Simple salts of abundant metals (Fe, Bi, and Ti) supported on montmorillonite as efficient and recyclable catalysts for regioselective intramolecular and intermolecular hydroalkoxylation reactions of double bonds and tandem processes

Abstract

The transfer of catalytic hydroalkoxylation reactions of olefins from homogeneous to heterogeneous conditions has been studied using two types of solid catalysts, namely montmorillonite (MMT) doped with metal cations and metal nanoparticles supported on oxides. In the case of intramolecular reactions, 38–99% yields of cyclic ethers have been obtained using Fe-MMT and Bi-MMT both in CH₃NO₂ and dimethyl carbonate (DMC) compared with other supported metal salts or metal nanoparticles. In the case of more challenging intermolecular reactions, conversions up to 72% and yields up to 54% were obtained with metal-doped MMT as well, such as Fe-, Bi-, and Ti-MMT. In this paper, we detail the substrate scope and limitations for both classes of reactions and tandem processes, their transposition in flow and some mechanistic insights concerning the active species, in processes identified as truly heterogeneously catalysed. As a general trend, it was observed that trisubstituted double bonds allowed the best results both in intra- and intermolecular reactions. The transfer of homogeneous catalysts onto heterogeneous ones in the case of Fe-MMT and Bi-MMT was successful and even allowed enhanced catalytic activities in the case of Bi-MMT.

---

Introduction

Metal-based catalysts, and in particular transition metal-based catalysts, have been successfully used for decades in a myriad of organic transformations.^1–4 Most of the developments of metal catalysis in fine chemical transformations have been designed in organic solvents that allow, in the same phase, pre-catalyst, catalyst, intermediate(s), substrate(s) and product solubilisation (homogeneous catalysis). Despite many assets, homogeneous metal-based catalysis rarely goes beyond the laboratory scale. The main reasons are related to cost (metal and ligands), tedious recycling operations, scale-up issues, and in the case of the synthesis of active pharmaceutical principles, the presence of residual metal in the final product.⁵ So, for industrial applications, cost-effective solid catalysts are preferred, allowing an easy recycling, and/or amenability to continuous processes.⁶ However, solid catalysts have been so far mostly designed for commodity chemicals' synthesis, and do not offer the same degree of selectivity and efficiency when dealing with sophisticated reactions.

As a result, a significant part of current research in catalytic methodologies is thus directed towards the transfer of the reactions from homogenous to heterogeneous conditions.⁷ The most straightforward strategy probably consists in attaching homogeneous catalysts to inert supports, the latter being either organic (polymers, dendrimers, polysaccharides,…) or inorganic (mesoporous silica, metal oxides, aluminosilicates,…).^8–10

In the context of sustainable chemistry, a stronger stress is placed to design new catalysts being based on abundant metals and easily recyclable, most of the metals conventionally used in catalysis being under a threat of depletion within 50 years.¹¹ Metals such as iron, copper, and bismuth have thus attracted the attention of chemists, these metals being abundant and also having acceptable (eco)toxicity profiles.^12–16

For all these reasons, we have chosen to study simple salts of sustainable metals supported on clays such as MMT, a layered inorganic material with exchangeable cations in the interlamellar space, and metal nanoparticles supported on silica. MMT is a cheap, natural and abundant inorganic material with both Lewis and Brønsted acidic sites very attractive as a support for catalysts.¹⁷ To test these catalysts, we have selected the hydroalkoxylation of olefins: a very useful reaction widely studied in homogeneous catalysis, and rarely in heterogeneous catalysis.^18–36 Zeolites have been the most studied heterogeneous catalysts for this reaction and generally required elevated temperatures (85–150 °C) to proceed, sometimes with selectivity issues.^34,37,38 Lanthanide triflates could be used and recycled efficiently in intramolecular reaction at 120 °C upon their sequestration in ionic liquids.³⁶ Gold nanoclusters were also reported as efficient catalysts in intramolecular reactions, although limited to monosubstituted olefins.³⁰

In a preliminary communication, we reported our results in the intramolecular version of the reaction catalysed by Fe(III) and Bi(III) salts supported on MMT.³⁹ In the present paper, our intention is to provide full experimental studies, the scope and limitations of the intramolecular reaction, as well as new data for the intermolecular version, introducing Ti-MMT as catalyst, and a mechanistic rationale.

Results and discussion

Catalysts preparation and characterisation

Metal doped montmorillonites (M-MMT). M-MMT materials were prepared by impregnating solutions of chloride salts in MeOH with lyophilised MMT. Namely, BiCl₃, FeCl₃, CuCl₂, and TiCl₄ were added to a suspension of Na-MMT in MeOH in a 1 mmol g⁻¹ of metal salt/MMT ratio. Exchange reached the thermodynamic equilibrium within few hours as monitored by conductimetry using a Na⁺-selective electrode. After filtration and trituration in degased MeOH, M-MMT were dried extensively under reduced pressure for 24 h. ICP-MS titration of metal content in the modified clays showed between 1.8 and 4.9% w/w metal content (Cu-MMT 2.04%, Fe-MMT 2.67%, Bi-MMT 4.87%, Ti-MMT 1.81%). As witnessed by XPS analysis, the oxidation state of inserted metal ions remained identical except in the case of Fe where a Fe³⁺/Fe²⁺ 53/47 ratio was shown to be the thermodynamic resulting mixture of the process. SAXS analysis clearly showed metal insertion in the interlayer spacing, although in the case of Bi-MMT remaining BiOCl formed upon hydrolysis of BiCl₃ was simultaneously observed.

Silica supported metal nanoparticles (M NP@SiO₂). For the fabrication of silica supported metal/metal oxide nanoparticles (Table 1, entries 1–5), we used supercritical CO₂ (scCO₂) as the main solvent in a batch mode setup. A detailed description of the setup and experimental conditions can be found elsewhere.⁴⁰ In summary, in a stainless steel batch reactor, the silica support is mixed with a metal hexafluoroacetylacetonate precursor (5% w/w of metal loading) in the absence of H₂ (as reducing agent) but using EtOH, as co-solvent and reducing agent, together with a surfactant (hexadecylamine) typically used in wet chemistry for metal reduction. After reaction, a fast depressurization in parallel with CO₂ purging yielded dry clean powder in the form of metal/metal oxide nanoparticles deposited on the silica support (catalyst physicochemical properties and TEM images are presented in ESI†).

Table 1 Initial screening of catalysts and conditions for the intramolecular hydroalkoxylation of olefins^a

Entry	R1	R2	Catalyst	Solvent^b	T °C	Conversion	Yield
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed solvent (1 ml) and catalyst (1–5 mol% metal/substrate ratio).b A: either CH3NO2 or DCE. B: either toluene, CH3OH, heptane, CH3CN, dioxane, acetone, AcOEt, THF, H2O/THF (9 : 1), DMF, CH3NO2 or DCE.c Commercially available from Strem.
1	Me	H	Pt NPs/SiO2	A	40 °C	0%	—
2	Me	H	Pt–Cu NPs/SiO2	A	40 °C	0%	—
3	Me	H	Cu NPs/SiO2	A	40 °C	0%	—
4	Me	H	CuxO NPs/SiO2	A	40 °C	0%	—
5	Me	H	SnO2 NPs/SiO2	A	40 °C	0%	—
6	H	Ph	Au NPs/TiO2^c	A	60 °C	0%	—
7	H	Ph	Au NPs/TiO2^c	A	60 °C	0%	—
8	H	Ph	Ni-LDH	B	40 °C	0%	—
9	H	Ph	Mg-LDH	B	40 °C	0%	—
10	H	Ph	Fe-MMT	CH3NO2	40 °C	35%	35%
11	H	Ph	Bi-MMT	DMC	80 °C	100%	76%
12	H	Ph	Bi-MMT	CH3NO2	80 °C	100%	77%
13	H	Ph	Bi-MMT	CH3CN	80 °C	100%	76%
14	H	Ph	Fe-MMT	DMC	80 °C	100%	86%
15	H	Ph	Cu-MMT	DMC	80 °C	39%	36%
16	H	Ph	Cu-MMT	CH3NO2	80 °C	100%	81%
17	H	Ph	Fe-MMT	CH3NO2	80 °C	100%	83%
18	H	Ph	None	DMC	80 °C	0%	—

---

Intramolecular hydroalkoxylation

Intramolecular hydroalkoxylation of olefins is the most direct and atom-economical route to cyclic ethers. It is typically achieved following two distinct modes of activation and mechanisms (Scheme 1). In the former, the double bond is activated by a Brønsted or a Lewis acid facilitating the nucleophilic attack of the pendant hydroxyl group: carbenium intermediates are involved.^41,42 In the latter, transition metals are first coordinated to the double bond which then undergo the attack of the pendant hydroxyl group resulting in the formation of an organometallic intermediate featuring a carbon–metal bond eventually cleaved by protonolysis.¹⁹ These differences in terms of intermediates influence the regioselectivity and the ring-closure overall efficiency. Substrates with trisubstituted double bonds, leading to tertiary carbenium ions, are efficiently cyclised under Brønsted acid-type activation while substrates with a terminal monosubstituted double bond are more efficiently cyclised by transition metal-catalysts. Hybrid mechanisms based on Lewis acid-assisted Brønsted acid catalysis have been invoked in reactions with metal triflates where activation of the hydroxyl group by the Lewis acid provides an acidified proton further transferred to the double bond.²²

Scheme 1 Generic reaction pathways for H⁺- and metal-catalysed intramolecular hydroalkoxylation.

We initially screened a series of supported catalysts, either MMT-supported metal salts and metal oxides-supported transition metal nanoparticles to possibly observe the abovementioned types of activation on substrates 1a and 1b (Table 1).

The metal choice was guided in both cases by the frequency of appearance of those in homogenous catalysis for hydroalkoxylation reactions of multiple bonds: Pt, Cu, Au and Sn for NPs, Fe, Bi, and Cu for supported metal salts.

Unfortunately, in our reaction conditions, metal/metal oxides NPs of Pt, Pt–Cu, Cu, Cu_xO, Au and SnO₂ supported on SiO₂ or TiO₂ (entries 1–7), as well as Ni-LDH and Mg-LDH (layered double hydroxides, entries 8 and 9), were unreactive. MMT doped with metal cations such Fe(III) (Fe-MMT) exhibited a promising 35% yield in cyclised product 2b at 40 °C (entry 10). This result was further confirmed at 80 °C with Bi-MMT, Fe-MMT, and Cu-MMT allowing conversion rates to reach 100% and 76–86% isolated yields of 2b (entries 11–17). Control reactions were performed with pristine MMT, and in the absence of catalyst in CH₃NO₂ at 80 °C. In these conditions, no conversion of the starting material 1b was observed.

Recyclability of the catalysts was tested with Fe-MMT in DMC and CH₃NO₂ in the reaction conditions of entry 14 (Table 1). In DMC, the yield of 2b could be maintained 5 times without significant loss, while in CH₃NO₂, it rapidly decreased down to zero after 4^th cycle.

This significant difference of results only depending on the solvent used suggested that the metal could be washed out of MMT in polar CH₃NO₂ (μ = 3.46 D), while in DMC (μ = 0.91 D) leaching was avoided. As a result, we selected DMC for the transposition of the reaction in continuous flow systems. Satisfyingly, using a 25 cm long stainless steel column with 10 mm of internal diameter charged with Fe-MMT, 15.3 g of cyclised product 2a could be obtained in 1 h at a flow rate of 1 ml min⁻¹ of a 1a 2 M solution in DMC. The key parameter of the system turned out to be the temperature, all reactions conducted above 75 °C leading to complete conversions with this system.

We further evaluated the scope of the reaction in terms of substrates. With this type of catalysts, a mechanism involving carbenium ions was likely to occur. In such case, we expected that the reaction would be influenced by the class of the alcohol, for the reason of steric hindrance and the competition with acid-catalysed dehydration reaction, and the class of the double bond, densely substituted ones presenting increased steric hindrance but leading to more stable carbenium ions.

A series of unsaturated alcohols was thus prepared following several synthetic routes (see ESI† for details). Secondary homoallylic alcohols were obtained by Grignard reaction of prenylmagnesium bromide with the corresponding aldehydes. Methallyl ethers derived from glycerol were obtained in two steps from solketal (glycerol acetonide) to yield 1f (R₄ = H). Upon protection of the primary alcohol, 1g and 1h were obtained bearing an acetyl or a TBDMS group, respectively. Primary bis-homoallylic alcohols 1i–k were formed by a Wittig reaction of γ-butyrolactol with the corresponding phosphonium salts.⁴³ Allylation of phenol was performed with the desired allylbromide (allyl, prenyl, geranyl) in the presence of NaH.⁴⁴

Our optimised conditions based on Fe-MMT in DMC at 80 °C were thus tested firstly on secondary alcohols bearing a trisubstituted double bond (Table 2).

Table 2 Cyclisation of alcohols 1a, 1c–e^a

Entry	R3	Substrate	Time (h)	Conversion	Isolated yield
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed DMC (1 ml) and Fe-MMT (5 mol% metal/substrate ratio).b Due to the high volatility of product 2a the reaction was performed in DMC-d6 and the yield determined by ^1H-NMR using benzene as internal standard.c Along with 12% of dehydration product.
1	Me	1a	7	100%	98%^b
2	n-C5H11	1c	7	100%	89%
3	Ph–CH2–CH2	1d	7	100%	89%
4	p-CH3C6H4	1e	7	100%	38%^c

---

Intramolecular hydroalkoxylation of trisubstituted double bonds by secondary alcohols

With substrate 1a, the reaction was almost quantitative in 7 hours in cyclised product 2a, for which the yield was determined by ¹H-NMR by the method of external calibration to avoid evaporation during work-up (entry 1). Substrates 1c and 1d were cleanly cyclised to their corresponding products both in 89% yield (entries 2 and 3). When a tolyl substituent was introduced (substrate 1e), a competition with the dehydration reaction was observed. The cyclised product 2e was obtained in 38%, together with 12% of styrene derivative; we assumed that the mass loss was due to the polymerisation of the latter at 80 °C in this acidic medium (entry 4). A limitation of this reaction was thus identified when stabilised carbenium ion could be formed upon the loss of HO⁻ as exemplified by 1e possibly leading to a benzylic cation.

Intramolecular hydroalkoxylation of 1,1-disubstituted double bonds by secondary alcohols derived from glycerol

We next turned our attention to secondary alcohols bearing a disubstituted terminal double bond derived from glycerol. These unsaturated alcohols were obtained from solketal by methallylation/deprotection (1f, R₄ = H) and with an additional protection of the primary alcohol by an acetyl or a TBDMS group (1g and 1h, respectively). These substrates were tested under our optimised reaction conditions based on Fe-MMT in DMC at 80 °C (Table 3).

Table 3 Cyclisation of alcohols 1f–h^a

Entry	R	Substrate	Time (h)	Conversion	Isolated yield
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed DMC (1 ml) and Fe-MMT (5 mol% metal/substrate ratio).b Together with 31% of the corresponding dioxane.
1	H	1f	24	100%	51% (2 dias 1 : 1)^b
2	Ac	1g	22	100%	59% (2 dias 1 : 1)
3	TBDMS	1h	7	100%	Mixture

---

With these substrates featuring an oxygen atom on the tether on the allylic position, the reaction yielded 1,3-dioxolanes. This reaction could either proceed by isomerisation/cyclisation, involving the formation of an enol ether intermediate, in a process already described with α-methallyloxy carboxylic acids in the presence of Cu(OTf)₂ as the catalyst,⁴⁵ or by formation of a transient oxonium ion (Scheme 2). With unprotected substrate 1f, a 5 : 3 mixture of the corresponding 1,3-dioxolane and 1,3-dioxane was obtained upon the nucleophilic attack of either one or the other alcohol function. With the primary alcohol function blocked by an acetyl group (substrate 1g), only the 1,3-dioxolane was formed in 59% isolated yield. With TBDMS-protected substrate 1h however, the silyl protecting group was partially removed in our reaction conditions and a mixture of protected and non-protected 1,3-dioxolanes and 1,3-dioxanes was obtained.

Scheme 2 Possible mechanisms accounting for the formation of 1,3-dioxolanes 2f–g.

Intramolecular hydroalkoxylation of 1,2-disubstituted double bonds by primary alcohols

The catalytic system was then tested on substrates 1i–j featuring a combination of primary alcohol function and a 1,2-disubstituted double bond (Table 4).

Table 4 Cyclisation of alcohols 1i–j^a

Entry	R1, R5	Substrate	Solvent	Time (h)	Conversion	Isolated yield
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed solvent (1 ml) and Fe-MMT (5 mol% metal/substrate ratio).b T °C = 80 °C.c T °C = 40 °C.d 2i/2i′ = 95 : 5.
1^b	H, n-C5H11	(Z)-1i	DMC	72	92%	Mixture
2^b	H, Ph	(E,Z)-1j	DMC	72	82%	Mixture
3^c	H, n-C5H11	(Z)-1i	CH3NO2	48	4%	Trace
4	H, n-C5H11	(Z)-1i	CH3NO2	72	89%	62%^d
5	H, Ph	(Z)-1j	CH3NO2	2.5	100%	99%
6	H, Ph	(E)-1j	CH3NO2	24	100%	99%
7^b	Me, Me	(E,Z)-1k	DMC	24	72%	Mixture

---

Substrate 1i was commercially available as the (Z)-stereoisomer. Substrate 1j was prepared by Wittig olefination of γ-butyrolactol. However, since the reaction with benzyltriphenylphosphonium bromide delivered a mixture of (E)- and (Z)-1j, the stereoisomer (E)-1j was obtained independently in two steps. Treatment of 1-phenylprop-2-en-1-ol with triethylorthoacetate furnished ethyl (E)-5-phenylpent-4-enoate which was then reduced to (E)-1j using LiAlH₄ (see ESI† for details).

When (Z)-1i was put to react with Fe-MMT at 80 °C in DMC, the expected cyclisation was observed leading to 2-pentyltetrahydrofuran 2i (28% GC-MS yield) but the main product was the methyl carbonate formed upon reaction with DMC. Only trace amount of 2-butyltetrahydropyran 2i′ was formed (entry 1). With (E,Z)-1j under the same reaction conditions, the (Z)-isomer reacted faster and the formation of the methyl carbonate was again observed (36% GC-MS yield) while the cyclic ether 2j was formed in 46% yield (entry 2). By replacing DMC by CH₃NO₂, at 40 °C the reaction of 1i was very slow and only 4% conversion was observed after 48 hours (entry 3). However, at 100 °C the reaction proceeded with 89% conversion in 72 hours and delivered 2i in 62% yield and 95% regioselectivity compared with 2i′ (entry 4). With (Z)- and (E)-1j, the reaction was completed in shorter reaction times and quantitatively provided the cyclic ether 2j (entries 5 and 6). Substrate 1k was partially converted (72%) to a mixture of elimination products and isomers, as suggested by GC-MS analysis (entry 7). The double bond substitution was thus shown to impact significantly the reaction efficiency. The stabilisation of positively charged intermediates formed by activation of the double bond requires substitution by electron-donating groups. In the case of 1j, the presence of the phenyl group allowed to reach both high conversions and yields.

Intramolecular hydroalkoxylation of trisubstituted double bonds by phenols

We further tested the efficiency of the catalytic system on phenolic substrates 1l–n featuring mono- and trisubstituted double bonds (Table 5).

Table 5 Cyclisation of alcohols 1l–n^a

Entry	R6, R7	Substrate	M	Solvent	Time (h)	Conversion	Isolated yield of 2l–n
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed solvent (1 mL) and Fe-MMT (5 mol% metal/substrate ratio).b Quantitative recovery of the starting material.c Bi-MMT prepared from BiCl3 was used as catalyst.d Bi-MMT prepared from Bi(OTf)3 was used as catalyst.e Together with 39% of tricyclic product 3n formed upon tandem reaction.f Together with 42% of product 3n.g Together with 79% of product 3n.h Product 3n formed in 96% was the sole product.
1	H, H	1l	Fe	DMC	48	0%	—^b
2^c	H, H	1l	Bi	DMC	48	0%	—^b
3^d	H, H	1l	Bi	DMC	48	64%	50%
4^c	Me, Me	1m	Bi	DCE	24	100%	99%
5^c	Me, Me	1m	Bi	CH3NO2	24	100%	57%
6^c	Me, Me	1m	Bi	DMC	24	91%	90%
7	Me, Me	1m	Fe	DMC	24	100%	99%
8	Me, prenyl	1n	Fe	DMC	5	100%	40%^e
9^c	Me, prenyl	1n	Bi	DMC	24	100%	32%^f
10	Me, prenyl	1n	Fe	CH3NO2	5	100%	21%^g
11^c	Me, prenyl	1n	Bi	CH3NO2	24	100%	—^h

---

With substrate 1l featuring a monosubstituted double bond, no conversion was observed under our optimised conditions either using Fe-MMT or Bi-MMT (entries 1 and 2). Interestingly, by using a Bi-MMT prepared by impregnation of Bi(OTf)₃ instead of BiCl₃, a conversion of 64% was observed in DMC for an isolated yield of cyclised product 2l of 50% (entry 3). Substrate 1m, with the phenol ring ortho-substituted with a prenyl side chain, cyclised readily either with Fe-MMT or Bi-MMT in DMC or CH₃NO₂ (entries 4–7), the best result being obtained in our standard conditions with 100% conversion and a quantitative yield of 2m (entry 7). Finally, substrate 1n, an ortho-geranylated phenol, reacted faster than 1l and 1m reaching total conversion in 5 hours in our optimised conditions but to yield a mixture of the expected chromane 2n and a tricyclic compound 3n resulting from a tandem reaction involving both double bonds (entries 8–10). While with Fe-MMT, only a mixture was obtained after several attempts, with Bi-MMT (ex-BiCl₃) in CH₃NO₂, tricyclic product 3n was formed in 96% yield as the sole product (entry 11 and Scheme 3). We hypothesised that the tandem process might start by the activation of the remote double bond by the catalyst followed by the attack of the internal double bond and interception of the carbenium ion formed by the phenolic hydroxyl group. It is worth mentioning that olefins that should be formed upon proton elimination from these intermediates were not observed,⁴⁶ suggesting a fast ring closure to 2n.

Scheme 3 Tandem reaction of 1n with Bi-MMT as catalyst and possible mechanism.

The superiority of Bi-MMT over Fe-MMT in this specific process could be a consequence of the difference of ionic radii between Fe³⁺ and Bi³⁺, the larger Bi³⁺ being most likely to interact with the remote position of the substrate (Bi: 0.96–1.17 Å; Fe: 0.49–0.78 Å). To test our hypothesis, we engaged dienol substrate 1o, featuring similarly to 1n an internal trisubstituted double bond, but with a terminal primary hydroxyl group and a tetrasubstituted double bond. In our optimised conditions with Fe-MMT, only the intramolecular hydroalkoxylation was observed and delivered the corresponding cyclic ether 2o in 56% yield (89% conversion) in DMC and 71% yield in CH₃NO₂. With Bi-MMT in CH₃NO₂, a mixture of cyclic products was obtained without selectivity (Scheme 4). It is worth noting that in Brønsted acid-promoted reaction, 1o is the precursor of an industrially relevant odorant molecule, ambroxide, formed upon a tandem cyclisation process.⁴⁷

Scheme 4 Cyclisation of 1o to 2o.

Intermolecular hydroalkoxylation

Having in hand an efficient catalytic system for the intramolecular hydroalkoxylation of olefins, we next focused on the more challenging intermolecular version of the reaction. For this reaction, one has to address several hurdles compared with the intramolecular version, such as the difficulty to run a bimolecular reaction with two substrates of low affinity, typically a polar protic alcohol and an apolar hydrophobic olefin, and the energetic penalties due to an entropically disfavoured process. Consequently, most of the catalytic systems described in the literature report modest conversion rates and yields and reactions performed at elevated temperatures. For example, in homogeneous catalysis, a combination of Pd(II) complexes and Cu(II) salts was used for the hydroalkoxylation of styrene derivatives¹⁸ and vinylphenols.²⁰ Metal triflates such as Zr(II), La(III)⁴⁸ and Hg(II)⁴⁹ have been used for styrene and α-olefins hydroalkoxylation, respectively. Styrenes and α-olefins could also be converted into ethers by addition of simple alcohols with Au(III) chloride as catalyst in the presence of Cu(II) salts.⁵⁰ In heterogeneous catalysis, zeolite beta could be used to catalyse the ethoxylation of α-olefins in continuous flow,³⁸ while in batch reaction limonene and α-pinene could be functionalised with various C1–C5 alchols.²¹

In this context, we tested several MMT doped with metal cations and metal nanoparticles supported on inorganic matrix as possible catalysts for the reaction. Initial attempts on a typical α-olefin (1-decene) in various alcoholic solvents (MeOH, EtOH, n-PrOH, n-BuOH) or in organic solvents in the presence of alcohols with the catalysts of Table 1 were unsuccessful, the starting material being recovered unchanged. Positive control experiments of the catalysts activity were performed on the cyclisation of 1b to 2b for MMT doped with metals and dipolar cycloaddition reaction following a recently published procedure for Cu nanoparticles.⁵¹ The olefin class could significantly account for the absence of reactivity in our conditions and we thus moved towards trisubstituted double bonds such as in 1p and 1q–r obtained by alkylation of citronellol and Wittig olefination, respectively (Scheme 5) and oct-1-en-3-one 1s presenting an activated terminal double bond. Substrate 1p has been already studied for intermolecular hydroalkoxylation catalysed by Sn(OTf)₄ by some of us and could be useful to compare with a homogeneous system.²³

Scheme 5 Preparation of 1p, 1q and 1r.

For these substrates, we focused on the use of MMT doped with metal cations or nanoparticles, and proceeded to an initial screening (Table 6).

Table 6 Hydroalkoxylation of olefins 1p–r^a

Entry	Substrate	M (x)	Solvent, T °C, R11OH	Conversion^b	Product, yield^b
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed solvent (1 mL) and M-MMT stirred at the specified temperature for 48 h.b Determined by ^1H-NMR or GC-TCD by external calibration.c Used as the solvent.d Determined by GC-MS.
1	1p	Fe (5)	MeOH^c, reflux	63%	2pa, 54%
2	1p	Bi (5)	MeOH^c, reflux	72%	2pa, 45%
3	1p	Fe (5)	EtOH^c, reflux	40%	2pb, 16%
4	1p	Bi (5)	EtOH^c, reflux	24%	2pb, 13%
5	1p	Fe (5)	n-PrOH^c, reflux	37%	2pc, 26%^d
6	1p	Bi (5)	n-PrOH^c, reflux	30%	2pc, 17%^d
7	1p	Fe (5)	n-BuOH^c, reflux	29%	2pd, 19%^d
8	1p	Bi (5)	n-BuOH^c, reflux	79%	2pd, 8%^d
9	1q	Fe (5)	MeOH^c, reflux	27%	2qa, 9%
10	1q	Bi (5)	MeOH^c, reflux	17%	2qa, 14%
11	1q	Ti (5)	MeOH^c, reflux	33%	2qa, 22%
12	1s	Cu NPs (0.25)	MeOH^c, reflux	100%	Degradation

---

The results showed olefins conversions below 79% and yields of hydroalkoxylation product up to 54%. Substrate 1s led to degradation products (entry 12). Interestingly, comparing entries 1–3 with 9–11, the presence of the methoxy substituent (R₁₀) of 1p instead of a methyl group allows for better results in terms of both conversion and yield. This trend could be seen as an effect on the overall polarity of the electrophile, as well as an improvement of its coordination ability. In order to favour the contact between the polar protic nucleophiles with hydrophobic electrophiles, several attempts to use ionic liquids or surfactants were performed with M-MMT as catalysts but did not allow to improve conversions and yields (data not shown).

Finally we tested equimolar amounts of our recyclable Ti-MMT for the hydroalkoxylation of 1q–r (Table 7).

Table 7 Hydroalkoxylation of olefins 1q–r with equimolar Ti-MMT^a

Entry	Substrate	R11OH	Conversion^b	Product, yield^b
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed solvent (1 ml) and Ti-MMT stirred at the specified temperature for 48 h.b Determined by GC-TCD by external calibration.
1	1q	Et	36%	2qb, 36%
2	1r	n-Pr	45%	2rc, 45%
3	1r	n-Bu	42%	2rd, 42%
4	1r	Solketal	23%	2re, 23%

---

By increasing the molar fraction of Ti-MMT, improved yields could thus be obtained for hydrophobic olefins 1q–r at 70 °C even with higher alcohols such as n-propanol and n-butanol or solketal which are typically leading to low yields.

Mechanistic studies

A reasonable question when using supported catalysts is whether or not the process is truly heterogeneously catalysed. In our case, metal cations are simply anchored to the MMT support by electrostatic binding and could be displaced in the reaction medium. For example, the result presented in Fig. 1 with CH₃NO₂ as the solvent suggests a loss of a fraction of supported metal at each cycle, solubilised in such a polar solvant, resulting in an overall decrease of the solid catalyst efficiency. In DMC however, a solvent in which the activity was maintained, we hypothesised that most of metal cations remained bound to MMT. We further performed a hot filtration test to determine whether or not solubilised metal species could be active in solution. The test consisted in monitoring the reaction progress (product formation) until the conversion reached ca. 50% conversion. The whole reaction mixture was then filtered over a fritted glass heated at 80 °C and the filtrate was put to react again at the same temperature under monitoring. The results are presented in Fig. 2 for reactions of 1b with Fe-MMT and Bi-MMT in DMC.

Fig. 1 Recycling studies of Fe-MMT in DMC and CH₃NO₂.

Fig. 2 Hot filtration tests on reaction of 1b with Fe-MMT and Bi-MMT in DMC at 80 °C.

It is clear that upon removal of the solid catalyst, no additional product formation was observed, which endorsed a truly heterogeneous process. For the sake of comparison, we tested the activity of FeCl₃ and BiCl₃ in homogeneous conditions on the same reaction, these halide salts being the precursors of Fe-MMT and Bi-MMT, respectively (Scheme 6).

Scheme 6 Comparison between heterogeneous and homogeneous versions of the reaction.

With FeCl₃ (5 mol%) after 7 hours, the expected cyclised product 2b was formed in 76% yield, along with 5% of tetrahydronaphtalene derivative formed upon Lewis acid-catalysed Friedel–Crafts type addition of the phenyl ring to the double bond. With BiCl₃, substrate 1b was recovered unchanged while with Bi-MMT, 2b was formed in 76% yield, indicating the presence of active bismuth species within the solid catalyst. Supported catalysts Fe-MMT and Bi-MMT were thus superior to their homogeneous equivalents in terms of activity and selectivity.

Control reactions with solid catalytic materials such as pristine Na-MMT, MMT K10, Amberlyst and Nafion were performed on various substrates (Table 8). With Na-MMT, no conversion of starting materials 1a, 1e, 1f was observed (entries 2–4). With a physical mixture of Na-MMT and BiCl₃ (5 mol%) as catalyst, conversion was not complete and a complex mixture of products was obtained (entry 5), as well as with MMT K10 (entry 6). With acidic resins Amberlyst-15 and Nafion SAC-13, a good reaction profile was observed in terms of conversion and yields, but the formation of the undesired hydroarylation product 2b′ was observed with 17–18% selectivity (entries 7 and 8). The reaction could thus be observed under Brønsted catalysis, but the most efficient conditions and catalyst remained those described above with Fe-MMT. These data suggested a truly metal-catalysed rather than a proton-catalysed process.

Table 8 Comparison of Fe-MMT with solid catalysts

Entry	Catalyst	Substrate	Time (h)	Conversion^c	Products, selectivity^c
a Reaction conditions: substrate (0.25 mmol), anhydrous and degassed CH3NO2 (1 mL) at 60 °C. Fe-MMT: 25 mg; Amberlyst: 20 mg; Nafion: 20 mg.b Reaction conditions: substrate (0.5 mmol), anhydrous and degassed DMC (2 ml) at 80 °C. Na-MMT, MMT K10: 50 mg.c Determined by GC-TCD by external calibration.d Less than 5% yield of dehydrated product was observed by GC-MS.
1^a	Fe-MMT	1b	7	100%	2b/2b′ 100/0
2^b	Na-MMT	1a	24	0%	—
3^b	Na-MMT	1e	12	0%	—^d
4^b	Na-MMT	1f	4.5	0%	—
5^b	Na-MMT + 5 mol% BiCl3	1a	7	Mixture
6^b	MMT K10	1a	7	100%	Mixture
7^a	Amberlyst 15	1b	24	100%	2b/2b′ 83/17
8^a	Nafion SAC-13	1b	24	100%	2b/2b′ 82/18

---

The question of the characterisation of the coordination sphere of the cations is difficult to address on these solid catalysts. It is reasonable to imagine mixed species with one or two chlorine atoms replaced by silicates from the interlamellar surface of MMT (Scheme 7).

Scheme 7 Hypothetical metal species within MMT doped with metal cations.

In such hypothesis where the binding to a pseudo-surface brings a large degree of steric hindrance, the metal–oxygen bond lengths should be longer than expected and cations should therefore exhibit an enhanced Lewis acid character, as it is the case with metal salts with bulky anions of low nucleophilicity, e.g. triflates or triflimidates.^52,53

Conclusions

Efficient, cost-effective, and sustainable catalysts for hydroalkoxylation reactions of olefins could be obtained upon impregnation of Fe(III), Bi(III), and Ti(IV) salts in montmorillonite. For intramolecular reactions, a large number of substrates with various substitution patterns could be efficiently cyclised. For intermolecular reactions, the results were modests in terms of both conversions and yields but remained in the range of what is typically observed under homogeneous and heterogeneous catalysis. These catalytic systems were shown to proceed under truly heterogeneous catalysis and could be transferred in continuous flow conditions, the first step towards industrial transfer.

Experimental

¹H NMR and ¹³C NMR spectra were recorded on BRUCKER AC 200 (200 MHz). ¹H NMR spectra are reported as follows: chemical shift in ppm (δ) relative to the chemical shift of CDCl₃ at 7.26 ppm, integration, multiplicities (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet and br = broadened), and coupling constants (Hz). ¹³C NMR spectra reported in ppm (δ) relative to CDCl₃ at 77.16 ppm. Column chromatography was carried out on silica gel (spherical, neutral, 63–200 um, Geduran Si 60, Merck KGaA). GC-TCD analyses were carried out using a Shimadzu QP2010 plus gas chromatograph, under the following operation conditions: vector gas, He; injector temperature, 250 °C; detector temperature, 210 °C at 60 mA; split ratio, 1/20; total flow, 22.5 ml min⁻¹; Phenomenex Zebron ZB5MS column, polydimethylsiloxane (10 m, inside diameter 0.10 mm, film thickness 0.10 μm); temperature program, 80–200 °C at 10 °C min⁻¹ and 200 °C for 8 min. GC/MS analyses were performed by using a Shimadzu QP2010 gas chromatograph (conditions: carrier gas, He; injector and detector temperatures, 250 °C; injected volume, 0.5 μL; split ratio, 1/100; pressure, 180 kPa; SLB-5ms capillary column (thickness: 0.25 mm, length: 30 m, inside diameter: 0.25 mm); temperature program, 60–250 °C at 2 °C min⁻¹, and 250 °C), coupled to a mass selective detector. Mass spectra were obtained by electron ionisation at 70 eV, m/z 35–400, source temperature 250 °C; only the most abundant ions are given. High resolution mass spectrometry (HRMS) was performed at ERINI platform (Grasse, FRANCE) using a Waters APGC coupled with a Waters Xevo G2 QTOF spectrometer. Screening reactions were performed in a Carousel 12 Plus parallel synthesizer purchased from Radleys.

General procedure for cyclisation of unsaturated alcohols

In a Schlenk tube, substrate (1 mmol), Fe-MMT (5 mol% of metal, 100 mg of material) and DMC (2 ml) are introduced and the tube closed with a PTFE cap. The mixture is stirred at the desired temperature and the reaction monitored by TLC or GC-TCD. After completion, the mixture is filtered through a cotton wool pad. The filter is rinsed with diethyl ether and the solution concentrated at reduced pressure affording the crude cyclic ether which is then purified by flash chromatography over silica gel (petroleum ether/Et₂O).

2,2,6-Trimethyltetrahydro-2H-pyran 2a^22,42. ¹H NMR (DMC-d₆, 200 MHz): δ 3.78–3.54 (m, 1H), 1.75–1.53 (m, 2H), 1.52–1.25 (m, 4H), 1.17 (s, 3H), 1.13 (s, 3H), 1.04 (d, J = 6.1 Hz, 3H). ¹³C NMR (DMC-d₆, 50 MHz): δ 71.64, 66.65, 36.43, 33.99, 31.84, 22.53, 21.67, 20.59. MS (EI; 70 eV) 128(0) [M]⁺˙, 113(47), 95(7), 70(10), 59(100), 56(36), 43(80).

2,2-Dimethyl-5,5-diphenyltetrahydro-2H-pyran 2b¹⁹. ¹H NMR (CDCl₃, 200 MHz): δ 7.37–7.10 (m, 10H), 4.06 (s, 2H), 2.50–2.33 (m, 2H), 1.40 (dd, J = 7.3, 5.0 Hz, 1H), 1.23 (s, 6H). ¹³C NMR (CDCl₃, 50 MHz): 146.54, 128.21, 128.02, 126.05, 71.32, 69.06, 45.96, 32.65, 30.91, 26.46. MS (EI; 70 eV) 266(5) [M]⁺˙, 236(15), 180(100), 165(36), 152 (4); 129(2), 115(13), 103(4), 91(14), 73(8), 65(3), 43(10), 41(5).

2,2-Dimethyl-6-pentyltetrahydro-2H-pyran 2c. ¹H NMR (CDCl₃, 200 MHz): δ 3.50–3.41 (m, 1H), 1.67–0.94 (m, 14H), 1.19 (s, 3H), 1.17 (s, 3H), 0.87 (t, J = 6.8 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 71.57, 70.46, 37.10, 36.45, 32.10, 31.66, 25.39, 22.79, 22.13, 20.28, 14.22. MS (EI; 70 eV) 184(0) [M]⁺˙, 169(16), 151 (4), 126(3), 113(63), 99 (15), 95(61), 83(15), 69(38), 59(100), 56(72), 43(57). HRMS calculated for C₁₂H₂₃O (M − H)⁺: 183.1749; found: 183.1743. |Δ| = 3.3 ppm.

2,2-Dimethyl-6-(2-phenylethyl)-tetrahydro-2H-pyran 2d. ¹H NMR (CDCl₃, 200 MHz): δ 7.43–7.01 (m, 5H), 3.66–3.30 (m, 1H), 2.89–2.46 (m, 2H), 1.90–0.99 (m, 8H), 1.23 (s, 3H), 1.17 (s, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 142.64, 128.62, 128.23, 125.60, 77.80, 77.16, 76.53, 71.62, 69.19, 38.49, 36.39, 32.05, 31.76, 22.07, 20.18. MS (EI; 70 eV) 218(12) [M]⁺˙, 185(6), 157 (7), 140(25), 129(14), 113(34), 104(36), 91(100), 69(22), 56(14), 43(29). HRMS calculated for C₁₅H₂₃O (M + H)⁺: 219.1749; found: 203.1751. |Δ| = 0.9 ppm.

2,2-Dimethyl-6-(p-tolyl)tetrahydro-2H-pyran 2e. ¹H NMR (CDCl₃, 200 MHz): δ 7.25 (d, J = 8 Hz, 2H), 7.12 (d, J = 8 Hz, 2H), 4.55 (dd, J = 11.5, 1.9 Hz, 1H), 2.31 (s, 3H), 1.82–1.41 (m, 6H), 1.30 (s, 3H), 1.29 (s, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 141.33, 136.76, 129.04, 126.15, 72.90, 72.42, 36.19, 34.21, 32.13, 22.07, 21.24, 20.61. MS (EI; 70 eV) 204(14) [M]⁺˙, 189(5), 146 (25), 131(15), 121(100), 105 (5), 91(19), 84(4), 77(6), 69(6), 56(31), 43(11). HRMS calculated for C₁₄H₁₉O (M − H)⁺: 203.1436; found: 203.1436. |Δ| = 0.0 ppm.

cis/trans-(2-Isopropyl-1,3-dioxolan-4-yl)methanol 2f and cis,trans-2-isopropyl-1,3-dioxan-5-ol 2f′⁵⁴. [cis/trans-2f] ¹H NMR (CDCl₃, 200 MHz): δ 4.75 (d, J = 4.7 Hz, 1H), 4.66 (d, J = 4.6 Hz, 1H), 4.32–4.02 (m, 3H), 4.00–3.16 (m, 7H), 2.04 (br s, 2H), 1.94–1.70 (m, 2H), 0.97 (d, J = 5.6 Hz, 6H), 0.94 (d, J = 6.6 Hz, 6H). ¹³C NMR (CDCl₃, 50 MHz): δ 108.79, 108.65, 76.52, 76.29, 66.80, 66.55, 63.52, 62.76, 32.30, 31.85, 16.99, 16.95, 16.89, 16.80. MS (EI; 70 eV) 146(0) [M]⁺˙, 145(1), 115(3), 103(82), 97(7), 86(2), 71(5), 69(2), 57(100), 55(31), 43(24), 41(15) (I stereoisomer). MS (EI; 70 eV) 146(0) [M]⁺˙, 145(2), 115(4), 103(80), 97(10), 86(1), 71(6), 69(3), 57(100), 55(31), 43(26), 41(16) (II stereoisomer) [cis/trans-2f′]. ¹H NMR (CDCl₃, 200 MHz): δ 4.27 (d, J = 5 Hz, 1H), 4.20 (d, J = 5 Hz, 1H), 4.13 (t, J = 4.6 Hz, 2H), 4.01 (d, J = 11.2 Hz, 1H), 3.93–3.76 (m, 3H), 3.86 (d, J = 11.2 Hz, 1H), 3.54–3.45 (m, 1H), 3.36 (d, J = 10.9 Hz, 1H), 3.31 (d, J = 10.2 Hz, 1H), 2.02 (br s, 2H), 1.91–1.66 (m, 2H), 0.94 (d, J = 6.8 Hz, 6H), 0.92 (d, J = 6.8 Hz, 6H). ¹³C NMR (CDCl₃, 50 MHz): δ 106.35, 105.69, 71.88, 71.70, 64.28, 61.61, 32.82, 32.35, 17.31, 16.94. MS (EI; 70 eV) 146(0) [M]⁺˙, 145(1), 116(1), 103(100), 73(53), 57(63), 55(31), 43(42), 41(15) (I stereoisomer). MS (EI; 70 eV) 146(0) [M]⁺˙, 145(1), 116(2), 103(100), 73(54), 57(64), 55(33), 43(46), 41(17) (II stereoisomer).

cis/trans-(2-Isopropyl-1,3-dioxolan-4-yl)-methyl acetate 2g. ¹H NMR (CDCl₃, 200 MHz): δ 4.73 (d, J = 4.6 Hz, 1H), 4.66 (d, J = 4.4 Hz, 1H), 4.37–4.18 (m, 2H), 4.17–3.98 (m, 3H), 3.91 (dd, J = 8.3, 6.9 Hz, 1H), 3.74 (dd, J = 8.3, 4.9 Hz, 1H), 3.58 (dd, J = 8.4, 6.7 Hz, 1H), 2.08 (s, 3H), 2.06 (s, 3H), 1.94–1.67 (m, 2H), 0.97–0.88 (m, 12H). ¹³C NMR (CDCl₃, 50 MHz): 170.89, 170.84, 109.02, 108.51, 73.70, 73.53, 67.27, 67.19, 64.72, 64.30, 32.06, 31.77, 20.89, 20.86, 16.93, 16.87, 16.59, 16.55. MS (EI; 70 eV) 188(0) [M]⁺˙, 187(0), 145(43), 115(1), 97(4), 86(1), 71(3), 61(4), 57(18), 43(100), 41(7) (I stereoisomer). MS (EI; 70 eV) 188(0) [M]⁺˙, 187(1), 145(46), 115(2), 97(45), 86(1), 71(4), 69(2), 57(19), 43(100), 41(7) (II stereoisomer). HRMS calculated for C₉H₁₅O₄ (M − H)⁺: 187.0970; found: 187.0967. |Δ| = 1.6 ppm.

2-Hexyltetrahydrofuran 2i and 2-pentyltetrahydro-2H-pyran 2i′^42,55. [THF-ether/2i] ¹H NMR (CDCl₃, 200 MHz): δ 3.97–3.59 (m, 3H), 2.08–1.69 (m, 3H), 1.69–1.03 (m, 11H), 0.86 (t, J = 6.8 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 79.59, 67.70, 35.89, 31.97, 31.51, 29.55, 26.50, 25.84, 22.73, 14.19. MS (EI; 70 eV) 156(1) [M]⁺˙, 138 (1), 96 (1), 81(1), 71(100), 55(6), 43(31), 41(31) [THP − ether/2i′]. ¹H NMR (CDCl₃, 200 MHz): δ 4.01–3.88 (m, 1H), 3.40 (td J = 10.9, 3.6 Hz, 1H), 3.29–3.08 (m, 1H), 2.17–1.04 (m, 14H), 0.87 (t, J = 6.5 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 78.08, 68.63, 36.78, 32.12, 32.09, 26.40, 25.36, 23.76, 22.77, 14.19. MS (EI; 70 eV) 156(1) [M]⁺˙, 138 (1), 95 (1), 85(100), 67(16), 57(15), 43(21), 41(24).

2-Phenyltetrahydro-2H-pyran 2j⁴¹. ¹H NMR (CDCl₃, 200 MHz): δ 7.39–7.15 (m, 5H), 4.36–4.24 (m, 1H), 4.12 (dd, J = 11.4, 3.5 Hz, 1H), 3.76–3.46 (m, 1H), 2.01–1.46 (m, 6H). ¹³C NMR (CDCl₃, 50 MHz): 143.45, 128.38, 127.38, 125.95, 80.26, 69.11, 34.14, 26.02, 24.14. MS (EI; 70 eV) 162(54) [M]⁺˙, 161 (46), 144(2), 133(4), 129(4), 115 (6), 105 (100), 91(29), 77(45), 65(8), 55(26), 51(19), 41(31).

2-Methyl-2,3-dihydrobenzofuran 2l⁵⁶. ¹H NMR (CDCl₃, 200 MHz): δ 7.23–7.05 (m, 2H), 6.85 (dd, J = 7.4, 0.9 Hz, 1H), 6.77 (d, J = 7.9 Hz, 1H), 4.93 (ddq, J = 8.7, 7.7, 6.2 Hz, 1H), 3.32 (dd, J = 15.4, 8.8 Hz, 1H), 2.82 (dd, J = 15.4, 7.7 Hz, 1H), 1.48 (d, J = 6.2 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 159.63, 128.07, 127.15, 125.08, 120.28, 109.44, 76.52, 37.25, 21.88. MS (EI; 70 eV) 134(100) [M]⁺˙, 133(42), 119(66), 115(27), 113(34), 119(11), 107(100), 91(22), 77(25), 65(6), 51(13), 41(11).

2,2-Dimethylchromane 2m^57,58. ¹H NMR (CDCl₃, 200 MHz): δ 7.15–7.05 (m, 2H), 6.90–6.70 (m, 2H), 2.80 (t, J = 6.7 Hz, 2H), 1.83 (t, J = 6.7 Hz, 2H), 1.36 (s, 6H). ¹³C NMR (CDCl₃, 50 MHz): δ 153.48, 128.93, 126.73, 120.39, 119.08, 116.72, 73.56, 32.29, 26.37 (2), 21.94. MS (EI; 70 eV) 162(11) [M]⁺˙, 147(17), 133(17), 119(40), 107(100), 91(5), 77(8), 51(8), 41(9).

2-Methyl-2-(4-methylpent-3-enyl)chromane 2n and cis/trans-1,1,4a-trimethyl-2,3,4,4a,9,9a-hexahydro-1H-xanthene 3n⁵⁹. [2n] ¹H NMR (CDCl₃, 200 MHz): δ 7.06 (d, J = 7.3 Hz, 2H), 6.88–6.70 (m, 2H), 5.17–5.06 (m, 1H), 2.82–2.52 (m, 2H), 2.08–1.39 (m, 6H), 1.72 (s, 3H), 1.64 (s, 3H), 1.34 (s, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 154.09, 129.50, 127.33, 124.38, 121.22, 119.64, 117.41, 76.02, 39.66, 31.06, 25.80, 24.37, 22.40, 22.24, 17.71. MS (EI; 70 eV) 230(52) [M]⁺˙, 215(2), 187(20), 174(17), 161(27), 147(48), 133(20), 123(59), 107(94), 91(40), 81(40), 69(74), 55(16), 41(100) [cis-3n]. ¹H NMR (CDCl₃, 200 MHz): δ 7.05 (d, J = 7.1 Hz, 2H), 6.88–6.69 (m, 2H), 3.05 (dd, J = 17.7, 7.9 Hz, 1H), 2.76 (d, J = 17.7 Hz, 1H), 2.08–1.20 (m, 7H), 1.21 (s, 3H), 0.97 (s, 3H), 0.65 (s, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 154.64, 129.04, 126.81, 122.18, 119.94, 117.23, 75.36, 44.58, 41.80, 39.71, 34.12, 32.41, 27.16, 23.74, 21.54, 18.23. MS (EI; 70 eV) 230(31) [M]⁺˙, 214(4), 187(2), 173(1), 159(11), 145(17), 123(100), 107(39), 91(10), 81(30), 77(10), 67(12), 55(10), 43(9), 41(18) [trans-3n]. ¹H NMR (CDCl₃, 200 MHz): δ 7.09 (d, J = 7.1 Hz, 2H), 6.81–6.72 (m, 2H), 2.73 (dd, J = 16.2, 6.7 Hz, 1H), 2.57 (d, J = 16.4 Hz, 1H), 2.17–1.16 (m, 7H), 1.23 (s, 3H), 1.04 (s, 3H), 0.94 (s, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 154.38, 129.75, 127.23, 122.72, 119.74, 117.15, 76.29, 48.19, 41.63, 40.13, 33.49, 32.22, 23.37, 20.80, 19.96, 19.91. MS (EI; 70 eV) 230(54) [M]⁺˙, 215(14), 187(10), 173(3), 159(26), 145(37), 123(100), 107(92), 91(21), 81(31), 77(17), 67(13), 55(18), 43(14), 41(32).

2-Methyl-2-(2-(2,6,6-trimethylcyclohex-1-en-1-yl)ethyl)tetrahydrofuran 2o⁶⁰. ¹H NMR (CDCl₃, 200 MHz): δ 3.97–3.68 (m, 1H), 2.17–1.33 (m, 14H), 1.59 (s, 1H), 1.21 (s, 1H), 0.98 (s, 1H). ¹³C NMR (CDCl₃, 50 MHz): δ 137.12, 126.90, 82.88, 67.20, 41.13, 40.05, 36.83, 35.22, 32.92, 28.78, 26.24, 25.48, 23.63, 19.84, 19.69. MS (EI; 70 eV) 236(2) [M]⁺, 221(2), 203(2), 177(4), 161(2), 149(2), 136(18), 123(15), 121(25), 107(15), 95(17), 93(17), 85(100), 81(14), 69(9), 67(10), 55(14), 43(52), 41(24).

1,7-Dimethoxy-3,7-dimethyloctane 2pa²³. ¹H NMR (CDCl₃, 200 MHz): δ 3.39 (t, J = 6.7 Hz, 2H), 3.32 (s, 3H), 3.16 (s, 3H), 1.79–1.20 (m, 9H), 1.12 (s, 6H), 0.88 (d, J = 6.4 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 74.6, 71.2, 58.6, 49.1, 40.1, 37.7, 36.7, 30.2, 21.2, 19.6. MS (EI; 70 eV) 202(0) [M]⁺˙, 187(4), 138(11), 123(29), 109(14), 95(50), 81(95), 73(100), 55(94).

7-Ethoxy-1-methoxy-3,7-dimethyloctane 2pb²³. ¹H NMR (CDCl₃, 200 MHz): δ 3.50–3.25 (m, 7H); 1.70–1.00 (m, 18H); 0.82 (d, J = 6.4 Hz, 3H). MS (EI; 70 eV) 216(0) [M]⁺˙, 87(100), 59(59), 55(9), 45(13), 43(13), 41(9).

1-Methoxy-7-propyloxy-3,7-dimethyloctane 2pc²³. ¹H NMR (CDCl₃, 200 MHz): δ 3.39 (t, J = 6.7 Hz, 2H), 3.31 (s, 3H), 3.23 (t, J = 6.7 Hz, 3H), 1.80–1.10 (m, 11H), 1.06 (s, 6H), 0.90–0.75 (m, 6H). MS (EI; 70 eV) 230(0) [M]⁺˙, 215(1), 101(74), 83(9), 69(7), 59(100), 55(11), 45(16), 43(21), 41(14).

7-Butoxy-1-methoxy-3,7-dimethyloctane 1pd²³. ¹H NMR (CDCl₃, 200 MHz): δ 3.40–3.10 (m, 7H); 1.70–1.00 (m, 22H); 0.82 (d, J = 6.4 Hz, 3H). MS (EI; 70 eV) 244(0) [M]⁺˙, 229 (1), 115(55), 83(7), 69(7), 59(100), 55(10), 45(15), 43(10), 41(16)

2-Methoxy-2-methyldecane 2qa. ¹H NMR (CDCl₃, 200 MHz): δ 3.17 (s, 3H), 1.52–1.37 (m, 2H), 1.37–1.20 (m, 12H), 1.12 (s, 6H), 0.88 (d, J = 6.4 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 74.79, 49.20, 39.97, 32.06, 30.41, 29.79, 29.47, 25.12 (2), 24.03, 22.82, 14.26. MS (EI; 70 eV) 186(0) [M]⁺˙, 171(3), 73(100), 69(4), 55(8), 43(10).

2-Ethoxy-2-methyldecane 2qb. ¹H NMR (CDCl₃, 200 MHz): δ 3.29 (q, J = 7.0 Hz, 2H), 1.45–1.30 (m, 2H), 1.30–1.15 (m, 12H), 1.08 (t, J = 7.0 Hz, 3H), 1.07 (s, 6H), 0.81 (d, J = 6.4 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 74.59, 56.38, 40.29, 32.06, 30.40, 29.79, 29.47, 25.85 (2), 24.06, 22.82, 16.36, 14.25. MS (EI; 70 eV) 200(0) [M]⁺˙, 185(3), 87(100), 59(57), 43(16).

2-Propyloxy-2-methylnonane 2rc. ¹H NMR (CDCl₃, 200 MHz): δ 3.25 (t, J = 6.8 Hz, 2H), 1.66–1.38 (m, 4H), 1.38–1.19 (m, 10H), 1.13 (s, 6H), 0.90 (t, J = 7.3 Hz, 3H), 0.88 (d, J = 6.4 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 74.39, 62.93, 40.44, 32.04, 30.38, 29.51, 25.78 (2), 24.05, 23.99, 22.83, 14.26, 10.94. MS (EI; 70 eV) 200(0) [M]⁺˙, 171(4), 87(100), 69(13), 59(64), 43(19).

2-Butyloxy-2-methylnonane 2rd. ¹H NMR (CDCl₃, 200 MHz): δ 3.28 (t, J = 6.4 Hz, 2H), 1.61–1.34 (m, 6H), 1.34–1.18 (m, 10H), 1.12 (s, 6H), 1.01–0.81 (m, 6H). ¹³C NMR (CDCl₃, 50 MHz): δ 74.37, 60.91, 40.43, 32.96, 32.04, 30.38, 29.50, 25.76 (2), 24.04, 22.83, 19.65, 14.25, 14.13. MS (EI; 70 eV) 214(0) [M]⁺˙, 199(1), 173(22), 115(95), 101(68), 85(28), 71(31), 57(84), 43(100).

2,2-Dimethyl-4-((2-methylnonan-2-yloxy)methyl)-1,3-dioxolane 2re. ¹H NMR (CDCl₃, 200 MHz): δ 4.18–4.11 (m, 1H), 4.10–4.01 (m, 1H), 3.80–3.69 (m, 1H), 3.50–3.38 (m, 1H), 3.32–3.19 (m, 1H), 1.52–1.38 (m, 5H), 1.36 (s, 3H), 1.30–1.18 (m, 10H), 1.13 (s, 6H), 0.88 (t, J = 6.4 Hz, 3H). ¹³C NMR (CDCl₃, 50 MHz): δ 109.25, 75.36, 75.14, 67.68, 62.98, 40.35, 32.03, 30.32, 29.49, 26.95, 25.62, 25.56 (2), 24.01, 22.82, 14.25. MS (EI; 70 eV) 272(0) [M]⁺˙, 257(8), 173(19), 115(84), 101(65), 85(25), 71(28), 57(77), 43(100).

Acknowledgements

This work was supported by the University Nice Sophia Antipolis, the CNRS, and the ANR program CD2I (Nanocausys project, grant number 12-CDII-0010-02). We are grateful to Dr Charles Fehr (Firmenich, CH) for a kind gift of compound (E)-1o and Baptiste Giroire from ICMCB (Bordeaux, France) for material analyses.

Notes and references

1. Transition Metals for Organic Synthesis, ed. M. Beller and C. Bolm, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, 2004.
2. Catalysis without Precious Metals, ed. R. M. Bullock, Wiley-VCH, Weinheim, 2010.
3. Metal-catalyzed cross-coupling reactions, ed. A. D. Meijere and F. Diederich, Wiley-VCH, Weinheim, 2004.
4. Acid Catalysis in Modern Organic Synthesis, ed. H. Yamamoto and K. Ishihara, Wiley-VCH, Weinheim, 2008.
5. C. E. Garrett and K. Prasad, Adv. Synth. Catal., 2004, 346, 889.
6. M. J. Sabater, F. Rey and J. Lazaro, RSC Green Chem. Ser., 2010, 5, 86.
7. S. L. Wegener, T. J. Marks and P. C. Stair, Acc. Chem. Res., 2012, 45, 206.
8. Supported catalysts and their applications, ed. D. C. Sherrington and A. P. Kybett, Royal Society of Chemistry, Cambridge, 2001.
9. J. H. Clark, Acc. Chem. Res., 2002, 35, 791.
10. D. J. MacQuarrie, in Heterogenized Homogeneous Catalysts for Fine Chemicals Production, ed. P. Barbaro and F. Liguiro, Springer, New York, 2010, p. 1.
11. J. R. Dodson, A. J. Hunt, H. L. Parker, Y. Yang and J. H. Clark, Chem. Eng. Process, 2012, 51, 69.
12. W. M. Czaplik, M. Mayer, J. Cvengroš and A. J. von Wangelin, ChemSusChem, 2009, 2, 396.
13. G. K. S. Prakash, T. Mathew and G. A. Olah, Acc. Chem. Res., 2012, 45, 565.
14. M. Sankar, N. Dimitratos, P. J. Miedziak, P. P. Wells, C. J. Kiely and G. J. Hutchings, Chem. Soc. Rev., 2012, 41, 8099.
15. M. S. Holzwarth and B. Plietker, ChemCatChem, 2013, 5, 1650.
16. M. Lamblin, L. Nassar-Hardy, J.-C. Hierso, E. Fouquet and F.-X. Felpin, Adv. Synth. Catal., 2010, 352, 33.
17. G. B. B. Varadwaj and K. M. Parida, RSC Adv., 2013, 3, 13583.
18. K. M. Gligorich, M. J. Schultz and M. S. Sigman, J. Am. Chem. Soc., 2006, 128, 2794.
19. H. Qian, X. Han and R. A. Widenhoefer, J. Am. Chem. Soc., 2004, 126, 9536.
20. Y. Zhang and M. S. Sigman, Org. Lett., 2006, 8, 5557.
21. K. Hensen, C. Mahaim and W. F. Holderich, Stud. Surf. Sci. Catal., 1997, 105, 1133.
22. L. Coulombel, M. Rajzmann, J.-M. Pons, S. Olivero and E. Duñach, Chem.–Eur. J., 2006, 12, 6356.
23. P. Lemechko, F. Grau, S. Antoniotti and E. Duñach, Tetrahedron Lett., 2007, 48, 5731.
24. M. J. L. Tschan, C. M. Thomas, H. Strub and J.-F. Carpentier, Adv. Synth. Catal., 2009, 351, 2496.
25. T. Hirai, A. Hamasaki, A. Nakamura and M. Tokunaga, Org. Lett., 2009, 11, 5510.
26. B. D. Kelly, J. M. Allen, R. E. Tundel and T. H. Lambert, Org. Lett., 2009, 11, 1381.
27. J. Tan, Z. Zhang and Z. Wang, Org. Biomol. Chem., 2008, 6, 1344.
28. A. Dzudza and T. J. Marks, Chem.–Eur. J., 2010, 16, 3403.
29. F. Ke, Z.-K. Li, H.-F. Xiang and X.-G. Zhou, Tetrahedron Lett., 2011, 52, 318.
30. I. Kamiya, H. Tsunoyama, T. Tsukuda and H. Sakurai, Chem. Lett., 2007, 36, 646.
31. K. Komeyama, T. Morimoto, Y. Nakayama and K. Takaki, Tetrahedron Lett., 2007, 48, 3259.
32. K. Hensen, C. Mahaim and W. F. Hölderich, Stud. Surf. Sci. Catal., 1997, 105, 1133.
33. J. Schlüter, M. Blazejak and L. Hintermann, ChemCatChem, 2013, 5, 3309.
34. E. Pérez-Mayoral, I. Matos, P. Nachtigall, M. Položij, I. Fonseca, D. Vitvarová-Procházková and J. Čejka, ChemSusChem, 2013, 6, 1021.
35. S. Radhakrishnan, G. Thoelen, J. Franken, J. Degrève, C. E. A. Kirschhock and J. A. Martens, ChemCatChem, 2013, 5, 576.
36. A. Dzudza and T. J. Marks, Org. Lett., 2009, 11, 1523.
37. K. Hensen, C. Mahaim and W. F. Hoelderich, Appl. Catal., A, 1997, 149, 311.
38. S. Radhakrishnan, J. Franken and J. A. Martens, Green Chem., 2012, 14, 1475.
39. I. Notar Francesco, B. Cacciuttolo, M. Pucheault and S. Antoniotti, Green Chem., 2015, 17, 837.
40. O. Pascu, B. Cacciuttolo, S. Marre, M. Pucheault and C. Aymonier, J. Supercrit. Fluids, 2015, 105, 84.
41. L. Coulombel, I. Favier and E. Duñach, Chem. Commun., 2005, 2286.
42. L. Coulombel and E. Duñach, Green Chem., 2004, 6, 499.
43. R. Tannert, L.-G. Milroy, B. Ellinger, T.-S. Hu, H.-D. Arndt and H. Waldmann, J. Am. Chem. Soc., 2010, 132, 3063.
44. F. Bigi, G. Casiraghi, G. Casnati and G. Sartori, Synthesis, 1981, 1981, 310.
45. X. Chaminade, L. Coulombel, S. Olivero and E. Duñach, Eur. J. Org. Chem., 2006, 3554.
46. J. Godeau, F. Fontaine-Vive, S. Antoniotti and E. Duñach, Chem.–Eur. J., 2012, 18, 16815.
47. A. F. Barrero, J. N. Altarejos, E. J. Alvarez-Manzaneda, J. M. Ramos and S. A. Salido, J. Org. Chem., 1996, 61, 2215.
48. D. B. G. Williams, M. S. Sibiya and P. S. van Heerden, Fuel Process. Technol., 2012, 94, 75.
49. M. Livneh, J. K. Sutter and C. N. Sukenik, J. Org. Chem., 1987, 52, 5039.
50. X. Zhang and A. Corma, Chem. Commun., 2007, 3080.
51. B. J. Borah, D. Dutta, P. P. Saikia, N. C. Barua and D. K. Dutta, Green Chem., 2011, 13, 3453.
52. S. Antoniotti, S. Poulain-Martini and E. Duñach, Synlett, 2010, 2973.
53. S. Antoniotti, V. Dalla and E. Duñach, Angew. Chem., Int. Ed., 2010, 49, 7860.
54. D. Marton and N. Vanzan, Ann. Chim., 1989, 79, 491.
55. M. C. Hansen, X. Verdaguer and S. L. Buchwald, J. Org. Chem., 1998, 63, 2360.
56. T. T. Dang, F. Boeck and L. Hintermann, J. Org. Chem., 2011, 76, 9353.
57. L. A. Adrio and K. K. Hii, Chem. Commun., 2008, 2325.
58. V. Vece, J. Ricci, S. Poulain-Martini, P. Nava, Y. Carissan, S. Humbel and E. Dunach, Eur. J. Org. Chem., 2010, 6239.
59. S. Nakamura, K. Ishihara and H. Yamamoto, J. Am. Chem. Soc., 2000, 122, 8131.
60. R. L. Snowden, J. C. Eichenberger, S. M. Linder, P. Sonnay, C. Vial and K. H. Schulte-Elte, J. Org. Chem., 1992, 57, 955.

---

Footnote

† Electronic supplementary information (ESI) available: Characterisation of catalysts, starting materials and products. See DOI: 10.1039/c5ra25176a

---