and 4 were acting as dehydration catalysts.[18] Interestingly, for
4, this same effect was not observed when sufficient linalool
was present in solution. As a control, a 0.01 mol% solution of
catalyst 4 in linalool was prepared. After the reaction had pro-
ceeded to 44% conversion, the mixture was heated to 608C
for 16 h. No dehydration of the product was observed. It is
also important to note that catalyst decomposed through air
exposure was not active for the dehydration of the alcohol.
After complete conversion of linalool to 1, rapid stirring of the
reaction mixture in open air or, alternatively, active bubbling of
air into the reaction flask resulted in a color change from
green to brown–black. This oxidized mixture was much less
prone to dehydration reactions.
Table 2. Catalysts for the dehydration of 1.
Catalyst
Temp.
[8C]
Time
[h]
Products
1:ether:dimer:oligomer
MMT-K10
Nafion Sac-13
Pd(COD)Cl2
2m HCl
MgSO4
Benzoic acid
AlPO4
25
25
25
25
25
25
25
1
1
5:41:22:32
6:35:23:36
<1:14:66:23
0:(86):13[a]
16:84:0:0
16
1
16
16
16
8:66:21:4
19:76:5:0[b]
[a] The number in parentheses is the mass% of ethers and dimers com-
bined. [b] The use of a mixed AlPO4/MgSO4 catalyst (608C, 5 h, 40 torr;
1 torr=1.333ꢁ102 Pa) allowed for isolation of MCPD in 78% yield. The
pot residue from this reaction consisted of 10% 1 and 90% ethers.
Although the dehydration reaction appeared to be mediated
by the ruthenium catalyst, another possibility is that the cata-
lyst reacted with linalool, 1, or water to exchange alkoxide or
hydroxide ligands with the chloride ligands. This process
would release catalytic amounts of HCl, which could then lead
to dehydration of the alcohol. To investigate the extent to
which a Lewis acid would dehydrate 1, the alcohol was al-
lowed to react with the Lewis acids PdCl2(PhCN)2 and [Ru-
(COD)Cl2]n at room temperature in CDCl3. As a control, Pd0 (5%
Pd/C) was also evaluated as a catalyst for the dehydration of 1.
Interestingly, all of the catalysts converted 1 to similar mixtures
of dehydrated products comparable to those observed with
the metathesis catalysts. Further observation revealed that al-
though neat samples of 1 were stable indefinitely in closed
flasks at room temperature, NMR samples in CDCl3 slowly con-
verted to dehydrated mixtures, albeit at a much slower rate
than for the Lewis-acid-catalyzed reactions. Given the known
decomposition of chloroform to produce HCl and phosgene, it
seems likely that even this small amount of acid was sufficient
to promote the dehydration of the alcohol.
mers. From this result it was clear that in the case of strong
heterogeneous acid catalysts, oligomerization occurred more
rapidly than MCPD could be removed from the reaction flask.
To further investigate optimal dehydration conditions, a
series of weak Bronsted- and Lewis-acid catalysts were
screened to determine their activity in the selective dehydra-
tion of 1 (Table 2). Benzoic acid and dilute HCl were unselec-
tive and produced primarily ether, along with dimer and
trimer. Surprisingly, Pd(COD)Cl2 reacted almost quantitatively
and produced 66% dimer along with significant amounts of
trimer and tetramer. In the search for a milder dehydrating
agent, magnesium sulfate was employed as a catalyst and pro-
duced only ethers. In contrast to the other dehydration cata-
lysts that produced primarily one ether isomer, MgSO4 pro-
duced the two distinguishable ether isomers in nearly equal
amounts. This difference in isomer distribution is attributed to
the lack of suitable acid sites on the catalyst. In the absence of
these sites the reaction is driven by the coordination of water
to magnesium cations and is dependent on the auto-ionization
of the alcohol. Based on these initial screening results, an alu-
minum phosphate catalyst was prepared[19] and evaluated as a
dehydration catalyst. Under a variety of conditions, this catalyst
was selective for the production of only ethers, MCPD, and
dimers; no heavier oligomers were formed. Despite the favora-
ble product distribution, the conversion efficiency of this cata-
lyst was limited by the production of water in the dehydration
reaction. To overcome this hurdle, mixtures of AlPO4 with a
suitable drying agent were employed. An AlPO4/molecular
sieve catalyst resulted in a low overall yield of MCPD with for-
mation of an oligomeric mixture. In contrast, an AlPO4/MgSO4
catalyst permitted the direct conversion to MCPD. The opti-
mized catalyst allowed for a 78% isolated yield of isomeric
MCPD from 1.
Although the ruthenium catalysts showed some modest ac-
tivity for the partial dehydration of 1, more efficient and selec-
tive methods were sought to convert 1 to MCPD. Given the
rapid room-temperature conversion of MCPD to dimer, particu-
larly in the presence of acid catalysts, two distinct routes to
the dimer were conceived. In the first route, a solid acid cata-
lyst would be employed and the dehydration and dimerization
would occur in the same flask. In the second route, a dehydra-
tion catalyst of much lower acidity would be employed and
the reaction carried out under reduced pressure allowing the
volatile MCPD to be easily separated from the reaction mix-
ture. For the first route, heterogeneous solid acid catalysts
were employed to allow for easy isolation of the product.
Montmorillonite K10 (MMT-K10), an acid clay, and Nafion SAC-
13, a silica-supported perfluorinated cation exchange resin,
were screened for activity. Although both catalysts resulted in
high conversions (95% conversion in 1 h at ambient tempera-
ture), both yielded complex mixtures consisting of ether,
dimer, significant amounts of trimer, and other heavier oligo-
mers (Table 2). To try and trap MCPD prior to oligomerization,
the reaction was conducted with Nafion SAC-13 at 408C under
reduced pressure (40 torr). Although the isolated MCPD was
>90% pure, the yield was low and the reaction mixture rapid-
ly oligomerized to a thick orange oil composed of heavy oligo-
The dimer product distribution resulting from the room tem-
perature Diels–Alder cycloaddition of MCPD (Figure 3) is of sig-
nificant interest and is in part controlled by the starting com-
position of MCPD isomers. Dehydration of the alcohol with
AlPO4 at 608C yields 84% 2-methylcyclopentadiene (8) and
16% 1-methylcyclopentadiene (9), while 5-methylcyclopenta-
diene was not observed. This preference for 8 results from the
formation of a more stable tertiary carbocation compared to
ChemSusChem 2011, 4, 465 – 469
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
467