

or plasticizers,^{69, 70}but also in the fragrance industries.⁷¹

HG2/Si-VPI-5, given in Figure 1 A, shows a catalytic contribution

Despite the recent evaluation of different porous silica owing to soluble HG2.

materials, no fundamental was delivered since, to

unambiguously clarify the textural and structural requirements

of the mesoporous silica to perform fast and stable

metathesis. In search for the ideal catalyst support for RO-RCM

of cis-cyclooctene to macrocycles, we undertook a systematic

study to understand better the impact of the support

properties (e.g. pore size, pore morphology) on the catalytic

activity and product selectivity of supported HG2, as well as to

Figure 1. Two examples of a split-test with Ru leaching for A) HG2/Si-VPI-5 and without

Ru leaching for B) HG2/TUD-1. Blue curve: Liquid phase with the heterogeneous

catalyst; Red curve: liquid phase after filtration. Reaction conditions: 0.05 M cis-

assess surface properties that could substantially affect surface

adsorption affinity and catalyst stability such as catalyst

surface loading, silanol density and type. Furthermore, we cyclooctene; 5 mL hexane; 50 mg HG2/Si-VPI-5 (pre-treated at 50 °C, 0.12 wt% Ru) or

HG2/TUD-1 (pre-treated at 150 °C; 0.21 wt% Ru); 35 °C.

addressed typical heterogeneous catalysis aspects like stability

in

a continuous reaction and discussed mass transfer

In contrast, the porous amorphous silica materials like silica

gels and OMS are truly heterogeneous, confirming the

essential role of surface silanols to firmly anchor HG2. The

split-test of HG2/TUD-1 is illustrated in Figure 1B. Clearly, no

activity is measured in the filtrate. Besides showing stable

heterogeneous catalysis, the silica gels (Table 1, entry 5-6) are

reasonable active, with TOF_i’s ranging between 4 to 4.5 s^-1.10^-

². The lower values, compared to some OMS, may be due to

the presence of impurities like Na (up to 0.06 %) and Ca (0.1

%), which modify the surface properties, leading to

deactivation of HG2, as suggested in literature.⁷³

The current wealth of existing OMS should allow a deeper

analysis of the relationship between structural and surface

characteristics and the catalytic performance (Table 1, entry 7-

17). However, our experience reveals a blurred analysis due to

a multi-variance of parameters and therefore drawing general

conclusions on activity dependence of e.g., particle size

(d_particle) (Figure S1) and pore size (Figure S2), based on the

limitations, rather uncommon in metathesis studies, but

necessary to understand heterogeneous metathesis catalysis.

Based on the insight, an optimal support-catalyst system is

designed and further used to unravel the reaction pathways of

RO-RCM of cis-cyclooctene to cyclic oligomers under

heterogeneous catalysis conditions. The kinetic study clearly

shows kinetic dissimilarities between homo- and

heterogeneous reactions due to pore-related issues, all of this

being dependent on the type of support and on the active site

density. Finally, other metathesis types, for which no

oligomers and polymers are formed, are also illustrated with

the best supported HG2 catalyst.

Results and discussion

Screening of micro- and mesoporous silica

Various potential silica supports such as silica gels, zeolites and

OMS were explored to support HG2. The immobilized catalysts

2 | J. Name., 2012, 00, 1-3

This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

Page 3 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

whole data set, is not always straightforward. To verify close to 1, whereas τ = 3 for MCM-48.⁷⁸

ARTICLE

View Article Online

DOI: 10.1039/C5CY01897H

therefore the effect of pore size, three hexagonally pore-

TUD-1, a highly porous structure with interconnecting

ordered mesoporous silicates (MCM-41) having pore cage-free mesopores, attains a high activity close to SBA-15

diameters of respectively 2.6, 3.3 and 3.6 nm,⁷⁴were tested and MCM-41 (Table 1, Entry 17). Just like KIT-6 and MCM-48, it

(Table 1, entry 14-16). Comparable TOF_i’s, 7.5 to 8.5 s^-1.10^-2has a three-dimensional pore network, albeit a higher activity

were shown, whereas SBA-3 (Table 1, entry 10), a mesoporous is observed. This could likely be attributed to the high porosity

silica with similar pore architecture as MCM-41 and similar and pore structure of TUD-1, providing optimal features for

pore volumes and BET surface area, but a smaller pore fast intracrystalline diffusion. Awaiting for

a profound

diameter of 1.8 nm, showed a lower activity (TOF_i= 3 s^-1.10^-2). structural analysis of TUD-1, other factors cannot be excluded.

Mass transfer issues, depending on HG2 loading, seem to play Nevertheless, TUD-1 has often been acknowledged in

a role in cyclooctene metathesis and impact the observed rate literature for its superior catalytic activity, outperforming

and selectivity. Though this point will be addressed later, it is other ordered mesoporous silica.^79-81

fair to conclude at this stage that 2.5 nm pores (or larger) are

To conclude, metathesis of cis-cyclooctene with low

recommended to efficiently catalyze cis-cyclooctene loadings of HG2 on mesoporous silica is truly heterogeneous,

metathesis in presence of 0.2-0.4 wt% Ru loaded catalysts. but care has to be taken to balance diffusion and the internal

Besides pore diameter, morphology also plays a key role. To catalytic activity in order to perform catalysis in the kinetic

prove this, SBA-15 was synthesized in two different regime and thus to use the costly HG2 most efficiently. Pure

morphologies with comparable particle and pore size, and siliceous mesoporous silica are recommended, preferably with

loaded with equal contents of HG2: rope-like (Table 1, entry pores larger than 2.5 nm, absence of cage-like or bimodal pore

13) and fiber-like (Table 1, entry 9). Rope-like structures are systems, and a short diffusion path, which are morphology and

typical for SBA-15 and have straight channels, whereas fibre- pore architecture related. Rope-like SBA-15, spherical MCM-41

like structures have curving or orbicular channel structures. and TUD-1 are the support candidates of choice for cis-

Though being instrumental to hold the catalyst in the pores, cyclooctene metathesis, because of absence of diffusion

corrugated pores are unfavourable for molecular pore limitations. Their catalytic activity approaches that of the

diffusion.⁷⁵As rope-like SBA-15 performs 2.5 times better than homogeneous complex (Table 1, entry 1).

the fibre-like SBA-15, intraparticle diffusion is a critical

parameter. Additionally, as a large part of the pore volume Influence of solvent polarity on surface mobility and HG2 leaching

of rope-like SBA-15 is occupied by micropores (inside the pore

For all porous amorphous silica, irrespective of the particle

walls), not every catalytic complex might be accessible, which

size, pore size and pore ordering, HG2 remains grafted on the

then could lead to a lower TOF_i.

silica surface, as demonstrated above in the catalyst split-tests.

Notably, HG2/KIT-5 and HG2/SBA-16 (Table 1, entry 7 and

To further confirm the true heterogeneity of the catalyst, an

8) gave the lowest activity of the OMS (TOF_iof 1.5 and 2 s^-1.10^-

additional test was developed to examine the mobility of the

²respectively), despite their high surface area and large pore

complex under reaction conditions. The experiment is

size. These two silicas have 3D pore structures with a

illustrated and described in Figure 2 for silica gel, but the

pronounced cage-like pore system. As the large cages of 5-8

conclusion holds for the other mesoporous silicas.

nm allows formation of large oligomers/polymers of cis-

cyclooctene, rate retardation is likely a result of internal pore

entrance blockage. In addition, we also noticed that HG2

deactivates rapidly within the cage-like KIT-5 and SBA-16 silica,

as visually seen by the fast green-to-brown colour change. As

the colour variation occurred also in absence of substrate,

deactivation is likely caused by fast disproportionation of HG2

due to a high concentration of HG2 and dynamics in the large

cages.⁷⁶Assuming HG2 is preferentially located inside the

Figure 2. Mobility experiment to prove firm anchoring of HG2 on the silica support

cages and its volume covers a minor part of the total volume

(11 %)⁷⁷, the volumetric concentration in the pores of KIT-5 can

be 35 times higher compared to fibre-like SBA-15 (1000 mol Ru

m^-3vs. 29 mol Ru m^-3). HG2 is more evenly spread in fibre-like

SBA-15, indicating that the low activity of this catalyst is likely

caused by low diffusion through the corrugated pore system.

KIT-6 and MCM-48 (Table 1, entry 11 and 12), cubic Ia3d

structures with a 3D channel network, have similar TOF_i’s of 5

and 5.5 s^-1.10^-2respectively, which is lower than one-

dimensional hexagonal structures like MCM-41 and SBA-15.

Despite their 3D pore network, pore diffusion might be slower

due to a higher tortuosity (τ). For instance, linear hexagonal

structures like MCM-41 and SBA-15 (Table 1 , entry 13) have a τ

under varying solvent and substrate concentrations.

Ten green pellets of HG2/silica gel were placed in a glass vial

and one unloaded white pellet of silica gel was added. Three

identical sets of such eleven pellets were brought into contact

with a 5 ml toluene solution (A), a substrate mixture of 0.05 M

cis-cyclooctene in hexane (B), and a concentration of 0.8 M cis-

cyclooctene in hexane (C), respectively. After 4 hours of

shaking and solvent evaporation, vial A contained 11 coloured

pellets, whereas vial B and C contained one unloaded white

pellet next to ten green pellets. In other words, HG2 leaches

from the silica support in toluene and is redistributed among

the eleven pellets. Under apolar conditions (hexane), and

This journal is © The Royal Society of Chemistry 20xx

J. Name., 2013, 00, 1-3 | 3

Please do not adjust margins

Catalysis Science & Technology

Page 4 of 18

ARTICLE

Journal Name

View Article Online

Table 1. Characterization of the supports and activity of the immobilized catalysts.

DOI: 10.1039/C5C_hY01897H

Entry

Material

Ru

(wt%)

-

Split

test^e

D_particle

D_meso

Morpho-

logy

-

Pore architecture

V_meso

V_micro

BET

TOF_i

(µm)^f

(nm)^g

(cm³ g^-1)

(m² g^-1) (s^-1.10^-2)

1

HG2^c

-

13

-

4

4.5

1.5

2

3

5

5.5

8

8.5

7.5

10

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Silicalite-1

AlPO₄-5

Si-VPI-5

Silica Gel^a

Silica Gel^b

KIT-5

SBA-16

SBA-15

SBA-3

KIT-6

MCM-48

SBA-15

C12-MCM-41

C16-MCM-41

C18-MCM-41

TUD-1

0.008

0.001

0.12

0.35

0.30

0.35

0.16

0.29

0.28

0.26

0.34

0.24

0.29

0.33

0.21

-

0.54

0.73

1.1

n.d.

6

Coffin

n.d.

MFI topology

AFI topology

VFI topology

Non-ordered

Cubic Fm3m

Cubic Im3m

Hexagonal p6mm

Cubic Ia3d

350

-

Needles

Irregular

Fibre-like

Platelets

Irregular

Sphere

Rope-like

Sphere

Irregular

+

2000

50

10-25

4-25

2-5

n.d.

0.72

0.30

0.14

0.98

0.68

0.72

0.51

0.34

0.67

0.85

1.1

n.d.

0

0.37

0.17

0.12

0

0.13

0

0.51

0

375

406

1054

535

791

1476

700

1600

1097

690

1333

1446

417

7^d

5.5^d

10

1.8

6

2.5

5.5

2.6

3.3

3.6

8-15

1-5

2-5

0.3-0.7

0.5-1

0.3-0.5

5-20

Cubic Ia3d

Hexagonal p6mm

Disordered

0

0.04

^aGrace Silica Gel 239 pellets; ^bSilica Gel 60 (Davisil Grade); ^c0.33 mol% HG2 compared to cis-cyclooctene; ^dpore entrance diameter (cage diameter not experimentally

e

determined: KIT-5⁷⁷= 6.8-8.3 nm and SBA-16⁸²=5.1-7.2 nm). – split: homogeneous activity; + split: heterogeneous activity. Reaction conditions: 0.05M cis-

cyclooctene; 5 mL hexane; 50 mg HG2/ support (pretreated at 150 °C); 35 °C. The homogeneous reaction (entry 1) is performed in toluene instead of hexane for

solubility reasons. ^fD_particle= particle diameter. ^gD_meso= mesopore diameter. ^hError in TOF_idetermination is < 0.5 s^-1.10^-2.

independent of substrate concentration, HG2 remains this is the highest reported TON for the ROMP of non-purified

anchored on the support as no migration was observed cis-cyclooctene under ambient conditions with an immobilized

throughout the liquid reaction mixture. Physisorption of HG2 HG2 complex.

on silica is thus truly heterogeneous, provided that the solvent

is apolar with restricted solubility of HG2.

Thermal treatment of the support

Surface hydroxyl groups were reported essential to anchor

HG2, but this was not investigated.⁵⁸Thermal treatment of

Stability of the immobilized catalyst – A continuous experiment

The experiment in Figure 2 already showed a firm anchoring of silica materials removes physisorbed water molecules from the

HG2 on silica support under reaction circumstances, but a surface and changes the chemical composition by a surface

continuous experiment is a better measure of catalyst dehydroxylation process forming siloxane moieties.⁸³If surface

robustness. For the actual experiment, 110 mg of a 0.20 wt%

hydroxyls are important in HG2 immobilization, thermal

Ru loaded HG2/TUD-1 (pellets) were packed in a reactor treatment of the parent support material should have a

together with quartz-wool and glass beads. The catalyst bed substantial influence on catalyst immobilization. Therefore,

had dimensions of 0.94 x 3.5 cm. A solution of 0.4 mol L^-1cis- prior to HG2 immobilization, MCM-41 (3.3 nm pores) and TUD-

cyclooctene in hexane was pumped through the reactor at 1 were subjected to elevated temperature treatments, ranging

room temperature dosed at 38 mL g_cat^-1h^-1. The reactor outlet from 150 °C to 900 °C. Samples without treatment, kept at

was sampled at regularly times. The accumulated turnover 25°C, were used as a reference. The immobilization process

number (TON) is represented in Figure 3.

was conducted under identical conditions as before. The initial

activities of variously pre-treated HG2/TUD-1 and HG2/MCM-

41, measured as TOF_i, in metathesis of cis-cyclooctene are

compared in Figure 4. The corresponding kinetic profiles,

plotting conversion in time, are shown in the supporting

information (Figure S3 and S4). Thermally treating MCM-41

leads to a higher activity, the TOF_iincreasing with elevating

temperature. The activity gain is most pronounced after

heating at 150°C, while a further gradual increase is observed

up to 700°C. Overall, the TOF_iincreases from 2.5 s^-1.10^-2for the

untreated up to 10.5 s^-1.10^-2for the 700 °C treated MCM-41.

Treatment at 900 °C also lead to an active catalyst, however

considerable Ru leaching into solution is observed with this

catalyst. The same trend is observed for TUD-1, though the

preferred treatment here is below 400 °C. Higher treatment

temperatures for TUD-1 led to inferior catalytic results,

accompanied by a fast green-to-brown coloration of the

Figure 3. Accumulated TON in a continuous experiment of HG2/ TUD-1. To the right,

the set-up of the reaction with the direction of the feed flow.

Initially, about 40 % cyclooctene is converted, but the

conversion decreases to 25 % after 200 min. and faded to 3 %

at the end of the reaction after 540 min. At this point, a TON of

18000 was obtained, corresponding to a conversion of 39 gram

cyclooctene per gram catalyst. To the best of our knowledge,

4 | J. Name., 2012, 00, 1-3

This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

Page 5 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

catalyst. A spectroscopic study was performed to rationalize role of silanols, a plot of the total silanol density per Ru site

View Article Online

DOI: 10.1039/C5CY01897H

against the initial activity of HG2/TUD-1 and HG2/MCM-41 was

these thermal treatment effects.

The temperature-induced dehydration and dehydroxy- constructed in Figure 5. Untreated materials TUD-1 (25°C; with

lation of the silica surface was therefore monitored by (near) 63 OH per Ru) and MCM-41 (25°C; with 145 OH per Ru) are

infra-red (IR) spectroscopy by changes in the _O-Hvibration represented by the blue dots and contain physisorbed water.

(3800-3000 cm^-1) and ( _O-H(4800-4200 cm^-1) overtone Dehydration of these materials at 50 °C under vacuum (at 10

+

domain at different temperatures. The data of the IR mbar) or at 150 °C (at atmospheric pressure) removes this

spectroscopic study are collected in Figure S5.

water, but retains the chemisorbed silanols. The chemical

equivalence of these two treatments was confirmed by NIR

diffuse reflectance spectroscopy (Figure S6).⁸⁵As this

dehydration step led to the sharpest activity increase, as

visualized by the blue arrows in Figure 5, presence of water is

clearly disadvantageous for metathesis catalysis, but also, it

can be concluded that the type of silanol is less crucial for the

catalytic activity. As the observation of the typically green

colour indicates a stable HG2 on the untreated supports, their

lower activity is not caused by catalyst deactivation, but rather

polarity effects, retarding cis-cyclooctene diffusion in pores

filled with water is the cause of the slow catalysis. In fact, it

may be estimated by TGA and surface area measurements that

water covers about 20 % and 60 % of the mesopore surface of

MCM-41 and TUD-1, respectively.

Figure 4. Influence of a thermal treatment of the support on the TOF_iof HG2/TUD-1

(0.21 wt% Ru) and HG2/MCM-41 (0.30 wt% Ru). ¹This sample shows considerable Ru

leaching, therefore overestimating the contribution of heterogeneous TOF. Reaction

conditions: 0.05 M cis-cyclooctene; 5 mL hexane; 50 mg HG2/support; 35 °C.

Quantification of the silanol density was accomplished by

integrating the (ν+δ)_O-Habsorption bands, as the molar

integrated absorption coefficient of these bands ε_(ν+δ)O-Hdoes

not vary with the nature and concentration of silanol groups.⁸⁴

At room temperature the surface is completely hydroxylated,

together with a monolayer of physically adsorbed water, while

drying the support at 150 °C and atmospheric pressure (or 50

°C under vacuum) removes the hydrogen-bonded water and

leaves the hydroxyl groups intact. At higher temperature

treatments, geminal and vicinal silanols gradually disappear in

favour of isolated silanols and surface siloxanes.⁸³The silanol

density per gram of TUD-1 and MCM-41 together with the

ratio of silanol density per surface area, of the two catalysts,

i.e. TUD-1 to MCM-41, are represented in Table 2.

Figure 5. Influence of molar ratio of silanols to Ru on the initial activity of the ROMP of

cis-cyclooctene when physisorbed water is present (blue dots), or with

a fully

dehydrated support (red dots). Reaction conditions: 0.05 M cis-cyclooctene; 5 mL

hexane; 50 mg HG2/MCM-41 and TUD-1 (pre-treated at different temperatures;

loading of 0.30 wt% Ru for HG2/MCM-41 and 0.21 wt% Ru for HG2/TUD-1); 35 °C.

Table 2. Silanol densities (mmol OH g^-1) of TUD-1 and MCM-41 after

pretreatment at elevated temperatures.

25 °C

(mmol

OH g^-1)

150 °C

(mmol

OH g^-1)

400 °C

(mmol

OH g^-1)

550 °C

(mmol

OH g^-1)

700 °C

(mmol

OH g^-1)

Experimental verification of the water effect is accomplished

by intentionally adding the amount of water, that is removed

by the thermal treatment from 25 °C to 150 °C, to a water-free

HG2/MCM-41 catalyst (pre-dried at 150°C). Addition of water

indeed lowered the TOF_ifrom 8.5 s^-1.10^-2to 6.0 s^-1.10^-2in

accord to the above considerations. The red dots in Figure 5

represents the activity of thermo-treated HG2/TUD-1 and

HG2/MCM-41. The thermal treatment was varied between 50

°C (10 mbar vacuum) and higher temperatures (atmospheric

pressure). From the right to the left hand-side, the silanol

TUD-1

1.30

4.35

0.96

1.27

4.35

0.93

0.74

3.22

0.73

0.53

2.34

0.73

0.48

1.59

0.96

MCM-41

TUD-1:

MCM-41^a

^aratio OH density of TUD-1:MCM-41 (calculated in mmol.m^-2). TUD-1: 417 m² g^-1;

MCM-41: 1333 m² g^-1.

The silanol density per surface area of the parent material is density decreases as a consequence of a dehydroxylation

comparable for the two silica materials and corresponds to 2 process. Clearly, high activities are observed as soon as water

OH per square nm, which is in accord with literature.⁸³As is removed, leaving sufficient content of silanols, viz. at least

expected, the silanol density remains intact after heating until more than 40 OH per Ru. Treating the parent TUD-1 sample at

150°C, but decreases with elevating temperature, albeit temperatures higher than 400 °C substantially reduces the

differently for both silicas. Considering the potential anchoring catalytic activity, whereas a treatment of the MCM-41 support

This journal is © The Royal Society of Chemistry 20xx

J. Name., 2013, 00, 1-3 | 5

Please do not adjust margins

Catalysis Science & Technology

Page 6 of 18

ARTICLE

Journal Name

at 900 °C led the catalytic activity unchanged, albeit gram catalyst, as a measure of Si-NH_2,as well as the

View Article Online

accompanied with considerable complex leaching in the latter corresponding number of siloxanes and^Dt^Oh^Ie^{: 1}a^0.v¹e⁰³ra^9/g^Ce^5Cd^Yis⁰t¹a⁸n⁹⁷c^He

case. Neither of the two phenomena originate from a of silanols on the surface of TUD-1 and MCM-41, are

temperature-induced collapse of the pore structure, as N₂presented in Table 3. TUD-1, though treated at lower

physisorption proved an intact structure for TUD-1 treated at temperature (here: 550°C), shows a substantial amount of

700 °C and the characteristics of MCM-41 mainly remained strained siloxanes, whereas their abundancy is ten times less,

unchanged up to 900 °C (although some pore shrinkage even for a 900°C treated MCM-41 sample. The high content of

occurred at 900 °C) (Table S1), in accord with literature- reactive siloxanes on TUD-1 may therefore be considered to

reported data of Si-MCM-41 under the same heating cause deactivation of HG2. Though the thermally treated

conditions.⁸⁶As the decrease in activity for HG2/TUD-1 at MCM-41 being free of siloxanes, keeps HG2 intact in accord

400°C treatment is accompanied by a rapid green-to-brown with its green colour, the treatment at 900°C causes

colour change, we hypothesized that the presence of strained considerable leaching of the Ru complex. The lower affinity to

siloxane bridges causes catalyst deactivation. Such reactive HG2 is likely due to the lesser amount of remaining silanols,

siloxanes have indeed been described as reactive sites on silica isolated at too large distance from each other (2.5 nm).

surfaces in different reactions^{87, 88}, as well as being responsible

Table 3. NH₂contents on TUD-1 and MCM-41.

for decomposition of Ru complexes due to Ru-O-Si bond

Siloxanes^a

(% as OH)

3.8

Distance silanols

formation.⁷⁶The concentration of such siloxanes typically

increases with rising of the pretreatment temperatures, but

why the observed catalyst deactivation is only prominent for

TUD-1 is unclear. As a difference in siloxane reactivity may

explain this phenomenon, the presence and reactivity of

siloxanes were monitored.

NH₃reacted

(mmol g^-1)

0.048

(nm)

1.3

TUD-1

(550 °C)

MCM-41

(900 °C)

0.0037

0.1

2.5

^asiloxanes reacted with NH₃divided by total silanol density.

An easy chemical method to determine the strained

siloxanes involves its reaction with NH₃under continuous flow

at elevated temperature to form Si-NH₂, after passivating the

reactive surface hydroxyl groups with dichlorodimethylsilane

(DCDMS).^89-91Finally, hydrolysis of the unstable Si-NH₂in

water, forming a silanol and ammonia, and measuring of the

pH allows a good estimate of the reactive siloxanes. The

chemical surface transformations during the procedure were

monitored with FT-IR spectroscopy (Figure 6).

In conclusion, for stable metathesis with supported HG2, the

mesoporous silica should be treated at elevated temperatures,

preferably for MCM-41 and TUD-1 at 50 °C under vacuum (10

mbar) or at 150 °C under atmospheric pressure, prior to

immobilization, to remove physisorbed water and to conserve

a high amount of silanol anchors for immobilization. Therefore,

care has to be taken to prevent formation of reactive siloxane

bridges during thermal treatment of the support as to keep

the surface unreactive toward HG2. Remarkably, the necessity

of silanol groups for stabilization of HG2 is in contrast with

previous studies that reported the detrimental role of silanol

groups on the stability of the Grubbs 1^stgeneration complex.^50,

76, 94

Adsorption isotherms of HG2 on mesoporous silica

As a high affinity of the support for HG2 is desirable to prevent

decomposition and leaching, adsorption isotherms of HG2 for

150°C dried MCM-41 (3.3 nm pores) and TUD-1 were

determined and investigated (Figure 7). A raise of HG2

concentration in the sorption solution initially results in a

linear increase of immobilized HG2 (Figure 7A), whereby 4 out

of 5 HG2 molecules adsorb from the solution on the support,

with an adsorption rate of 4 µmol HG2 per gram per minute.

This corresponds to 1 and 3 µmol HG2 per mmol OH per

minute for MCM-41 and TUD-1, respectively. At higher

concentrations, higher loadings are observed on MCM-41

compared to TUD-1, which is in line with the higher surface

area and thus anchoring capacity of MCM-41 per silica weight.

A sharp deflection from linear uptake is observed for both

TUD-1 and MCM-41 at Ru equilibrium concentrations above

0.7 and 1.7 mM, respectively, the adsorption process becomes

thus less efficient at higher HG2 concentrations. Though pore

blockage by HG2 itself cannot be excluded as explanation for

Figure 6. IR spectra of A: TUD-1 treated at 150 °C; B: TUD-1 after passivation with

DCDMS; C: TUD-1 after passivation with DCDMS and reaction with NH₃.

TUD-1, pretreated at 150 °C, shows an intense vibration band

at 3745 cm^-1(ν_sOH) and

a broad shoulder at lower

wavenumbers due to hydrogen bonding phenomena (Figure

6A). After treatment with DCDMS, the 3745 cm^-1band

completely disappears at the expense of three new features at

2970, 2901 and 1460 cm^-1, corresponding to ν_s(CH₃), ν_as(CH₃)

and δ(CH₃) respectively (Figure 6B).⁹²Ammonia reaction with

siloxanes is evidenced by the characteristic absorption at 3536,

3452 and 1550 cm^-1, attributed to ν_as(NH₂), ν_s(NH₂) and δ(NH₂)

vibrations, respectively (Figure 6C).⁹³Subsequent hydrolysis

and pH measurements allows determination of the reactive

siloxane content. The amount of ammonia recuperated per

6 | J. Name., 2012, 00, 1-3

This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

Page 7 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

View Article Online

DOI: 10.1039/C5CY01897H

Figure 7. A) Carrier loading (wt%) of Ru on TUD-1 and MCM-41 at different concentrations of HG2 in toluene. The kinetic profile of immobilization is presented in the inset. For

both materials the adsorption equilibrium was reached after 30 minutes, and therefore the adsorption isotherm experiments were run for 45 minutes to obtain

thermodynamically sound HG2 uptake values. B) Modified adsorption isotherms of Ru on TUD-1 and MCM-41. Both supports are pre-treated at 150 °C prior to immobilization.

the decreasing uptake efficiency with increasing HG2 loading, with highly loaded HG2 complex, as long as HG2 instability and

the sharp distinctive shape of the isotherm of both materials pore diffusion limitations (PDL) are absent. The optimal active

rather points to the existence of two distinct adsorption sites, site density was therefore searched for by systematically

one with high affinity (occupied at low loadings) and the other increasing HG2 loadings, ranging from 0.1 to 2.5 wt% Ru, on

with low affinity for HG2 (at higher catalyst loadings). In the 150°C dried MCM-41 (3.3 nm pores) and TUD-1. The total

modified adsorption isotherm (Figure 7B) the uptake profile of catalyst weight in the catalytic tests remained constant (50

HG2 is normalized for the amount of silanols (and at the same mg). All immobilized catalysts were bright green and thus not

time for the surface area as MCM-41 and TUD-1 have equal affected by deactivation. Whereas both supports show

surface silanol densities of 0.003 mmol m^-2), rather than for increasing conversion rates with higher HG2 loadings (Figure

the weight of catalyst. The deflection of the affinity sets in at S7), the activity per catalytic complex (TOFi), as shown in

8.2.10^-5and 1.2.10^-4mmol HG2 m^-² (1.22.10^-6and 1.90.10^-6Figure 8, indicates a more steep decrease of the TOFi for

mol HG2 mmol^-1OH g) for MCM-41 and TUD-1, respectively, HG2/MCM-41. Such behaviour is less obvious for HG2/TUD-1

corresponding to 1.1 wt% Ru for MCM-41 and 0.5 wt% Ru for as only a slight drop of the TOFi to 9 s-1.10-2 is observed at 1

TUD-1. The total adsorption capacities at room temperature µmol HG2, corresponding to 2.1 wt% Ru.

were estimated to be 3.2 and 1.8.10^-4mmol HG2 m^-2for TUD-1

and MCM-41, respectively.

Since both TUD-1 and MCM-41 materials show

a

comparable OH surface density (Table 2), a difference in

surface uptake behaviour of the respective silica was not

expected. The plot nevertheless reveals a higher affinity of the

TUD-1 silica surface and thus thermodynamically more

favourable adsorption sites are present when compared to

that of MCM-41. ²⁹Si MAS NMR of the surface of both support

materials, pretreated at 150 °C, was therefore studied. By

distinguishing the signal for isolated and geminal silanols,

according to Ide et al.⁹⁵, clear structural differences were

observed: 150°C treated TUD-1 has a substantially higher

fraction of geminal silanols (Q₂/Q₃+Q₂= 0.19) compared to

similarly dried MCM-41 (Q₂/Q₃+Q₂= 0.10), suggesting that HG2

more favourably interacts with that silanol type. The amount

of geminal silanols exceeds the linear uptake of HG2 with a

factor 5 for TUD-1 and 3.5 for MCM-41, which corresponds to

a higher density of 5.7 .10^-4mmol m^-2TUD-1 compared to 3.1

.10^-4mmol m^-2on MCM-41. Awaiting for additional arguments

for the (enthalpic) difference in surface adsorption chemistry,

steric (entropic) factors due to space restrictions, most

pronounced in the smaller pores of MCM-41, cannot be

excluded.

Figure 8. Influence of the abundance of ruthenium per catalyst volume on the initial

activity (TOF_i). PDL= pore diffusion limitations. HE= heterogeneous activity; HO=

homogeneous activity. Reaction conditions: 0.05 M cis-cyclooctene; 5 mL hexane; 50

mg HG2/TUD-1 or HG2/MCM-41 (pre-treated at 150 °C); 35 °C.

Therefore, loadings up to 2 wt% are possible on TUD-1 without

losing catalyst efficiency. Indeed, the experimental verification,

where the TOF_iis compared between two reactions with

catalysts having a different catalytic loading, is called the Koros

and Nowak criterion. According to this criterion, PDL is

excluded whenever the turnover frequency is invariant of the

density of the active sites.⁹⁶This criterion is only valid under

the following conditions: 1) isothermal reactions with

negligible thermal gradients; 2) every added catalytic complex

should be identical to the previous one and represent one

catalytic site.⁹⁷As these conditions are satisfied in our work,

the rate of cyclooctene metathesis with HG2/TUD-1 up to 2

wt% Ru is determined by the catalytic reaction and not by

Influence of active site density on catalytic activity

Heterogeneous metathesis reactions are preferably carried out

This journal is © The Royal Society of Chemistry 20xx

J. Name., 2013, 00, 1-3 | 7

Please do not adjust margins

Catalysis Science & Technology

Page 8 of 18

ARTICLE

Journal Name

mass transport. Higher loadings though will render the agreement with experimental data. Thus, at concentrations

View Article Online

DOI: 10.1039/C5CY01897H

catalysis inefficient due to PDL and thus artificial lower than 0.21 M, mostly cyclic oligomers are formed in

underestimation of TOFs.

HG2/MCM-41 has a more steep activity decrease from 8.5 employment of the homogeneous HG2 complex for the

s^-1.10^-2at 1.5 µmol HG2 (0.30 wt% Ru) to 6 s^-1.10^-2at 12 µmol formation of

library of macrocycles was already

HG2 (2.4 wt% Ru), and seems thus more susceptible to mass demonstrated multiple times in literature.^103-105

presence of homogeneous metathesis catalysts. The successful

a

transport phenomena than HG2/TUD-1. In fact, this conclusion

From our interest in selectively synthesizing cyclic

is in line with the smaller pore size of MCM-41 (3.3 nm) oligomers, we evaluated this reaction at low monomer

compared to that of TUD-1 (8-15 nm). The decrease in activity concentrations with

a homogeneous HG2 catalyst and

could be attributed to lower effective cyclooctene compared the catalytic outcome (product selectivity) with that

a

concentration in the pores leading to a lower activity on one of a reaction in presence of the immobilized complex on

hand, or to a less accessible active Ru site caused by pore porous silica under batch conditions. The potential of this

obstruction on the other hand. The true mechanism behind reaction was proposed earlier by us¹⁰⁶and later by research

this activity decrease was unravelled by carefully investigating groups centred around BASF⁶⁶and described under

product selectivity and will be discussed in the next paragraph. continuous-flow conditions.⁶⁰As heterogeneous reactions are

Note that extrapolation of the catalytic activity towards often confronted with confinement and diffusion issues,

low Ru loading, thus free of PDL, shows fairly similar TOFs kinetic dissimilarities between homo- and heterogeneous

irrespective of the mesoporous silica type, the value being metathesis catalysed reactions may be expected.¹⁰⁷This will

close to that of the homogeneous HG2 catalyst. This means especially be likely when linear oligomeric and polymeric

that despite dissimilarities of affinity and precise anchoring of products are involved in heterogeneous ROMP reactions.

HG2, no essential impact on the catalytic activity between the Careful product analysis was therefore attempted during the

two supported HG2 catalysts is observed.

metathesis of cyclooctene in presence of homogeneous HG2

The original activity of the highly loaded HG2/MCM-41 (2.4 and HG2/MCM-41 (3.3 nm pores; 0.30 wt% Ru). Cyclic

wt% Ru) on the graph in Figure 8 seems erroneous, as a TOF oligomers up to the pentamer (C40) were analysed with GC,

well below the observed 6 s^-1.10^-2was expected due to PDL whereas the larger non-volatile macrocycli (C48-C56) together

contribution. A split test and ICP-AES analysis of the reaction with linear oligomers and polymers were monitored by GPC to

solution pointed to significant leaching of HG2 for this highly complete the mass balance. The molecular weight range of

loaded MCM-41. Subtracting the homogeneous activity of HG2 these fractions is provided in the method section (catalytic

gives a true heterogeneous contribution (HE) of 4.3 s^-1.10^-2; reactions) and extra information about GPC analysis is

which better fits the expectations. Besides confirming mass reported in the supporting information (Figure S8).

transfer issues with MCM-41, this experiment shows that

stably anchoring of high HG2 loadings are better accomplished

on TUD-1.

The experimental results of a homogeneous reaction (0.05

cis-cyclooctene) are presented in Figure 9. At low

M

cyclooctene conversion, mostly linear oligomers and small

cyclic oligomers are formed with a minor amount of linear

polymer. As the reaction proceeds, the linear polymer and

oligomer fraction is converted almost exclusively to cyclic

oligomers (Figure 9A). Extrapolation of the initial development

of the reaction to zero conversion indicates that linear

polymerization largely overrules the initial direct formation of

cyclic oligomers from cyclooctene (Figure 9A and B). The cyclic

oligomers are predominantly formed through back-biting of

the growing linear oligomer and polymer chains which, once

formed, are quickly converted. This kinetic reaction pattern

was similar to the ones observed with Grubbs II and Grubbs-

Nolan catalysts.¹⁰⁴The cyclic oligomer fraction maximizes at 46

% conversion (selectivity of 98 %), and the selectivity

decreases only slightly to 94 % at full conversion in favour of

larger cyclic oligomers (4 %) and linear oligomers (2 %) upon

reaching thermodynamic equilibrium (Figure 9B). The majority

of the cyclic fraction consists of C16-C56 cyclic oligomers and

the weight distribution is given in Figure 9C, which was

Kinetic dissimilarities between supported and non-supported HG2

The enthalpic driving force of cyclic molecules like cyclooctene

enables them to undergo irreversible ROMP metathesis,

thereby releasing ring-strain. The formation of linear polymers

is therefore accompanied by the formation of low-molecular-

weight cyclic oligomers, which originates from the direct cyclo-

oligomerization of cyclooctene or from polymer back-biting.⁹⁸

The final ring-chain equilibrium is, besides the catalyst and

ring-strain of the cycloolefin, mostly dependent on the

100

monomer concentration.^99,

In 1950, Jacobson and Stock-

mayer (J-S) developed a theory of ring-chain equilibria,

including a prediction of the statistical distribution of cyclic

structures at equilibrium. While at low monomer

concentrations mostly cyclic oligomers are formed, linear

polymers are formed at high concentrations. The highest

concentration whereby no linear polymers were formed at

equilibrium situation was defined as the critical monomer

concentration.¹⁰¹This model was later refined by Kornfield et

al., taking into account the ring-strain phenomenon, which was

neglected in the J-S theory.¹⁰²(For further theoretical

background, see supp. info). The critical concentration [M]c,∞

was hereby calculated to be 0.21 mol/L for cyclooctene, as in

consistent with equilibrium distributions obtained by different

104, 108

authors.^60,

The observed equilibrium ring distribution

was compared to the one predicted by the original J-S theory,

as indicated by the black bullet points in Figure 10C, which

foresees a decrease of the molar cyclic oligomer concentration

C_i(with degree of polymerization i) with increasing ring size

8 | J. Name., 2012, 00, 1-3

This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

Page 9 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

View Article Online

DOI: 10.1039/C5CY01897H

Figure 9. Homogeneous RO-RCM with cis-cyclooctene. A) Distribution of the cyclic oligomer/ linear oligomer/ linear polymer fraction in function of conversion. B) Yield and

selectivity of C16-C56 with increasing conversion. C) Distribution (wt%) of C16-C56 cyclic oligomers. Reaction conditions: 0.05 M cis-cyclooctene; 16 mL toluene; 3 mg HG2; 35 °C.

proportional to i^-5/2.The equilibrium distribution observed here instance, at a cyclooctene conversion of 5 %, the selectivity

clearly does not obey this theory, since the dimer C16 is towards cyclic oligomers decreases from 54 % at 0.025 M, to

overestimated and its concentration (on weight base) is 39 % for 0.1 M and 26 % for 0.3 M, in favour of (long) linear

surpassed by the trimer C24, likely due to enthalpic reasons as chains (Figure S11A). As polymer growth is depending on

C16 has a higher ring-strain compared to C24 (which is monomer concentration, in contrast with polymer back-biting,

neglected in the J-S theory).¹⁰²Furthermore the proportion of which is the main route to cyclic oligomers under these

larger cyclic oligomers exceed the predicted values. Based on reaction conditions, it is conceivable that a higher monomer

the experimental data, relative molar equilibrium constants K_iconcentration leads to a higher degree of polymerization

for cyclic oligomers with a degree of polymerization i were initially. Moreover, the lower the initial monomer

determined and compared to those obtained by the J-S theory concentration, the higher the probability of a direct cyclic

and the refined model of Kornfield and coworkers (Figure S9). oligomer formation. The ratio between the C16-C56 fraction

For cyclic oligomers larger than the trimer C24, the predicted likewise changes with increasing monomer concentration. In

slope of -2.5 is well approximated; the deviation of K_ifrom particular, the contribution of C40-C56 cyclic oligomers

theoretical values is larger for smaller cycles.

increases (at 5% conversion) (Figure S11B). Although the

The reaction was performed as well at other monomer selectivity towards C16-C56 is lower at higher concentrations

starting concentrations. At low concentrations, i.e. 0.025 and (0.1 and 0.3 M), its molar concentration is still higher

0.1 M, C16-C56 cyclic oligomers are formed almost exclusively compared to e.g. 0.05 M., enhancing the self-metathesis of

at equilibrium, reaching 99 and 91 % product selectivity, cyclic oligomers (C16+16=C32 or C16+C24=C40), thus affording

respectively. At higher concentrations (0.3 M), this fraction C40-C56-enriched concentrations.

drops to 65 %, which is reasonable since the initial monomer

Impacting the kinetic product distribution is not limited to

concentration exceeds the critical concentration of 0.21 M. A variation of the feed concentration, as the catalyst itself may

graphical representation of this ring-chain equilibrium at also contribute hereto. Kavitake et al. showed that with an

different concentrations is shown in Figure S10. The unsymmetrical NHC-modified catalyst a 2:1 ratio of C16:C24

distribution within the C16-C56 cyclic oligomer fraction seems could be obtained at low conversions, attributed to the dual-

also related to this critical concentration. The results are site configuration of the homogeneous catalyst.¹⁰⁴Depending

presented in Table 4 . At concentrations below [M]c,∞, ring on the steric confinement, the active catalyst selectively

distributions are identical, while at 0.3 M, larger cyclic discriminates between ring-closing and propagation. After 30

oligomers (C40-C56) are more prominent. This is due to the % conversion of 0.025 M cyclooctene, C16 and C24 are formed

fact that the critical concentration of cyclic oligomers increases with 57 and 27 % selectivity, respectively, while our observed

proportional to their ring size.¹⁰⁹

ratio of C16:C24 is much lower (1.5:1) at 20 % conversion of

0.025 M cyclooctene with the symmetrical HG2 catalyst.

Table 4. C16-C56 cyclic oligomer distribution at equilibrium starting from

The homogeneous conversion of the RO-RCM of cyclooctene

was compared to a heterogeneous reaction with HG2/MCM-

41 (0.30 wt% Ru), as displayed in Figure 10. Initially, linear

oligomers and polymers are formed next to a large fraction of

C16-C56 cyclic oligomers (Figure 10A). Compared to the

unsupported HG2, a higher initial C16-C56 selectivity was

obtained (56 % at a conversion of < 2%). As the reaction

proceeds, the linear polymer fraction reaches a maximum of

25 % at 33 % conversion and undergoes further back-biting to

cyclic oligomers at prolonged reaction times. At full

conversion, the reaction mixture comprises 87 % of C16-C56

cyclic oligomers and 8 % of linear oligomers (mass balance of

95 %), while larger cyclic oligomers (> C56) were neglected

because of their low abundance. The larger fraction of linear

different cis-cyclooctene concentrations.

C16

25

C24

29

30

C32

20

C40

14

17

C48

8

10

13

C56

4

5

6

0.025 M

0.05 M

0.1 M

23

28

0.3 M

18

25

Reaction conditions: 0.025–0.3M cis-cyclooctene; 16 mL toluene, 3 mg HG2; 35°C.

At low conversions, the effect of concentration on the product

distribution is even more outspoken. While at low monomer

starting concentrations cyclic oligomers are predominantly

formed through backbiting of (long) linear chains, as indicated

for 0.05 M in Figure 9A and B, the initial cyclic oligomer

concentration decreases at higher concentrations. For

This journal is © The Royal Society of Chemistry 20xx

J. Name., 2013, 00, 1-3 | 9

Please do not adjust margins

Catalysis Science & Technology

Page 10 of 18

ARTICLE

Journal Name

View Article Online

DOI: 10.1039/C5CY01897H

Figure 10. Reaction profile of a heterogeneous RO-RCM with cis-cyclooctene with HG2/MCM-41. A) Distribution of the cyclic oligomer/ linear oligomer/ linear polymer fraction in

function of conversion. B) Yield and selectivity of C16-C56 with increasing conversion. C) Distribution of C16-C56 cyclic oligomers (wt%). Reaction conditions: 0.05 M cis-

cyclooctene; 16 mL hexane; 160 mg HG2/MCM-41 (pretreated at 150 °C; 0.30 wt% Ru); 35 °C.

oligomers in presence of the supported HG2 at equilibrium As a consequence, C16 cyclic oligomers tend to undergo self-

situation, viz. 8 % vs. 2 %, could be ascribed to secondary metathesis, forming C32 (schematic representation in Figure

intermolecular chain transfer reactions. Hereby a rearrange- S12). A higher C32:C24 ratio is thus expected and also

ment of double bonds occurs between two different nearby observed with the immobilized catalyst, as supported by the

growing polymer chains, a phenomena that seems plausible in data in Figure 11.

the pores of a heterogeneous catalyst where adjacent Ru-

polymer chains can interact closely.¹¹⁰The previous

observations though were surprising as the formation of linear

polymers was not reported in earlier studies with

heterogeneous ROMP of cyclooctene, neither was this

60

observed in that extent in our homogeneous reaction.^59,

These results do however demonstrate the potential of

heterogeneous metathesis in depolymerization reactions of

e.g., polybutadiene to multiple unsaturated macrocycles.

Figure 10B displays the yield and selectivity to the cyclic

fraction in function of time, while the product distribution

Figure 11. Distribution of C16-C40 cyclic oligomers for

a homogeneous and

within the C16-C6 fraction is presented in Figure 10C. Although heterogeneous catalysed metathesis of cis-cyclooctene (0.05 M). Reaction conditions:

See Figure 9 and Figure 10.

the distribution practically equals the one of the homogeneous

reaction, the contribution of C48 is slightly higher

These kinetic differences between homo- and heterogeneous

HG2 catalysis disappear at full conversion as to obey the

thermodynamic equilibrium distribution.

homogeneously. A molar ratio of C40:48 of 1.3 was obtained

with the soluble HG2, and increased to 1.8 in presence of the

supported HG2.

The equilibrium situation attained with HG2/MCM-41 was

further compared with the homogeneous reaction at different

concentrations. Overall, a lower selectivity towards the cyclic

C16-C56 fraction is attained at equilibrium with the supported

HG2, in favour of linear oligomers, which is clearly visible in

Figure 12. (Eq. distribution of cyclic oligomers in Figure S13).

While at 0.025 and 0.05 M this linear fraction is rather small, it

Most outspoken dissimilarities between the homo- and

heterogeneous catalysed reaction, i.e. the higher initial C16-

C56 cyclic oligomer selectivity and the presence of linear

polymer up to 25 %, are derived from the peculiar metathesis

catalysis occurring in the pores of the catalyst. Diffusion

problems of cyclooctene into the pores (vide supra, Figure 8)

lead to a lower effective cyclooctene concentration in the

pores, stimulating the direct formation of cyclic oligomers

instead of linear chains. The C16-C56 fraction reaches

therefore a higher initial selectivity of 60 % (5 % conversion) in

comparison with the unsupported HG2. The presence of

diffusion problems of cyclooctene into the pores implies that

the same is valid for the diffusion of products out of the pores

into the bulk. As a result, the formed cyclic oligomers tend to

react further to linear oligomers and polymers. Indirect

formation of cyclic oligomers through back-biting is hindered

due to pore confinement effects and the linear chains

accumulate in the pores.

Pore diffusion also creates a product shift within the C16-

C56 cyclic oligomer fraction. The difficult diffusion of

cyclooctene (C8) in and the diffusion of larger cyclic oligomers

out of the pores induces a lower C8:C16 ratio inside the pores.

Figure 12. Equilibrium distribution of a homo- and heterogeneous metathesis of cis-

cyclooctene at different monomer concentrations. Inset: GPC chromatogram of a

heterogeneous 0.1

M reaction. LO: linear oligomer; HET: heterogeneous; HOM:

homogeneous. Reaction conditions: See Figure 9 and Figure 10.

10 | J. Name., 2012, 00, 1-3

This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

Page 11 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

reaches a value of 20 % after reaction with 0.1 M cyclooctene kinetic product distribution (Figure S15). Both_Vip_ewh_Ae_rn_tico_lem_Oe_nln_ina_e

DOI: 10.1039/C5CY01897H

(GPC, inset Figure 12) with a corresponding C16-C56 selectivity confirm a non-equilibrium state after reaction with highly Ru

of 70 % (mass balance of 90 %). Similarly to the reaction with loaded MCM-41.

0.05 M, kinetic dissimilarities were observed at low

Next, to further examine the relationship between the

cyclooctene conversions of reactions with 0.025 M and 0.1 M product distribution and pore restrictions, two other

cyclooctene. Results are summarized in Table 5. On one hand heterogeneous catalysts with more open pore systems were

the maximal yield of linear polymer increased with investigated as well, i.e. TUD-1 and KIT-5. Recalling the data of

concentration, and due to diffusion problems of cyclooctene in Table 1, HG2/KIT-5 has a cage-like pore system and showed a

the pores the initial selectivity towards C16-C56 cyclic much lower TOF_icompared to HG2/MCM-41, while HG2/TUD-

oligomers increased at 0.1 M. This effect was less outspoken at 1 has a three-dimensional pore system with 8-15 nm pores and

0.025 M. Product shifts within the C16-C56 fraction on the showed slightly higher TOF_i. The selectivity to C16-C56 cyclic

other hand were likewise observed at 0.025 and 0.1 M; a lower oligomers and linear chains was compared between the three

C16:C24 and C24:C32 ratio at a concentration of 0.025 M is catalysts and presented in Figure 13 . As shown before, TUD-1

apparent, while the contribution of C40 increased at 0.1 M. with its open structure and containing 0.21 wt% of Ru,

Kinetic distribution plots (analogue to Figure 11) were set up experiences almost no pore diffusion limitation, and indeed

for these concentrations in Figure S14.

the reaction pattern of the cyclic oligomers closely resembles

that of the homogeneous reaction (Figure 9A). The formation

of the C16-C56 fraction with HG2/KIT-5, in contrast, bear

resemblance to the pattern of HG2/MCM-41, but only attains

a final C16-C56 selectivity of 75 %. More surprisingly, almost

no linear chains are noticed. Obstruction of polymer in the

cages is the most obvious explanation impeding the polymer of

diffusing out of the silica particle. Only 75 % of the mass was

analysed in solution at the end of reaction, while the deficient

part was found on the spent catalyst using TGA analysis.

Table 5. Selectivity towards C16-C56 cyclic oligomers at 5 % conversion at

different monomer concentrations.

0.025M

55

53

0.05M

52

60

0.1M

39

45

Homogeneous

Heterogeneous

Reaction conditions: See Figure 9 and Figure 10.

Based on previous results, it can be deduced that the presence

of diffusion issues in confined systems may have a significant

contribution to the observed product distribution. Either

better mass transport in a more open pore architecture, like in

TUD-1, or lower Ru loadings should therefore bring the

catalytic outcome of the HG2 supported system silica closer to

the homogeneous one.

The impact of the Ru loading on the product selectivity was

analysed first, since Figure 8 indicated that the catalytic

activity per Ru decreases with increasing Ru density on MCM-

41. Reactions were therefore performed with 0.3, 0.6 and 1.2

wt% Ru on MCM-41 and their product distributions were

compared. At low cyclooctene concentrations, the

contribution of linear polymer augments with higher Ru-

loaded catalysts (at 20-30 % conversion). As was mentioned

above, cyclic oligomers are mainly formed through polymer

back-biting in absence of space restriction. But inside the

pores of MCM-41, polymer accumulation occurs because the

back-biting is slow due to pore restrictions, followed by a slow

diffusion of the cyclic oligomers larger than cyclooctene out of

the pores, meanwhile also reacting to linear polymers. With

increasing Ru loading, the probability of cyclic oligomers to

react further into linear polymer is therefore enhanced. As a

result, pore obstruction makes the active site less accessible

for cyclooctene resulting into lower TOF_i’s (vide supra, Figure

8). At the end of the reaction significant amounts of carbon

residue, expectedly hold inside the catalyst pores, were

analysed by TGA. Existence of such carbon residue prevents

the reaction of reaching a product distribution according to

the thermodynamic equilibrium, and this was indeed

manifested in the product distribution. At full conversion, the

yield of C16-C56 cyclic oligomers only amounts to 71 % with

0.6 wt% Ru and 42 % with 1.2 wt% Ru; moreover, the

perceived distribution within the C16-C56 fraction resembles a

Figure 13. Selectivity towards the C16-C56 fraction (upper figure) and linear chains

(lower figure) with HG2/TUD-1 (0.21 wt%), HG2/MCM-41 (0.30 wt%) and HG2/KIT-5

(0.31 wt%). Reaction conditions: See Figure 10.

The main differences between homo- and heterogeneous

metathesis of cyclooctene are summarized in the reaction

scheme in Figure 14. Linear polymerization dominates the

direct formation of cyclic oligomers under homogeneous

conditions, causing cyclic oligomers to be formed primarily

through back-biting of the linear polymer chains. This effect is

most outspoken at high cyclooctene concentration.

Heterogeneous reactions with low Ru loadings supported on

silica with large pores and an open pore architecture like TUD-

1, proceed similarly, and therefore sufficiently high contact

times are advised for reactions with these catalysts in

continuous fixed-bed plug flow reactors.

This journal is © The Royal Society of Chemistry 20xx

J. Name., 2013, 00, 1-3 | 11

Please do not adjust margins

Catalysis Science & Technology

Page 12 of 18

ARTICLE

Journal Name

Heterogeneous reactions under conditions of pore metathesis of methyl oleate (Table 6, ent r_Vy_iew6_A)_rticlt_eo_Onli9_ne-

DOI: 10.1039/C5CY01897H

a valuable intermediate for the

restriction (or high Ru loading), like in case of HG2/MCM-41, octadecene-1,18-dioate,

exhibit large differences hereto. They perform under production of polymers and fine chemicals, was also very

diffusional control leading to slower reaction rates. Because of successful, even at low catalyst concentration. Cross-

the diffusional regime, mixtures of linear and cyclic oligomers metathesis of methyl-10-undecenoate was accomplished at 50

are formed, which only turn into comparable cyclic oligomers °C, yet 2.5 mol% Ru was needed to reach full conversion (Table

at full conversion. Yet, the reaction proceeds slower (per Ru) 6, entry 7). CM of terminal olefins is known to be more

and therefore these catalysts are less interesting for challenging, as it lacks the ring-strain release of ROMP and the

production of cyclic rings from cyclooctene.

entropic driving force of RCM.¹¹¹Finally, the cross-metathesis

of butyl vinyl glycolate (Table 6, entry 8) was conducted. This

novel renewable substrate can be derived in the methyl ester

form from glycolaldehyde and tetroses.^2,59,60Transesteri-

fication with n-butanol makes this compound more soluble in

hexane. The dimeric product, which is formed in high yields at

50 °C with 1.2 mol% ruthenium, can be

a valuable

intermediate in the production of new bio-based polymers.⁴

Conclusion

2^ndgeneration Hoveyda-Grubbs (HG2) was immobilized on a

series of porous silica materials and tested in the metathesis of

cis-cyclooctene. High density of silanols are required to firmly

anchor HG2, geminal types showing the highest affinity. The

affinity of the silica surface for HG2 reduces with HG2 loading.

Loadings above 2 wt% should therefore be avoided as they

cause HG2 leaching. Cages in the silica pore structure should

also be avoided as they root to HG2 degeneration.

The metathesis reaction proceeds under chemical regime,

and therefore efficient in HG2, as long as the pore dimensions

are large enough, the pore structure sufficiently open and the

loading of HG2 not too high. Under such circumstances,

cyclooctene is converted to linear polymers, which undergo

conversion to the desired cyclic oligomers through back-biting.

Reactions rates may be as high as that of the homogeneous

reaction and product distributions are identical under

conditions of high contact times. If the above criteria are not

fulfilled, diffusional control will govern the reaction rate and

the product selectivity, leading to slower reaction rates and

mixtures of linear polymers and cyclic oligomers. Only under

very high contact times, the thermodynamic product

distribution, mainly containing cyclic oligomers, is obtained.

Presence of water seems not detrimental to HG2, but its

presence in the pores retards the metathesis reaction.

Therefore, water removal is advised, prior to HG2

immobilisation. Care has to be taken though to keep the

silanol density high to ensure firm anchorage of HG2 and to

avoid formation of reactive siloxanes, which chemically react

with and destroy HG2.

Figure 14. Proposed reaction scheme for the ring-chain distribution of cyclooctene with

HG2 (green) and HG2/MCM-41 (red) under steady state circumstances. Route A: direct

formation of cyclic oligomers from cyclooctene. Route B: polymerization of cyclooctene

to linear chains (n < m). Route C: back-biting of the polymer chain to cyclic oligomers.

Use of HG2/TUD-1 in several metathesis reaction types

Being the best catalyst of choice, HG2/TUD-1 was further

exploited in ROMP (Table 6, entry 1-3), ring-closing metathesis

(RCM) (Table 6, entry 4-5) and cross-metathesis (CM) reactions

(Table 6, entry 6-8) of other substrates. For cyclooctadiene

(Table 6, entry 2), the initial TOF was lower than its mono-

unsaturated counterpart (Table 6, entry 1), but cyclic

oligomers were selectively produced. The ROMP of

norbornene (Table 6, entry 3) was conducted at a lower

concentration of 0.005 M due to its high reactivity as a

consequence of the presence of internal bridges. Full

conversion was already reached after 15 minutes. The catalyst

also showed good activities in the ring-closing metathesis

reactions of diethyl diallyl malonate (Table 6, entry 4) and 1,7-

octadiene (Table 6, entry 5), as high TOF_i’s were obtained and

full conversion was reached after 90 minutes. These two

substrates were subjected to additional catalytic tests to

evaluate the catalysts robustness in RCM reactions at low

catalyst loadings. With 0.5 M of diethyl diallylmalonate and

1,7-octadiene, 0.06 mol % HG2/TUD-1 (0.21 wt% Ru) at 35 °C,

TONs of 1290 and 1350 were obtained, respectively. Self-

Overall, TUD-1 with its large pores and open structure,

loaded with 0.2 wt% Ru (1.2 wt% HG2), was recognized as the

most privileged catalyst to perform ROMP of cyclic olefins.

With this immobilized catalyst, a TON of 18.000 was reached in

a

continuous experiment under non-optimized ambient

conditions. This catalyst performed also excellent in other

metathesis types. In particular the catalyst appears performant

for conversion of biobased substrates like methyl oleate and

butyl vinyl glycolate, indicating that immobilized metathesis

12 | J. Name., 2012, 00, 1-3

This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

Page 13 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

Table 6. Various metathesis reactions (ROMP, RCM and CM) performed with HG2/TUD-1.^a

View Article Online

DOI: 10.1039/C5CY01897H

Entry

Substrate

Product

Time [min]

(T [°C])

90 (35)

Conv.[%]

[TOF (h^-1)]

96 [360]

Conv. Split

15 min [%]

40

Conv. Split

End [%]

40

1

2

90 (35)

15 (35)

99 [126]

99 [342]

13

12

58

3^b

57 ^f

4

90 (35)

99 [306]

98 [468]

32

52

33

52

5

6^e

180 (35)

90 (50)

90 (60)

83 [470]

98 [138]

90 [330]

51

86

88

7^c

86 ^g

87 ^g

8^d,e

^a0.05 M substrate; internal standard n-undecane; 5 mL hexane; 0.4 mol % Ru; 35 °C. ^b0.005 M substrate; ^c2.5 mol % Ru; ^d1.2 mol % Ru. ^eReactions were performed in

nonane under reduced pressure (700 mbar) to efficiently remove ethylene. ^fSplit test after 5 min. ^gSplit test after 10 min.

catalysts can play an important role in the conversion of measurements (NIR DRS) were recorded on an Agilent Cary

112-115

biomass like oils

and biomass derived olefins^{103, 104,116}to 5000 spectrophotometer. Samples were placed in a quartz

commodity chemicals like polymer building blocks.⁴

tube with window, and were dried at 50 °C under vacuum (10

mbar) or at 150 °C (atmospheric pressure), before

measurement. ²⁹Si MAS NMR spectra were recorded on a

Bruker AMX300 spectrometer (B₀=7.0 T). At this field, the

Experimental

Synthesis and characterization of the support

The following silica materials were synthesized according to resonance frequency of ²⁹Si is 59.6 MHz. The samples were

literature: TUD-1,¹¹⁷MCM-41,⁷⁴SBA-15 (rope morphology),¹¹⁸packed in a 4 mm Zirconia rotor. Tetramethylsilane was used

SBA-15 (fiber morphology),¹¹⁹MCM-48,¹²⁰KIT-6,¹²¹SBA-3,¹²²as chemical shift reference. 2056 scans were accumulated with

SBA-16,⁸²KIT-5,⁷⁷Si-VPI-5,¹²³Silicalite-1,¹²⁴and AlPO₄-5.¹²⁵a recycle delay of 120 s. The spinning frequency of the rotor

Silica Gel 239 (Grace) and Silica Gel 60 (Davisil grade, Sigma- was 6000 Hz. For determination of the presence of reactive

Aldrich) were used as received. Scanning electron microscope strained siloxanes on MCM-41 and TUD-1, the supports were

(SEM) images were recorded on

a JEOL JSM-6010 JV thermally pretreated under a nitrogen atmosphere at 900 °C

microscope. Before measurements, the materials were coated and 700 °C, respectively. The dried samples were reacted with

with gold using a JEOL JSC-1300 sputter. Crystallinity of the dichlorodimethylsilane (DCDMS) under nitrogen atmosphere

mesoporous materials was verified with small angle X-ray to cover all surface hydroxyls, followed by drying under

scattering (SAXS) on a SMART 6000 diffractometer with Cu vacuum. The silanol-deactivated powders were introduced in a

source and a 2D CCD detector. Determination of the textural U-tube reactor and brought in contact with a NH₃flow (3 mL/s)

parameters of the support was done by nitrogen physisorption and heated to 500 °C, maintaining this temperature for 6 h to

on a Micrometrics TriStar 3000 Surface and Porosity Analyzer. react with strained siloxanes.¹²⁹To verify the covering of

The pore size is calculated according to non-local density surface hydroxyls by DCDMS and siloxane reaction with NH₃,

functional theory methods^126-128and pore volumes were FT-IR spectra were taken as described above. TGA analyses

determined with the t-plot method. Quantification of the silica were applied on a TGA Q500 (TA Instruments). Determination

silanols was done by FT-IR spectroscopy. Spectra were of the water amount in the pores was done under a flow of dry

recorded on a Nicolet 6700 spectrometer, equipped with a nitrogen from room temperature to 500 °C at a heating rate of

DGTS detector and KBr beam splitter (256 scans, resolution of 3 °C min^-1.

2 cm^-1). Self-supporting wafers were placed in a vacuum IR-cell

and dehydrated under vacuum for 1 h at 50 °C prior to

Immobilization support and characterization heterogeneous

catalyst

measurements. The dry samples were heated progressively at

Immobilization of the Hoveyda-Grubbs 2^ndgeneration catalyst

5 °C min^-1and kept at each temperature for 45 minutes. All of

(HG2) (Sigma-Aldrich, 97 %) was performed in an inert

the samples (except the one dried at 50 °C) were measured at

nitrogen atmosphere. The precatalyst was dissolved in toluene

150 °C. The molar integrated absorption coefficient ε_(ν+δ)OH

used was 0.16 cm µmol^-1.⁸⁴Near-infrared diffuse reflectance

(Acros Organics, 99.5 %) and added to the pre-dried silica

J. Name., 2013, 00, 1-3 | 13

Please do not adjust margins

Catalysis Science & Technology

Page 14 of 18

ARTICLE

Journal Name

source. Typically, a 1.57 mmol L^-1solution of HG2 in toluene (5 85 %) in methanol to quickly deactivate the catalyst, the other

View Article Online

DOI: 10.1039/C5CY01897H

mL) was added to 0.2 g of a silica support, resulting in a sample was filtered with a 0.45 µm PTFE filter to retain the

loading of 0.2-0.3 wt% Ru. The suspension is stirred at room immobilized catalyst and was further stirred at the same

temperature for 45 minutes, filtered and washed thoroughly reaction temperature. The conversion of the filtered sample

with hexane (Chem-lab, 99 %). After drying, a green powder is was compared to the quenched sample at the end of the

obtained, which is stored at -20 °C to guarantee stability. reaction. For continuous reactions, a glass reactor was charged

Determination of the amount of Ruthenium immobilized on with the catalyst (110 mg of HG2/TUD-1, 0.20 wt% Ru; mixed

the support was performed with UV-VIS spectroscopy analysis with unloaded pellets of TUD-1) and kept in place by quartz-

of the toluene solution before and after immobilization, using wool and glass beads. Cyclooctene (0.4 mol L^-1) was pumped at

a Shimadzu UV-1650PC spectrophotometer. ICP-AES was a rate of 80 mL h^-1over the catalyst bed at room temperature.

performed to measure the amount of Ruthenium in solution

Analysis: GC analysis of the volatile products was carried out

and to confirm the amount of Ruthenium immobilized on the on an Agilent 6890 GC equipped with a flame-ionization

support. Analyses were conducted on a Jobin Yvon Emmission detector (FID) and separated over a HP-5 capillary column.

Ultima ICP with argon plasma. Measurements of the atomic Identification of the reaction products was done with GC-MS

emission spectra of ruthenium were done at 240.727 nm. on an Agilent 6890N GC with an Agilent 5973N Mass Selective

Before measurements, the samples were dissolved in a light Detector, separated over a HP-5 column. GPC analysis was

acidic solution (3 % HNO₃in water). Measurements of the performed for the separation of bigger oligomers, using a

immobilized catalysts after reaction to determine the Waters e2695 Separations Module and a Waters 2414 RI

remaining carbon fraction were performed with TGA under a detector. The stationary phase consists of a Varian M-Gel 3

flow of dry oxygen. The weight loss was monitored from room mixed column. A 1 mL min^-1flow of THF was used. Polystyrene

temperature to 600 °C at a heating rate of 3 °C min^-1. standards were used for calibration. Molecular weight

Determination of the amount of water in the pores was done determined fractions: cyclic oligomers, i.e. C16-C56 oligomers

under a flow of dry nitrogen from room temperature to 500 °C of cyclooctene (M_w= 220 – 770 g/mol) or > C56 if mentioned;

at a heating rate of 3 °C min^-1.

linear oligomers, i.e. short linear fragments of cyclooctene

(M_w= 200 – 3000 g/mol); linear polymer (M_w> 3000 g/mol).

¹H-NMR spectra were recorded on a Bruker Avance 300MHz

spectrometer using DMSO as a solvent.

Catalytic reactions

The following substrates were used: cis-cyclooctene (Acros

Organics, 95 %), cis,cis-1,5-cyclooctadiene (Sigma Aldrich, ≥99

%), norbornene (Sigma-Aldrich, 99 %), diethyl diallylmalonate

(Sigma Aldrich, 98 %), methyl oleate (Sigma-Aldrich, 99 %),

methyl 10-undecenoate (Merck, ≥ 96 %), 1,7-octadiene (Alfa

Aesar, 97 %) and methyl(D,L)2-hydroxy-3-butenoate (TCI

chemicals, > 96 %). For batch reactions, glass reactor vials (10

mL) were charged with the catalyst under an atmosphere of

nitrogen (except for a catalyst with a non-thermally pretreated

support, e.g. MCM-41 at 25 °C). Stirring tests with increasing

amounts of catalysts pointed out that an ideal stirring is only

obtained when using 50 mg or less (Figure S16) per 5 mL

solvent. With higher amounts, the contact times needed to

reach the same conversion increases due to deficient stirring.

Therefore standard reactions were performed with 50 mg of

catalyst. The substrate was dissolved in hexane (5 mL) and

added to the glass reactor. The mixture was stirred at 35 °C,

unless mentioned otherwise. n-Undecane (Acros Organics, 99

%) was added as an internal standard whenever required.

Ethyl vinyl ether (Sigma Aldrich, 99 %) was used as a

terminating agent prior to filtration/centrifugation of the

heterogeneous catalyst when taking samples. For cross-

metathesis and ring-closing metathesis reactions of terminal

olefins an additional balloon filled with argon was used to

dilute ethylene formed during reaction when hexane was used

as a solvent, while for high-boiling compounds the reaction

was performed in hexane under reduced pressure (700 mbar)

to remove ethylene. To verify the heterogeneity of the

catalyst, a split-test was carried out. After 15 minutes, 2

samples were taken from the reaction mixture. One was

quenched with potassium 2-isocyanoacetate (Sigma Aldrich,

Acknowledgements

A.D and B.S thank FWO-Flanders for its financial support

(Project G.0229.12).The authors are grateful to Johan Martens

for general financial support through Methusalem. Kristof

Houthoofd is thanked for NMR measurements and

interpretation of the data, Iris Cuppens for ICP-AES, Stef

Kerkhofs and Gert Brabants for the SAXS measurements and

Walter Vermandel for his help with the continuous

experiments. Ivo Stassen is thanked for the NLDFT calculations.

Notes and references

1

1. K. C. Nicolaou, P. G. Bulger and D. Sarlah, Angew.

Chem. Int. Ed., 2005, 44, 4490-4527.

2

3

2. R. H. Grubbs, Tetrahedron, 2004, 60, 7117-7140.

3. R. H. Grubbs and S. Chang, Tetrahedron, 1998, 54

,

4413-4450.

4

4. M. Dusselier, P. Van Wouwe, S. De Smet, R. De

Clercq, L. Verbelen, P. Van Puyvelde, F. E. Du Prez and

B. F. Sels, ACS Catalysis, 2013, 3, 1786-1800.

5

6

7

5. C. Samojlowicz, M. Bieniek and K. Grela, Chem.

Rev., 2009, 109, 3708-3742.

6. P. Schwab, R. H. Grubbs and J. W. Ziller, J. Am.

Chem. Soc., 1996, 118, 100-110.

7. J. C. Mol, J. Mol. Catal. A: Chem., 2004, 213, 39-

45.

14 | J. Name., 2012, 00, 1-3

Please do not adjust margins

Page 15 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

8

8. H. Clavier, K. Grela, A. Kirschning, M. Mauduit and 32

S. P. Nolan, Angew. Chem. Int. Ed., 2007, 46, 6786-

32. K. H. Park, S. Kim and Y. K. Chung,_ViB_ewu_All_r._ticK_leo_Ore_nla_inn_e

Chem. Soc., 2008, 29, 2057-2060.

DOI: 10.1039/C5CY01897H

6801.

33

34

33. A. Monge-Marcet, R. Pleixats, X. Cattoën and M.

Wong Chi Man, J. Mol. Catal. A: Chem., 2012, 357, 59-

66.

34. S. T. Nguyen and R. H. Grubbs, J. Organomet.

Chem., 1995, 497, 195-200.

9

9. H. D. Maynard and R. H. Grubbs, Tetrahedron

Lett., 1999, 40, 4137-4140.

10. G. C. Vougioukalakis, Chem. Eur. J., 2012, 18

10

11

12

13

14

,

8868-8880.

11. M. S. Sanford, J. A. Love and R. H. Grubbs, J. Am. 35

Chem. Soc., 2001, 123, 6543-6554.

12. M. Ulman and R. H. Grubbs, The Journal of 36

Organic Chemistry, 1999, 64, 7202-7207.

13. J. C. Conrad and D. E. Fogg, Curr. Org. Chem.,

2006, 10, 185-202.

14. R. Buffon, A. Choplin, M. Leconte, J. M. Basset, R. 37

35. S. C. Schürer, S. Gessler, N. Buschmann and S.

Blechert, Angew Chem Int Ed, 2000, 39, 3898-3901.

36. I. Karamé, M. Boualleg, J.-M. Camus, T. K.

Maishal, J. Alauzun, J.-M. Basset, C. Copéret, R. J. P.

Corriu, E. Jeanneau, A. Mehdi, C. Reyé, L. Veyre and C.

Thieuleux, Chem. Eur. J., 2009, 15, 11820-11823.

37. A. G. M. Barrett and S. M. R. Cramp, Richard S.,

Touroude and W. A. Herrmann, J. Mol. Catal., 1992, 72

,

Org. Lett., 1999, 1, 1083-1086.

L7-L10.

38

39

38. G. Borja, R. Pleixats, R. Alibés, X. Cattoën and M.

W. C. Man, Molecules, 2010, 15, 5756-5767.

39. X. Elias, R. Pleixats, M. W. C. Man and J. J. E.

Moreau, Adv. Synth. Catal., 2007, 349, 1701-1713.

40. D. Fischer and S. Blechert, Adv. Synth. Catal.,

2005, 347, 1329-1332.

15

16

15. M. Chabanas, A. Baudouin, C. Copéret and J.-M.

Basset, J. Am. Chem. Soc., 2001, 123, 2062-2063.

16. R. P. Saint-Arroman, M. Chabanas, A. Baudouin, C.

Copéret, J.-M. Basset, A. Lesage and L. Emsley, J. Am. 40

Chem. Soc., 2001, 123, 3820-3821.

17

18

19

20

17. M. P. Conley, W. P. Forrest, V. Mougel, C. Copéret 41

and R. R. Schrock, Angew. Chem., 2014, 126, 14445-

41. S. B. Garber, J. S. Kingsbury, B. L. Gray and A. H.

Hoveyda, J. Am. Chem. Soc., 2000, 122, 8168-8179.

42. A. Keraani, C. Fischmeister, T. Renouard, M. Le

Floch, A. Baudry, C. Bruneau and M. Rabiller-Baudry, J.

Mol. Catal. A: Chem., 2012, 357, 73-80.

43. J. S. Kingsbury, S. B. Garber, J. M. Giftos, B. L.

Gray, M. M. Okamoto, R. A. Farrer, J. T. Fourkas and A.

H. Hoveyda, Angew Chem Int Ed, 2001, 40, 4251-4256.

44. J. Lim, S. S. Lee, S. N. Riduan and J. Y. Ying, Adv.

Synth. Catal., 2007, 349, 1066-1076.

14448.

42

43

18. V. Mougel, C. B. Santiago, P. A. Zhizhko, E. N.

Bess, J. Varga, G. Frater, M. S. Sigman and C. Copéret, J.

Am. Chem. Soc., 2015, 137, 6699-6704.

19. T. S. Halbach, S. Mix, D. Fischer, S. Maechling, J.

O. Krause, C. Sievers, S. Blechert, O. Nuyken and M. R.

Buchmeiser, J. Org. Chem., 2004, 70, 4687-4694.

20. J. O. Krause, S. H. Lubbad, O. Nuyken and M. R.

44

45

Buchmeiser, Macromol. Rapid Commun., 2003, 24

,

45. J. Lim, S. Seong Lee and J. Y. Ying, Chem.

Commun., 2010, 46, 806.

875-878.

21

22

23

21. J. O. Krause, O. Nuyken, K. Wurst and M. R. 46

Buchmeiser, Chem. Eur. J., 2004, 10, 777-784.

22. D. Bek, N. Žilková, J. Dědeček, J. Sedláček and H. 47

46. S. Varray, R. Lazaro, J. Martinez and F. Lamaty,

Organometallics, 2003, 22, 2426-2435.

47. Q. Yao, Angew Chem Int Ed, 2000, 39, 3896-3898.

48. H. Zhang, Y. Li, S. Shao, H. Wu and P. Wu, J. Mol.

Catal. A: Chem., 2013, 372, 35-43.

Balcar, Top. Catal., 2009, 53, 200-209.

23. P. Nieczypor, W. Buchowicz, W. J. N. Meester, F.

P. J. T. Rutjes and J. C. Mol, Tetrahedron Lett., 2001, 42

48

49

,

49. Q. T. Easter, V. Trauschke and S. A. Blum, ACS

7103-7105.

Catalysis, 2015, 5, 2290-2295.

24

25

26

27

28

29

30

31

24. K. Vehlow, S. Maechling, K. Köhler and S. 50

Blechert, J. Organomet. Chem., 2006, 691, 5267-5277.

25. L. Yang, M. Mayr, K. Wurst and M. R. Buchmeiser, 51

Chem. Eur. J., 2004, 10, 5761-5770.

50. H. Staub, F. Kleitz and F.-G. Fontaine, Microporous

Mesoporous Mater., 2013, 175, 170-177.

51. H. Yang, Z. Ma, T. Zhou, W. Zhang, J. Chao and Y.

Qin, ChemCatChem, 2013, 5, 2278-2287.

52. Q. Li, T. Zhou and H. Yang, ACS Catalysis, 2015, 5,

26. D. Bek, H. Balcar, N. Žilková, A. t. Zukal, M. 52

Horáček and J. i. ꢀejka, ACS Catalysis, 2011,

1

, 709-718.

2225-2231.

27. L. Li and J.-l. Shi, Adv. Synth. Catal., 2005, 347

,

53

53. C. Copéret and J. M. Basset, Adv. Synth. Catal.,

2007, 349, 78-92.

54. H. Balcar and J. ꢀejka, Coord. Chem. Rev., 2013,

257, 3107-3124.

55. H. Balcar and J. ꢀejka, Macromolecular Symposia,

2010, 293, 43-47.

56. M. P. Conley, C. Copéret and C. Thieuleux, ACS

1745-1749.

28. S. Randl, N. Buschmann, S. J. Connon and S. 54

Blechert, Synlett, 2001, 10, 1547-1550.

29. M. Mayr, B. Mayr and M. R. Buchmeiser, Angew 55

Chem Int Ed, 2001, 40, 3839-3842.

30. J. P. Gallivan, J. P. Jordan and R. H. Grubbs, 56

Tetrahedron Lett., 2005, 46, 2577-2580.

31. K. Melis, D. E. De Vos, P. A. Jacobs and F. 57

Verpoort, J. Mol. Catal. A: Chem., 2001, 169, 47-56.

Catalysis, 2014, 4, 1458-1469.

57. F. B. Hamad, C. Kai, Y. Cai, Y. Xie, Y. Lu, F. Ding, Y.

Sun and F. Verpoort, Curr. Org. Chem., 2013, 17, 2592-

2608.

J. Name., 2013, 00, 1-3 | 15

Please do not adjust margins

Catalysis Science & Technology

Page 16 of 18

ARTICLE

58

Journal Name

58. B. Van Berlo, K. Houthoofd, B. F. Sels and P. A. 82

Jacobs, Adv. Synth. Catal., 2008, 350, 1949-1953.

82. F. Kleitz, T.-W. Kim and R. Ryoo, Langmuir , 2006,

View Article Online

DOI: 10.1039/C5CY01897H

22, 440-445.

59

59. J. Cabrera, R. Padilla, M. Bru, R. Lindner, T. 83

Kageyama, K. Wilckens, S. L. Balof, H.-J. Schanz, R. 84

Dehn, J. H. Teles, S. Deuerlein, K. Müller, F. Rominger

and M. Limbach, Chem. Eur. J., 2012, 18, 14717-14724.

60. M. Bru, R. Dehn, J. H. Teles, S. Deuerlein, M. Danz, 85

83. L. T. Zhuravlev, Colloids Surf., A, 2000, 173, 1-38.

84. J.-P. Gallas, J.-M. Goupil, A. Vimont, J.-C. Lavalley,

B. Gil, J.-P. Gilson and O. Miserque, Langmuir, 2009, 25

5825-5834.

85. A. A. Christy, Advanced Materials Research, 2013,

650, 66-71.

,

60

I. B. Müller and M. Limbach, Chem. Eur. J., 2013, 19

,

11661-11671.

86

87

88

89

86. K. Wan, Q. Liu and C. Zhang, Mater. Lett., 2003,

57, 3839-3842.

87. S. Shioji, M. Hanada, Y. Hayashi, K. Tokami and H.

Yamamoto, Adv. Powder Technol., 2007, 18, 467-483.

88. P. Der Voort, J. Chem. Soc., Faraday Trans., 1990,

86, 3747-3750.

89. K. Unger, Porous Silica, Elsevier Science, 1979.

90. D. W. Sindorf and G. E. Maciel, J. Am. Chem. Soc.,

1983, 105, 3767-3776.

61

62

63

64

65

61. W. Solodenko, A. Doppiu, R. Frankfurter, C. Vogt

and A. Kirschning, Aust. J. Chem., 2013, 66, 183.

62. H. Balcar, T. Shinde, N. Žilková and Z. Bastl,

Beilstein J. Org. Chem., 2011, 7, 22-28.

63. H. Yang, Z. Ma, Y. Wang, Y. Wang and L. Fang,

Chem. Commun., 2010, 46, 8659.

64. T. Shinde, N. Žilková, V. Hanková and H. Balcar, 90

Catal. Today, 2012, 179, 123-129.

65. J. Pastva, K. Skowerski, S. J. Czarnocki, N. Žilková, 91

91. B. Fubini, V. Bolis, A. Cavenago and P. Ugliengo, J.

Chem. Soc., Faraday Trans., 1992, 88, 277-289.

92. J. P. Blitz, C. C. Meverden and R. E. Diebel,

Langmuir, 1998, 14, 1122-1129.

93. B. A. Morrow, I. A. Cody and L. S. M. Lee, J. Phys.

Chem., 1976, 80, 2761-2767.

J. ꢀejka, Z. Bastl and H. Balcar, ACS Catalysis, 2014,

3227-3236.

4,

92

66

67

66. United States Pat., US20120165588, 2012.

67. Z. Jia and M. J. Monteiro, J. Polym. Sci., Part A: 93

Polym. Chem., 2012, 50, 2085-2097.

68

68. H. R. Kricheldorf, J. Polym. Sci., Part A: Polym. 94

Chem., 2010, 48, 251-284.

69. United States Pat., US2553651, 1951.

94. H. Staub, R. Guillet-Nicolas, N. Even, L. Kayser, F.

Kleitz and F.-G. Fontaine, Chem. Eur. J., 2011, 17, 4254-

4265.

95. M. Ide, M. El-Roz, E. De Canck, A. Vicente, T.

Planckaert, T. Bogaerts, I. Van Driessche, F. Lynen, V.

Van Speybroeck, F. Thybault-Starzyk and P. Van Der

Voort, PCCP, 2013, 15, 642.

96. J. D. Robert, in Catalysis for the Conversion of

Biomass and Its Derivatives, eds. M. Behrens and A. K.

Datye, The Max Planck Research Library for the History

and Development of Knowledge, Berlin, 2013, pp. 255-

286.

69

70

71

70. United States Pat., US3465016, 1969.

95

71. C. Sell, in The Chemistry of Fragrances: From

Perfumer to Consumer, ed. C. S. Sell, The Royal Society

of Chemistry, Cambridge, 2006, ch. Ingredients for the

modern perfumery industry, pp. 95-106.

96

72

72. M. K. Samantaray, J. Alauzun, D. Gajan, S.

Kavitake, A. Mehdi, L. Veyre, M. Lelli, A. Lesage, L.

Emsley, C. Copéret and C. Thieuleux, J. Am. Chem. Soc.,

2013, 135, 3193-3199.

73

74

75

73. Information about impurities silica gels: Catalog 97

Resins & Media p.5, Sigma-Aldrich.

74. B. Pauwels, G. Van Tendeloo, C. Thoelen, W. Van 98

Rhijn and P. Jacobs, Adv. Mater., 2001, 13, 1317-1320.

75. H. Zhang, J. Sun, D. Ma, X. Bao, A. Klein- 99

Hoffmann, G. Weinberg, D. Su and R. Schlögl, J. Am.

97. D. Luss, Diffusion-Reaction Interactions in Catalyst

Pellets, Taylor & Francis, Texas, 1986.

98. S. Monfette and D. E. Fogg, Chem. Rev., 2009,

109, 3783-3816.

99. H. Höcker, W. Reimann, L. Reif and K. Riebel, J.

Mol. Catal., 1980, 8, 191-202.

Chem. Soc., 2004, 126, 7440-7441.

100

100. L. Reif and H. Hoecker, Macromolecules, 1984, 17,

76

77

76. S. Polarz, B. Völker and F. Jeremias, Dalton

Transactions, 2010, 39, 577-584.

77. F. Kleitz, D. Liu, G. M. Anilkumar, I.-S. Park, L. A.

952-956.

101

101. H. Jacobson and W. H. Stockmayer, The Journal of

Chemical Physics, 1950, 18, 1600-1606.

Solovyov, A. Shmakov, N. and R. Ryoo, The Journal of 102

Physical Chemistry B, 2003, 107, 14296-14300.

78. H.-J. Kim, K.-S. Jang, P. Galebach, C. Gilbert, G. 103

Tompsett, W. C. Conner, C. W. Jones and S. Nair,

102. Z. R. Chen, J. P. Claverie, R. H. Grubbs and J. A.

Kornfield, Macromolecules, 1995, 28, 2147-2154.

103. A. Blencowe and G. G. Qiao, J. Am. Chem. Soc.,

2013, 135, 5717-5725.

78

79

Journal of Membrane Science, 2013, 427, 293-302.

79. Z. Shan, J. C. Jansen, L. Marchese and T.

Maschmeyer, Microporous Mesoporous Mater., 2001,

48, 181-187.

104

104. S. Kavitake, M. K. Samantaray, R. Dehn, S.

Deuerlein, M. Limbach, J. A. Schachner, E. Jeanneau, C.

Copéret and C. Thieuleux, Dalton Transactions, 2011,

40, 12443.

80

81

80. S. Telalovic, A. Ramanathan, G. Mul and U. 105

Hanefeld, J. Mater. Chem., 2010, 20, 642-658.

81. R. Maheswari, R. Anand and G. Imran, J. Porous

Mater., 2012, 19, 283-288.

105. S. M. Rountree, M. C. Lagunas, C. Hardacre and P.

N. Davey, Applied Catalysis A: General, 2011, 408, 54-

62.

16 | J. Name., 2012, 00, 1-3

Please do not adjust margins

Page 17 of 18

Catalysis Science & Technology

Please do not adjust margins

Journal Name

ARTICLE

106

107

108

106. B. Van Berlo, J. Dijkmans, K. Houthoofd, M. 131

Tromp, I. Hermans, P. Jacobs and B. F. Sels, Detroit,

U.S.A., 2011.

107. R. Klaewkla, M. Arend and W. F. Hoelderich, in

Mass Transfer - Advanced Aspects, ed. H. Nakajima,

2011.

View Article Online

DOI: 10.1039/C5CY01897H

108. H. Höcker, W. Reimann, K. Riebel and Z.

Szentivanyi, Die Makromolekulare Chemie, 1976, 177

,

1707-1715.

109

110

111

112

109. E. Thorn-Csányi and K. Ruhland, Macromol. Chem.

Phys., 1999, 200, 1662-1671.

110. C. W. Bielawski and R. H. Grubbs, Prog. Polym.

Sci., 2007, 32, 1-29.

111. A. K. Chatterjee, T.-L. Choi, D. P. Sanders and R. H.

Grubbs, J. Am. Chem. Soc., 2003, 125, 11360-11370.

112. A. Nickel, T. Ung, G. Mkrtumyan, J. Uy, C. Lee, D.

Stoianova, J. Papazian, W.-H. Wei, A. Mallari, Y. Schrodi

and R. Pederson, Top. Catal., 2012, 55, 518-523.

113. J. C. Mol and R. Buffon, Journal of the Brazilian

113

Chemical Society, 1998,

9, 1-11.

114

115

114. J. C. Mol, Green Chemistry, 2002, 4, 5-13.

115. R. Kadyrov, C. Azap, S. Weidlich and D. Wolf, Top.

Catal., 2012, 55, 538-542.

116

117

116. Y. Yang, X. Wei, F. Zeng and L. Deng, Green

Chemistry, 2015.

117. J. C. Jansen, Z. Shan, T. Maschmeyer, L. Marchese,

W. Zhou and N. v. d. Puil, Chem. Commun., 2001, 713-

714.

118

118. K. Flodström and V. Alfredsson, Microporous

Mesoporous Mater., 2003, 59, 167-176.

119. D. Zhao, Science, 1998, 279, 548-552.

120. K. Schumacher, M. Grün and K. K. Unger,

Microporous Mesoporous Mater., 1999, 27, 201-206.

119

120

121

122

123

121. A.

Vinu,

N.

Gokulakrishnan,

V.

Balasubramanian, S. Alam, M. P. Kapoor, K. Ariga and T.

Mori, Chem. Eur. J., 2008, 14, 11529-11538.

122. O. A. Anunziata, A. R. Beltramone, M. L. Martínez

and L. L. Belon, J. Colloid Interface Sci., 2007, 315, 184-

190.

123. S. del Val, T. Blasco, E. Sastre and J. Perez-

Pariente, J. Chem. Soc., Chem. Commun., 1995, 731-

732.

124

125

126

127

124. C. Shao, X. Li, S. Qiu, F.-s. Xiao and O. Terasaki,

Microporous Mesoporous Mater., 2000, 39, 117-123.

125. M. Fang, H. Du, W. Xu, X. Meng and W. Pang,

Microporous Mater., 1997, 9, 59-61.

126. P. I. Ravikovitch, G. L. Haller and A. V. Neimark,

Adv. Colloid Interface Sci., 1998, 76–77, 203-226.

127. M. Jaroniec, M. Kruk, J. P. Olivier and S. Koch, in

Stud. Surf. Sci. Catal., eds. G. K. K.K. Unger and J. P.

Baselt, Elsevier, 2000, vol. Volume 128, pp. 71-80.

128. J. Landers, G. Y. Gor and A. V. Neimark, Colloids

Surf., A, 2013, 437, 3-32.

129. T. Asefa, M. Kruk, N. Coombs, H. Grondey, M. J.

MacLachlan, M. Jaroniec and G. A. Ozin, J. Am. Chem.

Soc., 2003, 125, 11662-11673.

128

129

130

J. Name., 2013, 00, 1-3 | 17

Please do not adjust margins

Catalysis Science & Technology

Page 18 of 18

The ideal support characteristics for immobilization of the Hoveyda-Grubbs 2 catalyst

were defined in the metathesis of cyclooctene and the reaction mechanism to cyclic

oligomers was unraveled.

Article Doi

Immobilized Grubbs catalysts on mesoporous silica materials: Insight into support characteristics and their impact on catalytic activity and product selectivity

DOI: 10.1039/c5cy01897h

Source and publish data:

Authors:

Article abstract of DOI:10.1039/c5cy01897h

Full text of DOI:10.1039/c5cy01897h

Products guided by the article

R&D Labs maybe for 23579-19-3

Relevant to this article

Hot Product